Conservative management following closed reduction of traumatic anterior dislocation of the shoulder

  • Conclusions changed
  • Review
  • Intervention

Authors

  • Nigel CA Hanchard,

    Corresponding author
    1. Teesside University, Health and Social Care Institute, Middlesbrough, Tees Valley, UK
    • Nigel CA Hanchard, Health and Social Care Institute, Teesside University, Middlesbrough, Tees Valley, TS1 3BA, UK. n.hanchard@tees.ac.uk.

    Search for more papers by this author
  • Lorna M Goodchild,

    1. The James Cook University Hospital, South Tees Hospitals NHS Foundation Trust, Department of Physiotherapy, Middlesbrough, Tees Valley, UK
    Search for more papers by this author
  • Lucksy Kottam

    1. The James Cook University Hospital, South Tees Hospitals NHS Foundation Trust, Department of Orthopaedics, Middlesbrough, UK
    Search for more papers by this author

Abstract

Background

Acute anterior dislocation, which is the most common type of shoulder dislocation, usually results from an injury. Subsequently, the shoulder is less stable and is more susceptible to re-dislocation, especially in active young adults. This is an update of a Cochrane review first published in 2006.

Objectives

To assess the effects (benefits and harms) of conservative interventions after closed reduction of traumatic anterior dislocation of the shoulder. These might include immobilisation, rehabilitative interventions or both.

Search methods

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (September 2013), the Cochrane Central Register of Controlled Trials (CENTRAL) (2013, Issue 8), MEDLINE (1946 to September 2013), EMBASE (1980 to Week 38, 2013), CINAHL (1982 to September 2013), PEDro (1929 to November 2012), OTseeker (inception to November 2012) and trial registries. We also searched conference proceedings and reference lists of included studies.

Selection criteria

Randomised or quasi-randomised controlled trials comparing various conservative interventions versus control (no or sham treatment) or other conservative interventions applied after closed reduction of traumatic anterior dislocation of the shoulder.

Data collection and analysis

All review authors independently selected trials, assessed risk of bias and extracted data. Study authors were contacted for additional information. Results of comparable groups of trials were pooled.

Main results

We included three randomised trials and one quasi-randomised trial, which involved 470 participants (371 male) with primary traumatic anterior dislocation of the shoulder reduced by various closed methods. Three studies evaluated mixed populations; in the fourth study, all participants were male and 80% were soldiers. All trials were at some risk of bias but to a differing extent. One was at high risk in all domains of the risk of bias tool, and one was at unclear or high risk in all domains; the other two trials were deemed to have predominantly low risk across all domains. Overall, reflecting both the risk of bias and the imprecision of findings, we judged the quality of evidence to be "very low" for all outcomes, meaning that we are very uncertain about the estimates of effect.

The four trials evaluated the same comparison - immobilisation in external rotation versus internal rotation - and each of our three primary outcomes (re-dislocation, patient-reported outcome measures (PROMs) for shoulder instability and resumption of activities) was reported by one or more of the trials, with two-year or longer follow-up. Pooling was possible for "re-dislocation" (three trials) and for aspects of "resumption of sport/activities at pre-injury level" (two trials).

There was no evidence to show a difference between the two groups in re-dislocation at two-year or longer follow-up (risk ratio (RR) 1.06 favouring internal rotation, 95% confidence interval (CI) 0.73 to 1.54; P value 0.77; 252 participants; three trials). In a low-risk population, with an illustrative baseline risk of 247 re-dislocations per 1000, these data equate to 15 more (95% CI 67 fewer to 133 more) re-dislocations per 1000 after immobilisation in external rotation. In a medium-risk population, with an illustrative baseline risk of 436 re-dislocations per 1000, the data equate to 26 more (95% CI 118 fewer to 235 more) re-dislocations after immobilisation in external rotation.

Nor was evidence found to show a difference between the two groups in return to pre-injury levels of activity at two-year or longer follow-up (RR 1.25 favouring external rotation, 95% CI 0.71 to 2.2; P value 0.43; 278 participants; two trials). In a low-risk population, with an illustrative baseline risk of 204 participants per 1000 returning to pre-injury levels of activity, this equates to 41 more (95% CI 59 fewer to 245 more) participants per 1000 resuming activity after immobilisation in external rotation. In a high-risk population, with an illustrative baseline risk of 605 participants per 1000 returning to pre-injury levels of activity, this equates to 161 more (95% CI 76 fewer to 395 more) participants per 1000 resuming activity after immobilisation in external rotation.

One trial reported that the difference between the two groups in Western Ontario Shoulder Instability Index scores, analysed using non-parametric statistics, was "not significant (P = 0.32)". Of our secondary outcomes, pooling was possible for "any instability" (two trials) and for important adverse events (three events, two trials). However, adverse event data were collected only in an ad hoc way, and it is unclear whether identification and reporting of such events was comprehensive. No report addressed participant satisfaction or health-related quality of life outcome measures.

There was no evidence confirming a difference between the two positions of immobilisation in any of the primary or secondary outcomes; for each outcome, the confidence intervals were wide, covering the possibility of substantial benefit for each intervention.

Authors' conclusions

Numerous conservative strategies may be adopted after closed reduction of a traumatic anterior dislocation of the shoulder, and many warrant investigation. However, our review reveals that evidence from randomised controlled trials is only available for a single approach: immobilisation in external rotation versus immobilisation in the traditional position of internal rotation. Moreover, this evidence is insufficient to demonstrate whether immobilisation in external rotation confers any benefit over immobilisation in internal rotation.

We identified six unpublished trials and two ongoing trials that compare immobilisation in external versus internal rotation. Given this, the main priority for research on this question consists of the publication of completed trials, and the completion and publication of ongoing trials. Meanwhile, increased attention to other interventions is required. Sufficiently powered, good quality, well reported randomised controlled trials with long-term surveillance should be conducted to examine the optimum duration of immobilisation, whether immobilisation is necessary at all (in older age groups particularly), which rehabilitative interventions work best and the acceptability to participants of different care strategies.

Résumé scientifique

La prise en charge conservatrice suite à une réduction fermée de la luxation antérieure traumatique de l'épaule

Contexte

La luxation antérieure aiguë, qui est le type le plus courant de la luxation de l'épaule, résulte généralement d'une blessure. Par la suite, l'épaule est moins stable et plus susceptible d'une nouvelle luxation, en particulier chez les jeunes adultes actifs. Ceci est une mise à jour d'une revue Cochrane publiée pour la première fois en 2006.

Objectifs

Évaluer les effets (bénéfiques et délétères) des interventions conservatrices après une réduction fermée de la luxation antérieure traumatique de l'épaule. Celles-ci peuvent inclure l'immobilisation, les interventions de rééducation ou les deux.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre spécialisé du groupe Cochrane sur les traumatismes ostéo-articulaires et musculaires (septembre 2013), le registre Cochrane des essais contrôlés (CENTRAL) (2013, numéro 8), MEDLINE (de 1946 à septembre 2013), EMBASE (de 1980 à la semaine 38, 2013), CINAHL (de 1982 à septembre 2013), PEDro (de 1929 à novembre 2012), OTseeker (des origines à novembre 2012) et les registres d'essais. Nous avons également effectué des recherches dans les actes de conférence et les références bibliographiques des études incluses.

Critères de sélection

Essais contrôlés randomisés ou quasi randomisés comparant différentes interventions conservatrices par rapport à un témoin (absence de traitement ou traitement fictif) ou à d'autres interventions conservatrices après une réduction fermée de la luxation antérieure traumatique de l'épaule.

Recueil et analyse des données

Tous les auteurs de la revue ont indépendamment sélectionné les essais, évalué le risque de biais et extrait les données. Les auteurs des études ont été contactés pour obtenir des informations supplémentaires. Les résultats des groupes d'essais comparables ont été combinés.

Résultats principaux

Nous avons inclus trois essais randomisés et un essai quasi randomisé, qui portaient sur 470 participants (371 hommes) atteints de luxation antérieure traumatique de l'épaule réduite par diverses méthodes fermées. Trois études évaluaient des populations mixtes ; dans la quatrième étude, tous les participants étaient des hommes et 80 % étaient des soldats. Tous les essais présentaient un risque de biais mais à des degrés divers. L'un était à risque élevé dans tous les domaines de l'outil de risque de biais, et un essai était à risque incertain ou élevé dans tous les domaines ; les deux autres essais ont été considérées à risque principalement faible dans tous les domaines. Dans l'ensemble, au regard des risques de biais et de l'imprécision des résultats, nous avons estimé que la qualité des preuves était « très faible » pour tous les critères de jugement, ce qui signifie que nous avons beaucoup d'incertitudes concernant les estimations d'effet.

Les quatre essais évaluaient une même comparaison - l'immobilisation dans la rotation externe par rapport à la rotation interne - et chacun de nos trois critères de jugement principaux (nouvelle luxation, mesures de résultats rapportés par les patients (PROM) pour l'instabilité de l'épaule, reprise des activités) a été rapporté par au moins un essai, avec un suivi d'au moins deux ans. Le regroupement était possible pour « nouvelle luxation » (trois essais) et pour certains aspects de « reprise du sport / des activités au niveau d'avant la blessure » (deux essais).

Il n'y avait aucune preuve permettant de démontrer une différence entre les deux groupes en termes de nouvelle luxation à deux ans ou lors d'un suivi plus long (risque relatif (RR) 1,06 en faveur de la rotation interne, intervalle de confiance (IC) à 95 % 0,73 à 1,54 ; P = 0,77 ; 252 participants ; trois essais). Dans une population à faible risque, avec un risque de base illustratif de 247 nouvelles luxations pour 1 000, ces données reviennent à 15 (IC à 95 % de moins 67 à plus 133) nouvelles luxations en plus pour 1 000 suite à l'immobilisation dans la rotation externe. Dans une population à risque moyen, avec un risque de base illustratif de 436 nouvelles luxations pour 1 000, les données reviennent à 26 (IC à 95 % de moins 118 à plus 235) nouvelles luxations en plus suite à l'immobilisation dans la rotation externe.

De même, aucune preuve n'a été trouvée pour démontrer une différence entre les deux groupes dans la reprise des niveaux d'activité d'avant la blessure à deux ans ou lors d'un suivi plus long (RR 1,25 en faveur de la rotation externe, IC à 95 % 0,71 à 2,2 ; P = 0,43 ; 278 participants ; deux essais). Dans une population à faible risque, avec un risque de base illustratif de 204 participants sur 1 000 revenant à leurs niveaux d'activité d'avant la blessure, cela correspond à 41 (IC à 95 % de moins 59 à plus 245) participants de plus sur 1 000 reprenant leurs activités après l'immobilisation dans la rotation externe. Dans une population à haut risque, avec un risque de base illustratif de 605 participants sur 1 000 revenant à leurs niveaux d'activité d'avant la blessure, cela correspond à 161 (IC à 95 % de moins 76 à plus 395) participants de plus sur 1 000 reprenant leurs activités après l'immobilisation dans la rotation externe.

Un essai a rapporté que la différence entre les deux groupes en termes de scores sur l'indice de l'instabilité de l'épaule de Western Ontario (WOSI), analysés à l'aide de statistiques non paramétriques, était « non significative (P = 0,32) ». De nos critères de jugement secondaires, le regroupement était possible pour « toute instabilité » (deux essais) et pour les événements indésirables graves (trois événements, deux essais). Cependant, les données sur les événements indésirables ont été recueillies seulement de manière ad hoc, et il est difficile de savoir si l'identification et la notification de ces événements était exhaustive. Aucun rapport n'a examiné la satisfaction des participants ou des mesures de résultats sur la qualité de vie liée à la santé.

Il n'y avait aucune preuve confirmant l'existence d'une différence entre les deux positions d'immobilisation dans aucun des critères de jugement principaux ou secondaires ; pour chaque critère de jugement, les intervalles de confiance étaient larges, comprenant la possibilité d'un bénéfice substantiel pour chaque intervention.

Conclusions des auteurs

De nombreuses stratégies de conservation peuvent être adoptées après une réduction fermée d'une luxation antérieure traumatique de l'épaule, et plusieurs méritent d'être étudiées. Cependant, notre revue révèle que des preuves issues d'essais contrôlés randomisés ne sont disponibles que pour une seule approche : l'immobilisation dans la rotation externe par rapport à l'immobilisation dans la position traditionnelle de la rotation interne. De plus, ces preuves sont insuffisantes pour démontrer si l'immobilisation dans la rotation externe confère un quelconque avantage par rapport à l'immobilisation dans la rotation interne.

Nous avons identifié six essais non publiés et deux essais en cours qui comparent l'immobilisation dans la rotation interne versus externe. De ce fait, la première priorité pour les recherches sur cette question concerne la publication des essais achevés et l'achèvement et la publication des essais en cours. En attendant, d'autres interventions requièrent plus d'attention. Des essais contrôlés randomisés de bonne qualité, à la puissance statistique suffisante et bien documentés, avec une surveillance à long terme, doivent être réalisés pour examiner la durée optimale d'immobilisation, la nécessité même de l'immobilisation (en particulier chez les personnes âgées), les interventions de rééducation fonctionnant le mieux ainsi que l'acceptabilité de différentes stratégies de soins pour les participants.

Plain language summary

Non-surgical management after non-surgical repositioning of traumatic anterior dislocation of the shoulder

Acute anterior shoulder dislocation is an injury in which the top end of the upper arm bone is pushed out of the joint socket in a forward direction. Afterwards, the shoulder is less stable and is prone to re-dislocation or subluxation (partial re-dislocation), especially in active young adults. Initial treatment involves putting the joint back in place. This is called 'closed reduction' when it is done without surgery. Subsequent treatment is often conservative (non-surgical) and generally involves placement of the injured arm in a sling or in another immobilising device followed by specific exercises.

