Interventions for treating proximal humeral fractures in adults

  • Conclusions changed
  • Review
  • Intervention

Authors


Abstract

Background

Fractures of the proximal humerus are common injuries. The management, including surgical intervention, of these fractures varies widely. This is an update of a Cochrane review first published in 2001 and last updated in 2010.

Objectives

To review the evidence supporting the various treatment and rehabilitation interventions for proximal humeral fractures.

Search methods

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register, the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE and other databases, and bibliographies of trial reports. The full search ended in January 2012.

Selection criteria

All randomised controlled trials pertinent to the management of proximal humeral fractures in adults were selected.

Data collection and analysis

Two people performed independent study selection, risk of bias assessment and data extraction. Only limited meta-analysis was performed.

Main results

Twenty-three small randomised trials with a total of 1238 participants were included. Bias in these trials could not be ruled out. Additionally there is a need for caution in interpreting the results of these small trials, which generally do not provide sufficient evidence to conclude that any non-statistically significant finding is 'evidence of no effect'.

Eight trials evaluated conservative treatment. One trial found an arm sling was generally more comfortable than a less commonly used body bandage. There was some evidence that 'immediate' physiotherapy compared with that delayed until after three weeks of immobilisation resulted in less pain and potentially better recovery in people with undisplaced or other stable fractures. Similarly, there was evidence that mobilisation at one week instead of three weeks alleviated short term pain without compromising long term outcome. Two trials provided some evidence that unsupervised patients could generally achieve a satisfactory outcome when given sufficient instruction for an adequate self-directed exercise programme.

Six heterogeneous trials, involving a total of 270 participants with displaced and/or complex fractures, compared surgical versus conservative treatment. Pooled results of patient-reported functional scores at one year from three trials (153 participants) showed no statistically significant difference between the two groups (standardised mean difference -0.10, 95% CI -0.42 to 0.22; negative results favour surgery). Quality of life based on the EuroQol results scores from three trials (153 participants) showed non-statistically significant differences between the two groups at three time points up to 12 months. However, the pooled EuroQol results at two years (101 participants) from two trials run concurrently from the same centre were significantly in favour of the surgical group. There was no significant difference between the two groups in mortality (8/98 versus 5/98; RR 1.55, 95% CI 0.55 to 4.36; 4 trials). Significantly more surgical group patients had additional or secondary surgery (18/112 versus 5/111; RR 3.36, 95% CI 1.33 to 8.49; 5 trials). This is equivalent to an extra operation in one of every nine surgically treated patients.

Different methods of surgical management were tested in seven small trials. One trial comparing two types of locking plate versus a locking nail for treating two-part surgical neck fractures found some evidence of better function after plate fixation but also of a higher rate of surgically-related complications. One trial comparing a locking plate versus minimally invasive fixation with distally inserted intramedullary nails found some evidence of a short-term benefit for the nailing group. Compared with hemiarthroplasty, tension-band fixation of severe injuries using wires was associated with a higher re-operation rate in one trial. Two trials found no important differences between 'polyaxial' and 'monaxial' screws combined with locking plate fixation. One trial produced some preliminary evidence that tended to support the use of medial support locking screws in locking plate fixation. One trial found better functional results for one of two types of hemiarthroplasty.

Very limited evidence suggested similar outcomes from early versus later mobilisation after either surgical fixation (one trial) or hemiarthroplasty (one trial).

Authors' conclusions

There is insufficient evidence to inform the management of these fractures. Early physiotherapy, without immobilisation, may be sufficient for some types of undisplaced fractures. It remains unclear whether surgery, even for specific fracture types, will produce consistently better long term outcomes but it is likely to be associated with a higher risk of surgery-related complications and requirement for further surgery.

There is insufficient evidence to establish what is the best method of surgical treatment, either in terms of the use of different categories of surgical intervention (such as plate versus nail fixation, or hemiarthroplasty versus tension-wire fixation) or different methods of performing an intervention in the same category (such as different methods of plate fixation). There is insufficient evidence to say when to start mobilisation after either surgical fixation or hemiarthroplasty.

Résumé scientifique

Interventions pour le traitement des fractures proximales de l'humérus

Contexte

Les fractures proximales de l'humérus sont des blessures courantes. La prise en charge de ces fractures, y compris par une intervention chirurgicale, varie considérablement. Cette revue est une mise à jour d'une revue Cochrane publiée pour la première fois en 2001 et mise à jour pour la dernière fois en 2010.

Objectifs

Examiner les éléments de preuve justifiant les différents traitements et les interventions de réadaptation pour les fractures proximales de l'humérus.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre spécialisé du Groupe Cochrane sur les traumatismes ostéo-articulaires et musculaires, le registre Cochrane des essais contrôlés, les bases de données MEDLINE, EMBASE et d'autres bases de données, ainsi que dans les bibliographies des rapports d'essais. La recherche complète a pris fin en janvier 2012.

Critères de sélection

Tous les essais contrôlés randomisés relatifs à la prise en charge des fractures proximales de l'humérus chez les adultes ont été sélectionnés.

Recueil et analyse des données

Deux personnes ont effectué la sélection des études, l'évaluation du risques de biais et l'extraction des données, de façon indépendante. Il n'a été effectué qu'une méta-analyse limitée.

Résultats principaux

Vingt-trois petits essais randomisés avec un total de 1238 participants ont été inclus. Dans ces essais, il s'est avéré impossible d'exclure la présence de biais. En outre, il est nécessaire d'être prudent dans l'interprétation des résultats de ces petits essais, qui, en règle générale, n'apportent pas suffisamment d'éléments de preuves pour conclure que toute décision non statistiquement significative est la « preuve d'un effet nul ».

Huit essais ont évalué un traitement conservateur. Un essai a mis en évidence qu'une écharpe est généralement plus confortable qu'un bandage du corps lequel est utilisé moins souvent. Certaines preuves indiquent que la kinésithérapie « immédiate » comparativement à la kinésithérapie n'ayant lieu qu'après trois semaines d'immobilisation a entraîné moins de douleur et une récupération potentiellement meilleure chez les personnes atteintes de fractures sans déplacement ou d'autres fractures « stables ». De même, on a constaté que la mobilisation à une semaine au lieu d'une mobilisation à trois semaines atténuait la douleur à court terme sans compromettre les résultats à long terme. Deux essais ont apporté des preuves que les patients livrés à eux-mêmes peuvent, en règle générale, obtenir un résultat satisfaisant lorsqu'on leur donne les instructions nécessaires pour leur permettre d'effectuer en autonome un programme d'exercices de rééducation.

Six essais hétérogènes, impliquant un total de 270 participants souffrant de fractures avec déplacement et/ou complexes, comparaient un traitement chirurgical versus un traitement conservateur. Les résultats regroupés des scores fonctionnels rapportés par les patients à un an, qui sont issus de trois essais (153 participants), n'ont montré aucune différence statistiquement significative entre les deux groupes (différence moyenne standardisée = -0,10, IC à 95 % de -0,42 à 0,22 ; résultats négatifs en faveur du traitement chirurgical). Pour la qualité de vie, d'après les scores EuroQol obtenus dans ces trois essais (153 participants), aucune différence statistiquement significative n'a été observée entre les deux groupes, à trois points temporels jusqu'à 12 mois. Cependant, les résultats EuroQol regroupés à deux ans (101 participants), qui sont issus de deux essais exécutés simultanément dans le même centre sont significativement en faveur du groupe avec traitement chirurgical. Aucune différence significative n'a été observée entre les deux groupes au plan de la mortalité (8/98 versus 5/98 ; RR = 1,55, IC à 95 % de 0,55 à 4,36 ; 4 essais). Un nombre de patients significativement plus élevé dans le groupe avec traitement chirurgical a subi une intervention chirurgicale supplémentaire ou secondaire (18/112 versus 5/111 ; RR = 3,36, IC à 95 % de 1,33 à 8,49 ; 5 essais). Cela correspond à une opération supplémentaire réalisée chez un patient sur neuf patients traités chirurgicalement.

Différentes méthodes de prise en charge chirurgicale ont été testées dans sept petits essais. Un essai comparant deux types de plaques à verrouillage versus clou huméral à verrouillage pour le traitement des fractures du col chirurgical à deux fragments a apporté des éléments de preuve d'une meilleure fonction après une ostéosynthèse par plaque, mais aussi d'un taux plus élevé de complications liées à l'intervention chirurgicale. Un essai comparant une plaque à verrouillage versus une fixation minimalement invasive avec insertion distale de clous centromédullaires a trouvé des preuves d'un bénéfice à court terme pour le groupe avec enclouage. Comparativement à l'hémiarthroplastie, la fixation par haubanage des lésions sévères à l'aide de cerclages a été associée à un taux plus élevé de ré-interventions dans un essai. Deux essais n'ont trouvé aucune différence importante entre les vis « polyaxiales » et « monoaxiales » associées à une ostéosynthèse par plaque à verrouillage. Un essai a produit certaines preuves préliminaires qui tendent à justifier l'utilisation de la mise en place de vis verrouillées dans la zone de la corticale médiale dans le cadre d'une ostéosynthèse par plaque. Un essai a mis en évidence de meilleurs résultats fonctionnels pour l'un des deux types d'hémiarthroplastie.

Les éléments de preuve suggérant des résultats similaires pour une mobilisation précoce comparativement à une mobilisation plus tardive après une ostéosynthèse chirurgicale (un essai) ou une hémiarthroplastie (un essai), étaient limités.

Conclusions des auteurs

Les éléments de preuve sont insuffisants pour permettre d'orienter la prise en charge de ces fractures. Une kinésithérapie précoce, sans immobilisation, peut être suffisante pour certains types de fractures sans déplacement. Il est difficile de déterminer si la chirurgie, même pour des types de fractures spécifiques, produira toujours de meilleurs résultats à long terme, mais elle est susceptible d'être associée à un risque plus élevé de complications liées à la chirurgie et de nécessiter une nouvelle intervention.

Il n'existe pas suffisamment de preuves pour établir quelle est la meilleure méthode de traitement chirurgical, pour ce qui est de l'utilisation des différentes catégories d'intervention chirurgicale (par exemple, ostéosynthèse par plaque versus clou, ou hémiarthroplastie versus ostéosynthèse par haubanage associé à un cerclage) ou pour les différentes méthodes de réalisation d'une intervention de la même catégorie (par exemple, les différentes méthodes d'ostéosynthèse par plaque). Il n'existe pas suffisamment de preuves pour indiquer à quel moment il y a lieu de commencer la mobilisation après une ostéosynthèse chirurgicale ou une hémiarthroplastie.

Plain language summary

Interventions for treating proximal humeral (top end of upper arm bone) fractures in adults

Fracture of the top end of the upper arm bone is a common injury in older people. It is often called a shoulder fracture. The bone typically fractures (breaks) just below the shoulder, usually after a fall. Most of these fractures occur without breaking of the skin. Often the injured arm can be supported in a sling until the fracture heals sufficiently to allow shoulder movement. More complex fractures may be treated surgically. This may involve fixing the fracture fragments together by various means. Alternatively, the top of the fractured bone may be replaced (half 'shoulder' replacement: hemiarthroplasty), or sometimes together with the joint socket (total 'shoulder' replacement). Physiotherapy is often used to help restore function.

This review includes evidence from 23 small randomised trials with a total of 1238 participants. All trials had weaknesses that could affect their results.

Eight trials evaluated conservative (non-surgical) treatment. One trial found an arm sling was generally more comfortable than a less commonly used body bandage. There was some evidence that 'immediate' physiotherapy compared with physiotherapy delayed until after three weeks of immobilisation resulted in less pain and faster recovery in people with 'stable' fractures. Similarly, there was evidence that mobilisation at one week instead of three weeks alleviated pain in the short term without compromising long term outcome. Two trials provided some evidence that patients could generally achieve a satisfactory outcome when given sufficient instruction to pursue exercises on their own.

Six trials, involving 270 participants with more severe fractures, compared surgical versus conservative treatment. Pooled results from three trials for patient-reported measures of function and quality of life were inconclusive. There was no difference between the two groups in mortality. However, more surgical group patients had additional or secondary surgery. This was equivalent to an extra operation in one of every nine surgically treated patients.

Seven trials tested different methods of surgical treatment. There was weak evidence of some differences (e.g. in complications) between some interventions (e.g. different devices or different ways of using devices).

There was very limited evidence suggesting similar outcomes for early versus later mobilisation after either surgical fixation or hemiarthroplasty.

Overall, there is some evidence to support earlier arm movement for less serious fractures. Otherwise, there is not enough evidence to determine the best treatment, including surgery, for these fractures.

Résumé simplifié

Interventions pour le traitement des fractures proximales de l'humérus (extrémité supérieure de l'os du bras) chez les adultes

La fracture de l'extrémité supérieure de l'os du bras est une blessure courante chez les personnes âgées. Elle est souvent appelée fracture de l'épaule. L'os, en règle générale, se fracture (casse) juste en dessous de l'épaule, le plus souvent après une chute. La plupart de ces fractures surviennent sans rupture de la peau. Souvent, le bras fracturé peut être soutenu dans une écharpe jusqu'à ce que la fracture se consolide suffisamment pour permettre le mouvement de l'épaule. Les fractures plus complexes peuvent être traitées chirurgicalement, ce qui peut impliquer la fixation des fragments osseux entre eux par divers moyens. En variante, la partie supérieure de l'os fracturé peut être remplacée (remplacement d'une seule partie de l'articulation de l'épaule : hémiarthroplastie), ou parfois avec le logement de l'articulation (remplacement total de « l'épaule »). La kinésithérapie est souvent utilisée pour aider à restaurer la fonction.

Cette revue comprend les éléments de preuve issus de 23 petits essais randomisés avec un total de 1238 participants. Tous les essais présentaient des faiblesses susceptibles d'avoir une incidence sur leurs résultats.

Huit essais ont évalué un traitement conservateur (non chirurgical). Un essai a mis en évidence qu'une écharpe est généralement plus confortable qu'un bandage du corps lequel est utilisé moins souvent. Certaines preuves indiquent que la kinésithérapie « immédiate » comparativement à la kinésithérapie n'ayant lieu qu'après trois semaines d'immobilisation a entraîné moins de douleur et une récupération plus rapide chez les personnes atteintes de fractures « stables ». De même, on a constaté que la mobilisation à une semaine au lieu de trois semaines atténuait la douleur à court terme sans compromettre les résultats à long terme. Deux essais ont fourni des preuves que les patients peuvent généralement obtenir un résultat satisfaisant lorsqu'on leur donne les instructions nécessaires pour poursuivre leurs exercices de rééducation tout seul.

Six essais portant sur 270 participants ayant des fractures plus sévères comparaient un traitement chirurgical versus un traitement conservateur. Les résultats regroupés de trois essais concernant les mesures rapportées par les patients de la fonction et de la qualité de vie n'étaient pas concluants. Aucune différence n'a été observée entre les deux groupes au plan de la mortalité. Toutefois, les patients du groupe avec traitement chirurgical ont subi une intervention chirurgicale supplémentaire ou secondaire. Cela correspondait à une opération supplémentaire réalisée chez un patient sur neuf patients traités chirurgicalement.

Sept essais ont testé différentes méthodes de traitement chirurgical. Quelques preuves légères indiquant certaines différences (par exemple, en termes de complications) ont été observées entre certaines interventions (par exemple, dispositifs différents ou différentes manières d'utiliser les dispositifs).

Les éléments de preuve suggérant des résultats similaires pour une mobilisation précoce comparativement à une mobilisation plus tardive après une ostéosynthèse chirurgicale ou une hémiarthroplastie étaient limités.

Globalement, il existe quelques preuves en faveur d'une mobilisation précoce du bras pour les fractures de moindre gravité. Sinon, il n'y a pas assez de preuves pour déterminer le meilleur traitement, y compris la chirurgie, pour ces fractures.

Notes de traduction

Traduit par: French Cochrane Centre 10th January, 2013
Traduction financée par: Minist�re Fran�ais des Affaires sociales et de la Sant�

Background

Description of the condition

Proximal humeral fractures account for approximately six per cent of all adult fractures (Court-Brown 2006). Their incidence rapidly increases with age, and women are affected about three times as often as men (Court-Brown 2006; Lind 1989). Many patients who sustain a proximal humeral fracture are old and their bones are osteoporotic. Court-Brown 2001 found that 87% of these fractures in adults resulted from falls from standing height. Palvanen 2006 found that the incidence of osteoporotic-related fractures of the proximal humerus in Finland had tripled between 1970 and 2002 to 105 per 100,000 people aged 60 or above.

Most proximal humeral fractures are closed fractures in that the overlying skin remains intact. The most commonly used classification of shoulder fractures is that of Neer (Neer 1970). Neer considered four potential segments of the proximal humerus - the articular part, the greater tuberosity, the lesser tuberosity and the humeral shaft. These may be affected by fracture lines but are only considered as a 'part' if displaced by more than one centimetre or 45 degrees angulation from each other. Fractures, regardless of the number of fracture lines present, which did not meet the criteria for displacement of any one segment with respect to the others were considered 'undisplaced' and categorised as one-part fractures. Neer's other categories, two-part, three-part and four-part fractures all involved the displacement of some or all of the above four segments. Each of these fracture types may be potentially associated with an anterior or posterior humeral head dislocation.

At initial presentation, it may be difficult to delineate the exact pattern of the fracture even with sophisticated imaging. In any event, this may not correlate with the extent to which the vascularity (blood supply) of the humeral head is compromised. The vascularity of the proximal humerus is a secondary focus of another widely used classification system for these fractures, the AO classification system (Muller 1991), which was updated in conjunction with the OTA classification in 2007 (Marsh 2007). There are three main types (A, B, C), which in turn are further divided into three groups, each with a further three subgroups. Type A fractures are "extra-articular, unifocal, with intact vascular supply"; type B fractures are "extra-articular, bifocal, with possible vascular compromise"; and type C fractures are "articular, with a high likelihood of vascular compromise" (Robinson 2008).

Many proximal humeral fractures are not displaced or only minimally displaced. Neer's estimate (Neer 1970) that approximately 85% of all proximal humeral fractures are "undisplaced", in that no bone fragment is displaced by more than one centimetre, or angulated by more than 45 degrees is often cited (Koval 1997). However, a lower figure of 49% was reported in a prospective study of over 1000 proximal humeral fractures (Court-Brown 2001).

Description of the intervention

Conservative treatment is generally the accepted treatment option for minimally displaced fractures, and frequently used for people with displaced fractures too. Conservative treatment usually involves a period of immobilisation, such as in an arm sling, followed by physiotherapy and exercises.

Surgery is usually reserved for displaced and unstable fractures and those with more complicated fracture patterns. Surgical interventions include:

  • Closed reduction and percutaneous stabilisation using pins or wires

  • External fixation

  • Open reduction and plating: for example, buttress plates, angle blade plates and proximal humeral locking plates

  • Open reduction and fixation using a tension-band principle

  • Intramedullary nailing either antegrade or retrograde insertion; nowadays, intramedullary nails are 'locked' into place, generally using screws

  • Hemiarthroplasty (replacement of the humeral head)

  • Total shoulder replacement (replacement of the entire joint; thus both the 'ball' (humeral head) and 'socket' (glenoid)). This includes reverse polarity arthroplasty, where the joint polarity is reversed such that the ball is on glenoid side and the socket on the humeral side.

Post-operative treatment generally involves a period of immobilisation followed by physiotherapy and exercises.

How the intervention might work

Immobilisation of the injured limb helps to maintain fracture stability and to provide pain relief during healing. However, there is a risk of the shoulder becoming stiff and painful with substantial reduction of function. Subsequent physiotherapy and exercises aim to restore function and mobility of the injured (or operated) arm. Malunion of proximal humeral fractures may result in impingement or compromised function of the muscles inserting into the proximal humerus.

After reduction or repositioning of the fractured parts, surgical fixation using various techniques aims to stabilise the reduced fracture and restore joint mechanics. Surgical stabilisation of the fracture may also allow earlier movement of the shoulder and elbow, preventing stiffness. Surgeons have often followed Neer's premise (Neer 1975) that in a four-part fracture head avascular necrosis is virtually guaranteed and have offered their patients a replacement arthroplasty, where the humeral head alone, or in combination with the socket, is replaced by artificial parts. An exception is often made for a specific type of four-part fracture, the valgus impacted four-part fracture, not mentioned in Neer's classification. This fracture, where the fractured parts are compressed towards each other, is less likely to lead to avascular necrosis of the humeral head, provided the lateral displacement of the head fragment is not excessive (Jakob 1991; Resch 1997). Bone quality also influences the appropriateness of any intervention and hence the long term clinical outcome. Furthermore, the patient's frailty may lead to a low rehabilitation drive and delay any recovery from both the initial trauma and any subsequent management.

Why it is important to do this review

Proximal humeral fractures are increasing in incidence, particularly in older people, and the short and long term consequences for individuals with these injuries and society are substantial (Palvanen 2006). There is considerable variation in practice, both in terms of definitive treatment such as surgical treatment for displaced fractures (Guy 2010) and rehabilitation (Hodgson 2006). The last two versions of this review noted the insufficiency of the evidence to inform practice, but also located ongoing trials that potentially could help address this deficiency (Handoll 2007; Handoll 2010). This update continues the systematic review of the evidence for managing these fractures.

Objectives

This review aims to determine the most appropriate treatment for fractures of the proximal humerus in skeletally mature people (adults).

We aimed to examine the evidence from randomised and quasi-randomised controlled trials for the effects (benefits and harms) of different treatment, including rehabilitation, interventions in adults with fractures of the proximal humerus. We defined a priori the following broad objectives:

  • To compare different methods of conservative treatment (including rehabilitation)

  • To compare surgical versus conservative treatment

  • To compare different methods of surgical treatment

  • To compare different methods of rehabilitation after surgical treatment

We planned to study the outcomes in different age groups (initially, under versus over 65 years) and for different types of proximal humeral fractures.

Methods

Criteria for considering studies for this review

Types of studies

All randomised or quasi-randomised (method of allocating participants to a treatment that is not strictly random; e.g. by hospital record number) trials which compared two or more interventions in the management of fractures of the proximal humerus in adults.

Types of participants

Patients of either sex who had completed skeletal growth, with a fracture of the proximal humerus. Stratification was planned by fracture type (e.g. based on the Neer classification (Neer 1970)) and by age (under versus over 65 years) if possible. Trials including children were included provided the proportion of children was small.

Types of interventions

Conservative and surgical interventions, as exemplified in Description of the intervention, used in the treatment and rehabilitation of fractures of the proximal humerus. Pharmacological trials were excluded.

Types of outcome measures

The primary focus is on long term functional outcome, preferably measured at one year or more.

Primary outcomes
  1. Functional outcomes: Patient-reported measures of upper-limb function (e.g. the Disability of the Arm, Shoulder, and Hand questionnaire (DASH) and other validated shoulder rating scales), activities of daily living and health related quality of life scores.

  2. Serious adverse events (e.g. death, deep infection, avascular necrosis, complex regional pain syndrome), and need for substantive treatment, such as an operation.

Secondary outcomes
  1. Composite scores of subjective and objectively rated function and overall outcome (e.g. Constant and Murley's score (Constant 1987); Neer's rating (Neer 1970)).

  2. Pain.

  3. Upper limb strength and range of movement.

  4. Less serious complications / adverse events of limited duration and impact (e.g. superficial infection, transient paraesthesia, skin irritation).

  5. Patient satisfaction with treatment, including cosmetic outcomes.

  6. Anatomical outcomes, e.g. radiological deformity.

Economic outcomes: each trial report was reviewed for costs and resource data, such as length of hospital stay and number of outpatient attendances, that would enable economic evaluation.

Search methods for identification of studies

Electronic searches

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (January 2012), the Cochrane Central Register of Controlled Trials (in The Cochrane Library 2012, Issue 1), MEDLINE (1966 to January week 2 2012), EMBASE (1988 to 2012 week 03), CINAHL (Cumulative Index to Nursing and Allied Health Literature)(September 2006 to 20 February 2012), AMED (Allied and Complementary Medicine) (1985 to April 2012), and PEDro - Physiotherapy Evidence Database (June 2012). The first two sections of the Cochrane optimal MEDLINE search strategy for randomised trials (Higgins 2006) were combined with the subject specific search (Appendix 1). Search strategies for the Cochrane Central Register of Controlled Trials, EMBASE, CINAHL and AMED can also be found in Appendix 1. For this update, the search results for the five databases were limited from 2010 onwards. Details of the search strategies used for previous versions of the review are given in Handoll 2007 and Handoll 2010. No language or publication restrictions were applied.

We searched the WHO International Clinical Trials Registry Platform Search Portal (April 2012), Current Controlled Trials (April 2012), and the UK National Research Register (NRR) Archive (August 2010) to identify ongoing and recently completed trials.

Searching other resources

We searched the reference lists of articles. We also included the findings from handsearches of the British Volume of the Journal of Bone and Joint Surgery supplements (1996 to 2006) and electronic searches of the Journal of Bone and Joint Surgery - British Volume and conference proceedings from 2006 onwards (26 April 2012) (see Appendix 1). We searched abstracts of the British Elbow and Shoulder Society annual meetings (2001 to 2011), the American Orthopaedic Trauma Association annual meetings (1996 to 2011), American Academy of Orthopaedic Surgeons annual meetings (2005 to 2006), the British Trauma Society Annual Scientific Meeting (2012), and the 53rd Congress of The Nordic Orthopaedic Federation 2006. We also included handsearch results from the final programmes of SICOT (1996 & 1999) and SICOT/SIROT (2003), the British Orthopaedic Association Congress (2000, 2001, 2002 and 2003), and various issues of Orthopaedic Transactions and supplements of Acta Orthopaedica Scandinavica.

Up to 2007, we scrutinised weekly downloads of "Fracture" articles in new issues of Acta Orthopaedica Scandinavica (subsequently Acta Orthopaedica); American Journal of Orthopedics; Archives of Orthopaedic and Trauma Surgery; Clinical Orthopedics and Related Research; Injury; Journal of the American Academy of Orthopaedic Surgeons; Journal of Arthroplasty; Journal of Bone and Joint Surgery (American and British Volumes); Journal of Orthopedic Trauma; Journal of Trauma; Orthopedics from AMEDEO.

Data collection and analysis

Selection of studies

Eligible trials were selected by one author (HH) from the outputs of the search strategies listed above. Additionally, independent screening of the results from several databases and conferences proceedings were additionally carried out by the two other review authors. The initial decisions of trial eligibility were based on citations and, where available, abstracts and indexing terms. Full articles were obtained and, where necessary to ascertain trial methods and status, one author (HH) sent requests for information to trial investigators. Trials appearing to involve random or quasi-random allocation of treatment inventions for proximal humeral fractures in adults were put forward for consideration by all of the review authors listed on the byline for the particular version of the review. Study inclusion was by consensus of all listed review authors.

Data extraction and management

A data extraction tool was developed, piloted and independently completed by pairs of review authors (HH and BO; HH and KR) for each newly included trial. Details of the study methods, participants, interventions and outcome assessment and results were recorded. Any differences that were clearly not transcription errors were discussed between review authors. Data management and entry into RevMan was mainly by one author (HH) with some data entry by KR and checks made by all three authors of this update. When necessary, additional details of trial methodology or data, or both were requested from trialists.

Assessment of risk of bias in included studies

Risk of bias was independently assessed, without masking of the source and authorship of the trial reports, by pairs of review authors (HH and BO; HH and KR) for newly included trials. Between rater and between versions consistency in assessment was checked by HH at data entry. All inter-rater differences were resolved by discussion. We used the tool outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008a). This tool incorporates assessment of randomisation (sequence generation and allocation concealment), blinding (of participants, treatment providers and outcome assessors), completeness of outcome data, selection of outcomes reported and other sources of bias. We considered subjective and functional outcomes (e.g. functional outcomes, pain, clinical outcomes, complications) and 'hard' outcomes (death, reoperation) separately in our assessment of blinding and completeness of outcome data. We assessed two additional sources of bias: bias resulting from major imbalances in key baseline characteristics (e.g. age, gender, type of fracture); and performance bias such as resulting from lack of comparability in the experience of care providers.

Additionally, we assessed four other aspects of trial quality and reporting that would help us judge the applicability of the trial findings. The four aspects were: definition of the study population; description of the interventions; definition of primary outcome measures; and length of follow-up.

Measures of treatment effect

For each trial, risk ratios and 95% confidence intervals were calculated for dichotomous outcomes, and mean differences and 95% confidence intervals were calculated for continuous outcomes. Standardised mean differences rather than mean differences were used when pooling data from continuous outcome measures based on different scoring schemes.

Unit of analysis issues

We remained aware of potential unit of analyses issues arising from inclusion of participants with bilateral fractures, and presentation of outcomes, such as total complications, by the number of outcomes rather than participants with these outcomes. There was just one patient with bilateral fractures (Kristiansen 1988) but there was insufficient information to quantify the small difference this would have made to study findings. We avoided the second described unit of analysis problem, mainly by reporting on incidences of individual complications.

Dealing with missing data

We contacted trialists for missing information, including for denominators and standard deviations. We performed intention-to-treat analyses where possible. We did not impute missing standard deviations.

Assessment of heterogeneity

We planned to assess heterogeneity for pooled data from comparable trials by visual inspection of the analyses along with consideration of the chi² test for heterogeneity and the I² statistic (Higgins 2003). The main quantitative assessment of heterogeneity was based on the I² statistic where the following interpretation from the Cochrane Handbook was used: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% considerable heterogeneity (Deeks 2011).