After a comprehensive search, completed in September 2013, for randomised controlled trials that compared different methods of conservative management of these injuries we included only four trials, one of which was not truly randomised. These trials involved a total of 470 participants (371 male). All had primary traumatic anterior dislocation of the shoulder reduced by various closed methods. Three studies evaluated mixed populations; in the fourth study, all participants were male and 80% were soldiers. All trials were at some risk of bias (systematic errors that could lead to overestimation or underestimation of treatment effectiveness), with two trials in particular being at high risk of bias in a number of aspects. Overall, the quality of the evidence was very low, meaning that we are very uncertain about the direction and size of effect.

All four trials compared immobilisation of the arm in external rotation (when the arm is orientated outwards with the forearm away from the chest) versus immobilisation in internal rotation (the usual sling position, where the arm rests against the chest) following closed reduction. Investigators followed patients for at least two years. The results showed no difference between the two groups in any of our pre-defined outcomes. These included re-dislocations, scores on validated shoulder function questionnaires, return to pre-injury activity or sport, and any instability. Other pre-defined outcomes (patient satisfaction with the intervention, and health-related quality of life outcome data) were not reported. Adverse events were poorly recorded.

In our recommendations for future research, we point out the importance of completing and publishing the eight other trials making the same comparison as the four included trials. We also note that other important questions need to be studied, such as how long the shoulder should be immobilised for the best outcomes. In conclusion, current evidence from randomised controlled trials is insufficient to inform choices for conservative management following closed reduction of traumatic anterior dislocation of the shoulder.

Résumé simplifié

La prise en charge non chirurgicale après le repositionnement non chirurgical de la luxation antérieure traumatique de l'épaule

La luxation antérieure aiguë de l'épaule est une blessure dans laquelle l'extrémité supérieure de l'os du bras sort de son articulation vers l'avant. Par la suite, l'épaule est moins stable et est susceptible d'une nouvelle luxation ou d'une subluxation (une nouvelle luxation partielle), en particulier chez les jeunes adultes actifs. Le traitement initial consiste à remettre l'articulation en place. On parle de « réduction fermée » lorsque celle-ci est pratiquée sans chirurgie. Le traitement ultérieur est souvent conservateur (non chirurgical) et implique généralement l'immobilisation du bras blessé dans une écharpe ou un autre dispositif, suivie par des exercices spécifiques.

Après une recherche exhaustive effectuée en septembre 2013 pour les essais contrôlés randomisés comparant différentes méthodes de prise en charge conservatrice de ces blessures, nous avons uniquement inclus quatre essais, dont l'un n'était pas véritablement randomisé. Ces essais portaient sur un total de 470 participants (371 hommes). Tous présentaient une luxation antérieure primaire traumatique de l'épaule réduite par diverses méthodes fermées. Trois études évaluaient des populations mixtes ; dans la quatrième étude, tous les participants étaient des hommes et 80 % étaient des soldats. Tous les essais présentaient un risque de biais (erreurs systématiques qui peuvent conduire à une surestimation ou sous-estimation des l'efficacité du traitement), avec deux essais, en particulier, étant à risque élevé de biais dans un certain nombre d'aspects. Dans l'ensemble, la qualité des preuves était très faible, ce qui signifie que nous ne pouvons pas estimer avec certitude la direction et l'ampleur de l'effet.

Les quatre essais comparaient l'immobilisation du bras dans la rotation externe (lorsque le bras est dirigé vers l'extérieur et l'avant-bras détache de la poitrine) par rapport à l'immobilisation dans la rotation interne (la position habituelle dans l'écharpe, où le bras repose contre la poitrine) suite à une réduction fermée. Les investigateurs suivaient les patients pendant au moins deux ans. Les résultats n'ont montré aucune différence entre les deux groupes pour aucun de nos critères de jugement prédéfinis. Ceux-ci incluaient les nouvelles luxations, les scores sur la fonction de l'épaule de questionnaires validés, la reprise d'une activité ou d'un sport d'avant la blessure et toute instabilité. Les autres critères de jugement prédéfinis (la satisfaction des patients vis-à-vis de l'intervention et les données de résultat sur la qualité de vie liée à la santé) n'étaient pas rapportés. Les événements indésirables étaient mal consignés.

Dans nos recommandations pour les recherches futures, nous avons souligné l'importance de terminer et de publier les huit autres essais effectuant les mêmes comparaisons que les quatre essais inclus. Nous avons également noté que d'autres questions importantes doivent être étudiées, telles que la durée optimale de l'immobilisation de l'épaule pour les meilleurs résultats. En conclusion, les preuves actuelles issues d'essais contrôlés randomisés sont insuffisantes pour orienter les choix pour la prise en charge conservatrice suite à une réduction fermée de la luxation antérieure traumatique de l'épaule.

Notes de traduction

Traduit par: French Cochrane Centre 16th July, 2014
Traduction financée par: Financeurs pour le Canada : Instituts de Recherche en Santé du Canada, Ministère de la Santé et des Services Sociaux du Québec, Fonds de recherche du Québec-Santé et Institut National d'Excellence en Santé et en Services Sociaux; pour la France : Ministère en charge de la Santé

Laienverständliche Zusammenfassung

Nicht-operatives Management nach nicht-operativer Reposition (Wiedereinrichtung) einer traumatischen (verletzungsbedingten) vorderen Schulterluxation

Eine akute vordere Schulterluxation ist eine Verletzung, bei der das obere Ende des Oberarmknochens aus der Gelenkpfanne nach vorne herausgedrückt wird. Anschließend ist die Schulter weniger stabil und anfällig für eine Reluxation (erneute Luxation) oder Subluxation (teilweise Reluxation), insbesondere bei aktiven jungen Erwachsenen. Die Erstbehandlung besteht darin, das Gelenk wieder in seine ursprüngliche Stellung zu bringen. Dies wird als 'geschlossene Reposition' bezeichnet, wenn es ohne einen operativen Eingriff erfolgt. Die weitere Behandlung ist oft konservativ (nicht-operativ) und beinhaltet im allgemeinen die Lagerung des verletzten Armes in einer Schlinge oder einer anderen ruhigstellenden Vorrichtung, gefolgt von spezifischen Übungen.

Nach einer umfassenden, im September 2013 abgeschlossenen Suche nach randomisierten kontrollierten Studien (Studien, in denen die Teilnehmer zufällig zwei oder mehr verschiedenen Studiengruppen zugeteilt werden), die verschiedene Methoden des konservativen Managements dieser Verletzung miteinander verglichen, schlossen wir nur vier Studien ein, von denen eine nicht wirklich randomisiert war. Diese Studien umfassten insgesamt 470 Teilnehmer (371 männlich). Alle hatten eine erste traumatische vordere Schulterluxation, die mit verschiedenen geschlossenen Verfahren gerichtet wurde. Drei Studien untersuchten gemischte Patientengruppen; in der vierten Studie waren alle Teilnehmer männlich und 80% Soldaten. Alle Studien waren von einem Verzerrungsrisiko betroffen (systematische Fehler, die zu einer Überschätzung oder Unterschätzung der Behandlungseffektivität führen können); insbesondere zwei Studien waren bezüglich mehrerer Aspekte von einem hohen Verzerrungsrisiko betroffen. Insgesamt war die Qualität der Evidenz (des wissenschaftlichen Belegs) sehr niedrig, was bedeutet, dass wir bezüglich der Richtung und Größe des Effektes sehr unsicher sind.

Alle vier Studien verglichen eine Ruhigstellung des Armes in Außenrotation (wenn der Arm mit dem Unterarm vom Oberkörper weg nach außen gedreht ist) gegenüber einer Ruhigstellung in Innenrotation (die übliche Schlingenposition, bei der der Arm dem Oberkörper anliegt) nach geschlossener Reposition. Die Untersucher folgten den Patienten über mindestens zwei Jahre. Die Ergebnisse ergaben für keinen unserer vordefinierten Endpunkte (Zielkriterien) einen Unterschied zwischen den Gruppen. Die Endpunkte waren Reluxationen, die Ergebnisse validierter (wissenschaftlich fundierter) Schulterfunktions-Fragebögen, die Wiederaufnahme von vor der Verletzung durchgeführten Aktivitäten oder Sport, oder Instabilität. Zu anderen vordefinierten Endpunkten (Patientenzufriedenheit mit der Behandlung und Ergebnisdaten zur gesundheitsbezogenen Lebensqualität) wurden keine Angaben gemacht. Unerwünschte Ereignisse wurden nur dürftig erfasst.

In unseren Empfehlungen für zukünftige Forschung heben wir die Bedeutung des Abschlusses und der Veröffentlichung der acht weiteren Studien hervor, die denselben Behandlungsvergleich wie die vier eingeschlossenen Studien durchgeführt haben. Wir stellen zudem fest, dass andere wichtige Fragen untersucht werden müssen, wie zum Beispiel, wie lange die Schulter ruhiggestellt werden sollte, um die besten Ergebnisse zu erbringen. Schlussfolgernd ist die derzeitige Evidenz aus randomisierten kontrollierten Studien als Entscheidungshilfe für das konservative Management nach geschlossener Reposition einer traumatischen vorderen Schulterluxation unzureichend.

Anmerkungen zur Übersetzung

C. Braun, K. Ehrenbrusthoff, N. Jahnke, Koordination durch Cochrane Schweiz

Summary of findings(Explanation)

Summary of findings for the main comparison. Immobilisation in external rotation compared with immobilisation in internal rotation for traumatic anterior dislocation of the shoulder
  1. 1Based on Finestone (24 to 48 months) and Liavaag (24 months). Follow-up point unspecified in Taskoran 2010.
    2Most information is from studies at low or unclear risk of bias.
    395% confidence interval around pooled estimate of effect includes (1) no effect, (2) appreciable benefit favouring immobilisation in external rotation and (3) appreciable benefit favouring immobilisation in internal rotation.
    4High risk of bias for one or more key domains.
    595% confidence interval around pooled estimate of effect includes (1) no effect and (2) appreciable benefit favouring immobilisation in external rotation.
    6The proportion of information from studies at high risk of bias is sufficient to affect the interpretation of results.
    7Chi2 was statistically significant at P value 0.09, and the I2 statistic was 65% ("which may represent substantial heterogeneity").
    8Mean 25.6 months (Itoi 2007) and 29.1 months (Liavaag 2011).
    9Chi2 was statistically significant at P value 0.06, and the I2 statistic was 72% ("which may represent substantial heterogeneity").
    10Data on adverse events were gathered ad hoc, and it is unclear whether all such events would have been captured by this strategy.
    11Reporting of this outcome was ad hoc and possibly was not comprehensive.

Immobilisation in external rotation compared with immobilisation in internal rotation for traumatic anterior dislocation of the shoulder
Patient or population: patients with traumatic anterior dislocation of the shoulder
Settings: splints applied in emergency departments
Intervention: immobilisation in external rotation
Comparison: immobilisation in internal rotation
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No. of participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
Immobilisation in internal rotation Immobilisation in external rotation
Re-dislocation, with two-year or longer follow-up
Verified self-report in Liavaag 2011. Mechanism of data collection unclear in Finestone 2009 and Taskoparan 2010
Follow up: 24 to 48 months1
Low-risk population RR 1.06
(0.73 to 1.54)
252
(3 studies)
⊕⊝⊝⊝
very low 2,3
Assumed low risk based on Liavaag 2011. Assumed medium risk based on the median of Finestone 2009 and Taskoparan 2010
247 per 1000 262 per 1000
(180 to 380)
Medium-risk population
436 per 1000 462 per 1000
(318 to 671)
Validated participant-reported outcome measures for shoulder instability
Western Ontario Shoulder Instability Index (WOSI). Scale from 0 to 2100
See commentSee commentNot estimable174
(1)
See commentOnly non-parametric data: quality of evidence not computable. Result reportedly non-significant
Resumption of sport at any level, with two-year or longer follow-up
Direct examination or telephone interview
Follow-up: mean 25.6 months
Medium-risk population RR 1.13
(0.87 to 1.48)
109
(1 study)
⊕⊝⊝⊝
very low 4,5
Assumed risk based on single study reporting this outcome
633 per 1000 715 per 1000
(551 to 937)
Resumption of sport/activities, with two-year or longer follow-up
Direct examination, telephone interview (Itoi 2007) or postal questionnaire (Liavaag 2011)
Follow-up: 24 to 54 months
Low-risk population RR 1.25
(0.71 to 2.2)
278
(2 studies)
⊕⊝⊝⊝
very low 5,6,7
Assumed low risk based on Itoi 2007. Assumed high risk based on Liavaag 2011
204 per 1000 255 per 1000
(145 to 449)
High-risk population
605 per 1000 756 per 1000
(429 to 1000)
Participant satisfaction with the intervention—not reportedSee commentSee commentNot estimable-See commentNo study reported this outcome
Any instability (subluxation or re-dislocation), with two-year or longer follow-up
Direct examination, telephone interview (Itoi 2007) or questionnaire (Liavaag 2011)
Follow-up: 24 to 54 months8
Medium-risk population RR 0.82
(0.5 to 1.33)
322
(2 studies)
⊕⊝⊝⊝
very low 2,3,6,9
Assumed risk based the on median of the two studies reporting this outcome
484 per 1000 397 per 1000
(242 to 644)
Important adverse events
Mechanism of data collection unclear
37 per 1000 23 per 1000
(3 to 165)
RR 0.61
(0.08 to 4.46)
216
(2 studies)
⊕⊝⊝⊝
very low 3,6,10,11
Assumed risk based on median of the two studies reporting this outcome
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in the comments. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.
GRADE Working Group grades of evidence.
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

Background

Description of the condition

Dislocation of the shoulder occurs when the head of the humerus (the top end of the upper arm bone) is displaced out of the glenohumeral (shoulder) joint. The extent of dislocation varies from subluxation (partial dislocation) to full dislocation (where the joint surfaces completely lose contact). It is usually diagnosed by a combination of history, physical examination findings and imaging - most often radiography (x-ray), but more rarely some other imaging modality such as magnetic resonance imaging (MRI). The direction of dislocation varies, but in most primary (first-time) dislocations, the head of the humerus (the top of the upper arm bone that forms the ball of the shoulder joint) is displaced anteriorly (forwards) in relation to the glenoid fossa (the socket of the shoulder joint). The cause is usually trauma, typically during contact sports in adolescents and younger adults. In older adults, dislocation may result from a fall from standing height. Traumatic anterior dislocation accounts for 96% of all shoulder dislocations (Goss 1988).