Assessment of reporting biases

There are insufficient data thus far to merit the production of funnel plots to explore publication bias. The search for trials via conference proceedings and trial registration, together with the contacting of authors for information of trial status and progress has provided some insights on unpublished trials, which generally were abandoned because of poor recruitment.

Data synthesis

It was intended that, where the data allowed, the results of comparable groups of trials would be pooled using both fixed-effect and random-effects models; the selection of the model for presentation was to be determined by the consideration of the extent of the heterogeneity.

Subgroup analysis and investigation of heterogeneity

We set out a priori two subgroup analyses: by age groups (initially, under versus over 65 years) and by types of fracture (initially, 'undisplaced' versus displaced based on the Neer classification). To test whether the subgroups are statistically significantly different from one another, we planned to inspect the overlap of confidence intervals and perform the test for subgroup differences available in RevMan.

Sensitivity analysis

We planned sensitivity analyses based on aspects of trial methodology; specifically the risk of bias associated with inadequate concealment of allocation, and to explore the effects of missing data, particularly for dichotomous outcomes.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.

Results of the search

For this update (the full search was completed in January 2012, but some, mainly ongoing, trials were identified subsequently), we screened a total of 368 records from the following databases: Cochrane Central Register of Controlled Trials (32), MEDLINE (34), EMBASE (120), CINAHL (72), AMED (37), PEDRo (16), WHO Trials Registry (32) and Current Controlled Trials (25). Tallies from screening other sources were not kept or not available.

Overall, 18 new studies were identified. Of these, seven were included (Ockert 2010; Olerud 2011a; Olerud 2011b; Smejkal 2011; Voigt 2011; Zhang 2011; Zhu 2011), four were excluded (Carbone 2012; Edelson 2008; Liao 2009; Zhang 2010), five were placed in ongoing trials (Boyer; Engelsma; HOMERUS; Stanley; TPHF) and two await assessment (Battistella 2011; Fjalestad (RCT proposal)). Further information was obtained for several studies in the previous version (Handoll 2010); this included the one year follow-up report of functional outcome for Fjalestad 2010, and information resulting in the exclusion of Sinopidis 2010, a former ongoing study. Further consideration of Shah, which was listed as an ongoing study, and Pullen 2007, which was awaiting classification, led to their exclusion: it is very unlikely that any further information will be obtained for these trials, including whether they started. Also excluded was Parnes 2005, another study awaiting classification in Handoll 2010; there is currently insufficient evidence to support this being a randomised trial.

Summaries of the trial populations of past and the present versions of this review as well as the changes between updates are presented in Appendix 2 .

In all, 23 trials are now included, 14 trials are listed as ongoing, 26 trials are excluded and three are in Studies awaiting classification.

Included studies

All 23 included trials were published as full reports in journals, their availability ranging from 1979 (Lundberg 1979) to 2012 (Fjalestad 2010). Additional information via other publications, conference abstracts, trial registration details and communications from trial investigators were available for 11 trials (Agorastides 2007; Fjalestad 2010; Hodgson 2003; Hoellen 1997; Lefevre-Colau 2007; Ockert 2010; Olerud 2011a; Olerud 2011b; Voigt 2011; Zhang 2011; Zyto 1997); these sometimes preceded the availability of the main report. Details of study methods, participants, interventions and outcome measurement for the individual studies are provided in the Characteristics of included studies and summarised below.

Design

Twenty-two trials were randomised clinical trials, although six of these (Hoellen 1997; Kristiansen 1988; Kristiansen 1989; Lundberg 1979; Stableforth 1984; Wirbel 1999) provided no details of their method of randomisation and thus the use of quasi-randomised methods for sequence generation cannot be ruled out. Rommens 1993 was a quasi-randomised trial using alternation for treatment allocation. Livesley 1992 was double-blinded.

Sample sizes

The 23 included trials involved a total of 1238 patients. Study size ranged from 20 participants (Bertoft 1984) to 86 participants (Hodgson 2003). One trial (Kristiansen 1989) included one person with bilateral fractures; the treatment allocation for this participant was unclear.

Setting

The 23 included trials were single centre studies conducted in 10 different countries: Austria (1 trial); Belgium (1); China (2); Czech Republic (1); Denmark (2); France (1); Germany (4); Norway (1); Sweden (6) and UK (4). (Though essentially a single centre trial, the interventions in Hodgson 2003 were undertaken at two centres within an NHS Trust in the UK.) Details of the timing and/or duration of trial recruitment provided for 19 included trials (see the Characteristics of included studies) show Stableforth 1984 to have the earliest start date (1970) and longest period of recruitment (11 years).

Participants

The majority of participants in each trial were women (67% to 88% of their trial population). Most participants were aged 60 and above; two trials (Livesley 1992; Wirbel 1999) included a small number of children. Twelve trials set lower age limits. In seven of these (Fialka 2008; Fjalestad 2010; Hodgson 2003; Hoellen 1997; Olerud 2011a; Olerud 2011b; Voigt 2011), the age limit restricted the population to older adults; the most extreme was Hoellen 1997, where only people who were 65 years or over were included. Zyto 1997 specified that participants should be "elderly". Five trials (Bertoft 1984; Hodgson 2003; Livesley 1992; Lundberg 1979; Revay 1992) included only non or minimally displaced fractures, whereas 15 (Agorastides 2007; Fialka 2008; Fjalestad 2010; Hoellen 1997; Kristiansen 1988; Ockert 2010; Olerud 2011a; Olerud 2011b; Smejkal 2011;Stableforth 1984; Voigt 2011; Wirbel 1999; Zhang 2011; Zhu 2011; Zyto 1997) selected only people with displaced fractures. The majority of fractures were minimally displaced in Kristiansen 1989 and Rommens 1993. Lefevre-Colau 2007 included either minimally displaced or stable impacted fractures. Fractures were graded using the Neer classification system (Neer 1970) in 20 trials, together with the AO classification system in Fialka 2008, Lefevre-Colau 2007 and Smejkal 2011. A modification of the AO classification system as described in Wirbel 1999, and a specific classification system was not referred to in the remaining two trials (Bertoft 1984; Rommens 1993).

Interventions

Ten trials evaluated conservative treatment; though this was post-operative treatment in two of these. Six trials compared surgical with conservative treatment and seven compared two methods of surgery. A list of the comparisons, patient numbers and associated trials grouped according to the main objectives presented in the Objectives is given below.

Methods of conservative management (including rehabilitation)
Initial treatment, including immobilisation
  • "Immediate" physiotherapy within one week of fracture versus delayed physiotherapy after three weeks of immobilisation in a collar and cuff sling: Hodgson 2003 (86 participants).

  • Immobilisation in sling and body bandage for one week versus three weeks: Kristiansen 1989 (85 participants).

  • Physiotherapy started within three days of fracture versus delayed physiotherapy after three weeks of immobilisation in a sling: Lefevre-Colau 2007 (74 participants).

  • Gilchrist bandage versus "classic" Desault bandage: Rommens 1993 (28 participants).

Continuing management (rehabilitation) after initial conservative treatment involving sling immobilisation
  • Instructed self-physiotherapy versus conventional physiotherapy: Bertoft 1984 (20 participants); Lundberg 1979 (42 participants).

  • Swimming pool treatment plus self-training versus self-training alone: Revay 1992 (48 participants).

  • Apparatus supplying pulsed electromagnetic field versus dummy apparatus: Livesley 1992 (48 participants).

Surgical treatment versus conservative treatment

The currently available trials fall into two subcategories but are treated all together in Effects of interventions.

Fracture fixation versus conservative treatment
  • Percutaneous reduction and external fixation versus closed manipulation and sling: Kristiansen 1988 (30 participants).

  • Internal fixation using surgical tension band or cerclage wiring versus sling: Zyto 1997 (40 participants; three more were recorded in Tornkvist 1995).

  • Surgery involving open reduction and fixation with an interlocking plate device and metal cerclages versus conservative treatment starting with immobilisation of the injured arm in a modified Velpeau bandage: Fjalestad 2010 (50 participants).

  • Surgery involving open reduction and fixation with a PHILOS plate and nonabsorbable sutures versus conservative treatment starting with arm immobilisation in a sling: Olerud 2011a (60 participants).

Arthroplasty versus conservative treatment
  • Hemi-arthroplasty versus closed manipulation and sling: Stableforth 1984 (32 participants).

  • Humeral head replacement with the Global Fx prosthesis versus conservative treatment starting with arm immobilisation in a sling: Olerud 2011b (55 participants).

Different methods of surgical management
Comparisons of different categories of surgical intervention
  • Open reduction with internal fixation using a locking plate (LPHP or PHILOS) versus a locking nail (PHN): Zhu 2011 (57 participants).

  • Open reduction and internal fixation using a PHILOS plate versus the Zifko method of minimally invasive fixation with intramedullary K-wire insertion (distally inserted): Smejkal 2011 (61 participants).

  • Hemi-arthroplasty versus tension band wiring: Hoellen 1997 (30 participants); an additional nine participants were reported in Holbein 1999.

Comparisons of different methods of performing an intervention in the same category
  • Polyaxial versus monoaxial locking plate fixation. NCB-PH plate versus PHILOS plate: Ockert 2010 (76 participants); and HSP plate versus PHILOS plate: Voigt 2011 (56 participants).

  • Open reduction with internal fixation with PHILOS locking plate with or without the use of medial support locking screws: Zhang 2011 (72 participants).

  • Hemiarthroplasty using EPOCA prosthesis versus hemiarthroplasty using HAS prosthesis: Fialka 2008 (40 participants).

Continuing management (including rehabilitation) after surgical intervention
  • Immobilisation in sling for one week versus three weeks after percutaneous fixation: Wirbel 1999 (77 participants).

  • Early active assisted mobilisation (after two weeks) versus late mobilisation (after six weeks) after cemented hemiarthroplasty: Agorastides 2007 (59 participants).

Outcomes

Many trials in previous versions of this review preceded the availability of validated patient-reported outcome measures (e.g. DASH, Oxford Shoulder Score (Dawson 1996)) for assessing function. Data for these outcomes have become available from several trials (Fjalestad 2010; Olerud 2011a; Olerud 2011b; Voigt 2011) in this update. Otherwise, all trials except Ockert 2010 assessed functioning and pain, but usually reported these as part of a combined overall assessment, such as that of Neer (Neer 1970) and Constant (Constant 1987), that included other measures. Most trials reported on adverse events or complications. Exceptionally, Fjalestad 2010 reported on costs. Livesley 1992 did not provide outcomes split by treatment group.

Excluded studies

Brief details and reasons for exclusion for 26 studies are given in the Characteristics of excluded studies. It is noteworthy that 11 excluded studies were trials that were registered (usually in the now archived National Research Register, UK) but either did not take place (Mechlenburg 2009) or were abandoned due to lack of or poor recruitment (Brownson 2001; Dias 2001; Flannery 2006; Hems 2000; Sinopidis 2010; Wallace 2000; Welsh 2000) or perhaps both of these (Pullen 2007); or not put forward for publication due to compromised methods or data (Bing 2002; Martin 2000). Edelson 2008 also reported an abandoned randomised trial because of lack of patient consent.

Ongoing studies

Details of the 14 ongoing trials are given in the Characteristics of ongoing studies. Of particular note are the five ongoing multi-centre trials (Brorson; Guy; ProCon; ProFHER; TPHF), which together with a further single-centre trial (Helsinki), compare surgical versus non-surgical intervention in people with displaced fractures.

Studies awaiting classification

Three studies await classification: see details in the Characteristics of studies awaiting classification. The full report of Luo 2008, which tests acupuncture and reports limited findings at one month follow-up, is in Chinese and we will seek translation of this article for a future update. There are insufficient data for Battistella 2011, reported in a conference abstract only, and the third study was, in May 2012, still under consideration (Fjalestad (RCT proposal).

New studies found at this update

Seven trials, including a total of 437 participants, were newly included in this update. Two trials compared surgical with conservative treatment (Olerud 2011a; Olerud 2011b), and the other five trials compared different methods of surgery (Ockert 2010; Smejkal 2011; Voigt 2011; Zhang 2011; Zhu 2011).

Risk of bias in included studies

The risk of bias judgements on nine items for the individual trials are summarised in Figure 1 and described in the risk of bias tables in the Characteristics of included studies. A 'Yes' (+) judgement means that the authors considered there was a low risk of bias associated with the item, whereas a 'No' (-) means that there was a high risk of bias. The majority of assessments resulted in an 'Unclear' (?) verdict; this often reflected a lack of information upon which to judge the item (see Figure 2). However, lack of information on blinding for functional outcomes was always taken to imply that there was no blinding and rated as a 'No'.

Figure 1.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figure 2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

Seven trials (Bertoft 1984; Fjalestad 2010; Lefevre-Colau 2007; Olerud 2011a; Olerud 2011b; Smejkal 2011; Voigt 2011) were judged at low risk of selection bias resulting from adequate sequence generation and allocation concealment; and a further two trials (Hodgson 2003; Livesley 1992) also took adequate measures to safeguard allocation concealment. Based on its post-randomisation application of exclusion criteria, Ockert 2010 was judged at high risk of selection bias; as was Rommens 1993, which was a quasi-randomised trial using alternation.

Blinding

A low risk of detection bias for functional outcomes resulting from assessor and participant blinding was judged for Livesley 1992, which used sham controls. While several other trials reported blinded assessors, the lack of reporting of adequate safeguards and the lack of blinding of participants or care providers meant that the risk of bias was considered unclear. No blinding was reported or otherwise indicated in 14 trials.

Incomplete outcome data

Five trials (Fjalestad 2010; Hodgson 2003; Olerud 2011a; Olerud 2011b; Zhu 2011) were considered to be at low risk of bias from the incompleteness of data on functional outcomes. Eleven trials were deemed at high risk of bias, usually reflecting large losses to follow-up and post-randomisation exclusions.

Selective reporting

The lack of trial registration details and protocols hindered the appraisal of the risk of bias from selective reporting. Four trials (Agorastides 2007; Hoellen 1997; Livesley 1992; Zyto 1997) were considered at high risk of selective reporting bias.

Other potential sources of bias

Baseline characteristics

No trial was considered at high risk of bias because of confounding resulting from major imbalances in baseline characteristics. However, low risk of bias judgements were given for only seven trials (Kristiansen 1988; Lefevre-Colau 2007; Lundberg 1979; Olerud 2011a; Olerud 2011b; Wirbel 1999; Zyto 1997).

Care programmes

Risk of performance bias from important differences in care programmes other than the trial interventions or differences in the experience of care providers was judged either low (12 trials) or unclear, usually based on inadequate information, in the other 11 trials.

Effects of interventions

Where available, outcome data reported at final follow-up for individual trials are presented in the analyses.

Methods of conservative management

Initial treatment, including immobilisation

Four trials (Hodgson 2003; Kristiansen 1989; Lefevre-Colau 2007; Rommens 1993) reported outcome following initial treatment for mainly non or minimally displaced ('undisplaced') fractures.

Early mobilisation versus delayed mobilisation

Although three trials (Hodgson 2003; Kristiansen 1989; Lefevre-Colau 2007) compared early versus delayed mobilisation, the timing of the start of early mobilisation varied as did the nature and intensity of the physiotherapy provided (notably, the long (two hour) duration of individual physiotherapy sessions of Lefevre-Colau 2007). Lack of comparable outcome measurement and data precluded data pooling and so the results of the individual trials are presented separately below.

Hodgson 2003 compared commencing physiotherapy within one week of fracture versus delayed physiotherapy after three weeks of immobilisation in a collar and cuff sling in 86 people with undisplaced fractures. The results, presented in Hodgson 2007, for self-reported shoulder disability using the Croft Shoulder Disability Questionnaire (Croft 1994) show a tendency for less disability in the early mobilisation group at one year (severe disability: 13/42 versus 17/40; RR 0.73, 95% CI 0.41 to 1.30), continuing improvement and recovery between one and two years, and also reveal that, overall, a substantial proportion of participants continued to report some or severe disability at two years (see Analysis 1.1). Results at two years for eight of the 22 questions of the Croft questionnaire are shown in Analysis 1.2; these are presented to give an indication of the variety of problems experienced by these patients. Only the difference in the numbers reporting pain on movement was statistically significant, but this needs to be viewed in the context of the overall lack of statistically significant differences in other aspects of disability. There was, however, some evidence supporting a quicker recovery in the early group. Trial participants given early physiotherapy attended significantly fewer treatment sessions (see Analysis 1.3: mean difference (MD) -5.00 sessions; 95% confidence interval (CI) -8.25 to -1.75) until they and their physiotherapists agreed that independent shoulder function had been achieved. As can be seen in Analysis 1.4, participants of the early group had significantly better health-related quality of life scores at 16 weeks in two dimensions of the SF36 (role limitation physical: MD 22.20, 95% CI 3.82 to 40.58; and pain: MD 12.10, 95% CI 3.26 to 20.94). There were no statistically significant differences between the two treatment groups in the other six dimensions (e.g. physical functioning) of the SF36 at 16 weeks, and in all eight dimensions at one year. There were no complications arising from fracture displacement. The only recorded complication in the trial was a frozen shoulder in a participant of the delayed physiotherapy group (see Analysis 1.5). Shoulder function, relative to the unaffected shoulder, measured using the Constant score (Constant 1987) was statistically significantly better at eight and 16 weeks (see Analysis 1.6: mean difference in ratio affected/unaffected arm 0.16; 95% CI 0.07 to 0.25). Again, the differences at one year, though still favouring the early group, were not statistically significant (MD 0.07, 95% CI -0.03 to 0.17).

Kristiansen 1989, which tested the duration of immobilisation in a sling and body bandage (one week versus three weeks) in 85 people with mainly undisplaced fractures, provided insufficient follow-up data to allow any test for statistical significance. The authors reported that while pain, function and mobility at six months and over were similar in both groups, the patients who started early mobilisation at one week suffered less pain in the first three months than those who kept their bandaging for three weeks. One case of reflex sympathetic dystrophy occurred in each group (see Analysis 1.5).

Lefevre-Colau 2007 compared commencing physiotherapy within three days of fracture versus delayed physiotherapy after three weeks of immobilisation in a sling in 74 people with undisplaced or "stable" impacted fractures. Ten trial participants withdrew from the trial because of difficulties in reaching the hospital for treatment. Participants were discharged from physiotherapy at six months. Shoulder function measured using the Constant score was statistically significantly better in the early group at six weeks and three months (see Analysis 1.7); the differences at six months and end of treatment, though favouring the early group, were not statistically significant (MD 6.10, 95% CI -0.22 to 12.42). Although the early group had significantly reduced pain compared with the three weeks group by three months follow-up, there was no difference at six months (see Analysis 1.8). Active range of motion, measured relative to the opposite arm, also did not differ significantly between the two groups at six months (see Analysis 1.9). There were no cases of fracture non-union or displacement. One participant of each group received treatment for subacromial impingement (see Analysis 1.5). All participants attended at least 70% of the supervised physiotherapy sessions; and very few expressed dissatisfaction with their treatment (see Analysis 1.10).

Gilchrist versus the Desault bandage

Rommens 1993 compared the use of two types of bandage, the Gilchrist versus the Desault, worn for two to three weeks in 28 patients with undisplaced fractures. Reporting up until fracture consolidation, Rommens 1993 reported, without presenting data, that they had found no differences in the end result, either in terms of functional outcome or fracture healing. More people found the initial application of a Desault bandage uncomfortable and severe skin irritation prompted premature removal of the bandage in two people of this group (see Analysis 2.1). Pain during immobilisation was also reported to be greater in the Desault group. Slight displacement of the fracture in the first week was reported in two participants of the Gilchrist group (see Analysis 2.2). At fracture consolidation, patients' rating of their assigned bandage was significantly more favourable in the Gilchrist group (see Analysis 2.3 "Poor or bad rating by patient at fracture consolidation": 2/14 versus 8/14; risk ratio (RR) 0.25, 95% CI 0.06 to 0.97).

Continuing management (rehabilitation) after initial conservative treatment involving sling immobilisation

Two small trials (Bertoft 1984; Lundberg 1979) compared self-directed treatment following a course of instruction versus conventional physiotherapy during the 12 weeks following trauma in a total of 62 patients with non or minimally displaced (undisplaced) fractures. In both trials there were no statistically significant differences between those receiving instruction for exercises at home and those undergoing supervised physiotherapy in any of the outcomes recorded (see Analysis 3.1, Analysis 3.2, Analysis 3.3, Analysis 3.4, Analysis 3.5 and Analysis 3.6). It should be noted that since Lundberg 1979 did not report whether there had been any loss to long-term follow-up at an average of 16 months, the results for Neer's score presented in Analysis 3.5 are for illustrative purposes only.

Revay 1992, which included 48 participants with undisplaced fractures, reported that the addition of supervised exercises in a swimming pool to self-treatment did not enhance long term outcome. Participants of the control group (self-treatment only) were reported as having significantly better functional movements, joint mobility and activities of daily living at two and three month follow-up. However, there were no significant differences at one year. Revay 1992 suggested that those using the pool may have neglected their home exercises, but the authors did not evaluate compliance.

Livesley 1992, which included 48 patients with minimally displaced fractures, reported that there was no difference in outcome between the two groups (pulsed electromagnetic high frequency energy (PHFE) versus placebo) at any stage of the trial, but provided no quantitative data. All trial participants were reported as achieving a "good" result as converse to a "poor" one.

Surgical treatment versus conservative treatment

Six heterogeneous trials, with a total of 270 participants, evaluated surgical intervention for displaced fractures, generally three- or four-part fractures (Neer classification). Table 1 gives a brief summary of their characteristics. The close similarity in the methods of the two most recent trials (Olerud 2011a; Olerud 2011b) ensured that there are now data for pooling, including for the primary outcomes. This, however, presented a dilemma as to whether to restructure this section to address the main comparison (surgery versus conservative treatment) and report the outcomes in turn or, in keeping with the considerable clinical heterogeneity including in the choice of surgical intervention (fixation versus arthroplasty) and general lack of comparable outcome measures, to keep to the previous structure and present the results of the trials in turn. Mindful of future updates, with the eventual inclusion of data for potentially over 1000 participants from the current ongoing trials (Brorson; Guy; Helsinki; ProCon; ProFHER; TPHF) on this question, we took the first option.

Table 1. Surgical versus conservative treatment trials: brief characteristics
StudyParticipants
(Neer classification)
SurgeryConservative
(starting with)
Follow-up

Fjalestad 2010

 

50 patients with 3- or 4-part fractures

(Norway)

Open reduction and fixation with an interlocking plate device and metal cerclagesImmobilisation of the injured arm in a modified Velpeau bandage. Closed reduction in 8 patients.1 year so far (2 years in all)

Kristiansen 1988

 

30 patients with 31 with 2-, 3- or 4-part fractures
(Denmark)

Percutaneous reduction and external fixation

 

Closed manipulation and sling immobilisation2 years
Olerud 2011a60 patients with 3-part fractures (all had displaced surgical neck fracture)
(Sweden)
Open reduction and fixation with a PHILOS plate and non-absorbable suturesSling immobilisation2 years
Olerud 2011b55 participants with 4-part fractures
(Sweden)
Humeral head replacement with the Global Fx prosthesisSling immobilisation2 years
Stableforth 198432 participants with 4-part fracture
(UK)
Hemi-arthroplastyClosed manipulation and sling6 months
Zyto 1997 40 patients with 3- or 4-part fractures (3 others excluded)
(Sweden)
Internal fixation using surgical tension band or cerclage wiringSling immobilisation50 months
Primary outcomes

Pooled results of two different patient-reported functional scores reported by three trials (Fjalestad 2010; Olerud 2011a; Olerud 2011b) at 12 months follow-up showed no statistically significant difference between the two groups (standardised mean difference (SMD) -0.10, 95% CI -0.42 to 0.22; higher = worse function; 153 participants, see Analysis 4.1). Pooled DASH scores from Olerud 2011a and Olerud 2011b showed no statistically significant differences between the two groups at four months, or at one or two years (see Analysis 4.2) but there was a potential trend to a more favourable result after surgery (DASH (0 to 100: worst function) at 24 months: MD -7.43; 95% CI -16.26 to 1.41; 99 participants). As reported in Fjalestad 2012, Fjalestad 2010 found no significant differences between the two groups in the American Shoulder and Elbow Surgeons (ASES) scores at either six months or 12 months follow-up (ASES (0 to 24: best function), MD -0.70, 95% CI -4.52 to 3.12; see Analysis 4.3). (A full report of the two year follow-up and functional outcome data of Fjalestad 2010 is pending.) There were no statistically significant differences in subjective assessment of function between the two groups of Zyto 1997 at either one or three years (see Analysis 4.4).

Quality of life based on the EuroQol scores from three trials (Fjalestad 2010; Olerud 2011a; Olerud 2011b) showed non-statistically significant differences between the two groups at three time points (3 to 4 months, 6 months and 12 months; see Analysis 4.5). However, the pooled results at two years from the two Olerud trials were in favour of the surgical group (MD 0.15; 95% CI 0.05 to 0.25, 101 participants; see Analysis 4.5). A separate breakdown of the results from Fjalestad 2010, which include the number of QALYs (quality of life years), showed no differences in any quality of life outcomes for this trial (see Analysis 4.6).

There was no significant differences between the two groups in mortality (8/98 versus 5/98; RR 1.55, 95% CI 0.55 to 4.36; 4 trials, see Analysis 4.7). Where reported, none of the deaths was related to their fracture or treatment. Notably, the two deaths that occurred within three months of surgery in Fjalestad 2010 were people with underlying health problems. In Zyto 1997, eight of the 11 missing participants had died at 50 months, but no information on group allocation or causes of death was provided.

Significantly more surgical group patients had additional or secondary surgery (18/112 versus 5/111; RR 3.36, 95% CI 1.33 to 8.49; 5 trials; see Analysis 4.8). In Fjalestad 2010, treatment failure resulting in a operation occurred in four surgical group participants, one of whom had re-fixation plus bone grafting at six months and three whose implants were removed because of screw penetration into the joint space; and one conservatively treated patient, who had surgery because of fracture redisplacement at two weeks. In Kristiansen 1988, the three cases of treatment failure were the removal of pins due to infection in one surgical group participant and a change of method resulting from a poor initial fracture reduction in two conservative group participants. The reasons for re-operations in the surgical group of Olerud 2011a were deep infection (2 cases), non-union (1), impingement (2), avascular necrosis (1), screw penetration into joint (1) and stiffness (2). One conservatively-treated patient in Olerud 2011a had surgery because of impingement. Not included in this analysis is another conservatively-treated patient with non-union who abstained from surgery partly because of a late diagnosis of axillary nerve palsy. The reasons for additional surgery in Olerud 2011b were screw penetration of the joint (for one patient treated with a locking plate), stiffness and impingement and displaced greater tuberosity respectively in three surgical group patients, and for complete displacement of the humeral shaft without bony contact in one conservatively treated patient. Not included in this analysis is another conservatively-treated patient who refused surgery for a non-union. In Stableforth 1984, one surgical group participant had their prothesis removed because of a deep infection. There were no data in the six trials on serious adverse events such as complex regional pain syndrome requiring extensive, non-surgical, therapy.

The numbers of people in each group with one or more adverse events or complications were not available for any of the six trials. Analysis 4.9 presents the available data for individual complications. Unsurprisingly, surgery-related complications (e.g. infection and screw penetration of the joint) were predominant in the surgery treatment group. None of the differences between the two groups in the radiologically detected outcomes of non-union, avascular necrosis and signs of osteoarthritis were statistically significant. For these three outcomes, data favouring surgery from Olerud 2011b needs to be seen in the context that these were only likely to be detected in conservative treated patients given that surgery involved the replacement of the humeral head. Additionally, some of these outcomes were without symptoms or minor in extent. For instance, both cases of nonunion in the conservative treatment group of Fjalestad 2010 were without symptoms.

Secondary outcomes

The differences between the two groups in the Constant scores at four different time points (4, 12, 24 and 50 months) were all non-significant (e.g. the analysis with the most data was for 12 months: MD 2.36, 95% CI -3.52 to 8.24; 152 participants, 3 trials; see Analysis 4.10). The same lack of differences between the two groups applied to the Constant scores of the injured arm in Fjalestad 2010 at both six and 12 months follow-up (see Analysis 4.11). Of those followed up to one year in Kristiansen 1988, fewer, but not statistically significantly fewer, participants of the surgical group had a poor or unsatisfactory rating of function (see Analysis 4.12: 3/11 versus 6/10; RR 0.45, 95% CI 0.15 to 1.35). In Stableforth 1984, significantly fewer participants of the prosthesis group needed some help with activities of daily living or had died by six months (see Analysis 4.13: 2/16 versus 9/16; RR 0.22, 95% CI 0.06 to 0.87).

Pooled results from two trials (Olerud 2011a; Olerud 2011b) showed some non-significant tendency to reduced pain at two year follow-up in the surgery group (MD -6.38; 95% CI -14.18 to 1.41; 101 participants; see Analysis 4.14). Nearly all trial participants in Stableforth 1984 had shoulder pain but significantly fewer in the prosthesis group reported constant pain that impaired sleep or function (see Analysis 4.14: 2/15 versus 9/15; RR 0.22, 95% CI 0.06 to 0.86). The categorisation of pain is not clear in the trial report nor whether pain was assessed for all participants. Assuming the latter is the case, then this result is no longer statistically significant when all those with more than occasional pain are included (4/15 versus 9/15; RR 0.44, 95% CI 0.17 to 1.13; analysis not shown). Zyto 1997, which provided a breakdown of the Constant score into the separate components (activities of daily living, pain, range of motion, strength), found no significant difference between the two groups in pain at 50 months (see Analysis 4.15).