A study based in the USA (Simonet 1984) found an overall adjusted incidence of initial traumatic shoulder dislocation of 8.2 per 100,000 person-years, and estimated the incidence of all traumatic shoulder dislocations - initial and recurrent - to be at least 11.2 per 100,000 person-years. Simonet 1984 was a relatively small epidemiological study that was limited to a single rural county. More recently, Zacchilli 2010 examined the incidence of traumatic shoulder dislocation in the whole USA population and reported that the overall adjusted incidence is more than twice as high as was previously believed (23.9, 95% confidence interval (CI) 20.8 to 27.0, per 100,000 person-years). Seventy-eight per cent of dislocations occur in men (overall incidence rate in males: 34.9 per 100,000 person-years, 95% CI 30.1 to 39.7; in females: 13.3 per 100,000 person-years, 95% CI 11.6 to 15.0) (Zacchilli 2010). Although shoulder dislocation is generally considered an injury of young adults, Rowe 1956 found that there were as many initial dislocations after age 45 as before age 45. Twenty per cent of 545 consecutive patients presenting with anterior shoulder dislocation were 60 years of age or older in Gumina 1997. Zacchilli 2010 found a bimodal age distribution with peaks in the third decade (corresponding to the peak male incidence) and the ninth decade (corresponding to the peak female incidence). This is broadly consistent with the results of another study, based in Sweden (Hovelius 1982), which found that 35 (1.7%) of 2092 randomly selected people 18 to 70 years of age had experienced shoulder dislocation. Although the overall male-to-female ratio was three to one, the ratio was nine to one in the 21 to 30 year age group and, conversely, one to three in the 50 to 70 year age group.

Once dislocation has occurred, the shoulder is less stable and is more susceptible to re-dislocation. Recurrent dislocation tends to be more common in younger adults. For instance, a 10-year follow-up evaluation found that 66% of those between 12 and 22 years of age at the time of their first dislocation had one or more recurrences, whereas this occurred in 24% of those between 30 and 40 years of age (Hovelius 1996). Gumina 1997 found single or multiple recurrences in 22% of those 60 years of age or older.

The nature and extent of damage to the soft tissue surrounding the shoulder joint from a traumatic anterior dislocation vary. Common presentations include the Bankart lesion, characterised by damage to the anteroinferior part of the glenoid labrum (the fibrocartilage rim that deepens the joint socket) and the capsule surrounding the joint (Bankart 1938), and the Hill-Sachs lesion, which involves a compression fracture of the humeral head, as well as damage to its overlying cartilage (Hill 1940).

Description of the intervention

Traditionally, a non-surgical (conservative) approach, comprising closed reduction, three to six weeks' immobilisation in a sling and a subsequent physiotherapy or physical therapy programme (O'Brien 1987), has been used for first time dislocations. Recent years have seen much interest in an alternative to the traditional immobilisation method, whereby the shoulder is immobilised in external (outward) rotation using a custom-made (Itoi 2003) or commercially manufactured (Sullivan 2007) splint.

Surgical intervention has generally been reserved for cases of chronic recurrence/instability. However, a Cochrane review (Handoll 2004; updated in 2009) comparing surgical with non-surgical treatment found some limited evidence supporting primary surgery for young adults, usually male, engaged in highly demanding physical activities who have sustained their first acute traumatic shoulder dislocation.

Our review considers the various approaches to post-reduction conservative treatment, such as the duration and position of sling immobilisation, the modalities used, and the timing and extent of rehabilitation interventions.

How the intervention might work

The aim of treatment for anterior dislocation is to restore a functional, painless and stable shoulder. The choice of treatment approach will be influenced by patient age and previous history of dislocation, occupation, level of activity, general health and ligamentous laxity and by expectations of patient adherence to a prescribed therapeutic regimen.

The aim of immobilisation is to allow healing. In this connection, some MRI and cadaveric studies of Bankart lesions (Dymond 2011; Itoi 2001; Kitamura 2005; Liavaag 2009; Miller 2004; Moxon 2010; Pennekamp 2006; Seybold 2009; Siegler 2010) have shown better and firmer repositioning of the peeled-away capsule when the shoulder is externally rotated than when it is internally (inwardly) rotated - the position naturally imposed by a sling. This has kindled and sustained interest in the possibility that external rotation immobilisation may improve healing, and consequently outcomes, in comparison with the traditional approach. However any immobilisation has potential disadvantages, and there is an argument for shortening its duration (Kiviluoto 1980) or forgoing it altogether (Hovelius 2008) to allow early restoration of movement, especially in the middle-aged to elderly, who are susceptible to immobilisational stiffness and frozen shoulder (Robinson 2012) but are less prone to re-dislocation than the young (Hovelius 1996).

Finally, various exercise interventions might theoretically increase functional stability by restoring shoulder joint proprioception (spatial awareness) in the shoulder joint and by retraining muscles to help maintain joint congruency (Karatsolis 2006); while motion-limiting braces might prevent re-dislocation by restricting shoulder movement in vulnerable directions (Murray 2013).

Why it is important to do this review

This is an update of a Cochrane review published in 2006 (Handoll 2006). Handoll 2006 included only preliminary data from one flawed, partly quasi-randomised trial (Itoi 2003), which compared the immobilisation positions of external and internal rotation and found no difference in terms of return to sport, re-dislocation or instability. Since then, there has been a proliferation of studies (laboratory-based (Limpisvasti 2008; Miller 2004), observational (Dymond 2011; Kitamura 2005; Moxon 2010; Pennekamp 2006; Seybold 2009; Siegler 2010) and RCTs (such as Finestone 2009; Liavaag 2011; Taskoparan 2010)) evaluating immobilisation in external rotation; there are also a number of registered trials. Questions surround other aspects of immobilisation as well (timing of application and duration, position and whether any immobilisation is better than none at all), rehabilitation (its general effectiveness, its relative effectiveness across different settings and the relative effectiveness of different packages and modes of delivery) and motion-limiting braces. These considerations illustrate the need for an updated review.

Objectives

To assess the effects (benefits and harms) of conservative interventions after closed reduction of traumatic anterior dislocation of the shoulder. These might include immobilisation, rehabilitative interventions or both.

Methods

Criteria for considering studies for this review

Types of studies

We considered any randomised or quasi-randomised (e.g. allocation by hospital record number or date of birth) controlled trials evaluating conservative treatment after anterior dislocation of the shoulder.

Types of participants

Individuals who have undergone closed reduction for traumatic anterior dislocation of the shoulder. Ideally, the acute anterior shoulder dislocation should have been confirmed by physical examination and radiography or another imaging modality such as MRI. We intended to consider the potential for misdiagnosis, such as a missed proximal humeral fracture, in trials in which the method of diagnosis was unspecified or in which diagnosis was based on physical examination alone.

Although we stated that we would include individuals of any age, we correctly anticipated finding no trials focused specifically on the management of traumatic anterior dislocation in children.

We excluded or would have excluded trials focusing on the treatment of participants with non-traumatic or habitual dislocations, or concomitant fractures or multiple trauma; and those focusing on management of neurovascular complications or postsurgical management. We stipulated that trials with mixed populations involving any one indication of the above would be considered for inclusion if the proportion of the latter (e.g. atraumatic dislocation) was clearly defined for each treatment group and was clearly small (< 10%), or if separate data for acute traumatic anterior dislocation were provided.

Types of interventions

We planned the following.

  1. To assess whether a difference exists between outcomes of different methods (including arm position) and durations (including none or intermittent) of postreduction immobilisation. However, we planned to exclude trials comparing variants (e.g. duration, position) or supplements to particular immobilisation techniques unless the general effectiveness of the method had been established.

  2. To assess whether a difference exists between outcomes of the provision of rehabilitation intervention (of any kind) versus no intervention. Examples of rehabilitation interventions include advice and education, active and passive mobilisation, proprioception and stabilisation exercises, scapular setting and trunk stability exercises. These may be used in combination or individually and may be applied in various ways and settings. Although these interventions are potentially available to all patients allocated the rehabilitation intervention, their actual application may vary according to the perceived needs of individual patients. We aimed to assess this separately for the provision of any rehabilitation (a) during immobilisation, and (b) after immobilisation.

  3. To assess whether a difference exists between outcomes of different types of rehabilitation interventions. Comparisons would have included different single modalities or different combinations of rehabilitation modalities. However, we planned to exclude trials comparing different techniques, timing (duration, frequency) and intensity of single rehabilitation modalities until the effectiveness of the modality itself had been established. We also would have excluded trials evaluating pharmacological interventions and trials testing interventions aimed solely at pain relief.

  4. To assess whether a difference exists between outcomes of different methods of delivering/providing various rehabilitation interventions. Comparisons would have included supervised therapy versus home exercises, different methods of supervised therapy (e.g. individual versus group instruction) and differences in the frequency and duration of rehabilitation. In the first instance, we did not plan to include comparisons of rehabilitation intervention delivered by individual professionals (e.g. doctors, physiotherapists, occupational therapists) with different levels or backgrounds of expertise or training.

Types of outcome measures

Primary outcomes

We sought the following outcome measures, which we prespecified as primary.

  1. Re-dislocation: separation of the joint requiring reduction and, ideally, verified

  2. Validated patient-reported outcome measures (PROMs) for shoulder instability (e.g. Oxford Shoulder Instability Score (Dawson 1999), Western Ontario Shoulder Instability Index (WOSI) (Kirkley 1998))

  3. Resumption of pre-injury activities (yes or no)

Secondary outcomes
  1. Participant satisfaction with the intervention

  2. Validated health-related quality of life outcome measures (e.g. EQ-5D (standardised measure of health outcome), Short Form (SF)-36)

  3. Any instability: subluxation (separation of the joint not requiring reduction) or subjective instability, either individually or grouped with dislocation as a composite outcome

  4. Important adverse events (not including re-dislocation or instability) that were plausibly attributable to post reduction management (e.g. persistent pain, frozen shoulder). Other adverse events were to be reported narratively

In addition, we intended to take note of any reports of service utilisation or resource use, for instance, length of hospital stay, outpatient attendance and the provision and nature of physiotherapy; and participants' adherence to their allocated interventions.

Timing of outcome measurement

Approximately one-third of re-dislocations occur within three months of the initial dislocation, and a further third between three and 12 months (Rhee 2009). We therefore proposed organising outcomes into the following time frames, with greatest importance attached to long-term reporting.

  1. Short-term: up to and including three months following dislocation

  2. Medium-term: greater than three months and up to and including 12 months following dislocation

  3. Long-term: greater than 12 months following dislocation

Search methods for identification of studies

Electronic searches

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (September 2013), the Cochrane Central Register of Controlled Trials (CENTRAL) (2013 Issue 8), MEDLINE (1946 to September Week 3 2013), MEDLINE In-Process & Other Non-Indexed Citations (September 2013), EMBASE (1980 to 2013 Week 38), CINAHL (1982 to September 2013), PEDro (Physiotherapy Evidence Database) (1929 to November 2012) and OTseeker (Occupational Therapy Systematic Evaluation of Evidence Database) (inception to November 2012). We also searched the WHO International Clinical Trials Registry Platform (November 2012) and the UK National Research Register (2005, Issue 3, now archived) for ongoing and recently completed trials. We applied no language restrictions.

In MEDLINE (Ovid Web), the subject-specific strategy was combined with the sensitivity-maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2011) and was modified for use in other databases. Search strategies for CENTRAL, MEDLINE, EMBASE and CINAHL can be found in Appendix 1. Details of the search strategies used previously are published in Handoll 2006.

Searching other resources

We checked reference lists of articles. We searched the conference proceedings of the British Elbow and Shoulder Society (2003 to 2012). We handsearched conference proceedings published in the supplements of the Journal of Bone and Joint Surgery - British Volume (now known as The Bone and Joint Journal) (January 2004 to January 2013) and Injury (January 2004 to January 2013).

Data collection and analysis

Selection of studies

Two review authors (LG, LK) independently assessed potentially eligible trials for inclusion, and assessments were checked by a third review author (NH). No disagreement was reported. Titles of journals, names of authors and names of supporting institutions were not masked at any stage.

Data extraction and management

All three review authors independently extracted data. The forms were piloted on an excluded study, and discrepancies were resolved through discussion.

Assessment of risk of bias in included studies

Risk of bias was independently assessed, without masking of the source and authorship of trial reports, by all three review authors. The assessment form was piloted on one trial. Between rater consistency in assessment was checked by one review author (NH) at data entry. All differences were resolved by discussion. We used the tool outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). This tool incorporates assessment of randomisation (sequence generation and allocation concealment), blinding (of participants and of treatment providers), blinding of outcome assessment, completeness of outcome data, selection of outcomes reported and other sources of bias. Considered among the latter were discrepancies in the level of skill or care with which compared interventions were applied (performance bias) and commercial sponsorship (because of the potential for reporting bias). In future updates, we may consider subjective outcomes (e.g. PROMs) and objective outcomes (re-dislocation) separately in our assessment of blinding of outcome assessment and completeness of outcome data.

Measures of treatment effect

When available and appropriate, quantitative data for outcomes listed in the inclusion criteria are presented graphically. Risk ratios (RRs) and 95% confidence intervals (CIs) were calculated for dichotomous outcomes. Had continuous data been available, we planned to calculate mean differences (MDs) and 95% CIs calculated for continuous outcomes.

Unit of analysis issues

We considered unit of analysis issues very unlikely but nonetheless were alert to the possibility of these in the included studies. None were identified.

Dealing with missing data

We approached study authors for missing data by email or, if no email address was available, by post. Email and postal return addresses were given. If answers to more than a few questions were requested, we provided a pro forma to ensure clarity and to minimise the burden on trial authors. To inform our risk of bias judgements of incomplete outcome data, we investigated the effects of missing dichotomous data by conducting worst-case analyses based on assessment of plausible event rates.

Assessment of heterogeneity

Heterogeneity between comparable trials was tested using a standard Chi² test and was considered statistically significant at P value < 0.1. When some indication of heterogeneity was noted, from visual inspection of the results or based on results of the Chi² test, we also quantified heterogeneity/inconsistency using the I² statistic (Higgins 2003). This we interpreted as follows: "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% represents considerable heterogeneity" (Section 9.5.2, Higgins 2011).

Assessment of reporting biases

In the event that a meta-analysis of a key outcome includes more than 10 studies in a future update, we will consider generating a funnel plot to explore the potential for publication bias.

Data synthesis

Results of comparable groups of trials were pooled for meta-analysis using the fixed-effect model and 95% CIs. When heterogeneity was present, we viewed the results of the random-effects model and presented these when appropriate.

Subgroup analysis and investigation of heterogeneity

We proposed, if possible, to undertake subgroup analyses by gender, as males are at much greater risk of re-dislocation (Robinson 2006). We also proposed to subgroup by age, using two thresholds: 25 years or younger versus 26 years or older; and 39 years or younger versus 40 years or older. We chose the former threshold because patients aged 25 years or younger are at much greater risk of re-dislocation (Robinson 2006), and the latter because of the markedly increased susceptibility of patients older than 40 years to postimmobilisation stiffness and secondary frozen shoulder (Robinson 2012); these events would be captured by PROMs (one of our primary outcomes) and “other adverse events” (one of our secondary outcomes). To test for differences between subgroups, we intended to inspect the overlap of confidence intervals and to perform the test for subgroup differences available in Review Manager software. We also planned but did not carry out separate outcome analyses of (1) participants who were physically active compared with those who were more sedentary, (2) physically active young adults engaged in highly demanding physical activities who have sustained primary anterior dislocation compared with others, (3) participants with a primary dislocation compared with those with a recurrent dislocation, and (4) participants with a specific lesion resulting from dislocation (e.g. a Bankart lesion) compared with those without. We anticipated that any subgroup differences would be in terms of size of effect (quantitative interaction) rather than direction of effect (qualitative interaction).

Sensitivity analysis

When appropriate, we performed sensitivity analyses to investigate various aspects of trial and review methodology. These included, when data were available, examining the effects of (1) removing trials at high risk of selection bias from inadequate allocation concealment or at high risk of detection bias from lack of blinded outcome assessment; (2) conducting worst-case analyses for trials with missing data; and (3) using fixed-effect versus random-effects models for pooling.

'Summary of findings' table and quality assessment of the evidence

We produced a 'Summary of findings' table for the only comparison tested so far in the review, and we used the GRADE approach to assess the quality of evidence related to each of the key outcomes listed in the Types of outcome measures (see Section 12.2, Schunemann 2011).

Results

Description of studies

Results of the search

Searches were carried out in two stages: August 2005 to October 2012; then October 2012 to September 2013 (see Appendix 1).

For this update, we screened a total of 812 records from the following databases: Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (18 records); Cochrane Central Register of Controlled Trials (58); MEDLINE (145); EMBASE (313); CINAHL (245); PEDro (22); OTseeker (5) and the World Health Organization (WHO) International Clinical Trials Registry Platform (6). We identified one potentially eligible study from handsearches. We reconsidered the nine studies deemed eligible or potentially eligible in the previous version of the review (Handoll 2006). (Results of previous searches, up to September 2005, are shown in Appendix 2.)

We attempted to contact authors of unpublished randomised controlled trials (RCTs) listed in the WHO International Clinical Trials Registry Platform for information on the current status of their respective trials and Kiviluoto 1980 for details of methods. We corresponded with Professor Itoi on his trials (Itoi 2003; Itoi 2007; Itoi 2013), Dr Liavaag on his trial (Liavaag 2011) and Professor Milgrom on his trial (Finestone 2009), and we attempted to contact Drs Miller and Whelan for details of their trials (Miller 2007; Whelan 2008).

In this update, the preliminary report by Itoi 2003 was excluded, although it had been included in the previous version of this review, in favour of the reportedly definitive Itoi 2007. Two newly identified studies were excluded (Chutkan 2012; Whelan 2010), as was Wakefield 2001, which was previously listed as an ongoing study.

Overall, there are now four included studies (Finestone 2009; Itoi 2007; Liavaag 2011; Taskoparan 2010), nine excluded studies (incorporating six carried forward from our previous search), seven studies awaiting classification and three ongoing studies. The results are displayed in a flow chart in Figure 1.

Figure 1.

Study flow diagram

Included studies

Full details of the individual studies are given in the Characteristics of included studies tables.

In total, four parallel, two-group RCTs were identified, including 470 participants (371 male), all with primary traumatic anterior dislocation of the shoulder reduced by various closed methods. Two of these studies were relatively large: Itoi 2007 and Liavaag 2011, with 198 and 188 participants, respectively. Finestone 2009 had 51 participants, and Taskoparan 2010 had 33.

Finestone 2009 was conducted in the Accident and Emergency Department of an Israeli university hospital; Itoi 2007, across 12 hospitals in Japan; Liavaag 2011, in 13 Norwegian hospital emergency departments; and Taskoparan 2010, in an emergency department in Turkey. Three studies evaluated mixed populations; in Finestone 2009, all 51 participants were male and 40 of these were soldiers.

All four studies compared post reduction immobilisation of the affected arm in external rotation ( the arm oriented outwards and the forearm away from the chest) versus immobilisation in internal rotation (the traditional sling arrangement, with the forearm rested across the abdomen). There was only minor variation in the degree of external rotation used (reportedly 10 degrees in Itoi 2007 and Taskoparan 2010, 15 degrees in Liavaag 2011 and 15 to 20 degrees in Finestone 2009) and in the duration of immobilisation (whether internal or external rotation) recommended (three weeks by Itoi 2007, Taskoparan 2010 and Liavaag 2011; four weeks by Finestone 2009). Participants were instructed to remove their immobiliser or sling only for showering.

Primary outcomes
Re-dislocation

Two studies prespecified re-dislocation as an outcome (Finestone 2009; Liavaag 2011). Taskoparan 2010 reported data on re-dislocation but did not prespecify this as an outcome.

Validated patient-reported outcome measures for shoulder instability

One study (Liavaag 2011) incorporated a patient-reported outcome measure for shoulder instability, the Western Ontario Shoulder Instability Index (WOSI), although only non-parametric data were reported.

Resumption of pre-injury activities (yes or no)

Two studies prespecified return to pre-injury sport or activity as an outcome (Itoi 2007; Liavaag 2011).

Secondary outcomes
Participant satisfaction with the intervention

None of the included studies incorporated data on participant satisfaction.

Validated health-related quality of life outcome measures (e.g. EQ-5D, SF-36)

None of the included studies incorporated generic health-related quality of life outcome data.

Any instability: subluxation (separation of the joint not requiring reduction) or subjective instability, either individually or grouped with dislocation as a composite outcome

Itoi 2007 and Liavaag 2011 prespecified re-dislocation or subluxation as a composite outcome, and Liavaag 2011 prespecified subluxation as a stand-alone outcome.

Serious adverse events including persistent pain

Although adverse events were mentioned in all of the reports, the trials did not have a priori strategies for defining or collecting these data.

Other outcomes

Adherence was the only other outcome collected by more than one of the included studies (Finestone 2009; Itoi 2007; Liavaag 2011).

Follow-up time points

Follow-ups were conducted at various time points, but only a two-year (or beyond, i.e. long-term) follow-up was common to Itoi 2007, Finestone 2009 and Liavaag 2011. Taskoparan 2010 did not specify when data were collected on re-dislocation or on adverse events.

Excluded studies

Nine studies were excluded, usually as the result of insufficient information and lack of response from study authors (Harper 2000; Kiviluoto 1980; Staply 2002; Wakefield 2001) or failure to meet our selection criteria (Chutkan 2012; Hovelius 1983; Whelan 2010; Xu 2003). Finally, the relationship between Itoi 2003, which was reported as a preliminary study, to Itoi 2007 is unclear, and, reflecting contradictory information received from the trial investigator, we cannot rule out the possibility that there were trial participants in common. Further details of these studies are given in the Characteristics of excluded studies tables.

Studies awaiting classification

Seven studies await classification. Of these, six were parallel, two-group RCTs comparing immobilisation in external rotation versus immobilisation in internal rotation: We have identified no published full reports related to any of these, and our efforts to contact the corresponding authors were unsuccessful. ISRCTN41070054 (with an initial target sample size of 50 (revised down to 38)), ISRCTN48254181 with a target sample size of 150 and NCT00707018 with a target sample size of 50 are all completed, according to the WHO International Clinical Trials Registry Platform; and long intervals have passed since the respective anticipated end dates of November 2010, May 2008 and June 2009. Seventy-two participants entered the trial by Kelly 2011, attenuating to 54 at one-year follow-up: Results from a two-year follow-up questionnaire were reported as pending in a 2011 report of the Institute of Orthopaedics (UK). Both Miller 2007 and Whelan 2008 are published abstracts, but neither provides sufficient information to stand alone. Miller 2007 reported interim results for 30 participants, but the total sample aimed for or achieved is unknown. The abstract by Whelan 2008 reported randomisation of 61 participants and only 6.5% loss to follow-up after an average of 18 months. Outcomes of the six studies comparing immobilisation in external rotation versus immobilisation in internal rotation included re-dislocation at one and two years (Kelly 2011) or at unspecified time points (ISRCTN41070054; ISRCTN48254181); patient-reported outcome measures for shoulder instability at one year (Kelly 2011; Miller 2007; NCT00707018) and two years (Kelly 2011; NCT00707018); time taken to resume pre-injury sport or other activities (NCT00707018); difficulties with the immobilisation or rehabilitation protocol at three months (Kelly 2011); disease-specific health-related quality of life data at two years (Whelan 2008); any instability at two years (NCT00707018; Whelan 2008); and participant adherence (Kelly 2011; Whelan 2008).

Also awaiting classification is Itoi 2013. This parallel, three-group RCT is published as a full report but compares supplements to a yet unproven method (immobilisation in external rotation). As such, this trial is not eligible for inclusion at this time.

Further details of these studies are given in the Characteristics of studies awaiting classification tables.

Ongoing studies

Three studies, all parallel, two-group RCTs, are ongoing (ACTRN12611001183976; NCT01111500; NCT01648335).

NCT01111500 with a target sample size of 50 and NCT01648335 with an unspecified target sample size are comparing immobilisation in external rotation versus immobilisation in internal rotation. Outcomes include re-dislocation at six months (NCT01648335), one year (NCT01111500) and two years (NCT01111500). The corresponding author of NCT01111500 estimates that her study will be completed in 2014-2015 (personal communication Dr Pelet, 21 January 2013); no corresponding author details are available for NCT01648335, but the entry in ClinicalTrials.gov (accessed through the WHO International Clinical Trials Registry Platform) anticipated an end date of July 2013.

ACTRN12611001183976 is a parallel, two-group RCT with a target sample size of 200 comparing two physiotherapy interventions: strengthening exercises versus pulsed ultrasound and massage, with graduated range of motion exercises common to both groups, after immobilisation in external rotation. The outcome measures are re-dislocation and patient reported outcome measures at six weeks, three months, one year and two years after first-time dislocation. Recruitment is slow, and in a personal communication on 13 November 2012, the corresponding author estimated that completion would take a further 10 years. Because the study is evaluating a supplement to a yet unproven intervention (immobilisation in external rotation), it is unclear how the study will be classified when completed.

Further details of these studies are given in Characteristics of ongoing studies tables.

Risk of bias in included studies

We evaluated risk of bias at study level (although we noted some variation within studies at the level of individual outcomes and therefore will expand the risk of bias tables to allow for this in future iterations). Risk of bias for the seven domains varied across included studies (see Figure 2; Figure 3), although all were at some risk.

Figure 2.

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

Figure 3.

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

Allocation

Selection bias is a consideration as some prediction of prognosis at the point of randomisation, particularly related to age and risk of re-dislocation, is possible and inferences might be drawn concerning individuals’ likely adherence to or acceptability of treatment protocols (Wood 2008). Finestone 2009 was at low risk of selection bias. At high risk were Itoi 2007, which probably lacked allocation concealment, and Taskoparan 2010, as the result of quasi-random sequence generation and lack of allocation concealment. Liavaag 2011, which used sealed envelopes, provided insufficient details to confirm allocation concealment and was judged to be at unclear risk of bias.

Blinding

The risk of performance bias was unclear in Finestone 2009 and Taskoparan 2010 as no information was provided on who applied the splints and gave initial instructions, and whether they were otherwise independent of the study. The risk of bias was high in Itoi 2007 and Liavaag 2011: in Itoi 2007, study co-authors applied the splints and gave initial instructions; and in Liavaag 2011, participants were informed of the preliminary results of Itoi 2007, which had favoured immobilisation in external rotation.

None of the study reports mentioned blinding of assessors. Despite this, Liavaag 2011 was judged to be at low risk of detection bias, because the primary outcome measure, re-dislocation, was clearly defined, and participant-reported re-dislocations were verified from medical records. The remaining studies and the secondary outcomes in Liavaag 2011 (subluxation or dislocation as a combined outcome, or subluxation alone) lacked such verification, leaving them vulnerable to detection bias. In addition, in Itoi 2007, although the primary outcomes (self-reported re-dislocation or subluxation) were each clearly defined, the latter was inevitably subjective; in Finestone 2009, a definition of "recurrent dislocation" was lacking; and in Taskoparan 2010, several outcomes were either susceptible to assessor influence or insufficiently defined to allow a judgement.