Reduced muscle strength and restricted mobility were less frequent in the prosthesis group survivors of Stableforth 1984 (see Analysis 4.17 and Analysis 4.18). Zyto 1997 found no difference between the two groups in strength ('power') at 50 months follow-up. Though statistically significant, the clinical relevance of the three point difference in the range of motion component of the Constant score is questionable (see Analysis 4.15).

Fjalestad 2010 found no differences at one year between the two groups in costs (see Analysis 4.19 and Analysis 4.20).

Different methods of surgical management

Comparisons of different categories of surgical intervention

Three trials (Hoellen 1997; Smejkal 2011; Zhu 2011) compared different methods of surgical management.

Open reduction with internal fixation using a locking plate versus a locking nail

Zhu 2011 compared open reduction with internal fixation using a locking plate (LPHP or PHILOS) versus a locking nail (PHN) in 57 participants with two-part surgical neck fractures. The American Shoulder and Elbow Surgeon's scores were significantly better in plate group at one year (MD 7.20; 95% CI 1.48 to 12.92) and three years (MD 4.00; 95% CI 0.01 to 7.99) (see Analysis 5.1). One participant of the nail group died of unrelated causes. While complications were not described in full, significantly more patients in the plate group had a complication (9/29 versus 1/28; RR 8.69, 95% CI 1.18 to 64.19; see Analysis 5.2). This included five patients in the plate group who had a re-operation for screw penetration into the joint surface of the humeral head. Zhu 2011 found a statistically significant difference in favour of the plate group in pain at one year but not at three years (see Analysis 5.3). There were no statistically significant differences between the two groups in the Constant scores at the two follow-up times (see Analysis 5.4) or in range of motion measures at either one year (not shown) or three years (see Analysis 5.5 and Analysis 5.6). Although the plate group had greater muscle strength at one year, the difference between the two groups was no longer statistically significant at three years (see Analysis 5.7). Both duration of surgery (MD 24.90 minutes, 95% CI 5.97 to 43.83 minutes) and blood loss were statistically significantly greater in the plate group (see Analysis 5.8). Consistent with the finding of an increased blood loss in the plate group, more people in this group had a blood transfusion but the difference between the two groups was not statistically significant (see Analysis 5.9).

Open reduction with internal fixation using a locking plate versus minimally invasive fixation with distally inserted intramedullary nails

Smejkal 2011 compared open reduction and internal fixation using a PHILOS plate versus the Zifko method of minimally invasive fixation with distally inserted intramedullary nails in 61 participants with two or three part fractures. Smejkal 2011 did not report patient-reported function or activities of daily living. Perhaps partly from difficulties in translation from Czech, but the account of the complications seemed incomplete, with no indication of how many required a re-operation. There was no significant difference between the two groups in the overall numbers of participants incurring a complication (11/28 versus 9/27; RR 1.18, 95% CI 0.58 to 2.38; see Analysis 6.1 The recorded nature of the complications reflected the type of implant, with four cases of screw protrusion in the plate group that resulted in impingement and migration of k-wires (Kirschner wires), a distal humeral fracture and a nerve injury in the Zifko group. Smejkal 2011 found no difference between the two groups in Constant scores relative to the healthy limb at a mean two years follow-up (MD -0.81%, 95% CI -7.45% to 5.83%; see Analysis 6.2). Three participants of each group had a 'poor' Constant score. Analysis 6.3 shows there were no statistically significant differences between the two groups in time to union (MD 2.10 weeks, 95% CI -2.25 to 6.45 weeks) or in a vaguely-described measure of time to recover normal upper limb function (27.2 versus 21.4 weeks; MD 5.80 weeks; 95% CI -0.16 to 11.76 weeks). Smejkal 2011 suggested that the greater time to recover in the plate group reflected greater impact of complications in this group. The duration of operation was significantly greater in the plate group (MD 44.74 minutes, 95% CI 32.23 to 57.25 minutes; see Analysis 6.4), but with a non-significant tendency for less X-ray exposure. The tendency for longer hospital stays for plate group patients did not achieve statistical significance (MD 1.20 days; 95% CI -0.34 to 2.74; see Analysis 6.5).

Hemi-arthroplasty versus tension band wiring

Hoellen 1997 compared humeral head replacement with an endoprosthesis (internally placed implant) against reduction and stabilisation of the fracture using tension band wiring. All 30 patients reported in Hoellen 1997 had four-part fractures. Patients with three-part fractures were also eligible according to a later report of the trial (Holbein 1999), which reported on 39 patients. However, until we obtain further information from the trialists, we will continue to report the results from Hoellen 1997. In Hoellen 1997, results for only 18 of the 30 trial participants were available at one year. There were no serious peri-operative or post-operative complications such as pulmonary embolism. No participants of the replacement group required further surgery compared with five participants of the osteosynthesis group (the wires displaced in four participants and the fracture completely dislocated in one participant): RR 0.09, 95% CI 0.01 to 1.51 (see Analysis 7.1 and Analysis 7.2). The mean Constant scores (minus the power component) for the 18 people available at one year follow-up were similar in the two groups (48 versus 49 points out of a maximum of 75). Though we have not obtained clarification on the inadequately reported results presented in Holbein 1999, these did not appear to differ in a major way from those in Hoellen 1997.

Comparisons of different methods of performing an intervention in the same category

Four trials (Fialka 2008; Ockert 2010; Smejkal 2011; Voigt 2011) compared different types or methods in the same intervention category (e.g. plating).

Polyaxial versus monoaxial locking plate fixation

Two trials (Ockert 2010; Voigt 2011) made this comparison. Ockert 2010, which reported on outcome for 66 patients with Neer two, three and four-part fractures, did not report on functional outcome. Voigt 2011 found no statistically significant differences at one year (48 patients with Neer three or four-part fractures) between the two groups in their DASH scores (see Analysis 8.1: RR 2.10, 95 CI -6.24 to 10.44), nor at three, six or 12 months in the 'simple shoulder test' (see Analysis 8.2). Neither trial found statistically significant differences between the two groups in participants having a re-operation (see Analysis 8.3: 8/54 versus 7/68; RR 1.43, 95% CI 0.56 to 3.67). In Ockert 2010, one participant of the polyaxial group had a loosened screw taken out at 10 weeks; one participant of each group had early hardware removal (at five months) because of subacromial impingement from poor plate positioning; and two monoaxial group participants had early hardware removal and a revision respectively because of intra-articular screw protrusion. In Voigt 2011, one person in each group had an early "prosthetic replacement" and three participants in the polyaxial group and one in the monoaxial group had refixation. The two other re-operated polyaxial group participants of Voigt 2011 had a corrective osteotomy and a screw removal respectively, while two other re-operated monoaxial group participants both had early implant removals.

Similarly, neither trial found statistically significant differences between the two groups in their other reported outcome measures. The available data are shown for death (see Analysis 8.4), participants with any or individual complications (see Analysis 8.6), the Constant score relative to the uninjured arm (see Analysis 8.5), range of motion (see Analysis 8.7) or duration of operation or fluoroscopy time (see Analysis 8.8).

Locking plate: use of medial support locking screws

Zhang 2011 tested the use of the medial support locking screws in 72 patients with Neer two, three or four-part fractures treated with open reduction with internal fixation using the PHILOS locking plate. In the medial support group, locking screws were introduced through the plate so as to run up the inferior portion of the humeral neck providing support to the calcar. In the control group, these screw holes were left empty. One participant in the medial screw group had early failure of fixation due to plate breakage compared with nine with early fixation failure (six varus collapse; three screw penetration) in the control group; however, this difference did not reach statistical significance (see Analysis 9.1: RR 0.15, 95% CI 0.02 to 1.11). Seven of these patients, including the patient in the medial screw group, consented to have a re-operation (RR 0.22; 95% CI 0.03 to 1.11). One patient in the medial screw group had asymptomatic osteonecrosis. The medial screw group had statistically significantly higher Constant scores (0 to 100: best score) at 31 month follow-up (see Analysis 9.2: MD 9.00, 95% CI 2.41 to 15.59).

Hemiarthoplasty: comparison of two types

Fialka 2008 compared two types of hemiarthroplasty, the EPOCA prosthesis versus the HAS prosthesis, which differ in a number of ways including the method of fixation of the tuberosities. Of the 40 trial participants, three had died and two were lost to follow-up at one year. Significantly better functional results, including range of motion, at one year were reported for EPOCA prosthesis group. The relative (compared with the patient's uninjured shoulder) individual Constant score results were 70.4% (range 38% to 102%) for the EPOCA group versus 46.2% (range 15% to 80%) for the HAS group (reported P = 0.001). Reported complications were two patients with deep infection in the EPOCA group, two patients with persistent pain scheduled for a reoperation in the HAS group (see Analysis 10.1), and a periprosthetic fracture that occurred in one of the three patients who had died by one year. Radiological findings, except for heterotopic ossification where there were contradictory data, are shown in Analysis 10.2. These tended to favour the EPOCA prothesis. Fialka 2008 noted some association between the bony resorption of the tuberosities and a decreased Constant outcome score. Results for range of motion are shown in Analysis 10.3.

Continuing management (including rehabilitation) after surgical treatment

Wirbel 1999 tested the duration of immobilisation (one week versus three weeks) before starting physiotherapy after closed reduction and percutaneous fixation of displaced fractures in 77 patients. Wirbel 1999 reported that there were no statistically significant differences between the two trial groups in their functional results, assessed using the Neer score, at three, six or at an average of 14.2 months. Data provided for unsatisfactory or worse outcome, as defined by the Neer score, at six months are consistent with this claim (see Analysis 11.1: 9/32 versus 10/32; RR 0.90, 95% CI 0.42 to 1.92). Premature removal of Kirschner wires because of loosening occurred in the five people in each group (see Analysis 11.2); these results, however, were not provided for the whole study population nor was it indicated in which groups the five people who underwent open revision or hemiarthroplasty belonged. Though similar numbers (3 versus 2) of people underwent removal of screws due to subacromial impingement after six months, the numbers of people in each group whose displaced tuberosity fractures were fixed with cannulated screws were not reported. Of the 21 participants followed up more than two years, one developed partial necrosis of the humeral head but was symptom-free and had a full range of motion of his affected shoulder.

Agorastides 2007 reported the findings of early active assisted mobilisation (after two weeks) versus late mobilisation (after six weeks) after cemented hemiarthroplasty in 49 of the 59 participants recruited in their trial. At one year follow-up, there were no significant differences between the two groups in function as rated by the Oxford shoulder score (see Analysis 12.1; MD -6.0, 95% CI -16.53 to 4.53; scale was 0 to 100) or the overall Constant score (see Analysis 12.2). Both nonunions occurred in the early group but none of the differences in radiologically-assessed outcomes between the two groups was statistically significant (see Analysis 12.3). The differences between the two groups at one year in elevation and external rotation were neither statistically nor clinically significant (see Analysis 12.4).

Discussion

Summary of main results

This review, which covers all non-pharmacological treatment and rehabilitation interventions for proximal humeral fractures in adults, now includes 23 single-centre trials involving a total of 1238 patients. With the increased availability of trials, this is the first version of the review for which some pooling of data was possible although limited to two comparisons. Additionally, several of the more recently included trials have reported patient-reported outcomes measures of function. However, despite these encouraging developments, the main result of this review continues to be the general lack of reliable evidence to inform the treatment of these common injuries.

The comparisons tested by the 23 included trials fell into four main treatment categories.

Methods of conservative management (including rehabilitation)

Conservative management, generally involving a period of arm immobilisation followed by physiotherapy, of (usually) minimally displaced fractures is the focus of eight trials. There was a general recognition of the impaired function and serious complications, such as complex regional pain syndromes, that could follow a proximal humeral fracture. For example, Bertoft 1984 noted that following injury there is a marked tendency for the capsule of the shoulder joint to contract and for the deltoid muscle to atrophy, leading to stiffness and inferior subluxation of the humeral head respectively.

Initial treatment, including immobilisation

When considering the extent and duration of initial immobilisation after a fracture, a balance is needed between the advantages of pain relief and avoidance of fracture displacement, and the consequences of immobilisation, notably joint stiffness and muscle atrophy. No data could be pooled from the three heterogeneous trials (Hodgson 2003; Kristiansen 1989; Lefevre-Colau 2007) comparing early versus delayed mobilisation for undisplaced or 'stable' fractures. Nonetheless, early mobilisation appeared to result in earlier recovery and less pain without compromising longer term outcome nor engendering serious complications including fracture displacement. There is limited evidence that the particular type of bandage used neither influences the time to fracture union nor the end functional result, although an arm sling was found to be generally more comfortable than a body bandage (Rommens 1993).

Continuing management (rehabilitation) after initial conservative treatment involving sling immobilisation

Two trials (Bertoft 1984; Lundberg 1979) investigated whether patients could undertake their own rehabilitation (instructed self-exercise) after receiving appropriate instruction with some monitoring, rather than with full supervision. Conversely, one trial (Revay 1992) studied the supplementation of self-treatment with supervised group training in a swimming pool. Their consensus that patients could generally achieve the desired end result with less supervision is not supported by sufficient evidence.

Livesley 1992 hypothesised that pain was associated with contracture of the capsule of the glenohumeral joint and that pulsed electromagnetic high frequency energy (PHFE) would reduce inflammation and swelling, improving the end functional result. However, the trial failed to provide any quantitative data to support or refute this hypothesis.

Surgical treatment versus conservative treatment

While the six trials, involving a total of 270 participants with displaced and/or complex fractures, making this comparison were heterogenous (e.g. in the choice of surgical intervention), the close similarity in the methods of the two most recent trials (Olerud 2011a; Olerud 2011b) ensured that there were data for pooling, particularly for the primary outcomes.

Pooled results of patient-reported functional scores at one year from three trials (153 participants) showed no statistically significant difference between the two groups. Pooled DASH scores from two trials (Olerud 2011a; Olerud 2011b) pointed to a longer-term trend to a more favourable result after surgery at 24 months. This was not found for subjectively assessed function at three years by another trial (Zyto 1997). Quality of life based on the EuroQol results scores from three trials (153 participants) showed non-statistically significant differences between the two groups at three time points up to 12 months. However, the pooled results (101 participants) at two years from the two Olerud trials were significantly in favour of the surgical group.

There was no significant differences between the two groups in mortality. Significantly more surgical group patients had additional or secondary surgery (18/112 versus 5/111; RR 3.36, 95% CI 1.33 to 8.49; 223 participants). This is equivalent to an extra operation in one in every nine surgically treated patients (NNT 9, 95% CI 6 to 25; risk difference 0.12, 95% CI 0.04 to 0.19). While the numbers of people in each group with one or more adverse events or complications were not available for any of the six trials, the available data for individual complications point to greater numbers of people with complications in the surgical treatment group reflecting surgery-related complications such as infection and screw penetration of the joint. None of the differences between the two groups in the radiologically detected outcomes of non-union, avascular necrosis and signs of osteoarthritis were statistically significant.

There were no significant differences between the two groups for the listed secondary outcomes, including Constant score results, pain, measures of functional impairment and costs (measured only in Fjalestad 2010).

Different methods of surgical management

Comparisons of different categories of surgical intervention

Three trials (Hoellen 1997; Smejkal 2011; Zhu 2011) compared different methods of surgical management.

Open reduction with internal fixation using a locking plate versus a locking nail

Zhu 2011 compared two locking plates versus a locking nail in 57 participants with two-part surgical neck fractures. The American Shoulder and Elbow Surgeon's scores were significantly better in the plate group at one and three year follow-ups. However, significantly more patients in the plate group had a complication (9/29 versus 1/28; RR 8.69, 95% CI 1.18 to 64.19). This included five patients in the plate group who had a re-operation for screw penetration into the joint. There were no statistically significant differences between the groups for secondary outcomes at three years follow-up.

Open reduction with internal fixation using a locking plate versus minimally invasive fixation with distally inserted intramedullary nails

Smejkal 2011, which made this comparison in 61 patients, found no statistically significant difference between the two groups in the overall numbers of participants incurring a complication or in the Constant scores at two years follow-up. While the differences between the groups in time to union, time to recover upper limb function and length of hospital stay did not attain statistical significance, in each case these were longer in the plate group.

Hemi-arthroplasty versus tension band wiring

Hoellen 1997 compared humeral head replacement with fracture fixation in 30 people with four-part fractures, but functional outcome data were only available for 18 participants at the one year follow-up. There were no differences between the two groups in the mean Constant scores or pain. Although all five reoperations occurred in the fixation group, the result was not statistically significant.

Comparisons of different methods of performing an intervention in the same category

Four trials (Fialka 2008; Ockert 2010; Smejkal 2011; Voigt 2011) compared different types or methods in the same intervention category (e.g. plating).

Polyaxial versus monoaxial locking plate fixation

Two trials (Ockert 2010; Voigt 2011), including 114 participants, made this comparison. Voigt 2011 found no statistically significant differences at one year between the two groups in their DASH scores nor at three, six or 12 months in the 'simple shoulder test'. Neither trial found statistically significant differences between the two groups in the numbers of participants having a re-operation (8/54 versus 7/68; RR 1.43, 95% CI 0.56 to 3.67). Similarly, neither trial found statistically significant differences between the two groups in their other outcome measures, namely death, participants with any or individual complications, Constant scores, range of motion, and duration of operation or fluoroscopy time.

Locking plate: use of medial locking screws

Zhang 2011, which tested the use of medial locking screws for PHILOS plate fixation in 72 patients, found fewer patients in the intervention group had early failure of fixation (1 versus 9); however, this difference did not reach statistical significance (RR 0.15, 95% CI 0.02 to 1.11). Seven of these patients, including the patient in the medial screw group, had a re-operation. The medial screw group had statistically significantly higher Constant scores at 31 month follow-up.

Hemiarthoplasty: comparison of two types

Fialka 2008 found significantly better functional, including range of motion, results at one year for the EPOCA prosthesis when compared with the HAS prosthesis in 40 people with four-part fractures. There were no statistically significant differences between the two groups in the incidence of individual adverse events.

Continuing management (including rehabilitation) after surgical intervention

The need for and duration of immobilisation before commencing physiotherapy after surgery for displaced fractures was tested in two small trials, both of which had potentially biased results. Wirbel 1999 provided very limited results on which to judge the claimed lack of difference in outcome following immobilisation for one versus three weeks after surgical fixation. Agorastides 2007 found no statistically significant differences in function (Oxford shoulder score at one year) and radiological outcome between participants mobilised after two weeks (which was current practice) after hemiarthroplasty versus those mobilised after six weeks. Thus, it is not possible to say whether early mobilisation could destabilise the fracture, or adversely affect the cuff repair or the hemiarthroplasty. Conversely, there is little evidence to determine if early mobilisation if early mobilisation offered any functional advantages.

Overall completeness and applicability of evidence

To inform consideration of applicability of the evidence from individual trials, we have increased the details given in the Characteristics of included studies on the study populations and interventions. Additionally, Table 2 shows our assessments for each trial of four aspects of relevance to ascertaining external validity: definition of the study population, description of the interventions, definition of primary outcome measures, and length of follow-up. Clearly unhelpful is where there are incomplete descriptions of study inclusion (six trials) and interventions (four trials). Four trials had less than one year follow-up: Lefevre-Colau 2007 (six months), Livesley 1992 (six months), Ockert 2010 (six months) and Rommens 1993 (until fracture consolidation - time unspecified). Despite the claims of longer follow-up, the results seemed to apply to six months at most in Stableforth 1984. In Wirbel 1999, though follow-up of 21 participants was more than two years, the main results applied to the set follow-up at six months. Our setting of our criterion to one year follow-up as acceptable is arbitrary and mainly reflects a reasonable timing for assessment of function. However, it should be noted that in terms of a full outcome assessment, data at one year follow-up must be considered preliminary results only given that complications such as avascular necrosis and device failure may not become evident until later.

Table 2. Assessment of items relating to applicability of trial findings
 Clearly defined study population?Interventions sufficiently described?Main outcomes sufficiently
described?
Appropriate timing of outcome measurement?
(Yes = ≥ 1 year)
Agorastides 2007Partial: exclusions not specified upfrontYesYesYes: 1 year
Bertoft 1984Partial: no exclusion criteria given (e.g. ability to understand instructions for exercises)YesYesYes: 1 year
Fialka 2008YesYesYesYes: 1 year
Fjalestad 2010YesYesYesYes: 2 years
Hoellen 1997Yes: but some question over fracture type in that the Holbein 1999 report included 3-part fractures tooYesYesYes: 1 year
Hodgson 2003YesYesYesYes: 2 years
Kristiansen 1988Partial: no exclusion criteria givenPartial: incomplete description of timing of sling use and care of external fixator pin sitesPartial: no description of measurement proceduresYes: 1 year
Kristiansen 1989Partial: no exclusion criteria givenPartial: although sling and body bandage are common expressions, some variation possiblePartial: no description of measurement proceduresYes: 2 years
Lefevre-Colau 2007YesYesYesPartial: 6 months
Livesley 1992Yes: although this included 4 patients under 20 years with epiphyseal fracturesYesYesPartial: 6 months
Lundberg 1979Partial: no exclusion criteria given (e.g. ability to understand instructions for exercises)YesYesYes: 1 year or above (mean: 16 months)
Ockert 2010Partial: exclusion criteria described in context of post-randomisation exclusions.YesYesPartial: 6 months
Olerud 2011aYesYesYesYes: 2 years
Olerud 2011bYesYesYesYes: 2 years
Revay 1992YesPartial: frequency of swimming sessions not statedYesYes: 1 year
Rommens 1993Yes: but to note that other fractures including rib (3 participants) were includedYesPartial: functional outcome assessment not described (sufficiently)No: only until fracture consolidation
Smejkal 2011YesPartial: Only minimal intra-operative details given and nothing regarding post-operative management including rehabilitationPartial: this may have been ‘lost in translation’ (Czech article)Yes: mean 2 years but range not stated (probably most / all > 1 year as recruitment had finished January 2010).
Stableforth 1984YesYesPartial: no description of measurement procedures, incomplete description of pain categoriesPartial: up to 6 months, then between 18 months to 12 years. This is too spread out. Most results applied to the 6 month follow-up.
Voigt 2011YesYesYesYes: 1 year
Wirbel 1999YesYesPartial: no description of measurement proceduresPartial: between 9 and 36 months; < 1 year in 10 participants. Main results applied to 6 months.
Zhang 2011YesYesPartial: Insufficient information on measurement of complications and timing of their measurement.Yes: All over 25 months (mean 30.8 months)
Zhu 2011YesYesYesYes: 1 and 3 years
Zyto 1997YesYesYesYes: 1 year, and 3 to 5 years

The measurement of outcome was variable, though generally comprehensive. In most of the older trials, there was frequent use of non-validated or, at best, partly validated scoring systems such as the Neer (Neer 1970) and Constant (Constant 1987) systems, but also of simple rating systems for individual outcomes. Validated schemes such as the Oxford Shoulder Score (Dawson 1996) and Shoulder Rating Questionnaire (L'Insalata 1997) for subjective assessment of symptoms and function were not available at the time for the trials in earlier version of this review. Nonetheless, some consideration of interobserver reproducibility and other aspects of validity was evident in the establishment of the Constant score and in two trials (Lundberg 1979; Zyto 1997). Non-validated outcome assessment schemes, often with arbitrary criteria for grading overall outcome (excellent, good, fair, poor), are probably best viewed as 'blunt' and flawed instruments. This needs to be noted when viewing the results of many of the older included trials; in particular Kristiansen 1989, whose outcome assessment is almost completely based on the Neer scoring system. More recent trials are generally better in this respect: Agorastides 2007 reports the Oxford Shoulder Score; Fjalestad 2010 recorded American Shoulder and Elbow Surgeons score; Hodgson 2003 used of the SF36 health survey and a validated scheme for self-reported disability resulting from shoulder problems (Croft 1994), and three trials (Olerud 2011a; Olerud 2011b; Voigt 2011) used the DASH (Disability of the Arm, Shoulder, and Hand) questionnaire.

The majority of the trials used Neer's fracture classification (Neer 1970). Problems, such as poor interobserver reproducibility and intraobserver reliability, with the classification of fractures according to the Neer and AO systems have been shown for both radiographs and computerised tomographic scans (Bernstein 1996; Brorson 2008; Sidor 1993; Siebenrock 1993; Sjoden 1997). This variation in the classification of fractures and hence diagnosis needs to be considered when interpreting the results of trials, both in respect to the comparability and composition of the intervention groups and in the applicability of the trial's findings. The limitations of the Neer classification scheme were also demonstrated by the identification of the valgus impacted four-part fracture as a separate category with a lower risk of avascular necrosis (Jakob 1991). Ideally a fracture classification system should act as a guide to treatment as well to enable the comparison of results from studies of patients with similar fracture patterns. However, other factors, such as osteoporotic bone, associated soft tissue injury and the patient's overall health and motivation, will also influence treatment choices and outcome. A recent study (Brorson 2012) looking at the agreement of surgeons' treatment recommendations in conjunction with the Neer classification concluded that the low observer agreement on the Neer classification may have less clinical importance than previously assumed. However, it noted that inter-observer agreement on treatment did not exceed moderate levels.

While it is possible that all 23 trials are relevant to current practice somewhere in the world, it is likely that some interventions are now rarely used. These include body bandages as tested in Rommens 1993: nowadays it is much more common practice to use either a 'collar and cuff' sling or a 'poly-sling' (these incorporate a chest strap that can be passed around the body). Additionally, the applicability of the findings from older trials, such as Stableforth 1984, is potentially less given subsequent changes in practice including the availability of new implants. These include locking plates, which are being increasingly used and promoted for these fractures (Thanasas 2009). In the last update (Handoll 2010), only Fjalestad 2010 evaluated the use of a locking plate. Locking plates were used in six out of the seven newly included trials. New materials are also being used, such as the widespread use of non-metallic materials instead of metal wires to achieve 'tension band fixation'.

Comments on individual comparisons

Methods of conservative management (including rehabilitation)

Initial treatment, including immobilisation

Though Hodgson 2003 provides strong evidence in favour of early physiotherapy, and avoiding routine immobilisation, in undisplaced two-part fractures, it is still a small study that might be affected by bias, particularly relating to the lack of blinding. A survey sent to senior hospital physiotherapists working directly with orthopaedic patients revealed large variation in rehabilitation, in particular with regards to routine immobilisation, duration of immobilisation and timing of first contact with a physiotherapist, within and between hospitals in the UK (Hodgson 2003a; Hodgson 2006). This points to the need for a similar but larger and preferably multi-centre trial testing the same comparison as Hodgson 2003 to confirm the results of this trial and examine their applicability.

As noted by McKee 2007 in his commentary on Lefevre-Colau 2007, the applicability of this trial is limited by the intensive physiotherapy regimen used in both groups. Both practically and financially the 32 two-hour sessions of physiotherapy may be difficult for patients and health providers; notably, 10 participants withdrew from the trial because of difficulties in attending. In contrast, the mean numbers of treatment sessions in Hodgson 2003 were nine and 14 respectively in the two groups.

Continuing management (rehabilitation) after initial conservative treatment involving sling immobilisation

Three trials in this category were based in Sweden and possible differences in conventional physiotherapy regimens within and between countries, then and now, need to be considered when considering the application of trial findings. If they work, self-instruction and home-based exercise programmes are attractive for patients and conserve health care resources. There is some evidence from a Cochrane review on fall prevention that elderly people, if well instructed and with intensive support (regular phone calls etc), can maintain a home-based exercise programme (Gillespie 2003; Gillespie 2009). However, there will still be some elderly patients with insufficient understanding or motivation to perform the required exercises.

Surgical treatment versus conservative treatment

Trials comparing surgical versus non-surgical interventions, or indeed different surgical interventions, risk losing currency as different implants and methods become available and fashionable. Previously (Handoll 2007), we observed that: "There are no randomised controlled trials comparing other forms of therapy which are now technically easier to perform given the improvements in intra-operative imaging. These include single and multiple, antegrade and retrograde nailing (Lin 1998; Wachtl 2000), external fixation with fine wires coupled to light-weight frames and percutaneous pin fixation of the head to the shaft coupled with tuberosity wiring (Ko 1996)." We observed that: "These may lower the risk of iatrogenic avascular necrosis (Resch 1997). However, aside from the valgus impacted four-part fracture with its reduced risk of avascular necrosis, many surgeons would not consider stabilisation for comminuted fractures and simply proceed to hemiarthroplasty, especially in older people. One key reason for this approach is the general recognition that hemiarthroplasty following failed fixation is technically difficult and the resulting outcome is usually less satisfactory (Naranja 2000; Sonnabend 2002)." However, this is now challenged by the more recent development of locking plating systems, which allow for stronger constructs and fixation of more complex fracture patterns in osteopenic bone with the potential for less soft-tissue stripping and compromise to the blood supply (Thanasas 2009). In summary, evolving technology (and marketing forces) mitigates against applying the findings of these types of trials. A greater emphasis is merited on the findings of the more recent trials in this comparison (Fjalestad 2010; Olerud 2011a; Olerud 2011b), all three of which have the added bonus of reporting patient-reported outcome measures of function and quality of life.

When considering the validity and applicability of surgical trials, account needs to be taken also of fundamental variations in surgical practice, including facilities and operator expertise. In particular, operator expertise and the linked issue of the surgical learning curve, play a pivotal role in the validity and applicability of surgical trial findings. It is this awareness that is behind the pragmatic decision in the ProFHER trial for surgeons to use methods with which they are familiar rather than stipulate the type of surgery.

Different methods of surgical management

Much of the recent research activity, in both published and registered trials, evaluates different types of surgery. We now distinguish between trials comparing different categories of surgical interventions (tested by three trials) and trials comparing different methods of performing an intervention in the same category (tested by four trials).

The variety of available implants in the same category can limit the applicability and usefulness of trials comparing different categories of surgical intervention by comparisons of specific implants. Nonetheless, the comparison by Zhu 2011 of one of two locking plates versus a locking nail is very pertinent in terms of providing a useful investigation of the appropriateness of the current trend from locking nails to plates. This trial, however, is too small to establish the superiority of one method over the other but it does provide some evidence of better function in the plate group at one year, and possibly for longer, but at a potentially greater risk of surgical complications and initially more invasive surgery. The comparison by Smejkal 2011 of a locking plate versus minimally invasive fixation with distally inserted multiple intramedullary nails (the Zifko method of minimally invasive fixation) is of relevance to current practice but, while data from Smejkal 2011 lend support to the use of Zifko nailing in terms of it being a less extensive surgical procedure with potentially an earlier recovery that plate fixation, there were inadequate data on longer term function and outcome. Hoellen 1997, a flawed trial with only one year follow-up, considered only one of several shoulder prostheses now available (the prosthesis was cemented in place) in their comparison with tension band wiring.

The two trials (Ockert 2010; Voigt 2011) comparing 'polyaxial' (where surgeons had greater control in their positioning of screws into the bone) versus 'monoaxial' locking plate fixation found no difference between the two methods. With six months follow-up and no report of functional outcome, Ockert 2010 contributed little to this question. Voigt 2011, which was a stronger trial but still insufficient to be conclusive, pointed out that the "majority of surgeons chose the same screw directions for the polyaxial screws as already exist in the fixed angle plate". Of note though is the differences in the types of screws in the two implants in Voigt 2011, which could in some respects alter the question. Zhang 2011 tested the use of medial support locking screws for fixation using the PHILOS locking plate. While Zhang 2011 did not provide conclusive evidence of clinical benefit of the enhanced stabilisation of this commonly used plate, the direction of effect is consistent with the theoretical advantages for medial support screws. Fialka 2008 compared two shoulder prostheses but although the authors ascribed the different functional outcomes to tuberosity fixation, other design differences may account for these results. These include a different stem finish and a more accurate recreation of pre-operative humeral geometry with the EPOCA prosthesis

Continuing management (including rehabilitation) after surgical intervention

The need for and duration of immobilisation before commencing physiotherapy after surgical treatment is likely to depend on the method of fixation or type of prosthesis; and also other factors such as bone quality. While neither trial found conclusive evidence for early mobilisation, it can also be observed that the evidence was inconclusive for later mobilisation too.

Quality of the evidence

The evidence base for this review, formed from 23 small heterogeneous trials, remains very limited. It is an encouraging find that several of the more recent trials have both improved study design (e.g. four of the seven newly included trials were judged at low risk of selection bias in terms of appropriate random sequence generation and allocation concealment) and reporting (e.g. including participant flow diagrams). The fact remains though that the majority of the included trials had serious shortcomings and were at high risk of bias that could affect the validity of their findings (see Figure 1). These shortcomings include use of quasi-randomised methods for treatment allocation (Rommens 1993), failure to present supportive data for claims of no differences (Livesley 1992), lack of assessor blinding (14 trials were considered at high risk) of bias, and incomplete data. The risk of bias resulting from a high loss to follow-up or exclusion of participants from the analyses was considered high in 11 trials, two of which were new to this version (Ockert 2010; Smejkal 2011. There is clearly a need for caution in interpreting the results of small trials which demonstrate 'no evidence of an effect' rather than 'evidence of no effect'. Insufficiencies in quantity and quality of the evidence preclude the drawing of robust conclusions for any of the comparisons evaluated by the included trials.

In the previous version of this review, we observed that the "scarcity of the evidence from randomised trials to inform practice for these common fractures is remarkable and depressing. Partly, it appears that a key problem for conducting trials, especially those involving surgery, is patient recruitment as shown by the abandonment of several, including multi-centre, trials as well as the slow and lower-than-planned recruitment in others". While the difficulties of patient recruitment remain, the increased research activity in this previously overlooked area is shown by the addition of seven new trials this update and the registration of five more ongoing trials. Especially noteworthy are the six, of which five are multi-centre, trials that are testing the important question of whether surgery gives better long term functional outcomes compared with conservative treatment. The published protocols for three of these trials (Brorson; ProCon; ProFHER) all show the use of robust methodology required to minimise bias.

Potential biases in the review process

While our search was comprehensive it is likely that we have failed to identify some randomised trials, particularly those reported only in abstracts or in non-English language publications. We may also have overlooked mixed-population trials that included proximal humeral fractures as a subgroup. However, we are almost certain that we have not overlooked trials that would provide the definitive evidence that could inform practice. It is clear, from the growing awareness and imperative of trial registration, that such trials are now in progress. Systematic processes were undertaken throughout the review and included the contact of trial investigators for clarification and missing data. Some mention of the dilemma presented in the pooling of data from clearly heterogeneous trials for the surgical treatment versus conservative treatment comparison is described in the Effects of interventions; this is, however, compatible with the overall question and notably the pooled analyses did not result in statistically significant heterogeneity.

Agreements and disagreements with other studies or reviews

Several systematic reviews, covering either all treatment (Lanting 2008), specific interventions (hemiarthroplasty: Kontakis 2008; Nijs 2009; locking plate for four-part fractures: Brorson 2011) or treatment questions (surgical versus conservative intervention: Nanidis 2010), for proximal humeral fractures have been published since 2007. All five reviews, which included evidence from a broader spectrum of study designs, have noted the limitations in the available evidence and called for well designed prospective, preferably randomised controlled, trials to inform practice.

Authors' conclusions

Implications for practice

Overall, there is insufficient evidence from randomised controlled trials to determine which interventions are the most appropriate for the management of different types of proximal humeral fracture.

Currently, most undisplaced proximal humeral fractures are treated conservatively. This generally involves a period of immobilisation followed by supervised physiotherapy. There is some good quality but still limited evidence that early physiotherapy, without routine immobilisation, is effective for undisplaced two-part fractures. There is also some limited evidence that short periods of arm immobilisation are acceptable and that, given adequate instruction, some patients may manage their own rehabilitation programme. However, careful selection and long term monitoring procedures should be put in place to check the outcome of these treatments.

Reduction and surgical fixation of displaced fractures attempt to restore the bony anatomy of the joint and hence joint mechanics. Often for older patients with more complex fractures, replacement of the humeral head is considered. However, while there is now some quality of life evidence in favour of surgery from two of the six trials, overall there is insufficient evidence to confirm that surgery gives better functional or quality of life outcomes compared with conservative treatment. Moreover, the complications associated with surgery need to be considered, especially given the evidence so far indicates that these lead to a greater risk of a further operation.

There is insufficient evidence to establish what is the best method of surgical treatment. It is possible that fixation using a locking plate rather than a locking nail of displaced two-part surgical neck fractures may have some benefit in terms of functional outcome but at a potentially higher risk of complications. In some types of severe injury, a hemiarthroplasty may yet turn out to be a better option than fracture fixation. There is insufficient evidence to establish what methods of plate fixation or hemiarthroplasty are best.

There is insufficient evidence to say when to start mobilisation after either surgical fixation or hemiarthroplasty.

Implications for research

This Cochrane review incorporates evidence from only 23 small single-centre randomised controlled trials of treatment of proximal humeral fractures. There are many issues that have not been addressed. There is a need for better information with regard to the optimal selection, timing and duration of all interventions. In particular, there is a need to determine if a simple undisplaced fracture should be immobilised, and if so, for how long, and the timing, type and extent of physiotherapy required. The urgent need to define more clearly the role and type of surgical intervention in the management of proximal humeral fracture should be addressed upon completion of trials that are currently underway. Any trials must take account of the important issues of method of randomisation, blinding and duration of follow-up. Such trials should use standard and validated outcome measures, including patient assessed functional outcomes, and also assess resource implications. They should also meet the CONSORT criteria for design and reporting of non-pharmacological studies (Boutron 2008).

This Cochrane review should be maintained and updated as further randomised controlled trials become available. The authors would be pleased to receive information about any other randomised controlled trials relating to the treatment of these fractures.

Acknowledgements

We thank Xavier Griffin, David Limb and Yemisi Takwoingi for their helpful feedback at editorial review. We thank Catherine Deering for patiently supplying several search downloads for this version, Joanne Elliott for advice on searching and Lindsey Elstub for her editorial support.

We thank the following for further information on their research in this area for this update: Stig Brorson, Tore Fjalestad, Per Olerud, Matt Smith, Christine Voigt and Lei Zhang.

The acknowledgements for former versions of the review are presented in Appendix 3.

Data and analyses

Download statistical data

Comparison 1. Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Shoulder disability: Croft Shoulder Disability Score1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Disability (1 or more problems) at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 Severe disability (5 or more problems) at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.3 Disability (1 or more problems) at 2 years1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.4 Severe disability (5 or more problems) at 2 years1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Croft shoulder disability score: individual problems at 2 years1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Pain on movement1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.2 Bathing difficulties1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.3 Change position at night more often1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.4 Disturbed sleep1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.5 No active pastimes or usual physical recreation1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.6 Lifting problems1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.7 Help needed1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.8 More accidents (e.g. dropping things)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3 Number of treatment sessions (until independent function achieved)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4 SF-36 scores: pain & physical dimensions1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 Physical functioning (0-100: excellent) at 16 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 Physical functioning (0-100: excellent) at 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.3 Role limitation physical (0-100: none) at 16 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.4 Role limitation physical (0-100: none) at 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.5 Pain (0-100: none) at 16 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.6 Pain (0-100: none) at 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Adverse events3 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
5.1 Frozen shoulder1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
5.2 Fracture displacement1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
5.3 Non-union1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
5.4 Reflex sympathetic dystrophy1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
5.5 Treated (injection) subacromial impingement1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6 Constant shoulder score (ratio of affected/unaffected arm)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
6.1 8 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6.2 16 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6.3 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7 Constant shoulder score (0 to 100: best)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
7.1 6 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.2 3 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.3 6 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.4 6 months: subjective assessment (0 to 35: best)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.5 6 months: objective assessment range of motion and strength (0 to 65: best)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8 Changes in pain intensity (mm) from baseline: 100 mm visual analogue scale (positive change = less pain)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
8.1 At 6 weeks1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8.2 At 3 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8.3 At 6 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
9 Range of motion at 6 months (degrees): difference between two shoulders1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
9.1 Abduction1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
9.2 Anterior elevation1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
9.3 Lateral rotation1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
10 Patient dissatisfied with treatment1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
Analysis 1.1.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 1 Shoulder disability: Croft Shoulder Disability Score.

Analysis 1.2.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 2 Croft shoulder disability score: individual problems at 2 years.

Analysis 1.3.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 3 Number of treatment sessions (until independent function achieved).

Analysis 1.4.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 4 SF-36 scores: pain & physical dimensions.

Analysis 1.5.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 5 Adverse events.

Analysis 1.6.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 6 Constant shoulder score (ratio of affected/unaffected arm).

Analysis 1.7.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 7 Constant shoulder score (0 to 100: best).

Analysis 1.8.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 8 Changes in pain intensity (mm) from baseline: 100 mm visual analogue scale (positive change = less pain).

Analysis 1.9.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 9 Range of motion at 6 months (degrees): difference between two shoulders.

Analysis 1.10.

Comparison 1 Early mobilisation (within or up to 1 week) versus immobilisation for 3 weeks, Outcome 10 Patient dissatisfied with treatment.

Comparison 2. Gilchrist bandage versus 'Classic' Desault bandage
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Problems with bandages1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Application of bandage was uncomfortable1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 Premature bandage removal1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Fracture displacement by 3 weeks1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3 Poor or bad rating by patient at fracture consolidation1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
Analysis 2.1.

Comparison 2 Gilchrist bandage versus 'Classic' Desault bandage, Outcome 1 Problems with bandages.

Analysis 2.2.

Comparison 2 Gilchrist bandage versus 'Classic' Desault bandage, Outcome 2 Fracture displacement by 3 weeks.

Analysis 2.3.

Comparison 2 Gilchrist bandage versus 'Classic' Desault bandage, Outcome 3 Poor or bad rating by patient at fracture consolidation.

Comparison 3. Instructed self physiotherapy versus conventional physiotherapy
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Pain at one year (scale 0 to 8: maximum pain)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
2 Severe or moderate pain at 3 months1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3 Requested change of therapy1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
4 Adverse events (frozen shoulder: 1 v 2; unexplained prolonged pain: 0 v 1)1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
5 Neer's rating (0 to 100: best) at mean 16 months (exploratory analysis)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
6 Active gleno-humeral elevation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
Analysis 3.1.

Comparison 3 Instructed self physiotherapy versus conventional physiotherapy, Outcome 1 Pain at one year (scale 0 to 8: maximum pain).

Analysis 3.2.

Comparison 3 Instructed self physiotherapy versus conventional physiotherapy, Outcome 2 Severe or moderate pain at 3 months.

Analysis 3.3.

Comparison 3 Instructed self physiotherapy versus conventional physiotherapy, Outcome 3 Requested change of therapy.

Analysis 3.4.

Comparison 3 Instructed self physiotherapy versus conventional physiotherapy, Outcome 4 Adverse events (frozen shoulder: 1 v 2; unexplained prolonged pain: 0 v 1).

Analysis 3.5.

Comparison 3 Instructed self physiotherapy versus conventional physiotherapy, Outcome 5 Neer's rating (0 to 100: best) at mean 16 months (exploratory analysis).

Analysis 3.6.

Comparison 3 Instructed self physiotherapy versus conventional physiotherapy, Outcome 6 Active gleno-humeral elevation (degrees).

Comparison 4. Surgery versus conservative treatment
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Functional scores at 12 months (higher = worse)3153Std. Mean Difference (IV, Fixed, 95% CI)-0.10 [-0.42, 0.22]
1.1 DASH (0 to 100: worst disability) (reversed)2105Std. Mean Difference (IV, Fixed, 95% CI)-0.19 [-0.57, 0.19]
1.2 ASES (0 to 24: best)148Std. Mean Difference (IV, Fixed, 95% CI)0.10 [-0.46, 0.67]
2 DASH (0 to 100: worst disability)2 Mean Difference (IV, Fixed, 95% CI)Subtotals only
2.1 at 4 months2106Mean Difference (IV, Fixed, 95% CI)0.91 [-7.00, 8.83]
2.2 at 12 months2105Mean Difference (IV, Fixed, 95% CI)-4.51 [-13.50, 4.48]
2.3 at 24 months299Mean Difference (IV, Fixed, 95% CI)-7.43 [-16.26, 1.41]
3 American Shoulder and Elbow Surgeons score (0 to 24: best)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3.1 at 6 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.2 at 12 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Activities of daily living1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
4.1 Unable to manage personal hygiene at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 Unable to comb hair at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.3 Unable to sleep on fractured side at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.4 Unable to carry 5 kg at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.5 Unable to manage personal hygiene at 50 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.6 Unable to comb hair at 50 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.7 Unable to sleep on fractured side at 50 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4.8 Unable to carry 5 kg at 50 months1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Quality of life assessment: EuroQol (0: dead to 1: best health)3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
5.1 at 3 to 4 months3156Mean Difference (IV, Fixed, 95% CI)0.01 [-0.02, 0.05]
5.2 at 6 months3156Mean Difference (IV, Fixed, 95% CI)0.04 [-0.01, 0.08]
5.3 at 12 months3153Mean Difference (IV, Fixed, 95% CI)0.04 [-0.01, 0.08]
5.4 at 24 months2101Mean Difference (IV, Fixed, 95% CI)0.15 [0.05, 0.25]
6 Quality of life assessment (Fjalestad 2010 data)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
6.1 15D at 3 months (0: death; 1: perfect health)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6.2 15D at 6 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6.3 15D at 12 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6.4 number of QALYs at 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6.5 numbers of QALYs at 1 year (- deaths)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7 Mortality4196Risk Ratio (M-H, Fixed, 95% CI)1.55 [0.55, 4.36]
8 Additional surgery (re-operation or secondary surgery)5223Risk Ratio (M-H, Fixed, 95% CI)3.36 [1.33, 8.49]
8.1 at 6 to 12 months3113Risk Ratio (M-H, Fixed, 95% CI)1.90 [0.54, 6.73]
8.2 at 2 years2110Risk Ratio (M-H, Fixed, 95% CI)5.83 [1.38, 24.70]
9 Adverse events / complications6 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
9.1 Infection6259Risk Ratio (M-H, Fixed, 95% CI)4.13 [0.91, 18.87]
9.2 Nerve injury / palsy3146Risk Ratio (M-H, Fixed, 95% CI)0.74 [0.20, 2.76]
9.3 Non union5223Risk Ratio (M-H, Fixed, 95% CI)0.48 [0.16, 1.40]
9.4 Avascular necrosis5213Risk Ratio (M-H, Fixed, 95% CI)0.65 [0.37, 1.16]
9.5 Post-traumatic osteoarthritis (signs of)3133Risk Ratio (M-H, Fixed, 95% CI)0.56 [0.21, 1.51]
9.6 Screw penetration into joint3160Risk Ratio (M-H, Fixed, 95% CI)11.49 [2.25, 58.76]
9.7 Impingement158Risk Ratio (M-H, Fixed, 95% CI)2.0 [0.19, 20.86]
9.8 Redisplacement resulting in an operation281Risk Ratio (M-H, Fixed, 95% CI)0.26 [0.03, 2.22]
9.9 Secondary dislocation or resorption of the greater tuberosity151Risk Ratio (M-H, Fixed, 95% CI)15.18 [0.92, 249.78]
9.10 Tuberosity displacement at 50 months129Risk Ratio (M-H, Fixed, 95% CI)0.15 [0.01, 2.71]
9.11 Wire penetration at 1 year138Risk Ratio (M-H, Fixed, 95% CI)3.0 [0.13, 69.31]
9.12 Fixation failure resulting in an operation150Risk Ratio (M-H, Fixed, 95% CI)3.0 [0.13, 70.30]
9.13 Refracture122Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.07, 14.05]
10 Constant scores (overall: 0 to 100: best score)4 Mean Difference (IV, Fixed, 95% CI)Subtotals only
10.1 at 4 months2107Mean Difference (IV, Fixed, 95% CI)-1.33 [-6.80, 4.14]
10.2 at 12 months3152Mean Difference (IV, Fixed, 95% CI)2.36 [-3.52, 8.24]
10.3 at 24 months2101Mean Difference (IV, Fixed, 95% CI)0.48 [-7.26, 8.22]
10.4 at 50 months129Mean Difference (IV, Fixed, 95% CI)-5.0 [-17.52, 7.52]
11 Constant scores (difference between injured and uninjured shoulder): Normal = 0.1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
11.1 at 6 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
11.2 at 12 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
12 Poor or unsatisfactory function at 1 year (Neer rating)1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
13 Dependent in activities of daily living (or dead) at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
14 Pain at 2 years: VAS (0 to 100: worst pain)2101Mean Difference (IV, Fixed, 95% CI)-6.38 [-14.18, 1.41]
15 Constant score at 50 months: overall and components1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
15.1 Overall score (0-100: best score)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
15.2 Pain (maximum score 15)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
15.3 Range of motion (maximum score 40)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
15.4 Power (maximum score 25)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
15.5 Activities of daily living (maximum score 20)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
16 Constant (often severe) pain at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
17 Failure to recover 75% muscle power relative to other arm (survivors) at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
17.1 Flexion1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
17.2 Abduction1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
17.3 Lateral rotation1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
18 Range of movement impairments in survivors at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
18.1 Flexion < 45 degrees1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
18.2 Unable to place thumb on mid spine (T12)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
18.3 Lateral rotation < 5 degrees1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
19 Costs at 1 year (Euros in 2005)  Other dataNo numeric data
20 Total costs including indirect costs (Euros) at 1 year1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
Analysis 4.1.

Comparison 4 Surgery versus conservative treatment, Outcome 1 Functional scores at 12 months (higher = worse).

Analysis 4.2.

Comparison 4 Surgery versus conservative treatment, Outcome 2 DASH (0 to 100: worst disability).

Analysis 4.3.

Comparison 4 Surgery versus conservative treatment, Outcome 3 American Shoulder and Elbow Surgeons score (0 to 24: best).

Analysis 4.4.

Comparison 4 Surgery versus conservative treatment, Outcome 4 Activities of daily living.

Analysis 4.5.

Comparison 4 Surgery versus conservative treatment, Outcome 5 Quality of life assessment: EuroQol (0: dead to 1: best health).

Analysis 4.6.

Comparison 4 Surgery versus conservative treatment, Outcome 6 Quality of life assessment (Fjalestad 2010 data).

Analysis 4.7.

Comparison 4 Surgery versus conservative treatment, Outcome 7 Mortality.

Analysis 4.8.

Comparison 4 Surgery versus conservative treatment, Outcome 8 Additional surgery (re-operation or secondary surgery).

Analysis 4.9.

Comparison 4 Surgery versus conservative treatment, Outcome 9 Adverse events / complications.

Analysis 4.10.

Comparison 4 Surgery versus conservative treatment, Outcome 10 Constant scores (overall: 0 to 100: best score).

Analysis 4.11.

Comparison 4 Surgery versus conservative treatment, Outcome 11 Constant scores (difference between injured and uninjured shoulder): Normal = 0..

Analysis 4.12.

Comparison 4 Surgery versus conservative treatment, Outcome 12 Poor or unsatisfactory function at 1 year (Neer rating).

Analysis 4.13.

Comparison 4 Surgery versus conservative treatment, Outcome 13 Dependent in activities of daily living (or dead) at 6 months.

Analysis 4.14.

Comparison 4 Surgery versus conservative treatment, Outcome 14 Pain at 2 years: VAS (0 to 100: worst pain).

Analysis 4.15.

Comparison 4 Surgery versus conservative treatment, Outcome 15 Constant score at 50 months: overall and components.

Analysis 4.16.

Comparison 4 Surgery versus conservative treatment, Outcome 16 Constant (often severe) pain at 6 months.

Analysis 4.17.

Comparison 4 Surgery versus conservative treatment, Outcome 17 Failure to recover 75% muscle power relative to other arm (survivors) at 6 months.

Analysis 4.18.

Comparison 4 Surgery versus conservative treatment, Outcome 18 Range of movement impairments in survivors at 6 months.

Analysis 4.19.

Comparison 4 Surgery versus conservative treatment, Outcome 19 Costs at 1 year (Euros in 2005).

Costs at 1 year (Euros in 2005)
StudyMeasureSurgeryConservative treatmentDifference (conclusion)
Fjalestad 2010Total health-care costsmean = 10,367mean = 10,946Abstract: "the mean difference in total health-care costs was 597 Euros in favour of surgery (95% CI = -5291, 3777)". No significant difference.
Fjalestad 2010Health-care + indirect costsmean = 23,953mean = 21,878Reformatted text: "Including indirect costs... the difference [was] 2,075 (95% CI = -15,949 to 20,100)". No significant difference, but favours the conservative group.
Analysis 4.20.

Comparison 4 Surgery versus conservative treatment, Outcome 20 Total costs including indirect costs (Euros) at 1 year.

Comparison 5. Locking plate versus locking intramedullary nail
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 American Shoulder and Elbow Surgeons (ASES) score (0 to 100: best)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 At 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 At 3 years1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Death, re-operation and adverse events1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Death1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.2 Any complication1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.3 Screw penetration into humeral head (all had re-operation)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.4 Heterotopic ossification1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.5 Infection1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.6 Osteonecrosis1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.7 Degenerative change of glenohumeral joint1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.8 Secondary varus collapse1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.9 Non-union1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3 Pain (VAS: 0 to 10: worst)  Other dataNo numeric data
4 Constant score (0 to 100: best)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 At 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 At 3 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Active range of motion (at 3 years)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
5.1 Forward elevation (degrees)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5.2 External rotation1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
6 Range of movement: internal rotation (level on spine)  Other dataNo numeric data
7 Strength of suprapinatus (relative to opposite side) % - at 3 years1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
7.1 At 1 year1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.2 At 3 years1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8 Operation times and blood loss1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
8.1 Duration of surgery (minutes)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8.2 Blood loss (ml)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
9 Intra-operative complication1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
9.1 Pneumothorax1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
9.2 Blood transfusion1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 5.1.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 1 American Shoulder and Elbow Surgeons (ASES) score (0 to 100: best).

Analysis 5.2.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 2 Death, re-operation and adverse events.

Analysis 5.3.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 3 Pain (VAS: 0 to 10: worst).

Pain (VAS: 0 to 10: worst)
StudyMeasureLocking plateLocking nailReported significance
Zhu 2011Pain at 1 year

median = 0.5

interquartile range: 1.8

n = 29

median = 1.0

interquartile range = 1.0
n = 26

P = 0.042
Zhu 2011Pain at 3 years

median = 0

interquartile range = 0.8
n = 26

median = 0

interquartile range = 1.0
n = 25

P = 0.642
Analysis 5.4.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 4 Constant score (0 to 100: best).

Analysis 5.5.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 5 Active range of motion (at 3 years).

Analysis 5.6.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 6 Range of movement: internal rotation (level on spine).

Range of movement: internal rotation (level on spine)
StudyMeasureLocking plateLocking nailReported significance
Zhu 2011At 1 year

mean location = T8

range = T4 to L2
n = 29

mean location = T9

range = T2 to buttock
n = 26

P = 0.443
Zhu 2011At 3 years

mean location = T8

range = T2 to buttock
n = 26

mean location = T8

range = T2 to buttock
n = 25

P = 0.636
Analysis 5.7.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 7 Strength of suprapinatus (relative to opposite side) % - at 3 years.

Analysis 5.8.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 8 Operation times and blood loss.

Analysis 5.9.

Comparison 5 Locking plate versus locking intramedullary nail, Outcome 9 Intra-operative complication.

Comparison 6. Locking plate versus intramedullary nails (Zifko method)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Complications and [slight] malunion1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Any complication1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 Malunion (usually slight)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Constant score (% of healthy limb) at mean 2 years1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3 Time to union and time to recover upper limb function (weeks)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3.1 Time to radiographic union1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3.2 Time to recover normal upper limb function1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Operation and fluoroscopic times1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 Duration of operation (minutes)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 X-ray exposure (minutes)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Length of hospital stay (days)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
Analysis 6.1.

Comparison 6 Locking plate versus intramedullary nails (Zifko method), Outcome 1 Complications and [slight] malunion.

Analysis 6.2.

Comparison 6 Locking plate versus intramedullary nails (Zifko method), Outcome 2 Constant score (% of healthy limb) at mean 2 years.

Analysis 6.3.

Comparison 6 Locking plate versus intramedullary nails (Zifko method), Outcome 3 Time to union and time to recover upper limb function (weeks).

Analysis 6.4.

Comparison 6 Locking plate versus intramedullary nails (Zifko method), Outcome 4 Operation and fluoroscopic times.

Analysis 6.5.

Comparison 6 Locking plate versus intramedullary nails (Zifko method), Outcome 5 Length of hospital stay (days).

Comparison 7. Hemi-arthroplasty versus tension band wiring (4 part fractures)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Re-operation at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2 Implant removal at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3 Pain at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
Analysis 7.1.

Comparison 7 Hemi-arthroplasty versus tension band wiring (4 part fractures), Outcome 1 Re-operation at 1 year.

Analysis 7.2.

Comparison 7 Hemi-arthroplasty versus tension band wiring (4 part fractures), Outcome 2 Implant removal at 1 year.

Analysis 7.3.

Comparison 7 Hemi-arthroplasty versus tension band wiring (4 part fractures), Outcome 3 Pain at 1 year.

Comparison 8. Polyaxial versus monoaxial screw insertion in plate fixation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DASH score at 12 months (0 to 100: greatest disability)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
2 Simple shoulder test (0 to 12: best outcome)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
2.1 At 3 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2.2 At 6 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2.3 At 12 months1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3 Re-operation2122Risk Ratio (M-H, Fixed, 95% CI)1.43 [0.56, 3.67]
3.1 By 6 months166Risk Ratio (M-H, Fixed, 95% CI)0.85 [0.15, 4.76]
3.2 By 1 year156Risk Ratio (M-H, Fixed, 95% CI)1.86 [0.59, 5.88]
4 Dead at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
5 Constant score at 12 months (% of contralateral limb)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
6 Complications (radiological assessment)2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
6.1 Any complication1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.2 Primary implant malposition1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.3 Secondary loss of reduction and screw perforation1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.4 Non-union / delayed union due to osteonecrosis (6 months)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.5 Avascular necrosis at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.6 Varus deformity (> 10 / ≥20 degrees)2 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.7 Greater tuberosity displacement1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.8 Screw cut-out (intra-articular)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
7 Range of motion (degrees) at 12 months1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
7.1 Flexion1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.2 Abduction1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.3 External rotation1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
7.4 Internal rotation1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8 Operation and fluoroscopic times1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
8.1 Duration of operation (minutes)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
8.2 Fluoroscopic time (minutes)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 8.1.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 1 DASH score at 12 months (0 to 100: greatest disability).