Incomplete outcome data

At low risk of attrition bias were Finestone 2009 (no loss to follow-up) and, in terms of its primary outcome, Liavaag 2011 (2.1% loss to follow-up). However, Itoi 2007 and the secondary outcomes of Liavaag 2011 were at high risk of attrition bias: for each, a worst-case scenario sensitivity analysis, assuming that approximately 50% of the dropouts in the external rotation group but none of the dropouts in the internal fixation group had an adverse outcome, resulted in a marked change in results. In Taskoparan 2010, also at high risk, there were large and unexplained losses to follow-up beyond six months.

Selective reporting

Liavaag 2011, the only included study for which a separately published a priori protocol was available, was at low risk of reporting bias, and Finestone 2009 was at unclear risk (although most key outcomes were reported). Itoi 2007 was at high risk (the start point and selection criteria varied across reports and may not have been determined prospectively, with related concerns regarding outcome assessment), as was Taskoparan 2010 (recurrent dislocation was not a prespecified outcome and the mechanisms and timings for gathering these data were not stated).

Other potential sources of bias

All of the reports specified their respective study eligibility criteria, although it was unclear in Itoi 2007 whether these had been prospectively established (see above). All studies had baseline comparability across groups.

Two of the four studies (Finestone 2009; Liavaag 2011) used commercially manufactured external rotation immobilisers from the outset, favouring standardisation, although Finestone 2009 did not specify the manufacturer. Itoi 2007 reported switching from ad hoc to commercially manufactured external rotation splints part way through the trial, with possible implications for positioning, consistency and adherence, and was marked down for this in the "other bias" section of the risk of bias tool. The report of Finestone 2009 included a conflict of interest statement specifying that no benefits had been or would be received from any commercial party. Liavaag 2011 did not include a conflict of interest statement, but their published protocol listed study sponsors and made no mention of the immobiliser's manufacturer. However, Itoi 2007 disclosed potentially substantial disbursements by the manufacturer of their immobilisers, raising the possibility of bias.

Neither Taskoparan 2010, who used ad hoc external rotation splints throughout, nor Itoi 2007 indicated how the intended position was determined. On this basis Taskoparan 2010 was marked "unclear" in the "other bias" section of the risk of bias tool. Itoi 2007 was categorised as "high risk" to account for uncertainties surrounding positioning and, as mentioned above, the switch to a commercially manufactured immobiliser mid-trial and the possibility of commercial influence.

The care programmes (i.e. other than immobilisation and related advice) appear to have been comparable for both groups in all studies.

Effects of interventions

See: Summary of findings for the main comparison Immobilisation in external rotation compared with immobilisation in internal rotation for traumatic anterior dislocation of the shoulder

All four trials compared immobilisation of the affected arm in external rotation versus internal rotation.

We did not organise outcomes within our planned framework of short term (up to and including three months following dislocation), medium term (greater than three months and up to and including 12 months) and long term (longer than 12 months) because all four primary studies reported outcomes immediately after immobilisation (some immobilisation-related adverse events) and over a two-year or longer follow-up period, by which time point > 85% of re-dislocations would be expected to have occurred (Robinson 2006).

The effects of interventions are reported below (also see Summary of findings for the main comparison). The quality of the evidence for all outcomes was judged to be 'very low'; this reflects in part the imprecision of the findings. This imprecision is more evident when considering the broad 95% confidence intervals (CI) for the illustrative absolute effects.

The proposed subgroup analyses were not undertaken. Although two included studies (Itoi 2007; Liavaag 2011) presented outcome data subgrouped by age, these pertained to different outcomes and different age categories and applied to small numbers. No usable subgroup data were presented in the primary studies related to gender, activity levels, recurrent versus first-time dislocations, nor the presence or absence of specific lesions resulting from dislocation (e.g. Bankart lesions).

Primary outcomes

Re-dislocation

Pooled data from three trials (Finestone 2009; Liavaag 2011; Taskoparan 2010) for re-dislocation over a two-year or longer follow-up showed very little difference between the two groups (RR 1.06 favouring internal rotation, 95% CI 0.73 to 1.54; P value 0.77; 252 participants; see Analysis 1.1). In a low-risk population, with an illustrative baseline risk of 247 re-dislocations per 1000, this equates to 15 more (95% CI 67 fewer to 133 more) re-dislocations per 1000 following immobilisation in external rotation. In a medium-risk population, with an illustrative baseline risk of 436 re-dislocations per 1000, this equates to 26 more (95% CI 118 fewer to 235 more) re-dislocations per 1000 following immobilisation in external rotation. A sensitivity analysis in which Taskoparan 2010 (the study at greatest risk of bias) was removed did not prove influential (RR 1.13 favouring internal rotation, 95% CI 0.77 to 1.67; analysis not shown).

Validated patient-reported outcome measures for shoulder instability

One study (Liavaag 2011) reported results at two years of the WOSI, a validated patient-reported outcome measure for shoulder instability, in 174 people. The data were not normally distributed, and therefore analysis was non-parametric. The median score was 238 (interquartile range 101 to 707) in the external rotation group versus 375 (interquartile range 135 to 719) in the internal rotation group. Lower scores are better in the WOSI (maximum score 2100), but the difference was reportedly "not significant (p = 0.32)". No further analysis was possible.

Resumption of pre-injury activities

Two studies (Itoi 2007; Liavaag 2011) addressed different aspects of this outcome.

Itoi 2007 found higher return to sport at any level in the external rotation group over a two-year or longer follow-up among individuals who had sustained their initial injury during sport (RR 1.13, 95% CI 0.87 to 1.48; 109 participants, see Analysis 1.2). In a medium-risk population, with an illustrative baseline risk of 633 participants per 1000 resuming sport at any level, this equates to 82 more (95% CI 82 fewer to 304 more) participants per 1000 resuming sport at any level following immobilisation in external rotation.

Itoi 2007 also reported return to sport at the pre-injury level over a two-year or longer follow-up and found in favour of external rotation, whereas Liavaag 2011 found no difference between the two groups in return to pre-injury level of activity in the same time frame. When pooled, these heterogeneous data (Chi2 was statistically significant at P value 0.09; I2 = 65%) favoured the external rotation group (RR 1.25, 95% CI 0.71 to 2.2; P value 0.43; 278 participants; see Analysis 1.3). In a low -risk population, with an illustrative baseline risk of 204 participants per 1000 returning to their pre-injury level of activity, this equates to 41 more (95% CI 59 fewer to 245 more) participants per 1000 resuming activities at a pre-injury level following immobilisation in external rotation. In a medium-risk population, with an illustrative baseline risk of 605 participants per 1000 returning to their pre-injury level of activity, this equates to 161 more (95% CI 76 fewer to 395 more) participants per 1000 resuming activities at a pre-injury level following immobilisation in external rotation.

Secondary outcomes

Participant satisfaction with the intervention

No studies considered this outcome.

Validated quality of life outcome measures

No studies considered these outcomes.

Any instability: subluxation or subjective instability (either individually or grouped with dislocation as a composite outcome)

Two studies (Itoi 2007; Liavaag 2011) reported "any instability," in the form of subluxation or re-dislocation, with two-year or longer follow-up. Itoi 2007 found in favour of external rotation, whereas Liavaag 2011 found no between-group differences. When pooled, these heterogeneous data (Chi2 was statistically significant at P value 0.06; I2 = 72%) favoured the external rotation group (RR 0.82, 95% CI 0.5 to 1.33; P value 0.42; 322 participants; see Analysis 1.4). In a medium-risk population, with an illustrative baseline risk of 484 instances of any instability per 1000, this equates to 87 fewer (95% CI 242 fewer to 160 more) participants with instability per 1000 following immobilisation in external rotation.

There were losses to follow-up of 20% in Itoi 2007 and 13% in Liavaag 2011 for this outcome, respectively posing "serious" and "intermediate" threats to validity (Straus 2005). A sensitivity analysis was therefore conducted, in which approximately half of losses to follow-up were imputed as events (worst-case scenario) in the external rotation groups but none of these losses were seen in the internal rotation groups. The pooled results of the sensitivity analysis favoured internal fixation (RR 1.06, 95% CI 0.84 to 1.34; see Analysis 1.5) and thus illustrate that these losses to follow-up could pose a serious threat to the validity of the trial findings.

Important adverse events

None of the included studies specified adverse events as a priori outcomes, but all reported them ad hoc. Finestone 2009 reported an axillary rash in two (8%) of 24 participants who had been immobilised in internal rotation. Itoi 2007 reported temporary stiffness, resolving within a month or two with range-of-motion exercises, in six (7%) of 85 participants immobilised in external rotation and "no other problems related to immobilization". However, we judged that none of these temporary adverse events could be rated as "important".

There were three important adverse events reported. Liavaag 2011 reported two "complications" at unspecified time points: one participant (1%) of 91 in the external rotation group had hyperaesthesia and moderate hand pain, and one of 93 (1%) in the internal rotation group had eighth cervical dermatome paraesthesia. The other important event occurred in Taskoparan 2010, in which one participant (6%) of 16 in the internal rotation group had "30° limitation in abduction and 10 degrees in internal rotation in the 6th and 12th months. This patient was 75 years old and had additional rotator cuff problems". Unfortunately, Taskopraran did not clarify whether the limitations referred to were active (as with rotator cuff tear) or active and passive (as with frozen shoulder). Pooled data favoured external rotation (RR 0.61, 95% CI 0.08 to 4.46; 216 participants; see Analysis 1.6). The imprecision of these results is clear from the absolute effect results: In a medium-risk population, these equate to 14 fewer (95% CI three fewer to 128 more) participants with an important adverse event per 1000 following immobilisation in external rotation.

Other outcomes

We intended to note any reports of service utilisation or resource use (e.g. length of hospital stay), outpatient attendance and the provision and nature of physiotherapy; and participants' adherence to their allocated interventions. Only the last of these was addressed by Finestone 2009, Itoi 2007 and Liavaag 2011.

In Finestone 2009, all participants reported "compliance with the protocol" except one (4%) of 27 in the external rotation group, who removed his splint two days early; Finestone 2009 did not consider this a deviation from protocol. In Itoi 2007, only 68 (80%) of 85 in the external rotation group and 50 (68%) of 74 participants in the internal rotation group wore their splints for the full three weeks. Of these, moreover, some only wore their splints part-time. In all, 61 of 85 participants (72%) in the external rotation group and 39 of 74 participants (53%) in the internal rotation group complied with the protocol. Liavaag 2011 reported that 63 of 93 participants (67.7%) in the external rotation group and 45 of 95 participants (47.4%) in the internal group used their splints for 16 hours or longer each day for at least 20 days.

Discussion

Summary of main results

This review now includes four trials with a total of 470 participants, of whom 412 were followed up for at least two years. There was only one comparison: immobilisation in external versus internal rotation for first-time, traumatic anterior dislocation of the shoulder. In relation to this comparison, there were data available for all of our three primary outcomes. Re-dislocation (Finestone 2009; Liavaag 2011; Taskoparan 2010) and resumption of activities (Itoi 2007; Liavaag 2011) enabled pooling for at least some aspects, and a validated patient-reported outcome measure (Western Ontario Shoulder Instability Index: WOSI) was presented by Liavaag 2011 but analysed using non-parametric statistics. Of our secondary outcome measures, the "any instability" outcome was addressed (Itoi 2007; Liavaag 2011), adverse events were reported by all, although on an ad hoc basis, and validated quality of life outcomes and participant satisfaction were not reported at all. No outcome had 95% confidence intervals sufficiently precise to demonstrate clear benefit for either position of immobilisation over the other. For this reason, and other reasons, the quality of the evidence was rated as "very low", meaning that "we are very uncertain about the estimate" (Summary of findings for the main comparison).

The main results of this review, therefore, are twofold. First, robust evidence for superiority of immobilisation in external rotation over traditional immobilisation in a sling (internal rotation) is lacking, and there is no justification for recommending any change in current clinical practice. Second, we found no includable evidence related to our other objectives.

Overall completeness and applicability of evidence

Despite our reasonably comprehensive search, we located very few potentially eligible studies and included only four. The shortage of randomised controlled trials for conservative treatment parallels that concerning the treatment of proximal humeral fractures (Handoll 2003) and the comparison of conservative versus surgical treatment for anterior dislocation of the shoulder (Handoll 2004), and is not unexpected.

Three of the included studies (Itoi 2007; Liavaag 2011; Taskoparan 2010) consisted of mixed-gender, general populations; while Finestone 2009 included males only (whether by chance or intention is unclear), most of whom were military recruits. There was a wide age range (12 to 90 years) overall, though Liavaag 2011 imposed a 40-years upper age limit and the age range in Finestone 2009 (17 to 27 years) reflected the special population from which their participants were recruited. Overall, the young and active as well as the relatively sedentary are represented.

We located three other trials, presently described in the Characteristics of ongoing studies tables, whose results should appear in subsequent updates of this review. Two of these (combined target sample size of at least 50) compare immobilisation in external versus internal rotation, and one (target sample size 200) compares two types of physiotherapy after immobilisation in external rotation. Another seven studies are described in Characteristics of studies awaiting classification tables. Six of these (combined target sample size of at least 383) compare immobilisation in external rotation versus immobilisation in internal rotation. The remaining study, in press at the time of writing, compares different durations of shoulder restriction (zero, three or six weeks) by a band after immobilisation in external rotation (Itoi 2013).