Analysis 8.2.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 2 Simple shoulder test (0 to 12: best outcome).

Analysis 8.3.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 3 Re-operation.

Analysis 8.4.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 4 Dead at 1 year.

Analysis 8.5.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 5 Constant score at 12 months (% of contralateral limb).

Analysis 8.6.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 6 Complications (radiological assessment).

Analysis 8.7.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 7 Range of motion (degrees) at 12 months.

Analysis 8.8.

Comparison 8 Polyaxial versus monoaxial screw insertion in plate fixation, Outcome 8 Operation and fluoroscopic times.

Comparison 9. Medial support screws versus control for locking plate fixation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Adverse events1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Early loss of fixation1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 Re-operation for early failure1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.3 Osteonecrosis (asymptomatic)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Constant score (0 to 100: best) at 2.5 years1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
Analysis 9.1.

Comparison 9 Medial support screws versus control for locking plate fixation, Outcome 1 Adverse events.

Analysis 9.2.

Comparison 9 Medial support screws versus control for locking plate fixation, Outcome 2 Constant score (0 to 100: best) at 2.5 years.

Comparison 10. Hemiarthoplasty: EPOCA prosthesis versus HAS prosthesis
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Adverse events1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Deep infection1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
1.2 Persistent pain - scheduled for reoperation1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2 Radiological assessment findings1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Resorption of tuberosities1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.2 Secondary dislocation of tuberosities1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.3 Superior migration of prosthesis1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.4 Anterior subluxations1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.5 Glenoid erosion1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
2.6 Aseptic loosening of stem1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3 Range of motion results at one year (degrees)  Other dataNo numeric data
Analysis 10.1.

Comparison 10 Hemiarthoplasty: EPOCA prosthesis versus HAS prosthesis, Outcome 1 Adverse events.

Analysis 10.2.

Comparison 10 Hemiarthoplasty: EPOCA prosthesis versus HAS prosthesis, Outcome 2 Radiological assessment findings.

Analysis 10.3.

Comparison 10 Hemiarthoplasty: EPOCA prosthesis versus HAS prosthesis, Outcome 3 Range of motion results at one year (degrees).

Range of motion results at one year (degrees)
StudyMeasure

EPOCA prosthesis

n = 18

HAS prosthesis

n = 17

Reported significance
Fialka 2008Active forward flexionmean = 109°
range = 30° to 150°
mean = 62°
range = 20° to 110°
P < 0.001
Fialka 2008Active abductionmean = 101°
range = 30° to 150°
mean = 62°
range = 30° to 100°
P = 0.001
Fialka 2008Active external rotation in 90° abductionmean = 30°
range = 0° to 60°
mean = 17°
range = 0° to 40°
P = 0.01
Fialka 2008Active external rotation in 90° abductionmean = 45°
range = 0° to 70°
mean = 13°
range = 0° to 40°
P = 0.001
Comparison 11. Post-operative (percutaneous fixation) immobilisation for 1 week versus 3 weeks
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Neer score ≤ 80 points (unsatisfactory or failure) at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2 Premature removal of Kirschner wires1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
Analysis 11.1.

Comparison 11 Post-operative (percutaneous fixation) immobilisation for 1 week versus 3 weeks, Outcome 1 Neer score ≤ 80 points (unsatisfactory or failure) at 6 months.

Analysis 11.2.

Comparison 11 Post-operative (percutaneous fixation) immobilisation for 1 week versus 3 weeks, Outcome 2 Premature removal of Kirschner wires.

Comparison 12. Post-operative (hemiarthroplasty) mobilisation: early (2 weeks immobilisation) versus late (6 weeks)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Oxford Shoulder Score at 1 year (adjusted: 0 to 100 best)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
2 Constant shoulder score (at 1 year)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
2.1 Overall score (0 to 100: best)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2.2 Pain component (0 to 15: best))1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2.3 Activities of daily living component (0 to 25: best)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2.4 Mobility component (0 to 40: best)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
2.5 Strength component (0 to 25: best)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
3 Radiological assessment findings1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Nonunion (with bone resorption)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3.2 Malunion1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3.3 Greater tuberosity migration (all had severe pain at 6 & 12 months)1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
3.4 Superior luxation of prosthesis1 Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Range of motion at 1 year1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 Elevation (degrees)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
4.2 External rotation (degrees)1 Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 12.1.

Comparison 12 Post-operative (hemiarthroplasty) mobilisation: early (2 weeks immobilisation) versus late (6 weeks), Outcome 1 Oxford Shoulder Score at 1 year (adjusted: 0 to 100 best).

Analysis 12.2.

Comparison 12 Post-operative (hemiarthroplasty) mobilisation: early (2 weeks immobilisation) versus late (6 weeks), Outcome 2 Constant shoulder score (at 1 year).

Analysis 12.3.

Comparison 12 Post-operative (hemiarthroplasty) mobilisation: early (2 weeks immobilisation) versus late (6 weeks), Outcome 3 Radiological assessment findings.

Analysis 12.4.

Comparison 12 Post-operative (hemiarthroplasty) mobilisation: early (2 weeks immobilisation) versus late (6 weeks), Outcome 4 Range of motion at 1 year.

Appendices

Appendix 1. Search strategies update 2012

The Cochrane Library (Wiley online library)

(Issue 1, 2012)

#1 MeSH descriptor Shoulder Fractures explode all trees (37)
#2 MeSH descriptor Humeral Fractures explode all trees (56)
#3 MeSH descriptor Humerus explode all trees (49)
#4 (shoulder* OR humer*) (3496)
#5 fract* (21021)
#6 (#3 OR #4) (3496)
#7 (#1 OR #2) (93)
#8 (#5 AND #6) (466)
#9 (#7 OR #8), from 2010 to 2012 (139)

(of which there were 32 clinical trials)

MEDLINE (OVID Online)

Database: Ovid MEDLINE(R) <1946 to January Week 2 2012>, Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations <January 20, 2012>, Ovid MEDLINE(R) Daily Update <January 20, 2012>

1   Shoulder Fractures/ (2035)
2   Humeral Fractures/ (5499)
3   ((humer$ or shoulder$) adj10 (fracture$ or fixat$)).tw. (6713)
4   or/2-3 (8826)
5   (proximal or neck$1 or sub?capital).tw. (260185)
6   and/4-5 (2101)
7   or/1,6 (3177)
8   Randomized controlled trial.pt. (318052)
9   Controlled clinical trial.pt. (83363)
10 Randomized Controlled Trials/ (76365)
11 Random Allocation/ (72791)
12 Double Blind Method/ (112237)
13 Single Blind Method/ (15567)
14 or/8-13 (534158)
15 Animals/ not Humans/ (3554274)
16 14 not 15 (494473)
17 Clinical trial.pt. (465893)
18 exp Clinical Trials as topic/ (248066)
19 (clinic$ adj25 trial$).tw. (209081)
20 ((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).tw. (117109)
21 Placebos/ (30346)
22 placebo$.tw. (136751)
23 random$.tw. (574268)
24 Research Design/ (64342)
25 or/17-24 (1170523)
26 25 not 15 (1085220)
27 26 not 16 (630474)
28 or/16,27 (1124947)
29 and/7,28 (176)
30 (2010$ or 2011$ or 2012$).ed. (1996349)
31 29 and 30 (34)

EMBASE (OVID Online)

Database: Embase <1980 to 2012 Week 03>

1   Humerus Fracture/ (6698)
2   ((humer$ or shoulder$) adj10 (fract$ or fixat$)).tw. (7561)
3   or/1-2 (9937)
4   (proximal or neck$1 or sub?capital).tw. (295789)
5   and/3-4 (2479)
6   exp Randomized Controlled Trial/ (296049)
7   exp Double Blind Procedure/ (102662)
8   exp Single Blind Procedure/ (14708)
9   exp Crossover Procedure/ (31692)
10 Controlled Study/ (3676250)
11 or/6-10 (3746256)
12 ((clinical or controlled or comparative or placebo or prospective$ or randomi#ed) adj3 (trial or study)).tw. (581615)
13 (random$ adj7 (allocat$ or allot$ or assign$ or basis$ or divid$ or order$)).tw. (141624)
14 ((singl$ or doubl$ or trebl$ or tripl$) adj7 (blind$ or mask$)).tw. (135740)
15 (cross?over$ or (cross adj1 over$)).tw. (58055)
16 ((allocat$ or allot$ or assign$ or divid$) adj3 (condition$ or experiment$ or intervention$ or treatment$ or therap$ or control$ or group$)).tw. (177626)
17 or/12-16 (872189)
18 or/11,17 (4168319)
19 limit 18 to human (2510151)
20 and/5,19 (512)
21 (2010$ or 2011$ or 2012$).em. (2563197)
22 20 and 21 (120)

CINAHL (EBSCO)

CINAHL SEARCH UPDATE 1/1/2010- 20 February 2012

S1 (MH "Shoulder Fractures") (448)
S2 (MH "Humeral Fractures") (752)
S3 (MH "Humerus/IN/SU") (270)
S4 TX ( humer* or shoulder* ) and TX ( fracture* or fixat* )  (2562)
S5 S2 or S3 or S4  (2675)
S6 TX proximal or neck or subcapital or sub-capital  (43912)
S7 S5 and S6  (659)
S8 S1 or S7  (911)
S9 (MH "Clinical Trials+") (134701)
S10 (MH "Evaluation Research+")  (17924)
S11 (MH "Comparative Studies")  (65322)
S12 (MH "Crossover Design")  (8821)
S13 PT Clinical trial  (68992)
S14 (MH“Random Assignment”) (31673)
S15 S9 or S10 or S11 or S12 or S13 or S14 (220821)
S16 TX ( clinical or controlled or comparative or placebo or prospective or randomised or randomized ) and TX ( trial or study )  (381585)
S17 TX random* and TX ( allocat* or allot* or assign* or basis* or divid* or order* ) (55154)
S18 TX ( singl* or doubl* or trebl* or tripl* ) and TX ( blind* or mask* )  (601444)
S19 TX crossover* or cross-over or "cross over"  (11205)
S20 TX ( allocat* or allot* or assign* or divid* ) and TX ( condition* or experiment* or intervention* or treatment* or therap* or control* or group* )  (68871)
S21 S16 or S17 or S18 or S19 or S20 (911326)
S22 S15 or S21  (966852)
S23 S8 and S22  (357)
S24 EM (2010 or 2011 or 2012)  (759278)
S25 23 and S24  (72)

AMED (OVID Online)

Search 2/4/2012

1. exp Shoulder/ (1082)
2. exp Humerus/ (86)
3. fractures bone/ (452)
4. (fract$ or break$ or broken or ruptur$).tw. (8691)
5. 1 or 2 (1149)
6. 3 or 4 (8691)
7. 5 and 6 (51)
8. ((humer$ or shoulder$) adj10 (fracture$ or fixat$)).tw. (113)
9. 7 or 8 (125) (7 articles on shoulder fractures, post 2008)
10. (proximal or neck$1 or sub?capital).tw. (4488)
11. 9 and 10 (37) (No articles on shoulder fractures, post 2008)

Other searches

Journal of Bone and Joint Surgery - British Volume and conference proceedings from 2006 onwards (April 2012). Searched using "randomised" and then checked through the 1782 citations using Find: "humer"

WHO International Clinical Trials Registry Platform Search Portal (April 2012)
Searched using: proximal and humer* and fracture* (32 records)

Current Controlled Trials (April 2012)
Searched using: proximal and humer* (25 records)

Appendix 2. Numbers and status of studies in the published versions of the review

VersionTrial statusChanges
Ist version
Issue 1, 2001
The original review had 9 included studies, 4 excluded studies and 6 studies listed as ongoing. 
2nd version (substantive update)
Issue 2, 2002
This update had 10 included studies, 9 excluded studies, 3 studies listed as ongoing and 1 study awaiting assessment.Of the four newly identified studies, one (Stableforth 1984) was included, one (Warnecke 1999) was excluded, one (Dias 2001) listed as ongoing, and the other (Martin 2000) placed in Studies Awaiting Assessment. Further information obtained from trialists resulted in the exclusion of four trials that had been previously listed as ongoing studies. Three (Brownson 2001; Hems 2000; Wallace 2000) of these had been set up as a multi-centre study to test the Halder nail (Halder 2001), and one (Welsh 2000) had been set up to compare surgical with conservative treatment.
3rd version (minor update)
Issue 3, 2002
As aboveNote: this update included some changes to the Discussion in response to comments received from an external reviewer.
4th version (substantive update)
Issue 4, 2003
This update had 12 included studies, 11 excluded studies, and 4 studies listed as ongoing.Of four newly identified studies, one (Wirbel 1999) was included, one (de Boer 2003) excluded, and two (Frostick 2003; Shah 2003) are listed as ongoing. The other newly included trial (Hodgson 2003) was formerly listed as an ongoing trial. A trial (Martin 2000), previously in 'Studies awaiting assessment', was excluded. Limited additional findings from newly identified trial reports were included for Hoellen 1997.
5th version (minor update)
Issue 2, 2007
This update had 12 included studies, 12 excluded studies, 5 studies listed as ongoing and 4 pending assessment.Six new studies were identified, one (Fjalestad 2007) was listed as ongoing, one (Flannery 2006) was excluded and the other four were placed in 'Studies awaiting assessment', pending further information.
6th version
(new citation update)
Issue 12, 2010
This update had 16 included studies, 18 excluded studies, 11 studies listed as ongoing and 4 pending assessment.Sixteen new studies were identified. Of these, one (Fialka 2008) was included, four (Gradl 2009; Mechlenburg 2009; Wan 2005; Yang 2006) were excluded, 10 (Brorson; Guy; Helsinki; HURA; Liverpool (renamed as Sinopidis 2010 in the next update); Loma Linda; Pelet; ProCon; ProFHER; Ring) were placed in ongoing trials and one (Luo 2008) awaits assessment. New reports or information resulted in the inclusion of three more trials (Agorastides 2007: former ongoing study Frostick 2003; Fjalestad 2010: former ongoing study Fjalestad 2007; and Lefevre-Colau 2007: formerly Lefevre-colau 2006, a study awaiting assessment); and the exclusion of two studies (Bing 2002: former ongoing trial Sharma 2000; Dias 2001: former ongoing trial Dias 2001 and study awaiting assessment Der Tavitian 2006).
7th version (new citation update) Issue x, 2012This update had 23 included studies, 26 excluded studies, 14 studies listed as ongoing and 3 pending assessment.Overall, 18 new studies were identified. Of these, seven (Ockert 2010; Olerud 2011a; Olerud 2011b; Smejkal 2011; Voigt 2011; Zhang 2011; Zhu 2011) were included, four (Carbone 2012; Edelson 2008; Liao 2009; Zhang 2010) were excluded, five (Boyer; Engelsma; HOMERUS; Stanley; TPHF) were placed in ongoing trials and two (Battistella 2011; Fjalestad (RCT proposal)) await assessment. Further information was obtained for several studies in the previous version (Handoll 2010); this included the one year follow-up report of functional outcome for Fjalestad 2010, and information resulting in the exclusion of Sinopidis 2010, a former ongoing study. Further consideration of Shah, which was listed as an ongoing study, and Pullen 2007, which was awaiting classification, led to their exclusion: it is very unlikely that any further information will be obtained for these trials, including whether they started. Also excluded was Parnes 2005, another study awaiting classification in Handoll 2010; there is currently insufficient evidence to support this being a randomised trial.

Appendix 3. Previous acknowledgements and contribution of authors

Acknowledgements

We thank Panos Thomas for his contribution to the protocol, and Linda Digance, Christopher Muller and Sonia Stewart for foreign translations. We thank Bill Gillespie, Peter Herbison, Leeann Morton, David Sonnabend, John Stothard, Marc Swiontkowski and Janet Wale for their help at editorial review of the first version. We thank David Sonnabend for sharing his observations on the second version, and Bill Gillespie for his suggestions for incorporating these into the third version. We thank Lesley Gillespie, Peter Herbison, Nicola Maffulli and Janet Wale for their help at editorial review of the fourth version. We thank Joanne Elliott, Bill Gillespie, Lindsey Shaw and Janet Wale for their contributions at editorial review of the fifth version. We thank also Lesley Gillespie for her help in developing the revised search strategy and trial retrieval for the updates and advice on presentation. We thank Laurent Audige of the AO-ASIF Foundation and Anette Blümle of the German Cochrane Centre for the supply of several trial reports.

For the sixth version: We thank Nigel Hanchard, Annie Herbert and David Limb for their helpful feedback at editorial review. We thank Joanne Elliott for patiently supplying several search downloads for this version and Lindsey Elstub for her editorial support. We thank Ling-Hsiang Chuang for her help with translations from Chinese. We thank Graham Tytherleigh-Strong for his early input on this update.

We thank Alastair Gibson for his major contributions to the first four versions of this review and Rajan Madhok for his major contributions to the first five versions of the review.

We thank the following for further information on their research in this area before the current update: Alison Armstrong, Cathy Booth, Stig Brorson, Peter Brownson, Piet de Boer, Joe Dias, Tore Fjalestad, Mark Flannery, Tim Hems, Stephen Hodgson, Inger Mechlenburg, Roo Kulkarni, Shea Palmer, Rajiv Sharma, Robin Turner and Lei Yang.

Helen Handoll's work on the first version of the review was supported by the Chief Scientist Office, Department of Health, The Scottish Office, UK. Her work on the first and second updates was supported by the East Riding and Hull Health Authority, UK.

Contribution of authors

Rajan Madhok (RM) and Panos Thomas initiated the review and wrote the protocol. Helen Handoll (HH) searched for trials and provided a set of potential studies for inclusion. Alastair Gibson (JNAG) and HH assessed trial quality, tabulated the data and were the main authors of first published version of the review. JNAG, HH and RM contributed to the final manuscript.

For the first and third (both substantive) updates, Helen Handoll initiated the update by extending the search for trials and relevant materials, contacting trialists and preparing the first drafts. JNAG, HH and RM assessed the newly identified trials and contributed to the final manuscripts. All authors contributed to rewording of the discussion in the second minor update (amendment).

For the fourth (minor) update, Helen Handoll initiated the update by extending the search for trials and relevant materials, contacting trialists and preparing the first drafts. RM performed study selection and contributed to the final manuscript.

For the fifth update (sixth version), Helen Handoll initiated the update by extending the search for trials and relevant materials, contacted trialists, revised text and tables to conform to new methodology and formatting requirements, performed risk of bias assessment for already included trials and prepared the first full draft. Both authors piloted forms, performed study selection, and assessed risk of bias and extracted data for the new included trials. Benjamin Ollivere provided feedback on interim drafts and contributed to the final manuscript.

What's new

DateEventDescription
22 October 2012New search has been performedIn this update, published in Issue 12, 2012, the following changes were made:
1. The full search was updated to January 2012, with other searches extended to June 2012.
2. Eighteen new studies were identified. Of these, seven were included, four were excluded, five were placed in ongoing studies and two await assessment. A new report was available for one already included trial, and contact with a trialist resulted in the exclusion of one study that had been previously listed as ongoing. In the absence of further information after attempts at contact, consideration of one former ongoing study and two studies formerly in 'Studies awaiting assessment' led to their exclusion.
3. Discussion revised and updated.
4. Changes made to the conclusions.
22 October 2012New citation required and conclusions have changed1. Conclusions changed to accommodate findings of the new comparisons.
2. Change in authorship.

History

Protocol first published: Issue 1, 1996
Review first published: Issue 1, 2001

DateEventDescription
1 November 2010New search has been performedIn this update, published in Issue 12, 2010, the following changes occurred:
1. The full search was updated to March 2010; with other searches extended to August 2010.
2. Sixteen new studies were identified, of which one was included, four were excluded, 10 were placed in ongoing trials and one awaits assessment. New reports or information resulted in the inclusion of three more trials and the exclusion of two studies that had been identified previously.
3. Review methods and formatting were updated.
4. Background and Discussion revised and updated.
5. Changes made to the conclusions.
1 November 2010New citation required and conclusions have changed1. Conclusions changed to accommodate findings of the new comparisons.
2. Change in authorship.
5 August 2008AmendedConverted to new review format.
28 September 2007New search has been performedThe fourth update (Issue 2, 2007) included the following:
1. Trial search extended from May 2003 to September 2006.
2. Identification of six new studies: one of which was placed in 'Ongoing studies', one was excluded and the other four are in 'Studies awaiting assessment', pending further information or publication.
3. Various adjustments were made to text, tables and presentation of the analyses to conform to revised methodology and the Cochrane Style Guide: the 'Synopsis' was amended to a 'Plain language summary'; the 'Abstract' was shortened; the 'Objectives' were reworded; methodological quality scores of individual criteria are no longer summed; all totals were removed from the Analyses (Forest plots) and the number of Analyses were reduced by presenting similar outcome measures (e.g. complications) together from the same trial.
There was no change to the conclusions of the review.
12 August 2003New search has been performedThe third update (Issue 4, 2003) included the following:
1. Trial search extended from November 2001 to May 2003.
2. Inclusion of two new trials, one of which had been listed as ongoing.
3. Inclusion of two new ongoing trials.
4. Exclusion of four trials previously listed as ongoing.
5. One trial, previously in pending, was excluded.
6. Addition of limited findings from newly identified trial reports of an already included trial.
7. The conclusions of the review were slightly modified to include the possibility of immediate physiotherapy, without immobilisation, for some types of undisplaced fractures.
8 May 2002AmendedThe second update (Issue 3, 2002) included some changes to the Discussion in response to comments received from an external reviewer.
15 February 2002New search has been performedThe first update (Issue 2, 2002) included the following:
1. Trial search extended from July 2000 to November 2001.
2. Inclusion of one new trial.
3. Inclusion of one new ongoing trial.
4. Exclusion of four trials previously listed as ongoing.
5. One trial excluded and another placed in pending.
6. Addition of material from a newly available epidemiological study and commentary on a newly available systematic review.
There was no change to the conclusions of the review.

Contributions of authors

For the sixth update (seventh version), Helen Handoll initiated the update by extending the search for trials and relevant materials, contacted trialists, performed most of data entry and prepared the first full draft. All three authors screened and selected studies, and assessed risk of bias and extracted data for the newly included trials. Katie Rollins also entered data into RevMan. Both Benjamin Ollivere and Katie Rollins provided feedback on interim drafts and contributed to the final manuscript.

Helen Handoll is the guarantor of the review.

The summaries of the contributions of authors for previous versions of the review are presented in Appendix 3.

Declarations of interest

None known. Helen Handoll is a member of the trial management group of an ongoing trial (ProFHER); arrangements will be made for independent review of this trial when completed.

Sources of support

Internal sources

  • University of Teesside, Middlesbrough, UK.

External sources

  • No sources of support supplied

Differences between protocol and review

Update in 2012

Statement added to Types of participants clarifying the inclusion of trials with a small proportion of children.

A new secondary outcome was added to Types of outcome measures (Composite scores, whether validated or not, of subjective and objectively rated function and overall outcome). This was to distinguish explicitly between validated measures of patient-reported function and activities of daily living and other commonly used composite scores such as the Constant score.

Examples of the secondary outcome 'Other complications' added.

Update in 2010

Most of the changes to methods in Issue 12, 2010 reflected the uptake of new methodology and reporting as described in the Handbook (Higgins 2008b). These include risk of bias assessment and more explicit reporting of data analysis and collection. Types of outcome measures have been revised to define primary and secondary outcomes. Patient-reported measures of upper-limb function and a separate category for serious adverse events have been added.

Update in 2007

The order of the main categories of outcome measures was altered in Issue 2, 2007 to reflect the greater priority given to functional and clinical outcomes.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Agorastides 2007

MethodsRandomised using sequentially numbered sealed envelopes
Assessor blinding: stated for Constant Shoulder Assessment and Oxford scores at 6 and 12 months
Loss to follow-up at 1 year: 10 (all exclusions: 4 wrong prosthesis; 1 pathological fracture; 1 deep infection requiring further procedure; 2 initial greater tuberosity malpositioning; 2 did not attend follow-up visits)
ParticipantsRoyal Liverpool Hospital, Liverpool, UK
Period of study recruitment: October 2002 to October 2003
59 patients with displaced proximal humeral fractures, 3-part or 4-part or articular fractures who were treated with cemented hemiarthroplasty. Isolated non-pathologic fractures < 6 weeks old. Physiologically old patients with poor bone quality. Informed consent.
Exclusion criteria: no extra information
Of 49: 39 female, 10 male; mean age 70 years, range 34 to 85 years
Interventions

Intervention started post surgery (mean 10 days; range 1 to 30 days after injury)
1. Early active assisted mobilisation (after 2 weeks). Arm kept in sling in neutral rotation for 2 weeks; only pendulum and elbow exercises allowed. Between weeks 3 and 6, progressed to active-assisted exercises; from week 7, to active exercises.
2. Late mobilisation (after 6 weeks). Arm kept in sling in neutral rotation for 6 weeks; only elbow exercises allowed. From week 7 to week 12, progressed from pendulum to active-assisted exercises; from week 13, to active exercises.

Both mobilisation protocols were supervised by a team of specialist shoulder physiotherapists

OutcomesLength of follow-up: 1 year; also assessed at 2 and 6 weeks, and 3 and 6 months (coinciding with outpatient visits)
Oxford shoulder score
Constant shoulder score (mobility, strength, pain, activities of daily living)
Range of motion: elevation, external and internal rotation
Complications
Radiological assessment: greater tuberosity migration; superior luxation of prosthesis
NotesThe early mobilisation regimen represented normal practice at the hospital.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo description of method: "Patients were randomly allocated"
Allocation concealment (selection bias)Unclear risk"Randomization took place in the operating theater after the procedure, by use of sequentially numbered, sealed envelopes."
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear risk"At the 6- and 12-month visits, an independent blinded observer completed the Constant Shoulder Assessment and Oxford scores." However, care providers and participants were not blind to allocation and assessment of complications was not blinded either.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo accounting of these, but lack of blinding unlikely to affect reporting of these.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskIncomplete account of participant flow, with exclusion of 10 participants from the analyses.
Incomplete outcome data (attrition bias)
Death, reoperation
High riskNo accounting of these outcomes, but incomplete account of participant flow, with exclusion of 10 participants from the analyses.
Selective reporting (reporting bias)High riskNo protocol available. May have been stopped early, greater tuberosity migration not specifically listed in brief trial entry in the National Research Register (UK).
Balance in baseline characteristics?Unclear riskIncomplete data to back up claims of lack of baseline differences as these given only for 49 (10 excluded) but a 5 year difference in mean age (72 versus 67 years).
Free from performance bias?Unclear riskAlthough 3 upper limb surgeons performing the operations agreed to the same procedures a different uncemented prosthesis was used in 4 subsequently excluded participants.
"Both mobilization protocols were supervised by a team of specialist shoulder physiotherapists."

Bertoft 1984

MethodsUse of permutation table, single-blind, independently administered
Assessor blinded
Loss to follow-up at 1 year: 7/20 (2 excluded)
ParticipantsCentral hospital, Vasteras, Sweden
Period of study recruitment: not stated
20 patients with non or minimally displaced proximal humeral fractures (7 had fracture of the greater tubercle); sling for 10 days.
Exclusion criteria: no information
17 female, 3 male; mean age 64 years, range 50 to 75 years
InterventionsInterventions started 10 to 12 days post injury, after removal of sling.
1. Instructed self exercise: patients instructed to train 5 to 10 minutes, 4 to 5 times daily. They had three training sessions (day 1, weeks 3 & 8 post injury)
2. Conventional physiotherapy: 9 sessions (average 20 to 30 minutes), 1 to 2 times each week, over 10 to 12 weeks. No thermoelectrotherapy.
Assigned: 10/10
Completed ( > 1 year): 7/6
OutcomesLength of follow-up: 1 year; also assessed at 3, 8, 16 & 24 weeks
Range of motion: forward flexion (graph), abduction, internal & external rotation
Functional movements: placing hand on neck, placing hand on back
Pain: when placing hand on neck: combing hair (graph)
Isometric muscle strength: vertical & horizontal pushing
Change of treatment requested
NotesThe 2 excluded participants were in the control group: 1 died and 1 underwent an operation.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskMention of "permutation table" and "randomized controlled" trial
Allocation concealment (selection bias)Low risk"A third person was responsible for the randomization procedure and kept the key to the permutation table"
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear risk"A second physiotherapist examined the patients. She did not know to which group the patient belonged, and the patients were instructed not to tell her." However, there is no guarantee of blinding and, for practical reasons, neither participants nor care provider were blinded
Blinding (performance bias and detection bias)
Death, reoperation
Low riskLack of blinding unlikely to affect assessment of these outcomes
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskParticipant flow provided but large loss to follow-up (7/20 = 35%).
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskParticipant flow provided
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Unclear riskIncomplete data to back up claims of lack of baseline differences but a 4 year difference in mean age between groups (66 versus 62 years).
Free from performance bias?Low riskNo indication of performance bias.

Fialka 2008

MethodsMethod of randomisation: referral to random list and randomisation timed at surgery
Assessor blinding: no
Loss to follow-up at 1 year: 5/40 (3 deaths, 2 lost to follow-up)
ParticipantsVienna General Hospital, Austria
Period of study recruitment: not stated - lasted 22 months
40 patients with acute 4-part (Neer) proximal humeral fractures (type C: AO/ASIF classification), aged > 50 years, no history of previous problems in either shoulder, informed consent
Exclusion criteria: concomitant vascular or neurological injuries of involved limb; prior operative procedures; neurologic or mental disorders; or drug abuse
30 female, 10 male; mean age 75 years; of 35: range 56 to 88 years
InterventionsSurgery started 7.3 days of injury (0 to 26 days). General anaesthesia used in all cases. Stems were cemented in place and bone grafting was performed using cancellous bone from patient's humeral head.
1. Hemiarthroplasty using EPOCA prosthesis (Argomedical). Fixation of tuberosities using wire cables threaded through a medial and lateral hole in the stem.
2. Hemiarthroplasty using HAS prosthesis (Stryker). Fixation of tuberosities using transosseous braided sutures tied to lateral fin of the stem.
Same general rehabilitation protocol used for both groups: shoulder kept for 2 weeks in immobiliser to prevent active external rotation, passive movement for 15 minutes per day by physiotherapist to avoid contractures and shoulder stiffness. Then, active range of motion increased to horizontal level. Active external rotation initiated after another 2 weeks.
Assigned: ?/?
Completed (at 1 year): 18/17
OutcomesLength of follow-up: 1 year; also assessed at 12 days, 3 & 6 weeks, and 6 months
Functional assessment (individual Constant score, where results were relative to patient's unaffected shoulder)
Range of motion (active forward flexion, abduction, external rotation)
Radiological assessment: resorption of tuberosities, superior migration of prosthesis, anterior subluxations, glenoid erosion, aseptic stem loosening, secondary dislocation of the tuberosities, heterotopic ossification
Deep infection
Periprosthetic fracture
Reoperation & scheduled for reoperation (persistent pain)
Mortality
NotesDifferences between the two prostheses include the type and position of fixation of the tuberosities and the volume of the stem in the metaphyseal area, thus allowing different amounts of additional (autologous) cancellous bone grafting.
The data for heterotopic ossification were contradictory and not used here.
Request for information sent to contact trialist on 19 February 2010.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"The random list was designed to finally produce 2 groups of equal size."
Allocation concealment (selection bias)Unclear risk"Each surgeon was informed at the beginning of the operation as to which implant had randomly been selected."
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo blinding.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskLack of blinding unlikely to affect assessment of these outcomes. Standardisation of assessment.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskThe group allocation and baseline data were not provided for 5 participants: 2 lost to follow-up and 3 who had died. Standard deviations not provided.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskGroup allocation not provided for those who had died.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Unclear riskIncomplete baseline data (5 excluded) to confirm baseline comparability of those in analysis.
Free from performance bias?Low riskNo indication of performance bias: a "general rehabilitation protocol was used for all patients regardless of the type of implant."; each of the 4 participating surgeons was experienced in joint replacement surgery.

Fjalestad 2010

MethodsMethod of randomisation: use of computer software by independent hospital statistician; block size 12; use of numbered opaque sealed envelopes
Assessor blinding: no, but assessment by two independent physiotherapists
Loss to follow-up at 1 year: 2/50 (2 deaths)
ParticipantsOlso University Hospital, Oslo, Norway
Period of study recruitment: May 2003 to May 2008
50 patients with displaced proximal humeral fractures, AO group B2 or C2 (displaced 3-part and 4-part fractures) who were admitted to hospital. Malposition was at least 45° angular deviation in the true frontal (inclination) or transthoracic radiographic projections, regardless of whether the fracture was impacted or not. The greater or lesser tuberosity had to be displaced at least 10 mm. Furthermore, the displacement between the head and metaphyseal/diaphyseal main fragments should not exceed 50% of the diaphyseal diameter. Age 60 years or over. Informed consent. Resident in Oslo.
Exclusion criteria: non-Scandinavian ethnicity, previous history of injury or illness of the injured or contralateral shoulder, injury of the other part of the humerus or the contralateral upper extremity, alcohol- or drug abuse, dementia or neurological disease and severe cardiovascular disease that would contraindicate surgery.
44 female, 6 male; mean age 73 years, range 60 to 88 years
InterventionsInterventions (and randomisation) started after hospital admission. (On admission to the hospital, all patients were immobilized in a modified Velpeau bandage.)
1. Surgery: operation occurred within the first week after admission to hospital. Open reduction and fixation using a minimally open deltopectoral approach with an interlocking plate device (Locking Compression Plate (LCP) of the AO basic type, Synthes, Switzerland) and metal cerclages to secure the tuberosities. Operation was performed under fluoroscopic control. Then immobilisation in a modified Velpeau bandage until self-exercises and instructed physiotherapy was started on the third postoperative day.
2. Conservative treatment: All patients stayed in the hospital for at least 1 day and received the same instructions from the physiotherapist as those allocated to surgery. If the displacement between the head and metaphyseal fragment (main fragments) exceeded 50% of the diaphyseal diameter (subsequent to randomisation), closed reduction was performed in the operating room under general anaesthesia within 48 hours of admission. Immobilisation in a modified Velpeau bandage for 2 weeks before self exercises and instructed physiotherapy started on day 15.
The same self-training programme and instructed physiotherapy programme used for both groups, although the conservative treatment group started 12 days later. Both groups progressed to strengthening exercises at the 6 week time point. Physical therapy and self-exercise were recommended for at least 6 months.
Assigned: 25/25
Completed (at 1 year): 23/25
OutcomesLength of follow up: 2 years; also assessed at 2, 8, 12, 26 and 52 weeks
Constant shoulder score (both shoulders) (3, 6 & 12 months)
ASES (American Shoulder and Elbow Surgeons) questionnaire (sports domains not included - maximum 24 points) (6 & 12 months)
Quality of life score: Harri Sintonen 15D instrument (sexual function domain not included)
Mortality
Fixation failure or redisplacement - subsequent operation
Radiographic outcomes including avascular necrosis (score 2 = no changes; 1 = changes to normal trabecular organisation < 50% of humeral head; 0 = > 50% or partial collapse)
Check for axillary nerve injury
Health economic outcomes, including direct (cost of surgery; cost of hospital stays) and indirect costs (sick leave, family use of time to assist patient)
Length of stay in acute hospital and hospital rehabilitation centre
Notes

Information on the trial received December 2006 from Dr Tore Fjalestad.
Currently only some results for one year follow-up are published. Communication from Dr Tore Fjalestad in April 2010 indicated that the two year follow-up was likely to be finished during 2010. Further information from Dr Tore Fjalestad in April 2012 indicated that the two year follow-up had been submitted to another journal (estimated publication during 2012).

More details on conservative treatment were provided in Fjalestad 2012. Tore Fjalestad also provided in an email (April 2012) the following clarification on the use of closed reduction for 8 conservatively treated participants (this had not been described in the protocol): "The primary X-rays were assessed for classification and decision-making for closed reduction. Those eight patients had a new radiographic examination after allocation to conservative treatment and after the procedure in the operating room, to confirm an acceptable position of the fragments. If not acceptable, the patients had to be treated with ORIF. Surprisingly, only one patient demonstrated unacceptable re-displacement after two weeks, and was analyzed according to intention-to-treat principle in the conservative group at one year."

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The [randomisation] procedure was designed by the statistician at the hospital research centre using the computer software S-PLUS 6.0 for Windows 2002 ... Randomisation was based on equal blocks of length 12, with the exception of the last one, which was interrupted due to 50 patients."
Allocation concealment (selection bias)Low risk"Randomisation was performed by means of consecutively numbered and sealed non-translucent envelopes containing each participant’s allocation to surgery or to conservative treatment." Independent statistician.
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskTwo trained physiotherapists performed the 15D interviews. The physiotherapists were not blinded to group assignment. No provider or participant blinding.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskLack of blinding unlikely to affect assessment of these outcomes, but may affect decisions for subsequent surgery.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Low riskParticipant flow diagram provided and intention-to-treat analysis conducted.
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskParticipant flow diagram provided and intention-to-treat analysis conducted.
Selective reporting (reporting bias)Unclear riskTrial registered after completion. Small discrepancies in trial inclusion criteria.
Balance in baseline characteristics?Unclear riskStatistically non significant imbalance in gender (5 versus 1 males) and baseline quality of life scores (higher in surgical group).
Free from performance bias?Low riskAll the operations were performed by three surgeons experienced in the procedure performed.

Hodgson 2003

MethodsRandomised using sequentially numbered sealed envelopes
Assessor blinding: yes, on review of patients at home or clinic appointment
Loss to follow-up at 1 year: 4 (1 death); at 2 years: 12 (3 deaths)
ParticipantsRoyal Hallamshire Hospital, Sheffield, UK
Period of study recruitment: November 1998 to April 2000
86 patients, over 40 years old, with minimally displaced 2-part fractures (Neer), including isolated fractures of the greater tuberosity
Exclusion criteria: inability to understand written or verbal information
70 female, 16 male; mean age 70 years
InterventionsIntervention started: at arrival at A&E.
1. Early physiotherapy (within 1 week of the fracture). Most patients were seen by a physiotherapist at clinic the day after their fracture. Patients received a sling for comfort but were instructed to take their arm out of the sling and to perform gradual, assisted movements of the upper limb.
2. Late physiotherapy after 3 weeks of immobilisation in a collar and cuff sling.
Both groups received same rehabilitation programme. First 2 weeks: education and instruction for home exercises; weeks 2 to 4: progression to full passive flexion and light functional exercises; week 4: start of progressive functional exercises. Discharge when both patient and physiotherapist thought independent shoulder function was achieved.
OutcomesLength of follow-up: 2 years, also 8 and 16 weeks and 1 year
Functional assessment (Constant score)
Patients' perceived health status: SF36 (physical function, physical role limitation, pain); Croft shoulder disability questionnaire
Complications
Number of physiotherapy treatment sessions
NotesInformation on this trial received from Mr Hodgson on several occasions. This included draft report of the 2 year follow-up and notice of their plan to extend follow-up to 5 years.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: "using sequentially numbered sealed envelopes we randomly allocated patients"
Allocation concealment (selection bias)Low risk"using sequentially numbered sealed envelopes we randomly allocated patients". Also from phone conversation (08/08/2001): "physio opened envelopes when details entered on envelope".
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskBlinded assessor of function but patients and care providers were not blinded.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskLack of blinding unlikely to affect assessment of these outcomes.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Low riskA full account of loss to follow-up provided. While 14% at 2 years (12/86), it was under 5% (4/86) at 1 year.
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskParticipant flow provided.
Selective reporting (reporting bias)Unclear riskTrial registration was incomplete and differed slightly from final reports.
Balance in baseline characteristics?Unclear riskMore males in the early mobilisation group (11 versus 5).
Free from performance bias?Low riskPerformance bias seemed unlikely.

Hoellen 1997

MethodsRandomisation method unknown
Assessor blinding: not stated
Loss to follow-up at 1 year: 12/30 (3 deaths)
ParticipantsUniversity Clinic Ulm, Germany
Period of study recruitment: 1/12/1994 to 30/06/1996 in Hoellen 1997 report (to 31/08/1998 in Holbein 1999 report)
30* patients with 4-part fractures (Neer). *see Notes.
Exclusion criteria: age < 65 years, > 14 days since fracture, rheumatoid arthritis, previous shoulder injury, terminally ill
24 female, 6 male; mean age 74 years
InterventionsInterventions started within 14 days of fracture.
1. Hemi-arthroplasty (Global prosthesis, DePuy, US) - cemented
2. "Minimal osteosynthesis": tension band wiring - 2 pins + figure of 8 wire
All were given low dose heparin for DVT prophylaxis. The same post-operative treatment was used in both groups. A Glichrist bandage was used for temporary rests. Passive moving exercises started from first postoperative day, with active exercises postponed until after 6 weeks. Referral to rehabilitation clinic for 3 to 4 weeks post discharge.
Assigned: 15/15
Completed (1 year): 9/9
OutcomesLength of follow-up: 1 year
Functional assessment (Constant score)
Mobility (component of Constant score)
Pain (ditto)
Power
Haematoma
Infection
Implant failure
Medical complications
Re-operation
Time on ward
Discharge location
Mortality
NotesThe plan for longer term follow-up was announced in the Hoellen 1997 trial report. Further abstracts and a trial report (Holbein 1999) were identified for the review update (Issue 4, 2003). Holbein 1999 reported on 39 patients (19 versus 20), with 3- and 4-part fractures, 31 (?/?) of whom had been followed up for 1 year and 24 (?/?) for 2 years. Requests (June 2003) for further information, including for denominators, resulted in the discovery that both Dr Holbein and Dr Hoellen were no longer at Ulm.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: prospective randomised trial
Allocation concealment (selection bias)Unclear riskNo details: prospective randomised trial
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo blinding.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskLack of blinding unlikely to affect assessment of these outcomes
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskParticipant flow provided but large loss to follow-up (12/30 = 40%); and potential exclusions.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskParticipant flow provided but large loss to follow-up (12/30 = 40%). Serious outcomes though are less likely to be missed.
Selective reporting (reporting bias)High riskInsufficient information to judge this but the pragmatic removal of the power component of the Constant score was post hoc. Also non addressed difference in trial inclusion criteria between the two reports of this trial.
Balance in baseline characteristics?Unclear riskNo information on baseline characteristics of the two treatment groups but inclusion criteria rule out some confounders.
Free from performance bias?Unclear riskSame post-operative treatment but in all there is insufficient information to assess performance bias.

Kristiansen 1988

MethodsMethod of randomisation: unknown, "randomly selected"
Assessor blinding: unlikely
Loss to follow-up at 1 year: 10/31 (4 failed to attend, 2 died, 4 excluded)
ParticipantsRigshospitalet, Copenhagen, Denmark
Period of study recruitment: not stated
30 patients with 31 displaced 2-, 3- and 4-part proximal humeral fractures (Neer).
Exclusion criteria: no information
22 female, 9 male; age range 30 to 91 years
InterventionsInterventions started: not stated.
1. Surgery: Percutaneous reduction (using Steinmann pin under image intensifier control) and external fixation (2 half pins with continuous threads into humeral head and 2 or 3 pins into the humeral shaft, and neutralising bar applied; Steinmann pin removed)
2. Conservative treatment: closed manipulation under general anaesthesia & sling
Assigned: 15/16
Completed (at 1 year): 11/10
OutcomesLength of follow-up: 12 months; also assessed at 3 & 6 months
‘Treatment failure’: poor reduction, pin removal due to loosening
Non-union
Quality of fracture reduction: good, fair, poor
Functional overall score: excellent, satisfactory, unsatisfactory, poor. Neer (without anatomical section)
Complications: avascular humeral head necrosis, deep infection, radiographic pseudarthrosis, refracture
Reoperations
Mortality
NotesIn both groups, functional exercises were started under instruction during the first week.
Excluded participants were: 1 treatment failure (deep infection) in the surgical group; and 2 treatment failures (poor reduction) and 1 refracture in the conservative treatment group.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: "randomly selected for treatment"
Allocation concealment (selection bias)Unclear riskNo details: "randomly selected for treatment"
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo blinding reported.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskUnlikely to be affected by lack of blinding
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskExclusion of data for participants with treatment failure and early refracture from 12 month review. Large loss to follow-up (10/31 = 32%).
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskParticipant flow provided.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Low riskNo information on the patient with bilateral fractures but a relatively minor unit of analysis issue.
Free from performance bias?Unclear riskNo information on operator competence/expertise.

Kristiansen 1989

MethodsMethod of randomisation: unknown
Assessor blinding: yes at 2 year follow-up
Loss to follow-up at 2 years: 46/85 (18 deaths, 28 non-attenders)
ParticipantsHvidovre University Hospital, Denmark
Period of study recruitment: 1983
85 patients with proximal humeral fractures; 74% minimally displaced (Neer).
Exclusion criteria: no information
60 female, 25 male; median age 72 years (1 week group), 70 years (3 weeks group)
InterventionsInterventions started immediately or after closed or open manipulation.
1. One week immobilisation in sling and body bandage.
2. Three weeks immobilisation in sling and body bandage.
At the end of immobilisation, instructions were given to perform Codman's pendulum exercises as well as active movements of the elbow and hand.
Assigned: 42/43
Completed (at 2 years): 18/21
OutcomesLength of follow-up: 2 years; also assessed at 1, 3, 6 & 12 months
Overall score (Neer without anatomic section)
Mobility: overall from Neer score (range of motion: flexion, extension, abduction, internal & external rotation)
Function: overall from Neer score (strength, reaching, stability)
Pain: overall from Neer score (none to disabling)
Reflex sympathetic dystrophy
NotesPost immobilisation for both groups: instructions given for Codman’s pendulum exercises as well as active movements of elbow and hand.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: "Random allocation to immobilization for 1 to 3 weeks was performed"
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskOnly claimed for outcome assessors at final follow-up: "The 2-year follow-up examination was blind, as the examiners had no knowledge of the period of immobilization."
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo blinding but may not have affected appraisal of mortality (which was not split by treatment group).
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskLarge loss to follow-up (46/85 = 54%). Numbers given for those available at follow-up but incompletely reported data: only medians.
Incomplete outcome data (attrition bias)
Death, reoperation
High riskAlthough numbers given for those available at follow-up, only overall mortality data provided (extracted from graph).
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Unclear riskAlthough there appeared to be comparability between treatment groups in age and gender, the percentage of minimally displaced fractures (79% versus 70%: 33/42 versus 30/43) differed between the two groups and no information was available on the numbers who had open manipulation (thus entailing surgery).
Free from performance bias?Unclear riskLack of information to judge on performance bias.

Lefevre-Colau 2007

MethodsRandomised using block randomisation (under supervision of a statistician) and telephone to an independent researcher with patient details.
Assessor blinding: yes
Loss to follow-up at 6 months: 10 (all had difficulties in travelling to the hospital for scheduled sessions)
ParticipantsCochlin Hospital, Paris, France
Period of study recruitment: October 2002 to March 2005
74 patients, over 20 years old, with non-operatively treated impacted ("stable") fractures, including minimally displaced (1-part fracture); 2-part (surgical neck or greater tuberosity (1)); and 3-part (surgical neck and greater tuberosity) (Neer). (AO classification also given). Written consent.
Exclusion criteria: pre-existing shoulder pathology, neurological upper limb disorder, indication for shoulder surgery, multiple injuries, high-energy trauma, or difficulties with language or unable to understand rehabilitation programme or other treatment information.
54 female, 20 male; mean age 63 years
Interventions

Intervention started with 72 hours after fracture.
1. Early mobilisation: active rehabilitation begun within 72 hours of fracture: 2 hour sessions supervised by a physiotherapist, 5 times a week. Progressing from physical techniques to manage pain, then passive motion, performed by physiotherapist, in a) abduction, with arm suspension and patient supine (session 1); passive range of motion in forward elevation with the patient in a lateral supine position (session 2), with addition of external rotation with the patient in a seated position at session 8. After 3 weeks, sessions occurred twice a week without arm suspension. Patients wore a sling between sessions for 4 to 6 weeks, depending on the level of pain. After 6 weeks, active range of motion was begun during weekly sessions. Strengthening began at 3 months in twice-monthly sessions. Patients underwent a total of 32 sessions.
2. Usual care, starting with 3 weeks of sling immobilisation. Then 2 hour sessions supervised by a physiotherapist 4 times a week for 4 weeks. Passive mobilisation in all planes without arm suspension was performed by physiotherapist. Patients kept their arm in a sling between sessions for 1 to 3 additional weeks, depending on pain level. Then sessions were scheduled 2 times weekly for 5 weeks. Active range-of-motion exercises began after 6 weeks. After 9 weeks of rehabilitation, sessions occurred twice monthly until 6 months. Each patient underwent a total of 33 sessions.

Patients used oral analgesics to manage pain. After 4 to 6 weeks, patients were advised to perform daily exercises at home. Patients were discharged from the study at 6 months.

OutcomesLength of follow-up: 6 months, also 6 weeks and 3 months
Functional assessment (Constant score: split into subjective and objective components)
Pain
Patient satisfaction
Range of motion: abduction, anterior elevation, lateral rotation
Complications: non-union (0); fracture displacement (0); treatment (injection) for subacromial impingement syndrome
Compliance
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Block randomization involved choosing randomly from among blocks of lengths 4 and 2 to prevent the risk of predictability."
Allocation concealment (selection bias)Low risk"After completion of the trial entry details, an independent researcher responsible for treatment allocation was contacted by telephone."
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear risk"Outcome measures were recorded by two physicians, including one of the authors (F.F.), who were blinded to the treatment assignments." However, care providers and participants were not blinded to allocation.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNot reported.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskData were unavailable from 10 participants (5 in each group) who were lost to follow-up because of difficulties in travelling to the hospital. Their characteristics were reported not to differ from those who attended.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskNot reported.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this; retrospective trial registration.
Balance in baseline characteristics?Low risk 
Free from performance bias?Low riskRehabilitation was standardised and "delivered by physiotherapists who were experienced in the field".

Livesley 1992

MethodsMethod of randomisation: unknown, double-blind
Assessor blinding: likely as code only broken at end of trial
Loss to follow-up at 6 months: 3/48
ParticipantsMansfield District General Hospital, Mansfield, UK
Period of study recruitment: November 1988 to May 1990
48 patients with minimally displaced humeral neck fractures (all Neer Group 1); 4 had epiphyseal fractures
Exclusion criteria: able to co-operate with treatment and attend daily therapy for the first 10 working days.
37 female, 11 male; age range 11 to 85 years
InterventionsInterventions started on average 8.6 days since injury, upon referral to physiotherapy department.
1. Pulsed high frequency electromagnetic field (‘Curapulse’), 30 minutes/day for first 10 working days. (Intensity setting 3, pulse repetition frequency 35, maximum pulse power 300 watts.)
2. Dummy apparatus (deactivated machine).
Assigned: 22/26
Completed (at 6 months): 21/24
OutcomesLength of follow-up: 6 months; also assessed at 1 & 2 months
No data provided in report
Range of movement of glenohumeral & scapulothoracic joints
Pain scores, at rest, on movement, analgesia requirement
Muscle wasting and strength
Overall functional assessment score
Subjective opinion of treatment
Overall estimation of treatment (a ‘good result’)
Time to discharge
NotesAll patients received the same standardised physiotherapy regimen.
No data provided in report for comparison between the two interventions.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details provided: "patients were randomized into two groups"
Allocation concealment (selection bias)Low risk"double-blind", and randomisation code was only broken at end of the trial period to permit analyses
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Low risk"double-blind", use of sham control
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo report of these outcomes
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskAlthough loss to follow-up reported, no results were presented for the trial groups.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskNo report of these outcomes
Selective reporting (reporting bias)High riskResults not presented.
Balance in baseline characteristics?Unclear riskBaseline comparability. However, although the article claims "patients ... were referred to the physiotherapy department without delay", the ranges for average time from injury to start treatment were 0 to 17 days (active) and 0 to 27 days (sham).
Free from performance bias?Unclear risk"Standardized physiotherapy regimen". However, although the article claims "patients ... were referred to the physiotherapy department without delay", the ranges for average time from injury to start treatment were 0 to 17 days (active) and 0 to 27 days (sham).

Lundberg 1979

MethodsMethod of randomisation: unknown
Assessor blinding: no, but mention of independent assessors
Loss to follow-up at 3 months: 0/42; not known for final assessment.
ParticipantsGavle, Sweden
Period of study recruitment: not stated
42 patients with undisplaced proximal humeral fractures (all Neer Group 1) fixed with a sling; 13 had avulsion of the greater tuberosity.
Exclusion criteria: no information
37 female, 5 male; mean age 65 years
InterventionsInterventions started 7 days post injury, after removal of sling.
1. Instructed self exercise: patients instructed to train 5 to 10 minutes, 4 to 5 times daily. They had 3 visits (day 1, and 1 & 3 months) to physiotherapist for instructions and checks. At 1 month, patients were told how to extend their exercises to same level as in physiotherapy group.
2. Conventional physiotherapy: 9 visits (average 20 to 30 minutes) between 2 to 3 months; patients encouraged to continue exercise at home. At about 4 weeks, treatment was intensified.
Assigned: 20/22
Completed (at 3 months): 20/22; (at mean 16 months): ?/?
OutcomesLength of follow-up: > 1 year (mean 16 months); also assessed at 1 & 3 months
Range of movement: abduction, shoulder elevation - active & passive
Pain (insignificant, moderate, severe), longstanding
Lifting power of shoulder
Frozen shoulder (secondary)
Neer score (at final evaluation) including failure category
Hand grip strength
NotesNo indication in the report of any loss to follow-up at last follow-up (> 1 year), but cannot be assumed.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details of method: "In all, 42 patients were randomly assigned into two groups."
Allocation concealment (selection bias)Unclear riskNo details of method: "In all, 42 patients were randomly assigned into two groups."
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo blinding, although independent assessment claimed: "Examination was made by physicians and physiotherapists independently at 1 month and 3 months.."
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo reported.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskFull data provided for 1 and 3 months follow-up; but denominators not stated for long-term (mean 16 months) follow-up
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskData not reported for these outcomes
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Low riskNo major imbalances in baseline characteristics
Free from performance bias?Low riskNo indications of performance bias.

Ockert 2010

Methods

Method of randomisation: used closed envelopes

Assessor blinding: unknown

Loss to follow-up: 10 patients excluded from analysis following randomisation; 6 with polytrauma, 2 with neurologic deficiency and 2 (1 versus 1) who were converted to shoulder arthroplasty intraoperatively. There was no mention of group allocation at randomisation or evaluation in the paper - these (8 versus 2) were notified after contact with the lead trial investigator.

Participants

Ludwig-Maximilians University, Munich, Germany

Period of study recruitment: August 2006 to July 2008

76 patients, aged over 18 years, with displaced proximal humeral fractures with displacement > 1 cm and angulation of fragments > 45 degrees (Neer criteria).

Exclusion criteria: poly-traumatised patients, neurologic deficit or intra-operative conversion to shoulder arthroplasty. (Paper noted there were no open or pathological fractures.)

Of 66: 48 female, 18 male; mean age 68 years, range 29 to 92 years

Interventions

1. Polyaxial angular stable plate fixation (Non-contact bridging – Proximal Humerus (NCB-PH)). Polyaxial plating allows a range of 0 to 15-degree angle off-centre. After insertion, a threaded screw cap locks the axis of the screw.
2. Monoaxial angular stable plate fixation with Proximal Humeral Internal Locking System (PHILOS) Synthes GmbH. Monoaxial locking plate technique is characterized by fixed divergent and convergent screw orientation due to threaded screw holes.

A deltopectoral approach was used for open reduction and internal fixation of all fractures. All patients received prophylactic intravenous antibiotic immediately before surgery. "The postoperative rehabilitation protocol included immediate passive and active-assisted range of motion (ROM) up to 60-degree angle of abduction and elevation without forced external rotation for 6 weeks. Full ROM with active exercises was started 6 weeks after operation.”

Assigned: 39/37

Completed: 29/37

Outcomes

Length of follow-up: 6 months (X-rays 1 day, 6 weeks, 3 months and 6 months)

Secondary varus displacement (> 10 degrees)

Delayed union (due to osteonecrosis)
Intra-articular screw cut out

Re-operation: revision surgery and early hardware removal

Infection (none)
Neurovascular injuries (none)

NotesRequest for information sent to Dr Ockert on 2 June 2012. Repeated on 8 June 2012, in email Peter Biberthaler regarding identification and further information on ongoing trial referred to in conference abstract (Biberthaler 2009) - it seems highly likely that the ongoing trial was this trial. However, this was not clear from email from Ben Ockert on 18 June 2012; this also provided details on the method of randomisation, the numbers allocated and analysed in each group
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"consecutive patients ... were prospectively randomized". No description of sequence generation.
Allocation concealment (selection bias)Unclear risk"consecutive patients ... were prospectively randomized". Contact from trialist revealed they "used closed envelope technique for randomization". (Exclusion criteria appeared to be applied post-randomisation.)
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo mention of blinding. Radiographic assessment performed by two trained radiologists twice in separate sessions 8 weeks apart. Consensus decision for osteonecrosis and implant related failure. Criteria for healing stated.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo mention of blinding, but unlikely to affect this.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High risk

"Follow-up rate was 71% of all radiographs taken 1 day, 6 weeks, 3 months, and 6 months after surgery."

Numbers of patients allocated or assessed by intervention group provided after personal communication. Post-randomisation exclusions (10/76 = 13%),was imbalanced (8 versus 2) and other loss to follow-up not accounted for.

Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskAs above. Paper described cases of revision surgery and early removal of metalwork; however, group allocation not given. Information provided subsequently.
Selective reporting (reporting bias)Unclear riskNo protocol available.
Balance in baseline characteristics?Unclear risk"The fracture types were equally distributed in both study groups." However, this applied to 66 participants. Does not state how many patients in each group or compare demographics.
Free from performance bias?Low risk

Six experienced surgeons performed the surgery: "In advance of this study, all surgeons were trained in the respective monoaxial and polyaxial locking plate system”.

Same antibiotic regimen and post-op management.