Thus the included, ongoing and unclassified studies reveal an exclusive emphasis in research to date on immobilisation in external rotation, usually as a comparator against immobilisation in internal rotation, but occasionally as the standard immobilisation in studies comparing different post immobilisation regimens. No studies addressed other interventions of interest, involving comparisons of different durations of immobilisation, rehabilitation versus no rehabilitation and different variants of rehabilitation. These other issues remain relevant, as does the question of whether immobilisation is necessary at all for older people, who are at a much reduced risk of recurrence but at greater risk of shoulder stiffness (de Boer 2005). It is notable that but for the discovery of a report (published in 1999) that provides a full description of the method of treatment allocation, one well-cited trial (Hovelius 1983) testing duration of immobilisation might have been included (see Characteristics of excluded studies).

Also, some key outcomes are lacking in the primary research. Among the important outcomes that have not been directly addressed is participant satisfaction, including the acceptability of allotted interventions. In several of the included studies, participants' adherence to the protocol has served as a proxy for the latter aspect, but the atypical (mainly military) population in Finestone 2009 and the high risk of performance bias in Itoi 2007 and Liavaag 2011 confound this. Intuitively, one might expect maintenance of external rotation at the shoulder to be cumbersome. Additionally, one might reasonably expect a heightened risk of accidents. Yet no a priori strategies are in place to collect these or other adverse event data. Other important omissions are very long-term outcomes, notably persistent pain and arthropathy, as well as quality of life, using validated measures, at all time points.

The Cochrane review comparing surgical versus non-surgical treatment for anterior dislocation of the shoulder found some evidence from randomised controlled trials to support primary surgery in young adults (usually male) engaged in highly demanding physical activities who have sustained their first acute traumatic shoulder dislocation (Handoll 2004, updated in 2009). No evidence was available for other categories of participants. The review suggested that randomised controlled trials comparing good standard surgical intervention versus good standard conservative treatment for primary anterior shoulder dislocation are needed and anticipated that the present review of different conservative interventions including rehabilitation should provide important information in this area (Handoll 2004). The most important finding, however, of our review is the lack of good quality evidence to permit any recommendations for practice or to inform on the best comparator for trials of surgical versus non-surgical management. Conclusive evidence on the efficacy of immobilisation in external rotation would inform this but is presently lacking. In this light, it is a frustrating limitation of the present review that six completed studies comparing external versus internal rotation are classified as "awaiting classification".

Quality of the evidence

In aggregate, the quality of the evidence was "very low", mostly reflecting limitations in study design (all outcomes) and imprecision of point estimates (all outcomes) as well as inconsistency ("resumption of sport/activities at pre-injury level, with two-year or longer follow-up" and "any instability") and indirectness ("important adverse events"). This grading means that we are very uncertain about the estimates of effect (Summary of findings for the main comparison). A universally ad hoc approach to collecting data on adverse events hampers assessment of the potential for harms. In addition, poor reporting of how the desired position of external rotation was achieved in locally manufactured splints (Itoi 2007, Taskoparan 2010) and of the manufacturer's details of commercial splints (Finestone 2009) undermines the applicability of the results.

The previous version of our review (Handoll 2006) contained only one study (Itoi 2003). Itoi 2013 was a preliminary report whose findings favoured immobilisation in external rotation, although not statistically significantly so. We reported: "[this study] does not provide the robust and sufficient evidence required to establish clinical benefit. As well as starting off using quasi-randomised methods and then insufficient methods to guarantee allocation concealment, the outcome assessment of Itoi 2003 is notably flawed. For instance, trials should measure the recovery to pre-injury levels of all participants and actively record recurrence at set times. Holding the arm in external rotation is more hazardous and awkward for people and so some better monitoring of acceptability, adherence and accidents should also be done. Particularly given that external rotation immobilisers are now available on the commercial market it is important that good quality randomised controlled trials of external rotation versus internal rotation are conducted by independent investigators."

In the present version of the review, Itoi 2003 has been replaced by Itoi 2007. This was a comparatively large, and potentially influential, study of a sample reportedly distinct from Itoi 2003 (personal communication with Prof Itoi, 8 April 2011), with some methodological improvements over its predecessor. Thus there was randomisation, return to activity was evaluated in all participants and there was active recording of outcomes at or after two years, as well as reporting on adherence and adverse events. On the other hand, the start point and the selection criteria are inconsistent across reports (Itoi 2003; Itoi 2004; Itoi 2007) and may not have been determined prospectively, leading us to downgrade the study to an "unclear risk" of reporting bias. Furthermore, allocation does not appear to have been concealed (tending to cancel any benefit of randomisation). There were also serious problems with losses to follow-up, such that the favourable results were not robust to sensitivity analysis; there was vulnerability to performance bias and detection bias; and the risk of reporting bias was unclear.

Of the three other studies included, Taskoparan 2010 was very compromised methodologically, and its results should be viewed with particular circumspection. For this study, performance bias, attrition bias and other bias were rated "unclear risk" and all else "high risk". This was a small study, however, and the imprecision of its results meant that it was not influential in any meta-analysis. Finestone 2009 and Liavaag 2011 were somewhat stronger methodologically. Both reported adequate randomisation, although allocation concealment was unclear in Liavaag 2011. But neither had a low risk of performance bias (Finestone 2009: unclear risk; Liavaag 2011: high risk), and only one (Liavaag 2011) had a low risk of detection bias (Finestone 2009: high risk). Finestone 2009 had a low risk of attrition bias, however, as did Liavaag 2011 for its primary outcome. The combined re-dislocation or subluxation outcome in Liavaag 2011 was at high risk of attrition bias, but, as a secondary outcome, this did not affect the "low risk" grading given using the 'Risk of bias' tool. Liavaag 2011 was at low risk of reporting bias and Finestone 2009, unclear risk.

Potential biases in the review process

We strove throughout the development of this review to minimise any potential for biases. Thus, when feasible, we adhered throughout to the detailed a priori protocol, which detailed every aspect of the review's aims, objectives and methods, and noted any necessary changes in Differences between protocol and review. We conducted a reasonably comprehensive search without language restriction, and the key processes of study selection, data extraction and risk of bias evaluation were each independently performed by at least two review authors.

We contacted study authors for missing data and, when appropriate, methodological information. However, it has been shown empirically that unclear reporting of allocation concealment is itself associated with inflated effect estimates for subjective outcomes (Schulz 1995; Wood 2008); therefore, disregarding such instances would have meant ignoring a source of bias. We therefore made judgements on the satisfactoriness of allocation concealment based on the description given in the study report, narratively interpreting the grading given in relation to specific outcomes (i.e. subjective or objective) and did not contact authors for clarification of safeguards.

As previously indicated (see Overall completeness and applicability of evidence), we recognise the large number of studies awaiting classification as an unavoidable limitation of our review.

Agreements and disagreements with other studies or reviews

We identified three systematic reviews (Gibson 2004; Paterson 2010; Smith 2006) evaluating aspects of conservative care following closed reduction, that is, whether to immobilise, the position in which to immobilise, how long to immobilise and how to rehabilitate, in various combinations. The outcomes of interest were any outcomes (Smith 2006), recurrent dislocation (Gibson 2004; Paterson 2010), resumption of pre-dislocation activity (Gibson 2004; Smith 2006) and functional outcomes (Gibson 2004). All differed from our own methodologically in the inclusion of observational as well as experimental study designs.

Gibson 2004 concluded that "weak evidence" favoured immobilisation for three weeks over one week in terms of recurrence and functional outcomes, although those immobilised for one week returned to work sooner. However, among the studies supporting this was only one (Kiviluoto 1980) that was reportedly experimental, and this had been excluded by ourselves in this review. The conclusions of Smith 2006 were similar to our own, that is, it is unclear whether shoulders should be immobilised after anterior dislocation, for how long and whether such immobilisation should be in internal or external rotation. More optimistic were Paterson 2010, who concluded, "there is no benefit of conventional sling immobilisation for longer than one week ... in younger patients", but based this upon Hovelius 1983 and its follow-ups and Kiviluoto 1980; both studies were excluded from our review. Paterson 2010 also concluded that, "bracing in external rotation may provide a clinically important benefit over traditional sling immobilisation, but the difference in recurrence rates did not achieve significance with the numbers available". We would downplay this further, based on the results of our own review, for two reasons. First, our review is more current. It includes Liavaag 2011, a study of moderately high quality that failed to replicate the encouraging results of Itoi 2003 and Itoi 2007, both of which were at much greater risk of bias. Second, we believe that our review is more robust, as it excludes Itoi 2003, thereby avoiding the potential risk of including data from duplicate populations, and it identifies the potential for high risk of bias of Itoi 2007 by recognising that the start point and the selection criteria of this trial vary across reports and may not have been determined prospectively, raising a question of selective reporting bias.

Authors' conclusions

Implications for practice

All available evidence concerns comparison of immobilisation in external rotation, a relatively novel technique, versus immobilisation in internal rotation, the traditional approach to management following closed reduction of anterior dislocation of the shoulder. There is no evidence that either approach is more effective than the other, and none of the included studies collected data on adverse events with sufficient rigour to enable proper assessment of potential harms. There is no justification for recommending that practice be changed.

In general, there is a lack of evidence from randomised controlled trials to inform the choice of conservative management following closed reduction of traumatic dislocation of the shoulder.

Implications for research

There remains a need for sufficiently powered, high quality and appropriately reported (i.e. CONSORT-compliant) randomised controlled trials of key conservative treatment and rehabilitation options for traumatic anterior shoulder dislocation. Blinding of interventions is not easy to do, but concealed allocation should always be done and, when possible, blinded outcome measurement should be done, as these would improve the quality and validity of future results.

Trials should include evaluation of the use and duration of immobilisation (for example, one week versus three weeks) and timing for resumption of activities that may provoke a recurrence. For these trials, we suggest that some distinction should be made between young adults engaged in highly demanding physical activities, who are at greater risk of recurrence, and older adults who have lower risk of recurrence but are likely to have pre-existing degenerative disease of the rotator cuff.

Attention should be given to outcome assessment in future trials. The use of well-defined and validated functional outcome measures, including participant-derived quality of life measures, is preferable, and individual participant data should be made available. Other important outcomes include participant satisfaction and adverse events. Collection of adverse events data ad hoc, in the absence of pre-defined definitions or strategies for data collection, is a serious flaw that casts doubt on the completeness of data and hampers comparison of risks and benefits. In future studies, key categories of potential adverse events should be established prospectively, and an active strategy should be put in place for collecting these data. Categories should be reported on even if they contain no events.

To ensure that efforts are not duplicated, researchers should consult trial registries before planning further research in this area. This particularly relates to the comparison between immobilisation in external versus internal rotation, along with research that is complementary to these; in these areas, pressing priorities include completion of ongoing work and publication of completed work.

Acknowledgements

We are very grateful to Lesley Gillespie for developing the search strategy for this review, for searching the Cochrane Bone, Joint and Muscle Trauma Group's Specialised Register and for providing help during editorial processing; to Joanne Elliott for updating the search; and to Lindsey Elstub, Diane Horsley, Laura MacDonald and Helen Handoll for providing editorial support. We thank Mario Lenza and Piet de Boer for providing editorial comments. We additionally thank Helen Handoll for her contribution as lead author in the previous version of this review.

We thank Eiji Itoi, Sigurd Liavaag, Charles Milgrom, Stéphane Pelet, Alison Wakefield and Timothy Walker for providing information on their trials, Mavis Luya for chasing up information on another trial and Yan Gong and Mingming Zwang for translating the methods of a potentially eligible study that was published in Chinese.

Data and analyses

Download statistical data

Comparison 1. Immobilisation in external versus internal rotation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Re-dislocation,with two-year or longer follow-up3252Risk Ratio (M-H, Fixed, 95% CI)1.06 [0.73, 1.54]
2 Resumption of sport at any level, with two-year or longer follow-up1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3 Resumption of sport/activities at pre-dislocation level, with two-year or longer follow-up2278Risk Ratio (M-H, Random, 95% CI)1.25 [0.71, 2.20]
3.1 Resumption of sport at the pre-dislocation level (subgroup who dislocated during sport)1109Risk Ratio (M-H, Random, 95% CI)1.80 [0.94, 3.43]
3.2 Resumption of pre-injury activities (all participants)1169Risk Ratio (M-H, Random, 95% CI)1.02 [0.80, 1.29]
4 Any instability (subluxation or re-dislocation), with two-year or longer follow-up2322Risk Ratio (M-H, Random, 95% CI)0.82 [0.50, 1.33]
5 Any instability (subluxation or re-dislocation), with two-year or longer follow-up: 'worst-case' sensitivity analysis2386Risk Ratio (M-H, Fixed, 95% CI)1.06 [0.84, 1.34]
6 Important adverse events2216Risk Ratio (M-H, Fixed, 95% CI)0.61 [0.08, 4.46]
Analysis 1.1.

Comparison 1 Immobilisation in external versus internal rotation, Outcome 1 Re-dislocation,with two-year or longer follow-up.

Analysis 1.2.

Comparison 1 Immobilisation in external versus internal rotation, Outcome 2 Resumption of sport at any level, with two-year or longer follow-up.

Analysis 1.3.

Comparison 1 Immobilisation in external versus internal rotation, Outcome 3 Resumption of sport/activities at pre-dislocation level, with two-year or longer follow-up.

Analysis 1.4.

Comparison 1 Immobilisation in external versus internal rotation, Outcome 4 Any instability (subluxation or re-dislocation), with two-year or longer follow-up.

Analysis 1.5.

Comparison 1 Immobilisation in external versus internal rotation, Outcome 5 Any instability (subluxation or re-dislocation), with two-year or longer follow-up: 'worst-case' sensitivity analysis.

Analysis 1.6.

Comparison 1 Immobilisation in external versus internal rotation, Outcome 6 Important adverse events.

Appendices

Appendix 1. Search strategies (2005 to present)

Please note the searches were carried out in 2 stages:
First search: August 2005 to October 2012.
Supplementary search: October 2012 to September 2013.