Olerud 2011a

MethodsMethod of randomisation: opaque, sealed envelopes
Assessor blinding: no, but mention of independent surgeon
Loss to follow-up at 24 months: 7/60 (1 excluded themselves; 2 lost; 4 died)
ParticipantsStockholm Söder Hospital, Stockholm, Sweden
Period of study recruitment: April 2003 to March 2008
60 patients with an acute displaced (based on Neer's criteria) 3-part proximal humeral fractures (all had displaced surgical neck fracture, all bar one had a displaced greater tuberosity; the exception had a displaced lesser tuberosity). Age 55 or older with a fracture sustained after a low-energy trauma (e.g. a simple fall). Independent living conditions
Exclusion criteria: patients with a completely displaced shaft in relation to the head fragment or with a valgus impact fracture. Institutionalised, severe cognitive dysfunction (< 3 correct answers on a 10-item Short Portable Mental Status Questionnaire).
48 female, 11 male; mean age 74 years (1 patient excluded themselves), range 56 to 92 years (operations were performed within a mean of 6 (SD 4.1) days after the injury).
InterventionsInterventions (and randomisation) started after hospital admission.
1. Surgery: operation occurred at mean of 6.1 days of injury. Open reduction and fixation using a deltopectoral approach with a PHILOS plate (Synthes, Stockholm, Sweden) and with nonabsorbable sutures used to fix displaced / unstable lesser and/or greater tuberosity fractures. The reduction and position of the implant was checked with the aid of an X-ray image intensifier. (All patients had pre-operative antibiotics.) Post surgery, the arm was placed in a sling and patients were referred to a physiotherapist. The sling was used for 4 weeks; afterwards, the patients were allowed to use it at their own convenience. Pendulum exercises and passive elevation/abduction up to 90 degrees were started from the first postoperative day. After 4 weeks, the patients were allowed a free active range of movement.
2. Conservative treatment: arm immobilisation in a sling for 2 weeks, after which they were allowed to use it at their own convenience. After 2 weeks, the patients were referred to a physiotherapist, and pendulum exercises and passive elevation/abduction up to 90 degrees were started. After 4 weeks, they were allowed a free active range of movement.
Assigned: 30/30
Completed (at 2 years): 27/26
Outcomes

Length of follow up: 2 years
Constant shoulder score (both shoulders)
DASH (Disabilities of the Arm, Shoulder and Hand) questionnaire
Quality of life score: EQ-5D
Mortality
Pain
Range of motion: abduction, flexion
Fixation failure, redisplacement, non-union, malunion

Subsequent surgery (reasons including deep infection, etc)
Radiographic outcomes including avascular necrosis and osteoarthritis

Notes

Trial run concurrently with Olerud 2011b.

Additional information on randomisation and trial location obtained from Dr Olerud (April 2012). Pain data received May 2012.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

"After clearance by an anesthetist, the patients were randomized (independently prepared opaque, sealed envelopes) to open reduction and internal fixation with a locking plate or nonoperative treatment." trial report

"the patients were randomised by numbered sealed opaque envelopes drawn consecutively.

The envelopes were independently prepared and thoroughly mixed. After that the envelopes were numbered by another person. At the time of randomisation the envelopes were drawn in numerical order." personal communication

Allocation concealment (selection bias)Low riskSee above
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High risk

No assessor blinding, although "The final 24-month follow-up was performed by an independent orthopaedic surgeon not previously involved in the treatment."

No provider or participant blinding.

Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskLack of blinding unlikely to affect assessment of these outcomes, but may affect decisions for subsequent surgery.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Low risk"In the outcome analyses, all patients remained in their randomization group regardless of secondary procedures according to the intention-to-treat principle." Participant flow provided; no cause for concern.
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskAs above.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this. No protocol found.
Balance in baseline characteristics?Low riskNo imbalances: baseline comparability.
Free from performance bias?Low risk

"All operations in patients randomized to surgery were performed by 1 of 2 orthopaedic surgeons, both well experienced in shoulder surgery."

While all surgical patients were referred to a physiotherapist after their surgery and conservatively treated patients were referred after 2 weeks, this was unlikely to influence results. Otherwise, similar exercise / rehabilitation schedules.

Olerud 2011b

MethodsMethod of randomisation: opaque, sealed envelopes
Assessor blinding: no, but mention of independent surgeon
Loss to follow-up at 24 months: 6/55 (1 lost; 5 died)
ParticipantsStockholm Söder Hospital, Stockholm, Sweden
Period of study recruitment: April 2003 to March 2008
55 patients with an acute displaced (based on Neer's criteria) 4-part proximal humeral fractures (all had displaced surgical neck, greater and lesser tuberosity fractures). Age 55 or older with a fracture sustained after a low-energy trauma (e.g. a simple fall), no previous shoulder problems. Independent living conditions.
Exclusion criteria: patients with a completely displaced shaft in relation to the head fragment or with a valgus impact fracture. Institutionalised, severe cognitive dysfunction (< 3 correct answers on a 10-item Short Portable Mental Status Questionnaire).
47 female, 8 male; mean age 77 years, range 58 to 92 years
Interventions

Interventions (and randomisation) started after hospital admission.
1. Surgery: operation occurred at mean of 6.0 days of injury. Humeral head replacement using a deltopectoral approach with the Global Fx prosthesis (DePuy, Sollentuna, Sweden); this is a modular prosthesis with a fixed angle and a conventional head - it has 3 fins. Heavy nonabsorbable sutures were tagged on the bone tendon interface of both tuberosities.

Cancellous bone graft from the head fragment was placed between the shaft and the tuberosities. (All patients had pre-operative and 2 doses post-operative antibiotics.) Post surgery, the arm was placed in a sling and patients were referred to a physiotherapist. The sling was used for 6 weeks; afterwards, the patients were allowed to use it at their own convenience. Pendulum exercises and passive elevation/abduction up to 90 degrees were started from the first postoperative day. After 6 weeks, the patients were allowed a free active range of movement. Strengthening exercises were begun after 3 months.
2. Conservative treatment: arm immobilisation in a sling for 2 weeks, after which they were allowed to use it at their own convenience. After 2 weeks, the patients were referred to a physiotherapist, and pendulum exercises and passive elevation/abduction up to 90 degrees were started. After 4 weeks, they were allowed a free active range of movement.
Assigned: 27/28
Completed (at 2 years): 24/25

Outcomes

Length of follow up: 2 years
Constant shoulder score (both shoulders)
DASH (Diasabilities of the Arm, Shoulder and Hand) questionnaire
Quality of life score: EQ-5D
Mortality
Pain
Range of motion: abduction, flexion
Fixation failure, redisplacement, non-union, malunion

Subsequent surgery (reasons including non-union, etc)
Radiographic outcomes including avascular necrosis and osteoarthritis

NotesTrial run concurrently with Olerud 2011a.
Additional information on randomisation and trial location obtained from Dr Olerud (April 2012). Pain data received May 2012.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

"After clearance by an anesthesiologist, the patients were randomized (opaque sealed envelopes prepared independently) to a primary HA or nonoperative treatment." trial report

"the patients were randomised by numbered sealed opaque envelopes drawn consecutively.

The envelopes were independently prepared and thoroughly mixed. After that the envelopes were numbered by another person. At the time of randomisation the envelopes were drawn in numerical order." personal communication

Allocation concealment (selection bias)Low riskSee above
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High risk

No assessor blinding, although "The final 24-month follow-up was performed by an independent orthopaedic surgeon not previously involved in the treatment."

No provider or participant blinding.

Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskLack of blinding unlikely to affect assessment of these outcomes, but may affect decisions for subsequent surgery.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Low risk"In the outcome analyses, all patients remained in their
randomization group regardless of secondary procedures according to the intention-to-treat principle." Participant flow provided; no cause for concern.
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskAs above.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this. No protocol found.
Balance in baseline characteristics?Low riskNo imbalances: baseline comparability.
Free from performance bias?Unclear risk

"In patients randomized to surgery, all operations were performed by 1 of 2 orthopedic surgeons, both well experienced in shoulder surgery ..."

While all surgical patients were referred to a physiotherapist after their surgery and conservatively treated patients were referred after 2 weeks, this was unlikely to influence results. As was the differences in timing for free ROM (6 versus 4 weeks). However, it was only reported for the surgical group that strengthening exercises were begun after 3 months.

Revay 1992

MethodsRandomisation from closed envelopes
Assessor blinded
Loss to follow-up at 1 year: 1/48
ParticipantsDanderyd Hospital, Danderyd, Sweden
Period of study recruitment: not stated
48 patients with 2-, 3- or 4-part minimally displaced proximal humeral fractures (< 1 cm or < 45 degrees; Neer Group 1) treated conservatively with sling immobilisation for 1 week.
Exclusion criteria: patients with skin diseases and/or chlorine allergy, non-ambulatory
39 female, 9 male; mean age 66 years
InterventionsInterventions started 5 to 10 days post-injury after removal of sling.
1. Swimming pool training (30 minutes each session, up to 20 sessions maximum) in groups (6 to 8 patients) plus instructions for self training (see below).
2. Instructions for self-training: exercises to be performed at least 4 times a day for 10 to 15 minutes each time, use of hand on injured side for activities of daily living, advice on relaxation and resting positions.
Assigned: 25/23
Completed: ?/?
OutcomesLength of follow-up: 1 year; also assessed at 3 weeks, 2 & 3 months
Pain (analogue scale)
Activities of daily living: subjective assessment of 9 activities each rated on a 5 point scale
Functional scale: 6 point scale
Joint movement (abduction, flexion, internal rotation)
NotesMeans (probably) presented without standard deviations.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: "patients were randomized into two groups"
Allocation concealment (selection bias)Unclear riskInsufficient details of safeguards: "randomized and given instructions in a sealed envelope"
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear risk"All patients were examined by a physiotherapist who did not know which group each patient belonged to". However, no participant or care provider blinding nor mention of ways to prevent disclosure to assessor.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNot reported.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskThe treatment group of the participant lost to follow-up was not stated. Standard deviations not provided. Graphs only provided for female participants - denominators not provided for these.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskNot reported. The treatment group of the participant lost to follow-up was not stated.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Unclear riskBaseline data not provided for gender.
Free from performance bias?Unclear riskUncertainty if any compensatory advice given for the control group.

Rommens 1993

MethodsMethod of randomisation: alternation
Assessor blinding: unlikely
Loss to follow up at 3 weeks: 0/28
ParticipantsLeuven University Hospital, Belgium
Period of study recruitment: 1991
28 patients with acute 2- and 3-part proximal humeral fractures (but most were non or minimally displaced).
Exclusion criteria: those indicated for surgical intervention, age < 15 years, with multiple injuries or other fractures at same site
22 female, 6 male; mean age 69 years, range 25 to 100 years
InterventionsInterventions started immediately.
1. Gilchrist bandage, 2 to 3 weeks. The arm was bandaged with mesh type tubing and held by two slings: one round the shoulder and neck and the other which immobilised the distal part of the upper arm. (Bandage allowed wrist and hand exercises.)
2. Desault bandage, 2 to 3 weeks. Arm was immobilised to the chest using a circular elastic body bandage. (Some had one or more strips of plaster to stop the bandage slipping.)
Assigned: 14/14
Completed (at fracture consolidation): 14/14
OutcomesLength of follow-up: until fracture consolidation; also assessed at 1 & 3 weeks
Functional results: overall result, no data
Pain: patient questionnaire, 0 (none) to 100 (significant) scale
Displacement of fracture
Complication: skin irritation
Removal of bandage
Surgeon assessment of ease of application of bandage
Patient assessment of bandage
Notes

Two fractures in the Gilchrist group required reduction. Seven participants had other fractures: 3 in group 1 (2 rib, 1 vertebra); 4 in group 2 (1 ankle, 1 hip, 1 rib, 1 vertebra).

Trial reports in German; translation obtained.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High riskQuasi-randomised: alternation
Allocation concealment (selection bias)High riskAlternation
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo mention of blinding
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNot reported
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskWhile all participants were followed up and intention-to-treat analyses seemed to have been done, no data on function were presented nor were the criteria for judging fracture consolidation.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskNot reported
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Unclear riskSmall discrepancies (e.g. in other injuries or having fracture reduction) can have bigger consequences for small group sizes.
Free from performance bias?Unclear riskDifferences in care programmes cannot be ruled out.

Smejkal 2011

Methods

Method of randomisation: computer generated block randomisation with sealed envelopes

Assessor blinding: no mention in the paper

Loss to follow-up: 4 lost to follow-up and 2 died of breast cancer during the study period

Participants

University Hospital in Hradec Králové, Czech Republic

Period of study recruitment: January 2006 to January 2010

61 patients with AO type A2, A3, B1 and C1 (2 part and 3 part) proximal humerus fractures aged between 18 and 80 years able to give informed consent.

Exclusion criteria: open fracture, associated injury (AIS > 2), open growth plates, or patient's health would limit the extent of surgery.

45 females, 10 males; mean age 61 years, range 21 to 81 years

Interventions

Interventions started 0 to 24 days after injury.
1. Open reduction and internal fixation group: consisted of patients undergoing open reduction with angle-stable osteosynthesis using a PHILOS plate (Synthes, Switzerland).

2. Minimally invasive group: Zifko method of minimally invasive osteosynthesis with intramedullary K-wire insertion (distally inserted) – figure in article shows 8 wires inserted into humeral head along medullary canal.

Assigned: ?/? (total 61)

Completed: 28/27

Outcomes

Length of follow-up: mean 2 years

Days to operation

Constant-Murley score (relative to healthy limb)

Time to recover normal upper limb function

Complications
Time to radiographically assessed recovery
Anatomical position

X-ray exposure

Length of operation

Length of hospital stay

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

“The patients were randomised to the groups by a computer programme which facilitates the maintenance of homogeneity of the groups compared.”

Web-based translation implied use of random numbers and permuted blocks so as to get similar numbers on each group. Produced independently by a statistical company.

Allocation concealment (selection bias)Low riskThe sealed envelopes were created by a professional statistical company (Pharm test s. r. o., Hradec Králové): in accordance with randomization sheet each envelope their number and sealed inside information," zifko" or "LCP."  The sealed envelopes were opened sequentially.
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNot possible to blind patient / providers. No mention of outcome assessment.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskMay not affect assessment
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskIncomplete data (and group of 6 excluded participants not noted)
Incomplete outcome data (attrition bias)
Death, reoperation
High riskIncomplete data (and group of 2 deaths not stated)
Selective reporting (reporting bias)Unclear riskNo protocol
Balance in baseline characteristics?Unclear riskAside from age – no details or confirmation of this
Free from performance bias?Unclear riskNo details – including of surgeon’s experience

Stableforth 1984

MethodsMethod of randomisation: unknown, "randomly selected"
Assessor blinding: unlikely
Loss to follow-up at 18 months to 12 years: 2/32 (2 deaths)
ParticipantsBristol Royal Infirmary, Bristol, UK
Period of study recruitment: 1970 to 1981
32 patients with displaced 4-part proximal humeral fractures (Neer).
Exclusion criteria: impacted or minimally displaced fractures
25 female, 7 male; mean age 68 years, range 52 to 88 years
InterventionsInterventions started: within 5 days for surgery.
1. Neer prosthesis, uncemented
2. Closed manipulation
All placed in sling, mobilisation of hand encouraged, shoulder flexion rotation exercises after 2 to 3 days. Supervised physiotherapy for 3 to 6 months.
Assigned: 16/16
Completed (at 1 year): 15/15 (but totals given as 16/16 in tables in the trial report)
OutcomesLength of follow-up: stated as 18 months to 12 years; but also assessed regularly up to 6 months
Dependent in activities of daily living
Range of motion (flexion, medial rotation, lateral rotation)
Pain
Muscle strength (flexion, abduction, lateral rotation)
Complications: haematoma, cellulitis, deep sepsis, early shoulder stiffness
Mortality
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: "assigned by pre-arranged random selection"
Allocation concealment (selection bias)Unclear riskNo details: "assigned by pre-arranged random selection"
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNot blinded
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo blinding but may not have affected appraisal of mortality
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskLarge loss to follow-up (46/85 = 54%). Numbers given for those available at follow-up but incompletely reported data: only medians.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskSlight discrepancy in trial report that 2 deaths are reported, one in each group, but long term denominators are as at baseline.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this, but the protracted nature of this trial makes selective reporting more likely.
Balance in baseline characteristics?Unclear riskSurgical group on average 4.5 years younger, but uncertainties mainly reflect Inadequate information in terms of other co-morbidities and injuries for this broad category of patients.
Free from performance bias?Unclear riskInadequate information on care programme comparability especially given the protracted nature of the trial recruitment. However, one surgeon operated throughout.

Voigt 2011

Methods

Method of randomisation: drawing balls from a bag by an independent person

Assessor blinding: assessor blinding

Loss to follow-up at 12 months: 8/56 (did not complete follow-up: 2 deaths, 4 drop-outs, 2 excluded because of early secondary arthroplasty)

Participants

Friederikenstift Hospital Hannover, Hannover, Germany

Period of study recruitment: conducted over 18 month period (no dates)

56 patients with isolated Neer type 3- and 4-part proximal humeral fractures, aged > 60 years.

Exclusion criteria: fractures older than 2 weeks, open fractures, pathological fractures, refractures, neurologic disease and patients who would be clearly non-compliant (e.g. alcoholics, patients of no fixed address)

Of 48: 38 female, 10 male; Of 56: mean age 74 years, range 60 to 87 years

Interventions

Interventions started: at surgical fixation (time to surgery from injury not given)

1) Polyaxial locked screws: Humeral Suture Plate (HSP) (Arthrex, Naples, FL) with polyaxially locked screws. Screws were blunt-ended (considered better in the prevention of glenoid erosions in case of screw perforations).

2) Non-polyaxial (monoaxial) implant: Proximal Humerus Internal Locking System (PHILOS) (Synthes, Bettlach, Switzerland) with nonpolyaxially locked screws. (Screws were pointed in the PHILOS plate.)

All surgery performed under general anaesthesia using deltopectoral approach. Tuberosity fragments reduced with fibre wire, different approaches for head fragment depending on whether valgus or varus. Allocated plate positioned anatomically and fixed with a shaft screw.

Patients' shoulders were immobilised in a sling for 2 days. Then, active-assisted motion beyond 90 degrees flexion and abduction were initiated avoiding the provocation of pain. At 7 weeks, free range of motion was allowed.

Assigned: 25/31

Completed: 20/28 (at 12 months)

Outcomes

Length of follow-up: 12 months

Simple shoulder test
Disability of the Arm, Shoulder and Hand (DASH) score

Constant score (relative to contralateral limb)
Death
Complications
Re-operation

Range of active shoulder motion (flexion, abduction, internal rotation, external rotation)

Fracture healing - AP and axillary radiographs

Duration of operation

Fluoroscopy time

NotesAdditional information and clarification of 8 participants who did not complete follow-up and gender data for those who completed follow-up obtained from Dr Voigt (May 2012).
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk“The randomization technique was blinded by drawing balls from a bag: one ball for HSP and the other ball for PHILOS by an independent person.”
Allocation concealment (selection bias)Low risk“The randomization technique was blinded by drawing balls from a bag: one ball for HSP and the other ball for PHILOS by an independent person.”
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo mention of blinding of patients or personnel other than assessor blinding: "Follow-up evaluations postoperatively were performed in a standardized fashion by an independent trauma surgeon"
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskUnlikely to influence this.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskAlthough clarification on loss to follow-up (8 patients: 5 versus 3) received from author, the impact on the results for functional outcomes is unclear.
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskClarification received from author on the loss to follow-up: 2 were deaths and 2 were replacement arthroplasty.
Selective reporting (reporting bias)Unclear riskNo protocol provided
Balance in baseline characteristics?Unclear riskBalance in 3 versus 4 part fractures, probably age and pre-operative DASH. Incomplete data on gender, 2 versus 6 with diabetes (but no frozen shoulder).
Free from performance bias?Unclear riskNo details of surgeon experience.

Wirbel 1999

MethodsMethod of randomisation: unknown, "random allocation"
Assessor blinding: unlikely
Loss to follow up at 6 months: 13/77; also 14 months (9 to 36 months): 18/77
ParticipantsUniversity Hospital, Homburg/Saar, Germany
Period of study recruitment: January 1995 to March 1998
77 patients with displaced (separation exceeds 1 cm; fragment angulation > 30 degrees, or when tuberosity fragment is separated by > 3 mm) subcapital humeral fractures of type A1, A3, B and C1 (modified AO classification) treated by closed reduction and percutaneous fixation.
Exclusion criteria: Extensive local skin infection. Impacted fractures of type A2 (treated conservatively). Not fit enough to undergo anaesthesia and X-ray of affected shoulder in anterior-posterior plane. Closed reduction not feasible.
54 female, 23 male; mean age 63 years, range 6 to 89 years
InterventionsInterventions started post-operatively after percutaneous fixation (Kirschner wires plus in 38 cases, cannulated screws).
1. 1 week immobilisation in Gilchrist sling
2. 3 weeks immobilisation in Gilchrist sling
Active mobilisation of elbow from first post-operative day. Active and passive physiotherapy of the shoulder (optional continuous passive motion) after removal of sling. Removal of Kirschner wires after 4 to 6 weeks, with post-procedure continuation of active exercises.
Assigned: 38/39
Completed (at 6 months): 32/32
OutcomesLength of follow-up: 9 to 36 (mean 14 months) months (in 59 participants), but also assessed at 1, 3 and 6 months
Neer score
Complications: avascular necrosis, local infection/haematoma, premature removal of Kirschner wires, screw removal due to subacromial impingement
NotesShort report (1997) from conference proceedings gave interim results for 51 patients. Full report and some results provided by Dr Wirbel (February 2003).
Most of the results given in the trial report were either for the whole study population or split by basic AO fracture type.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo details: "a random allocation of patients in 2 groups was done"
Allocation concealment (selection bias)Unclear riskNo details: "a random allocation of patients in 2 groups was done"
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo mention of blinding.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNot blinded but less likely that these outcomes would be affected.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
High riskLimited data on function using a non-validated assessment instrument with a moderate loss to follow-up at 6 months (17/77 = 17%).
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskIncomplete data. Although loss to follow-up reported, reoperations were not sufficiently reported by treatment group.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Low riskNo indication of any major baseline imbalance.
Free from performance bias?Low riskNo indication of performance bias from differences in care programmes.

Zhang 2011

Methods

Method of randomisation: computer generated random numbers

Assessor blinding: likely, "independent" assessor at follow-up

Loss to follow-up: 4 patients within the first year after surgery due to moving out of the area and change of telephone number

Participants

The Third Affiliated Hospital of Whenzou, Whenzou, China

Period of study recruitment: October 2007 to September 2008

72 patients aged over 18 years with an acute closed 2, 3 or 4 part fracture (Neer classification) of the proximal humerus treated with open reduction and internal fixation using a locking plate.

Exclusion criteria: pathological fractures, primary or metastatic tumour and fracture with non-union.

Of 68 followed-up: 46 female, 22 male; mean age 63 years, range 32 to 78 years

Interventions

Interventions started: both at surgery, time from injury not stated

1. ORIF with PHILOS locking plates (Synthes, Switzerland). Standard deltopectoral approach; reduction enabled with a K-wire under fluoroscopy. Locking plate was placed 10 mm posterior to the intertubercular groove and 10 mm distal to the tip of greater tubercle. A cortical screw was inserted initially to fix the distal fragment. Four or five locking screws were used for the fixation of the proximal fragment. All proximal screws were inserted 5 mm below subchondral bone. One or two additional locking screws were inserted obliquely into the medio-inferior region of the humeral head in this group.

The tubercular fragments and rotator cuff tendon were fixed using Ethibond sutures. Autograft bone was used in comminuted fractures where there was a mass defect and for reconstruction of the medial support structures. Fracture reduction and screw length were finally assessed with fluoroscopy.

2. As above without medial support locking screws.

All patients received prophylactic intravenous antibiotics before the procedure. Passive abduction and clock-wise rotation exercises were allowed on the day after surgery. Active rehabilitation was started six weeks postoperatively.

Assigned: 32/40 (total: 72)
Completed (2+ years): 29/39

Outcomes

Length of follow-up: average 30.8 months (also 4, 8, 12 weeks, 6, 9 and 12 months and yearly)

Shoulder function (Constant shoulder score)

Union

Complications: osteonecrosis of the humeral head, early failure and loss of fixation
Re-operation

NotesPersonal contact (email 14/05/2012) clarified method of randomisation, group of patients who were lost to follow-up; and complications.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The patients were randomized into two groups for study according to computer-generated random numbers" (Group sizes, however, were unequal.)
Allocation concealment (selection bias)Unclear riskThere were insufficient safeguards (selection according to odd and even random numbers) to confirm allocation concealment.
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskIt is possible that the participants did not know which group they were in. The difference between the two interventions was not large.
“Complications, shoulder function and radiological measurement were recorded by an independent junior doctor (YJH) who did not participate in the surgery.”
Blinding (performance bias and detection bias)
Death, reoperation
Low riskThese outcomes are fairly robust regarding blinding.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskActive surveillance but missing data for 4 participants lost to follow-up. Personal correspondence gave details on complications.
Incomplete outcome data (attrition bias)
Death, reoperation
Unclear riskMissing data for 4 participants. Personal correspondence provided information on re-operations.
Selective reporting (reporting bias)Unclear riskInsufficient data to judge this. No protocol available.
Balance in baseline characteristics?Unclear riskAlthough the baseline characteristics of 68 participants were comparable, data were missing for 4 participants loss to follow-up.
Free from performance bias?Low risk“Operations were performed by two senior surgeons.” All participants received same rehabilitation.

Zhu 2011

MethodsMethod of randomisation: computer generated random numbers list reviewed by nurse before surgery
Assessor blinding: no, but mention of independent observer
Loss to follow-up at 3 years: 6/57 (5 lost; 1 died)
Participants

Beijing Ji Shui Tan Hospital, Beijing, China

Period of study recruitment: November 2004 to December 2006
57 skeletally mature patients with an acute 2-part surgical neck fracture of the proximal humerus (Neer's classification) treated surgically within 21 days of the injury. Patient consent.
Exclusion criteria: open physes, fracture and displacement involving the greater or lesser tuberosity or extension of the fracture line distally beyond the deltoid tubercle, associated musculoskeletal injuries to the same upper extremity, open fracture, and prior surgery on the affected shoulder.
Of 51 followed up: 34 female, 17 male; mean age 53 years

Interventions

Interventions started: surgery on average 9 days after injury (randomisation before surgery).

1. Open reduction with internal fixation using a locking plate: Locking Proximal Humeral Plate (LPHP; Synthes) or the Proximal Humeral Internal Locking System (PHILOS; Synthes). General anaesthesia combined with an interscalene block. Indirect reduction under image intensifier, with reduced fracture temporarily fixed by a Kirschner wire. After placement, position of the locking plate checked with the image intensifier intraoperatively, and the plate was fixed with locking screws. Finally, a thorough fluoroscopic screening was done to ensure that no screw was penetrating the articular surface of the humeral head.

2. Open reduction with internal fixation using a locking nail: the Proximal Humeral Nail (PHN; Synthes). An interscalene brachial plexus block was used. Nail was inserted under image control without reaming after the fracture was fully reduced. After insertion of the spiral blade and the distal locking screws, an end cap was screwed in to lock the spiral blade. The rotator cuff tendon and the deltoid were carefully repaired during wound closure.

The affected extremity was protected by a sling for six weeks postoperatively. Passive range-of-motion exercises, supervised by a physical therapist, were initiated on the first postoperative day. Active and active-assisted exercises began after six weeks, when early callus formation could be seen on radiographs. Strengthening exercises were started three months after the surgery.
Assigned: 29/28
Completed (at 3 years): 26/25

OutcomesLength of follow up: 3 years (also 1 year)
ASES (American Shoulder and Elbow Surgeons) score
Constant shoulder score (both shoulders)
Pain (VAS)
Mortality
Complications (overall, infection (none), heterotopic ossification, screw penetration, pseudothorax)
Re-operation
Range of motion (active flexion, external rotation, internal rotation)
Strength
Duration of surgery
Blood loss and transfusion
Radiographic outcomes including avascular necrosis, union, and degenerative change (osteoarthritis)
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Randomization was accomplished with use of a random numbers list generated by software and kept by the operating room nurse."
Allocation concealment (selection bias)Unclear risk"Before the surgery, the circulating nurse reviewed the random-numbers list. Patients who had been assigned an odd number were subsequently treated with a locking nail, and those who had been assigned an even number were managed with a locking plate."
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskNo mention of blinding. However:
"All of the follow-up physical examinations and radiographic evaluations were done by the same independent observer."
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo mention of blinding. Lack of blinding less likely to affect these outcomes.
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Low riskParticipant flow diagram provided; similar and modest losses in each group.
Incomplete outcome data (attrition bias)
Death, reoperation
Low riskDitto - data reported.
Selective reporting (reporting bias)Unclear riskInsufficient information to judge this.
Balance in baseline characteristics?Unclear riskNo indication of any major baseline imbalance in 51 participants followed up at 3 years but no data for 6 participants lost to follow-up.
Free from performance bias?Low riskAll surgical procedures performed by senior surgeon and comparable rehabilitation. Although general anaesthesia used only for the plate group, this was considered unlikely to affect the findings..

Zyto 1997

  1. a

    AO = Arbeitsgemeinschaft fur Osteosynthesefragen / Association for the Study of Internal Fixation (or ASIF)
    AVN = avascular necrosis
    A&E = accident and emergency
    MI = myocardial infarction
    PE = pulmonary embolism

MethodsMethod of randomisation: sealed envelopes
Independent assessor at final follow-up
Loss to follow-up at 3 years: 14/43 (8 deaths, 2 could not be traced, 1 hemi-prosthesis, 3 exclusions)
ParticipantsHuddinge University Hospital, Stockholm, Sweden
Period of study recruitment: April 1990 to February 1993
43 "elderly" patients with proximal humeral fractures (AO classification system: A 8; B 27; C 8) - see notes.
In trial report:
40 patients with displaced 3- or 4-part fractures (Neer).
Exclusion criteria: pathological fracture, high energy trauma, < 30% contact between humeral head and shaft, other fractures, impaired ability of patient to co-operate, relevant concomitant disease
35 female, 5 male; mean age 74 years
InterventionsInterventions started: surgery within 48 hours.
1. Internal fixation (cerclage wiring (8): or surgical tension band (14)) under general anaesthesia. Antibiotic therapy. Physiotherapy.
2. Conservative treatment: sling for 7 to 10 days. Then physiotherapy.
Assigned: 22/21; (20/20)
Completed (50 months): 15/14
OutcomesLength of follow-up: 3 to 5 years (listed as 50 months in trial report; patient questionnaire, clinical and radiological assessment); also after treatment and at 1 year:
Subjective assessment of function including ability to carry 5 kg, sleep on injured side, comb hair, perform personal hygiene
Constant score: overall shoulder function and components (pain, power, range of motion, activities of daily living)
Complications: deep infection, non-union, pulmonary embolism, avascular necrosis of humeral head
Mortality
Notes

Both groups had the same physiotherapy regimen.