The Cochrane Library (Wiley Online Library)

#1 MeSH descriptor: [Shoulder Dislocation] this term only  (88)    
#2 (shoulder* near dislocat*) or (shoulder* near sublux*) or (shoulder* near instability) or (shoulder* near unstable):ti,ab,kw (Word variations have been searched)  (201)    
#3 (glenohumeral next joint) or (glenohumeral next instability) or (glenohumeral next unstable):ti,ab,kw (Word variations have been searched)  (56)    
#4 (lesion* near/3 hill next sachs) or (lesion* near/3 bankart):ti,ab,kw (Word variations have been searched)  (20)    
#5 #1 or #2 or #3 or #4   (246)          
#6 immobilis* or immobiliz* or sling* or conservativ* or therap* or physiother* or rehabilitat* or mobilis* or mobiliz* or train* or exercis* or (physical next therap*):ti,ab,kw (Word variations have been searched)  (241953)    
#7 MeSH descriptor: [Exercise] this term only  (8288)    
#8 MeSH descriptor: [Rehabilitation] 3 tree(s) exploded  (12120)    
#9 Any MeSH descriptor with qualifier(s): [Rehabilitation - RH, Therapy - TH] in all MeSH products  (67572)    
#10 #6 or #7 or #8 or #9   (270647)          
#11 #5 and #10  (137)          
#12 SR-muskinj  (7174)          
#13 #11 not #12   (57) [trials]

Supplementary search (2012 to Sept 2013):
#13 #11 not #12 (1) [trials]

MEDLINE (Ovid Web)

1 Shoulder Dislocation/ (4573)
2 (shoulder$ adj3 (dislocat$ or sublux$ or instability or unstable)).tw. (3877)
3 (glenohumeral adj (joint or instability or unstable)).tw. (1843)
4 (lesion$1 adj (Hill Sachs or Bankart)).tw. (10)
5 or/1-4 (7063)
6 (immobilis$ or immobiliz$ or sling$ or conservativ$ or therap$ or physiother$ or rehabilitat$ or mobilis$ or mobiliz$ or train$ or exercis$ or physical therap$).tw. (2163712)
7 Exercise/ or exp Rehabilitation/ (197506)
8 (rh or th).fs. (1477224)
9 or/6-8 (3235953)
10 and/5,9 (2310)
11 Randomized controlled trial.pt. (339605)
12 Controlled clinical trial.pt. (85425)
13 randomized.ab. (242346)
14 placebo.ab. (135629)
15 Drug Therapy.fs. (1578525)
16 randomly.ab. (173863)
17 trial.ab. (251056)
18 groups.ab. (1136962)
19 or/11-18 (2939707)
20 exp Animals/ not Humans/ (3797751)
21 19 not 20 (2497281)
22 10 and 21 (268)
23 (200508* or 200509* or 200510* or 200511* or 200512* or 2006* or 2007* or 2008* or 2009* or 2010* or 2011* or 2012*).ed. (5375215)
24 22 and 23 (128)

Supplementary search (Oct 2012 to Sept 2013):
24 22 and 23 (17)

EMBASE (Ovid Web)

1 Shoulder Dislocation/ or Bankart Lesion/ (4516)
2 (shoulder$ adj3 (dislocat$ or sublux$ or instability or unstable)).tw. (4648)
3 (glenohumeral adj (joint or instability or unstable)).tw. (2234)
4 (lesion$1 adj (Hill Sachs or Bankart)).tw. (11)
5 or/1-4 (8191)
6 (immobilis$ or immobiliz$ or sling$ or conservativ$ or therap$ or physiother$ or rehabilitat$ or mobilis$ or mobiliz$ or train$ or exercis$ or physical therap$).tw. (2855947)
7 (rh or th).fs. (1359978)
8 Conservative Treatment/ or Physiotherapy/ or exp Exercise/ or Rehabilitation/ (287688)
9 or/6-8 (3834171)
10 and/5,9 (2810)
11 exp Randomized Controlled Trial/ (331217)
12 exp Double Blind Procedure/ (111497)
13 exp Single Blind Procedure/ (16539)
14 exp Crossover Procedure/ (35309)
15 Controlled Study/ (3890318)
16 or/11-15 (3969729)
17 ((clinical or controlled or comparative or placebo or prospective$ or randomi#ed) adj3 (trial or study)).tw. (652284)
18 (random$ adj7 (allocat$ or allot$ or assign$ or basis$ or divid$ or order$)).tw. (159069)
19 ((singl$ or doubl$ or trebl$ or tripl$) adj7 (blind$ or mask$)).tw. (148185)
20 (cross?over$ or (cross adj1 over$)).tw. (63302)
21 ((allocat$ or allot$ or assign$ or divid$) adj3 (condition$ or experiment$ or intervention$ or treatment$ or therap$ or control$ or group$)).tw. (199932)
22 or/17-21 (974005)
23 or/16,22 (4451730)
24 limit 23 to human (2701949)
25 and/10,24 (488)
26 (2005* or 2006* or 2007* or 2008* or 2009* or 2010* or 2011* or 2012*).em. (7902638)
27 25 and 26 (288)

Supplementary search (Oct 2012 to Sept 2013):
27 25 and 26 (25)

CINAHL (EBSCO)

S1  (MH "Shoulder Dislocation")  (725)
S2  TI ( (shoulder N3 dislocat*) or (shoulder N3 sublux*) or (shoulder N3 instability) or (shoulder N3 unstable) ) OR AB ( (glenohumeral N3 joint) or (glenohumeral N3 instability) or (glenohumeral N3 unstable) )  (1126)
S3  TI ( (glenohumeral N3 joint) or (glenohumeral N3 instability) or (glenohumeral N3 unstable) ) OR AB ( (glenohumeral N3 joint) or (glenohumeral N3 instability) or (glenohumeral N3 unstable) )  (591)
S4  TI ( (lesion* N3 Hill Sachs) or (lesion* N3 Bankart) ) OR AB ( (lesion* N3 Hill Sachs) or (lesion* N3 Bankart) )  (127)
S5  S1 or S2 or S3 or S4  (1618)
S6  TI ( (immobilis* or immobiliz* or sling* or conservativ* or therap* or physiother* or rehabilitat* or mobilis* or mobiliz* or train* or exercis* or physical therap*) ) OR AB ( (immobilis* or immobiliz* or sling* or conservativ* or therap* or physiother* or rehabilitat* or mobilis* or mobiliz* or train* or exercis* or physical therap*) )  (371609)
S7  (MH "Rehabilitation+")  (153920)
S8  MW rh OR MW th  (336486)
S9  S6 or S7 or S8  (683613)
S10  S5 and S9  (731)
S11  (MH "Clinical Trials+")  (149457)
S12  (MH "Evaluation Research+")  (18850)
S13  (MH "Comparative Studies")  (68744)
S14  (MH "Crossover Design")  (9756)
S15  PT Clinical Trial  (73931)
S16  (MH "Random Assignment")  (33559)
S17  S11 or S12 or S13 or S14 or S15 or S16  (240443)
S18  TX ((clinical or controlled or comparative or placebo or prospective or randomi?ed) and (trial or study))  (412564)
S19  TX (random* and (allocat* or allot* or assign* or basis* or divid* or order*))  (58804)
S20  TX ((singl* or doubl* or trebl* or tripl*) and (blind* or mask*))  (640484)
S21  TX ( crossover* or 'cross over' ) or TX cross n1 over  (12256)
S22  TX ((allocat* or allot* or assign* or divid*) and (condition* or experiment* or intervention* or treatment* or therap* or control* or group*))  73797 S23  S18 or S19 or S20 or S21 or S22  (974727)
S24  S17 or S23  (1034060)
S25  S10 and S24  (324)
S26  EM 2005 OR EM 2006 OR EM 2007 OR EM 2008 OR EM 2009 OR EM 2010 OR EM 2011 OR EM 2012  (2365064)
S27  S25 and S26  (208)

Supplementary search (Oct 2012 to Sept 2013):
S27 S25 and S26 (37)

Appendix 2. Previous searches (to September 2005)

We identified nine studies, only one (Itoi 2003) of which was included in the previous version of this review. Five studies were excluded, three of which are reported in full (Hovelius 1983; Kiviluoto 1980; Xu 2003) and two of which (Harper 2000; Staply 2002) were listed only in the National Research Register UK, and are now presumed abandoned. A report for Wakefield 2001 was apparently in preparation (personal communication, 25 June 2005), but this trial is now also presumed abandoned (see Characteristics of excluded studies table). Details of Pimpalnerkar 2008 (ISRCTN48254181) and Itoi 2006, which is now in press (Itoi 2013), are provided in the Characteristics of ongoing studies table.

What's new

DateEventDescription
1 March 2014New search has been performed

For this version of the review, published in 2014, Issue 4, the following changes were made.

1. The search was updated to September 2013.
2. Four new trials (Finestone 2009; Itoi 2007; Liavaag 2011; Taskoparan 2010) were included, and one previously included trial (Itoi 2003) was excluded.
3. In accordance with the policy of The Cochrane Collaboration, the review was updated to include new methodology, including assessment of risk of bias.
4. A 'Summary of findings' table was added.

1 March 2014New citation required and conclusions have changed

Although the review conclusions still point to an overall lack of evidence to inform practice, the inclusion of four new trials, all of which compared immobilisation in external rotation versus immobilisation in the traditional position of internal rotation, brought a specific focus to the conclusions.

Changes were made to the byline, with one new review author and removal of two previous review authors.

History

Protocol first published: Issue 4, 2004
Review first published: Issue 1, 2006

DateEventDescription
5 September 2008AmendedConverted to new review format.

Contributions of authors

The review update was developed by Nigel Hanchard (NH), Lorna Goodchild (LG) and Lucksy Kottam (LK). NH contacted authors of eligible trials to inform choices for study inclusion and to ask for extra information. All review authors performed trial selection, quality assessment and data extraction. NH compiled the first draft of the review and received critical feedback from LG and LK. Nigel Hanchard is the guarantor of the review.

Declarations of interest

Nigel CA Hanchard: none known
Lorna M Goodchild: none known
Lucksy Kottam: none known

Sources of support

Internal sources

  • University of Teesside, Middlesbrough, UK.

External sources

  • The Physiotherapy Research Foundation of the Chartered Society of Physiotherapy Charitable Trust (registered charity 279882), UK.

Differences between protocol and review

In line with recommendations of The Cochrane Collaboration, the review now assesses risk of bias.

For this version of the review, outcomes were revised and, to comply with current guidance of The Cochrane Collaboration, the number of primary outcomes was reduced. We used GRADE to judge the quality of the evidence and have incorporated a summary of findings table: Summary of findings for the main comparison.

As indicated in the Results, we anticipated organising outcomes within a framework of short term (up to and including three months following dislocation), medium term (greater than three months and up to and including 12 months) and long term (longer than 12 months). Imposition of this framework was not necessary because all included studies reported outcomes in a mutually compatible framework: during or shortly after immobilisation (some adverse events) or up to or beyond two years post dislocation.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Finestone 2009

Methods

Aim: "To ascertain whether immobilisation after primary traumatic anterior dislocation of the shoulder in external rotation was more effective than immobilisation in internal rotation in preventing recurrent dislocation in a physically active population"

Method of randomisation: computer-generated randomised list

Assessor blinding: no

Intention-to-treat analysis: no information

Loss to follow-up: none

Participants

Setting: Accident and Emergency Department of a University hospital, Israel

Period of data collection: 1 January 2004 to 31 December 2006 (additional information provided by contact author)

Participants: 51

Inclusion criteria: primary anterior dislocation of the shoulder; informed consent. It is unclear whether the exclusive recruitment of males was by chance or by intention

Exclusion criteria: injury sustained in a motor vehicle accident; concurrent fracture of the greater tuberosity

Sex: 51 male

Age: mean 20 years, range 17 to 27 years

Pre-injury status: 22/27 (81.5%) in Group A and 18/24 (75%) in Group B were soldiers (additional information provided by contact author)

Assigned: 27/24 (Group A/B)

Assessed: 27/24 (at final follow-up)

Interventions

Group A: immobilisation in 15 to 20 degrees of shoulder external rotation in a manufactured device (manufacturer's details not given), except when showering or changing clothes. No information was given on who applied the splint or gave the accompanying instructions

Group B: immobilisation in an internal rotation brace with a component encircling the lower arm to maintain elbow adduction, except when showering or changing clothes. No information was given on who applied the splint or gave the accompanying instructions

Both groups: after four weeks' immobilisation, participants began a standard regimen of physiotherapy (no further details were given). Participants with clinically stable shoulders on the basis of a negative supine apprehension test returned to full activity after three months

Outcomes

Reviews were conducted at two weeks, six weeks, three months, six months, and one, two, three and four years post injury

(Outcomes were not prespecified as primary or secondary)

Re-dislocation: it is unclear whether this was verified by reference to the medical notes, but this is likely to have been so, given the population in question. Identified re-dislocators were further evaluated: "Those who sustained re-dislocations were evaluated by either CT or MRI arthrography"

Adverse events (other than need for surgical intervention)

Adverse events (need for surgical intervention)

Adherence

Notes

At injury, 5/27 (18.5%) in Group A and 5/24 (20.8%) in Group B had sustained axillary nerve neuropraxia. This resolved within 10 weeks in all cases. There were no fractures of the glenoid rim.