Three patients excluded from 1995 data set (Tornkvist 1995) as, on review by Zyto and a radiologist, the patients did not have 3- or 4-part fractures (personal communication).

Zyto's response to a letter from H. A. Karladani admits that there may have been some inaccuracy in their classification of the fracture patterns but stressed that the Neer classification system was flawed and that other factors such as osteoporotic bone need to be considered too.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"randomised by sealed envelopes"
Allocation concealment (selection bias)Unclear risk"randomised by sealed envelopes" (at time of admission) No indication of safeguards.
Blinding (performance bias and detection bias)
Functional outcomes, pain, clinical outcomes, complications
High riskSome independent assessment by radiographer and potentially by main author but no blinding.
Blinding (performance bias and detection bias)
Death, reoperation
Unclear riskNo blinding but may not have affected appraisal of mortality (which was not split by treatment group).
Incomplete outcome data (attrition bias)
Functional outcomes, pain, clinical outcomes, complications
Unclear riskPost-randomisation exclusions and moderately large loss to follow-up (14/43 = 32%; (11/40 = 28%)).
Incomplete outcome data (attrition bias)
Death, reoperation
High riskOnly whole group data presented for deaths out of 40 participants.
Selective reporting (reporting bias)High riskInsufficient information to judge this but some post-randomisation exclusions and final follow-up performed by first author who does not appear in the earlier reports of the trial.
Balance in baseline characteristics?Low riskNo important imbalances in baseline characteristics.
Free from performance bias?Low riskNo indications of serious performance bias: surgery performed by orthopaedic specialists who were experienced in the surgical technique.

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Bing 2002

This was a randomised clinical trial (sealed envelopes - computer generated sequence - held in a box), recruitment 03/11/1997 to 14/01/1999, that compared Rush pins fixation with Polaris nail fixation of displaced two part fractures of the proximal humerus. Contact with a Dr Sharma in July 2000 revealed 65 of the 80 patients in the trial had reached 2 year follow-up. Abstract by Bing et al published in 2002 indicated 40 patients of whom 30 had been followed-up for one year. Information gained via Alison Armstrong from Grahame Taylor (one of the authors of the Bing abstract) indicated that there were some concerns about the extent of missing data. Both groups had a high reoperation rate to remove metal ware causing impingement. This trial has been excluded because of insufficient data.

It seems very likely, based on location and study dates, that the trial registration (Der Tavitian 2006) formerly awaiting classification is for this trial.

Bolano 1995No proximal humeral fractures in a randomised trial of humeral shaft fracture treatment.
Brownson 2001This is listed in the National Research Register as a multi-centre randomised trial of the management of displaced surgical neck and displaced shaft fractures of the humerus with the Halder humeral nail. Contact with Mr Brownson revealed this to be part of the trial run from Nottingham (see Wallace 2000) which had been abandoned. Mr Brownson indicated that the very specific inclusion criteria (2-part fractures with over 50% displacement) had reduced the potential sample size; patient consent had also been a problem.
Carbone 2012This is a prospective comparison of MIROS (Minimally Invasive Reduction and Osteosynthesis System®) versus traditional percutaneous pinning, each intervention being carried out at one of two hospitals in the same town in Italy. Not randomised.
Chapman 1997No proximal humeral fractures in a randomised trial of humeral shaft fracture treatment.
Chiu 1997No proximal humeral fractures in a quasi-randomised trial of humeral shaft fracture treatment.
de Boer 2003This is a multi-centre comparative study of locked internal fixators and non-operative treatment. Not randomised.
Dias 2001Trial abandoned. This randomised trial (random number sheets that are remotely administered) compared hemiarthroplasty versus fixation (generally suture reinforced with wires) versus conservative treatment (manipulation, sling for 2 weeks, then mobilisation) for 3- and 4-part fractures of the proximal humerus. Trial started in 2001, with one year follow-up (outcome was assessed by independent physiotherapists). Aimed for 90 to 100 participants, aged > 45 years. Contact with Alison Armstrong revealed that recruitment stalled at 11 patients (16 refusals) in 2008; centre stopped trial when it became a trial site for the ProFHER trial.
Edelson 2008Article mentions an abandoned randomised trial comparing "operative versus conservative care" which was unsuccessful "because patients insisted on proactively choosing rather than being assigned to a treatment group by lot"." No other details given.
Flannery 2006This is listed in the National Research Register as a randomised trial comparing conservative treatment and hemiarthroplasty for four-part fractures of the proximal humerus. Contact with Mr Flannery revealed his centre failed to recruit anyone into the trial. Mr Turner, the lead investigator of the multi-centre trial, involving the South Thames Shoulder and Elbow Group, confirmed that the trial was abandoned due to the inability to recruit patients.
Gradl 2009Prospective study involving 152 patients with unilateral displaced and unstable proximal humeral fractures treated either with an antegrade angular and sliding stable proximal interlocking nail or an angular stable plate. Not a randomised or quasi-randomised trial.
Hems 2000This is listed in the National Research Register as a randomised trial comparing conservative treatment and the Halder humeral nail for displaced fractures of the surgical neck and shaft of the humerus. Contact with Mr Hems revealed this to be part of the trial run from Nottingham (see Wallace 2000). Mr Hems indicated that they had had considerable difficulty in recruiting patients (only those with proximal humeral fractures were eligible in his centre) and had no results.
Liao 2009While the English abstract claims that "70 senile patients" were "randomly divided into three groups to receive different surgical methods" the distribution and characteristics (age and fracture type) of the patients in the three groups indicated serious selection bias and implied this was not a randomised trial. For example: "21 patients in the group A receiving Kirschner tension band or screw internal fixation, 37 patients in group B receiving internal fixation of locking proximal humeral plate, and 12 patients in group C receiving humeral head replacement." There was no reply to request for clarification from the lead author.
Martin 2000Contact with a trialist revealed that due to the discovery of problems with randomisation it was decided not to proceed with publication as the trial results could be compromised.
Mechlenburg 2009This was originally registered as a randomised controlled trial comparing a plate with a hemiarthroplasty. However, it is now registered as a prospective study of fixation with a PHILOS plate. Inger Mechlenburg confirmed that no patients had been included in the trial - the trial was abandoned because no funding was obtained.
Parnes 2005There has been no response from the lead author of this ‘trial’ (last contact attempted 8 June 2012), which appears to have been reported in a conference abstract only. In 2003, 50 patients with 3- and 4-part fractures and fracture dislocations of the proximal humerus were "random selected" for surgery (closed or open reduction and external fixation or hemiarthroplasty) or conservative treatment.  The very limited results are split descriptively according to three groups (2 reflecting the 2 different surgical methods). There is currently insufficient evidence to support this being a randomised trial.
Pullen 2007This is listed in the National Research Register Archive as a randomised trial comparing the T2 proximal humeral nail with the PHILOS plate system in patients with 2- or 3-part proximal humeral fractures. the recruitment target was 100 patients (between 01/09/2005 to 01/09/2007), and follow-up was 16 weeks. We have not located any other report of this study than the details provided in the National Research Register (UK) by, at that time, a Trauma and Orthopaedic Registrar who has now moved to another hospital. There was no response to a request for further information sent 8/6/2012. There is no indication that this study, which may not have started, will ever be reported.
Rodriguez-Merchan 95No proximal humeral fractures in a quasi-randomised trial of humeral shaft fracture treatment.
ShahThis is listed in the National Research Register Archive as a multi-centre randomised trial of the management of four part fractures of proximal humerus that compared hemiarthroplasty versus conservative treatment. The recruitment target was 200 patients, with a one year follow-up using the Constant-Murley shoulder score and Oxford Shoulder score. The listed start and end dates were 01/01/2003 and 01/02/2005. No details were received of the other centres in the very limited further information received from Mr Shah in April 2003. There was no response to a request for further information sent 13/11/06. There is no indication that this study, which may not have started, will ever be reported.
Sinopidis 2010This was registered as a randomised study of reverse shoulder prosthesis and hemiarthroplasty for elderly patients with proximal humeral fractures. However, the principal investigator left the hospital (and country) before it started and a contact at Liverpool (Matthew Smith) confirmed that the study was closed after this.
Wallace 2000This is listed in the National Research Register as a multi-centre randomised trial of the management of displaced surgical neck and displaced shaft fractures of the humerus with the Halder humeral nail. Contact with Prof Wallace's secretary revealed that the study had not gone ahead. The secretary mentioned three other sites (Halifax; Liverpool; and one in Scotland). No reason given. See Brownson 2001.
Wan 2005This is a mixed population trial evaluating additional mobilisation therapy that included other fractures (e.g. clavicular and scapular fractures) as well as proximal humeral fractures. This trial was excluded because separate proximal humeral fracture data were not reported and the contact author is unavailable.
Warnecke 1999A multicentre prospective study but not a randomised trial.
Welsh 2000This is listed in the National Research Register as a randomised comparison of operative and non-operative management of proximal humeral fractures. This trial was abandoned due to poor recruitment, mainly due to lack of patient consent.
Yang 2006Correspondence with the author revealed that this was not a randomised trial. The choice of surgery was dependent on the success of closed reduction.
Zhang 2010While the English abstract claims that "58 patients with 3 parts and 4 parts fractures of proximal humerus were randomly treated with AO locked compressive plates (LCP) or humeral head replacement." the characteristics ( fracture type) of the patients in the two groups indicated serious selection bias and implied this was not a randomised trial. Thus, 25 of 28 patients in the plate group had 3 part fractures (1 with a dislocation) and 3 had 4 part fractures (1 with dislocation) whereas 11 of 30 in the replacement group had 3 part fractures (2 with dislocation), 16 with 4 part fractures (4 with dislocation) and the other three had humeral head split fractures. There was no reply to request for clarification from the lead author.

Characteristics of studies awaiting assessment [ordered by study ID]

Battistella 2011

Methods"Randomized clinical study"
Participants54 patients (38 female, 26 male, mean age 61 years) with 2-part surgical neck fractures or 3-part valgus impacted fractures
Interventions

Surgery involving a titanium plate:

1. Minimally invasive fixation based on anterolateral deltoid split approach and percutaneous reduction
2. Open reduction and internal fixation by standard deltopectoral approach

OutcomesConstant score, instrumental activities of daily living, pain (VAS), range of motion, union, complications
NotesRequests for further information sent to Dr Battistella (8 and 14 May 2012) were unsuccessful.

Fjalestad (RCT proposal)

MethodsRandomised trial; 4 centres
"semi-blinded" for the physiotherapist who will examine the patients
Participants110 patients
Interventions

1. Reversed shoulder prosthesis

2. ORIF with PHILOS plate

OutcomesPrimary / secondary outcomes will be: Constant score, possibly the Oxford Shoulder score, a QoL score, the “15D”, and a short patient self-evaluation score in addition to the X-ray and CT 3D scans
NotesProposal for this new trial was being finalised in May 2012 (personal communication from Tore Fjalestad)

Luo 2008

MethodsPatients were randomly allocated via a random numbers table.
Participants60 patients (32 females, 28 males; age range: 39 to 62 years) treated operatively for fracture of the surgical neck of the humerus.
Interventions

1. Acupuncture (electroacupuncture and infrared radiation) plus passive exercise of the shoulder joint
2. Exercises only: passive exercise of the shoulder joint followed by active exercises

Treatment lasted 1 month.

OutcomesFollow-up: 1 month
shoulder pain score (VAS)
shoulder joint activity
NotesTrial in Chinese with English abstract. Translation of methods section (1.1) confirmed that this was a randomised trial.

Characteristics of ongoing studies [ordered by study ID]

Boyer

Trial name or titleRandomised study between intramedullary locking nails and locking plates for treatment of proximal humerus fractures (HUMERUS)
Official title: Randomised study between intramedullary locking nails and locking plates for treatment of proximal humerus fractures in patients after 40-year-old
MethodsSingle centre, randomised controlled trial. Unblinded.
Participants144 patients, aged between 40 and 85 years, with a type III or IV "cephalotuberosity" proximal humeral fracture (classification of Neer an DUPARC)
Interventions

1. Intramedullary nail (Multilock, Synthes)

2. Locking plate (SURFIX, Integra)

OutcomesFollow-up: 1 year
Primary outcome: Constant Score
Secondary outcomes: Quick DASH, complication (malunion, necrosis, infection)
Starting dateStart date: February 2012
Estimated completion date: November 2015
Contact informationDr Patrick Boyer
Group Hospitalier Bichat - Claude Bernard 46, rue Henri-Huchard
Paris
Ile de France
France 75018
patrick.boyer@bch.aphp.fr
Notes 

Brorson

Trial name or titleEffect of osteosynthesis, primary hemiarthroplasty, and non-surgical management for displaced four-part fractures of the proximal humerus in elderly: a multi-centre, randomised clinical trial
MethodsMulti-centre, randomised clinical trial (central randomisation unit)
Participants162 patients with displaced 4-part fractures of the proximal humerus
Interventions1. Hemiarthroplasty
2. Fixed-angle plate osteosynthesis
3. Non-surgical treatment
OutcomesFollow-up: 3 years (primary outcome: 1 year)
Primary outcome: Constant Disability Scale
Secondary outcomes: Oxford Shoulder Score, Short Form-36
Starting dateStart date: April 2009
End date: March 2012 (final date for primary outcome measure); * revised date not established (communication June 11 2012)
Contact informationDr Stig Brorson
Department of Orthopaedic Surgery
Herlev University Hospital
Herlev
Denmark
DK-2730
sbrorson@hotmail.com
Notes

Published protocol.

* Correspondence from Stig Brorson (June 11 2012) reveals a slower than anticipated recruitment,

Engelsma

Trial name or titleArthroplasty in three- and four-part proximal humerus fracture: hemi or reverse? Prospective multi-centre randomised clinical trial
MethodsMulti-centre, randomised controlled trial. Unblinded.
Participants52 patients aged over 65 years with displaced 3 or 4 part proximal humerus fractures who are candidates for primary shoulder arthroplasty.
Interventions

1. Aequalis fracture prosthesis

2. Aequalis reverse fracture prosthesis

Outcomes

Follow-up: 1 year

Primary outcome: Constant Shoulder Score

Secondary outcomes: DASH (Disabilities of the Arm, Shoulder, and Hand) Score, SF-12 questionnaire, Visual Analogue Score of pain

Starting date

Start date: July 2010

Estimated completion date: December 2014

Contact information

Yde Engelsma

Medisch Centrum Alkmaar

Wilhelminalaan 12

1815 JD Alkmaar

The Netherlands

y.engelsma@mca.nl

Notes 

Guy

Trial name or titleA multicentre prospective randomized control trial on the treatment of three and four part proximal humerus fractures in patients 70 years and older: comparing open reduction and internal fixation with non operative treatment
MethodsRandomised controlled trial: "randomly (like flipping a coin)"
Participants120 patients aged 70 years or over with a 3 or 4 part fracture.
Interventions1. Open reduction and internal fixation
2. Non-operative treatment (reduction and immobilisation)
OutcomesFollow-up: 1 year
Primary outcome: patients' functional shoulder scores as measured by the Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire
Secondary outcomes: functional and mental status instruments (i.e. SF-36/EQ-5D) used to assess the patient's health-related quality of life; re-operation rates; and the time required to return to pre-injury level of independence
Starting dateNovember 2010
Estimated completion date: October 2011
Contact informationContact: Raman Johal (raman.johal@vch.ca)
Principal investigator: Pierre Guy, University of British Columbia, Canada
NotesAs of registration update (17/02/2011) the study was recruiting. Start and end dates amended. 

Helsinki

Trial name or titleEffectiveness and cost-effectiveness of operative and conservative treatment of comminuted fractures of the proximal humerus. A randomised, controlled study.
MethodsRandomised single blind (outcomes assessor)
Participants

150 older patients with comminuted, displaced fractures of the proximal humerus.

Inclusion criteria: Age over 65 years; acute trauma with randomisation within 7 days of injury; 3- or 4-part fracture with > 5 mm dislocation of the anatomic neck (AO classification C1-2 for non-luxation fractures; C3 for luxation fractures)

Interventions1. PHILOS locking plate: open reduction of the fracture (and GH joint), internal fixation with the PHILOS locking plate. Tuberculum fragments are sutured to the plate with thick non-absorbable suture.
2. Global FX hemiarthroplasty: replacement of the humeral articular head with hemiprosthesis. Tubercles are sutured to the prosthesis with thick nonabsorbable sutures.
3. Conservative treatment: immobilisation in a supporting brace for 3 weeks, then increasingly active rehabilitation programme supported by a physiotherapist until 12 weeks of the injury.
OutcomesFollow-up: 24 months
Primary outcomes: Pain at rest and activity (Numeric Rating Scale), Constant score
Secondary outcomes: Simple Shoulder test (SST), Disabilities of the Arm, Shoulder and Hand (DASH), quality of life assessment (15D), subjective patient satisfaction, complications and costs.
Starting dateNovember 2010
Final follow-up date: November 2014
Contact information

Tuomas Lähdeoja, MD: tuomas.lahdeoja@hus.fi
Mika Paavola, MD: mika.paavola@hus.fi

Helsinki University, Helsinki, Finland

NotesAs of registration update (22/01/2012) the study was recruiting. Start and end dates amended. 

HOMERUS

Trial name or titleHemiarthroplasty versus osteosynthesis in humeral fractures (HOMERUS): A multicentre randomised trial
Official title: Three- and four-part fracture of the proximal humerus in the elderly. Angle stable locking compression plate osteosynthesis versus hemiarthroplasty
MethodsMulti-centre, randomised clinical trial. Single blinded (outcomes assessor).
Participants134 patients aged over 60 years with displaced 3 and 4 part proximal humeral fractures with more than 5 mm of dislocation in one of the fracture planes.
Interventions

1. Hemiarthroplasty

2. Angle stable locking compression plate osteosynthesis

Outcomes

Follow-up: 2 years

Primary outcome: DASH (Disabilities of the Arm, Shoulder, and Hand) Score

Secondary outcomes: VAS (Visual analogue score) for pain and patient satisfaction, Constant-Murley Score, SF-36, Radiographic evaluation, complications

Starting date

Start date: September 2010

Anticipated completion date: August 2012

Contact information

Dr PA Verbeek

Department of Orthopaedic Surgery

University Medical Centre Groningen (UMCG)

Groningen

The Netherlands

paulverbeek@gmail.com

Notes 

HURA

Trial name or titleA randomised clinical trial comparing a lateral minimally invasive approach and the traditional anterior approach for plating of proximal humerus fractures
MethodsRandomised, single blind (outcome assessors), clinical trial
Participants90 patients, with humeral surgical neck fracture, Neer II valgus-type, and Neer III.
Interventions1. Lateral minimally invasive approach (plate fixation)
2. Deltopectoral approach (plate fixation)
OutcomesFollow-up: 3, 6, and 12 weeks, and at 6, 12, 18 and 24 months
Primary outcome: Quick DASH
Secondary outcomes: SF-12v2 Questionnaire, Constant Shoulder Score, the Patient Scar Assessment Scale, complication rate
Starting dateStart date: November 2007
End date: January 2012
Contact informationMarie-France Poirer
Hopital Sacré-Coeur
Montreal
Quebec
Canada
H4J1C5
mariefrancepoirier@hotmail.com
NotesEntry for trial (clinicaltrials.gov) on April 10 2012, indicated that "The recruitment status of this study is unknown because the information has not been verified recently"

Loma Linda

Trial name or titleClinical outcome comparison between medial and lateral offset reverse shoulder arthroplasty
MethodsRandomised single blinded trial
Participants40 patients aged between the ages of 50 and 95 years of age who are a candidate for a reverse shoulder arthroplasty. This includes patients with rotator cuff tear arthroplasty, irreparable rotator cuff tears, significant proximal humerus fractures and malunions, and chronic proximal humerus dislocators.
InterventionsTornier Reversed shoulder arthroplasty:
1. Medial offset design
2. Lateral offset design
Outcomes

Follow-up: 2 years

Shoulder functional score
Pain scores
Radiological outcomes

Starting date

May 2010

End date: April 2012

Contact information

Wesley Phipatanakul, MD
wphip@hotmail.com

Principal investigator: Montri D Wongworawat, MD,
Loma Linda University Health Department of Orthopaedic Surgery, Loma Linda
Californa 92354
USA

NotesThe future inclusion of this mixed population trial will depend on the numbers of participants with proximal humeral fractures.
As of registration update (21/06/2011) the study was still recruiting  

Pelet

Trial name or titleEffectiveness of intensive rehabilitation on shoulder function after a fracture of the proximal humerus treated by locked plate. A prospective randomized study
MethodsRandomised clinical trial
Participants80 patients aged over 18 years treated by PHILOS locked plate system for unstable closed fracture of the proximal humerus (two-part and three-part fractures according to the Neer classification) within 7 days on injury.
Interventions

1. Early and intensive exercise programme
A thoraco brachial brace will be worn for 48 hours following the surgery and then removed for the remainder of treatment. Patients will then start the intensive rehabilitation programme without physical therapy. The exercise programme will be provided to the patient. The exercises consist of active and active assisted movements of the shoulder for a period of six weeks, limiting external rotation to 0°. Patients are encouraged to use their affected limb for daily activities. Strengthening exercises are started the 6th week following surgery and the full programme will be completed three months after surgery. Patients who wish can then continue their rehabilitation with a physiotherapist. The patient will complete a daily diary to validate the frequency and intensity of the exercises.

versus

1. Standard rehabilitation programme
The patient will wear the thoraco brachial brace for a period of four weeks following the surgery. It may be taken off for hygiene purposes and dressing up. After the four weeks, the patient will take the brace off permanently and begins an exercise programme, writing down the frequency and intensity of the exercises. Physiotherapy is allowed for the remaining part of the three months rehabilitation programme.

OutcomesLength of follow-up: 12 months
Primary outcome: Constant score (adjusted for age) at 6 months. A difference of 10 points is considered significant (standard deviation of 15 points).
Secondary outcomes: Reoperation, redisplacement, Constant score at 12 months, Dash, return to professional activities, pain, range of motion
Starting dateDecember 2009
Final follow-up date: December 2012 (completion: October 2013)
Contact informationHélène Côté, Reg. Nurse: helco3@hotmail.com
Stéphane Pelet, MD, PhD: spelet01@hotmail.com
Hopital de l'Enfant-Jésus, Canada
Notes 

ProCon

Trial name or titlePrimary hemiarthroplasty versus conservative treatment for comminuted fractures of the proximal humerus in the elderly (ProCon) - a multicenter randomized trial
MethodsRandomised trial: "variable block randomisation will be accomplished via a trial website"
Participants

Patients (65 years or older) with a comminuted proximal humeral fracture.

80 patients (65 years or older) with a comminuted proximal humeral fracture: three-part (Hertel classification type 9, 10, 11), four-part (Hertel type 12), anatomical neck (Hertel type 2), or split-head fractures of the humeral head

Interventions1. Hemiarthroplasty (Affinis® Fracture shoulder endoprosthesis)
2. Conservative treatment (collar and cuff for three weeks)
Outcomes

Follow-up: 1, 3 and 6 weeks, and 3, 6, 12 and 24 months

Primary outcome (Constant Score) and secondary outcomes (DASH, pain, radiographic healing, secondary intervention rates, complication rates, mortality rates, SF-36, and EQ-5D)
Costs for (in)formal healthcare consumption

Starting dateStart date: 15 June 2009
Planned end date: 31 December 2013
Contact informationDennis Den Hartog
Department of Surgery-Traumatology
Erasmus MC
University Medical Center
Rotterdam
P.O. Box 2040
3000 CA Rotterdam
The Netherlands
d.denhartog@erasmusmc.nl
NotesPublished protocol

ProFHER

Trial name or titlePragmatic multi-centre randomised trial of surgical versus non-surgical treatment for proximal fracture of the humerus in adults
MethodsMulti-centre randomised controlled trial
Participants250 patients, aged 16 or above, presenting to the participating trauma centre within 3 weeks of their injury with a radiologically confirmed displaced fracture of the proximal humerus involving the surgical neck
Interventions1. Surgery (fixation or joint replacement)
2. Non-surgical management (sling immobilisation)
OutcomesFollow-up: 6, 12 and 24 months
Primary outcome: Oxford Shoulder Score (12-item condition-specific questionnaire providing a total score based on the person's subjective assessment of pain and activities of daily living impairment) assessed at 6, 12 and 24 months via postal questionnaire.
Secondary outcomes: 12-item short form health survey (SF-12) and Euroqol (EQ-5D) for general health status data (at 6, 12 and 24 months); complications, including surgical complications (wound infection, implant failure, shoulder dislocation, septicaemia); early medical complications, i.e. chest infection, confirmed myocardial infarction or stroke, treated deep vein thrombosis and pulmonary embolism; mortality; subsequent referral for operation or substantive treatment; data for economic evaluation: NHS and societal costs
Starting dateStart date: 1/10/2009
End date: 28/02/2014
Contact informationProf Amar Rangan
Consultant Orthopaedic Surgeon
The James Cook University Hospital
Marton Road
Middlesbrough
TS4 3BW
United Kingdom
amar.rangan@stees.nhs.uk
NotesPublished protocol. Extension to recruitment period has resulted in a delay to the completion date (28/02/2014). Recruitment target achieved.

Ring

Trial name or titleEarly vs delayed physical therapy (exercises) for non-operatively-treated proximal humerus fractures: a prospective randomized trial
MethodsRandomised trial
Participants60 patients, aged 18 years or over, with non-operatively treated proximal humeral fractures
Interventions1. Physical therapy started immediately after diagnosis of injury
2. Physical therapy delayed until 3 weeks after diagnosis of injury
OutcomesFollow-up: 6 months
Primary outcome: shoulder flexion
Secondary outcomes: shoulder pain Likert scores; external and internal rotation; abduction; DASH and Constant scores
Starting dateStart date: February 2005
End date: December 2013
Contact informationProf David Ring
Director of Research
Hand Service
Massachusetts General Hospital
Boston
Massachusetts
USA
dring@partners.org
NotesChanges to NCT00438633 on 27 May 2008 seemed to indicate that, despite its official title, this is now a prospective cohort study (accessed: April 10 2012). David Ring confirmed it was still an RCT (April 16 2012).

Stanley

Trial name or titleMinimally invasive versus standard open reduction of proximal humerus fractures
Official title: A prospective randomised controlled trial comparing clinical and radiographic outcomes of the deltopectoral and limited deltoid splitting approaches for fixation of displaced proximal humeral fractures in a skeletally mature population
MethodsSingle-centre, randomised controlled trial. Computer generated random list. Unblinded.
Participants90 patients with displaced proximal humeral fractures
Interventions1. Minimally invasive approach with closed reduction and temporary fixation with k-wires, then the limited deltoid splitting approach (locking plate)
2. Standard open reduction and internal fixation via the deltopectoral approach (locking plate)
Outcomes

Follow-up: 1 year

Primary outcome: Constant Shoulder Score and DASH (Disabilities of the Arm, Shoulder, and Hand) Score

Secondary outcomes: radiographic outcomes, complications, union rate

Starting dateAnticipated start date: October 2010 (however, study is not yet recruiting
Contact information

Jeremy Stanley

14 Grand Drive

Remuera

Auckland 1050

New Zealand

jellystan@hotmail.com

NotesTrial was not recruiting (nor ethics approval) at time of registration (02/09/2010)

TPHF

  1. a

    AO = Arbeitsgemeinschaft fur Osteosynthesefragen / Association for the Study of Internal Fixation (or ASIF)
    LCP = Locking compression plate
    NRR = National Research Register

Trial name or titleTreatment of Proximal Humeral Fractures (TPHF)
Official title: A national, prospective, randomized, multicenter, controlled head-to-head comparison of conservative, plate fixation and prosthesis in treatment of displaced 2, 3, and 4 part fractures of proximal humerus of 60 years and older patients
MethodsMulti-centre, randomised clinical trial. Single blinded (outcomes assessor)
Participants290 patients aged over 60 years with displaced 2-, 3- and 4-part proximal humeral fractures
Interventions

1. PHILOS locking plate

2. Epoca prosthesis
3. Conservative management

Outcomes

Follow-up: 2 years

Primary outcome: DASH (Disabilities of the Arm, Shoulder, and Hand) score

Secondary outcome: EQ-5D Questionnaire

Starting date

Start date: January 2011

Estimated completion date: September 2016

Contact information

Antti Launonen

Tampere University Hospital

Tampere

Pirkanmaa

Finland, 33521

antti.launonen@pshp.fi

NotesAs of registration update (01/02/2011), trial was recruiting in Tampere University Hospital, but recruitment had not started in Oulu or Kuopio

Ancillary