Closed reduction was performed in the Accident and Emergency Department by the Milch technique (n = 41) or the Hippocratic method (n = 10; 7/27 = 25.9% in Group A and 3/24 = 12.5% in Group B (additional information provided by contact author))

Re-dislocations: these occurred at a mean of 13.8 months (range 4 to 43 months) after the first in Group A, and 12.4 months (range 4 to 46 months) after the first in Group B. Redislocators were evaluated by CT or MRI arthrography. 19/20 (95%) were found to have a Bankart lesion, one of which was bony

Adverse events (other than surgery): 2 participants in the internal rotation group developed axillary rashes (no further details were given)

Adverse events (surgery): surgery was undertaken in 2/27 (7.4%) in Group A and 4/24 (16.7%) in Group B. All participants who underwent surgery were soldiers (additional information provided by contact author)

Adherence: all participants reported adherence to the protocol except one in Group A, who removed his splint two days early. This was not counted as a protocol violation

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"We aimed to have 55% of the patients in the external rotation brace ... we made up prior to the study 100 patient booklets numbered 001 to 100. Each booklet had a sealed envelope which contained either a slip for external or internal rotation bracing. The slips were placed into their respective envelopes according to a computer generated randomised list based on the file number" (additional information provided by contact author)
Allocation concealment (selection bias)Low riskSee above
Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskIt is unclear who applied the splints and who gave initial instructions, and whether they were otherwise independent of the study, raising the possibility of performance bias
Blinding of outcome assessment (detection bias)
All outcomes
High riskThere is no mention of blinding, and the outcome of participant-reported "recurrent dislocation" is not explicitly defined. No mention is made of whether self-reported re-dislocations were verified from the medical notes
Incomplete outcome data (attrition bias)
All outcomes
Low riskParticipants appear to have been followed up to four years or the point of re-dislocation, whichever was sooner. "No patient was lost to follow up" (additional information provided by contact author)
Selective reporting (reporting bias)Unclear riskNo protocol is available, but most key outcomes were reported. The numbers with clinical instability at three months, based on the supine instability test, were not reported. Group data on the numbers of participants requiring surgery were not given in the paper but were provided on request by the contact author
Other biasLow riskNone identified. There is a conflict of interest statement ruling out involvement of any commercial organisation

Itoi 2007

Methods

Aim: "To report the two-year results of a study testing the hypothesis that immobilization in external rotation versus internal rotation would reduce the recurrence rate after initial traumatic anterior dislocation of the shoulder"

Method of randomisation: random number table

Assessor blinding: no

Intention-to-treat analysis: yes

Loss to follow-up: 39/198 (19.7%), comprising 19/104 (18.3%) in Group A and 20/94 (21.3%) in Group B

Participants

Setting: Akita Hospital and 11 other institutes across Japan

Period of data collection: October 2000 to March 2004

Participants: 198

Inclusion criteria: initial anterior dislocation caused by a substantial traumatic event; presentation within three days (including the day of dislocation); no associated fracture visible on routine radiographic examination; informed consent

Exclusion criteria: No information was given

Sex: 136 male

Age: mean 37 years, range 12 to 90 years

Subgroups:

By age < 20 years; 21 to 30 years; < 30 years; 31 to 40 years; > 41 years

By interval between dislocation and immobilisation: 0 days; 1 day; 2 days

By age and interval between dislocation and immobilisation: age < 30 years and interval 0 days; age < 30 years and interval 1 day; age < 30 years and interval 2 days

Pre-injury status: 71/104 (68.3%) in Group A and 64/94 (68.1%) in Group B sustained their injury in sport. Among participants aged < 30 years, 53/56 (94.6%) in Group A and 39/42 (92.3%) in Group B sustained their injury in sport

Assigned: 104/94 (Group A/B)

Assessed: 85/74 (at final follow-up)

Interventions

Group A: immobilisation in 10 degrees of external rotation, except when showering. Until November 2003, this was achieved using a wire-mesh splint covered with sponge and stockinette (see Itoi 2003 for details). Starting in November 2003, a prototype manufactured brace was used (Alcare, Tokyo, Japan). Both methods appeared to limit wrist mobility. The treating surgeons (possibly authors of the report) applied the splintage and gave the accompanying instructions

Group B: immobilisation in a sling and swathe except when showering. The treating surgeons (possibly authors of the report) applied the splintage and gave the accompanying instructions

Both groups: immobilisation commenced between zero and two days post dislocation and was continued for three weeks. Afterwards, rehabilitation comprised self-directed active and passive range-of-motion exercises. Avoidance of vigorous sports activities for at least three months was advised

Outcomes

Follow-up reviews were conducted at six months, one year and two years, face-to-face or by telephone

Prespecified as the primary outcome

Self-reported re-dislocation or subluxation: dislocation was defined as the humeral head being completely out of the glenoid until a reduction manoeuvre was performed or subluxation.

Implicitly secondary outcomes

Return to pre-injury sport: at any level (subgroup who sustained injury in sport)

Return to pre-injury sport: at pre-injury level (subgroup who sustained injury in sport)

Return to pre-injury sport: at any level (subgroup aged < 30 who sustained injury in sport)

Return to pre-injury sport: at pre-injury level (subgroup aged < 30 who sustained injury in sport)

Adherence: use of splint full- or part-time for full three weeks

Adeherence: use of splint full-time for full three weeks

Notes

Relationship between this and previous reports:

The full trial report of Itoi 2003 indicated that this was a preliminary study. A subsequent (2004) abstract stated that it was an interim report of an ongoing trial, referred to the random assignment of 96 participants and reported results for 80 participants. The selection criteria had changed in that participants with recurrent dislocations were now included, but humeral fractures were explicitly excluded, as were shoulders not immobilised within three days of injury. The mean follow-up time in the 2004 abstract was 12.4 months.

Correspondence with Prof Itoi (July 2005) revealed that the trial was still ongoing. Results for 131 participants, all with initial dislocations, were reported at an Instructional Course Lecture (Principles and procedures for shoulder instability: an international perspective: non-operative management of anterior instability) at the meeting of the American Academy of Orthopaedic Surgeons in 2005 (handout provided to authors by Prof Itoi)

Itoi 2007 reports results for 198 participants with initial dislocation, recruited from October 2000, and correspondence with Prof Itoi (8 April 2011) indicates that this study sample does not incorporate that of Itoi 2003. However, the start point and selection criteria of Itoi 2007 may not have been determined prospectively, contributing to the grading of “unclear” in the “selection bias” section of the risk of bias table

Method of reduction: manual reduction was by the elevation method (101 shoulders, 57 in Group A), the Hippocratic method (22 shoulders, 13 in Group A), the external rotation method (17 shoulders, 8 in Group A), the Kocher method (16 shoulders, 6 in Group A), the Stimson method (14 shoulders, 8 in Group A) and other methods (28 shoulders) (additional information provided by contact author)

External rotation brace: adoption of the Alcare brace, which was easier to don and doff than the wire-mesh splint, may have influenced adherence, as well as positioning and standardisation, part way through the study

Re-dislocations: mostly occurred during the first year (82% of those in Group A and 84% of those in Group B)

Adverse events: were not prespecified as an outcome. However, 6/85 (7.1%) participants in Group A had temporary stiffness of the involved shoulder, which "resolved within a month or two through the use of self directed range-of-motion exercises". 8/22 (36.4%) participants who had experienced recurrent dislocation or subluxation in Group A and 9/31 (29%) in Group B eventually underwent surgical stabilisation

Adherence: 61/85 (71.8%) of the participants followed up in Group A and 39/74 (52.7%) of those in Group B complied with the protocol

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Randomization was performed with the use of a random-number table created by the principal investigator. Co-investigators allocated the patients at their institutes with the use of this random number table"
Allocation concealment (selection bias)High riskThe random-number table was probably open: see above
Blinding of participants and personnel (performance bias)
All outcomes
High riskThere is a risk of performance bias, thus: "We and the other treating surgeons might have made a stronger effort to ensure adherence to the external rotation immobilization"
Blinding of outcome assessment (detection bias)
All outcomes
High riskThere is no mention of blinding, and although the primary outcomes, self-reported re-dislocation or subluxation, are both clearly defined, the latter is inevitably subjective
Incomplete outcome data (attrition bias)
All outcomes
High riskThere was attrition in both groups. Reasons were not provided. A sensitivity analysis identified this as a potentially serious source of bias
Selective reporting (reporting bias)High riskNo protocol is available. Key outcomes were reported, but the start point and selection criteria are inconsistent across reports and may not have been determined prospectively
Other biasHigh risk

There is disclosure of funding or grants of < $10,000 by Alcare to one or more of the authors in support of preparation for the research. There is disclosure of payments or benefits of < $10,000 by Alcare to one or more of the authors or their immediate families.

External rotation splints were ad hoc from October 2000 to November 2003 but commercially manufactured from November 2003 to March 2004. This has implications for positioning (neither Itoi 2003 nor Itoi 2007b specify how 10° was measured in the ad hoc splints), consistency and adherence; and therefore may impact external validity

Liavaag 2011

Methods

Aim: "To compare immobilization in internal rotation and external rotation after anterior shoulder dislocation"

Method of randomisation: block randomisation

Assessor blinding: no

Intention-to-treat analysis: yes

Loss to follow-up: 4/188 (2.1%) comprising 2/93 (2.2%) in Group A and 2/95 (2.1%) in Group B

Participants

Setting: Emergency Department of 13 hospitals in Norway

Period of data collection: January 2005 to February 2008

Participants: 188

Inclusion criteria: age 16 to 40 years; successful reduction of primary traumatic anterior dislocation documented on conventional radiographs

Exclusion criteria: fractured glenoid with a large bony defect (including > 20% of the glenoid rim) or a bony defect involving > 1/3 of the diameter of the glenoid fossa at the same level; fractured greater tuberosity with malalignment after repositioning; neuropathy related to dislocation or reduction; unwillingness or inability to participate

Sex: 153 male

Age: mean 26.8 years, SD 7.1 years

Subgroups: age 16 to 24 years; age 25 to 40 years

Pre-injury status: 44/93 (47.3%) in Group A and 39/95 (41.1%) in Group B sustained their injury in sport. The ratio of minor to major trauma was 52:41 in Group A and 48:47 in Group B. Two participants in each group had positive sulcus signs. Six participants in Group A and eight in Group B had other signs associated with generalised ligamentous laxity

Assigned: 93/95 (Group A/B)

Assessed: 91/93 (at final follow-up)

Interventions

Group A: immobilisation in an appropriately sized, manufactured shoulder immobiliser (15-degree Ultrasling ER, DonJoy, Vista, California) in 15 degrees of external rotation, except when showering. A line at the top of the immobiliser was parallel to the frontal plane when the arm was correctly positioned. The immobiliser incorporated the wrist. No information was given on who applied the immobilisation or gave the accompanying instructions

Group B: immobilisation in a collar and cuff or a sling and swathe, except when showering. No information was given on who applied the immobilisation or gave the accompanying instructions

Both groups: immobilisation commenced immediately after reduction and continued for three weeks. No information was given on rehabilitation

Outcomes

A standardised questionnaire was sent to participants two years post dislocation. Adherence data were gathered at three-week follow-up

Prespecified as the primary outcome

Self-reported (and verified) re-dislocation: dislocation was defined as the humeral head being completely out of the glenoid until a reduction manoeuvre was performed. Participants recorded the date of the first recurrent dislocation on the questionnaire and this was confirmed by reference to medical records

Prespecified as secondary outcomes

Self-reported subluxation: subluxation was defined as the humeral head being partially out of the glenoid

Self-reported recurrent instability: recurrent dislocation or subluxation

Resumption of pre-dislocation level of physical activity

Adverse events (other than need for surgical intervention): the questionnaire asked whether there were any persistent complaints caused by the shoulder dislocation (additional information provided by contact author)

Adverse events (need for surgical intervention)

Western Ontario Shoulder Instability Index (WOSI)

Adherence: daily immobilisation was recorded as "not at all," "up to 8 hours," "between 8 and 16 hours" and "more than 16 hours". The number of days on which immobilisation was used was also recorded. Participants who used immobilisation for > 16 hours daily for at least 20 days were labelled compliant

Notes

"There was no difference between the randomised groups in terms of demographic characteristics" at baseline

Physicians were allowed to choose the method of reduction according to personal preference. Details were not reported

Five deviations from protocol were analysed as randomised in the intention-to-treat analysis, as follows: Two participants in Group A were enrolled despite being ineligible (they were recurrent dislocators); one participant in Group B was enrolled despite being ineligible (this participant was a recurrent dislocator); two participants in Group A were incorrectly immobilised in internal rotation

Re-dislocations: mean time to re-dislocation was 10.5 months (range one week to two years) in Group A and 11.6 months (range two months to two years) in Group B

Self-reported recurrent instability: a discrepancy was noted. Participants with recurrent instability (re-dislocation or subluxation) are reported as a percentage of the full sample (N = 184), and yet "only 163 patients answered the question as to whether they had experienced subluxation". There appears to have been an assumption that non-responders had not experienced subluxation. Because the distribution of responses across the two groups is unknown, the data are not usable in a meta-analysis

WOSI data: were reported to be not normally distributed. The individual data were unavailable to us, precluding analysis

Adherence: 63/93 (67.7%) participants in Group A and 45/95 (47.4%) participants in Group B protocol were labelled compliant

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Block randomization was conducted for each hospital," but it is unclear how the sequences were generated
Allocation concealment (selection bias)Unclear risk"Concealed randomization was performed according to the method described by Altman, and sealed envelopes were used," but specific details were not given
Blinding of participants and personnel (performance bias)
All outcomes
High riskParticipants were informed of the preliminary results of Itoi 2007 (favouring immobilisation in external rotation), which may have influenced adherence
Blinding of outcome assessment (detection bias)
All outcomes
Low riskThere is no mention of blinding, but the primary outcome measure, re-dislocation, was clearly defined, and participant reports were verified by reference to their medical records
Incomplete outcome data (attrition bias)
All outcomes
Low riskFor the reportedly primary outcome, the proportion of missing data is insufficient to have a clinically relevant impact on estimates of effect
Selective reporting (reporting bias)Low riskA protocol was published (ClinicalTrials.gov Identifier: NCT00202735), and the study's prespecified primary outcome was reported
Other biasLow riskNone identified

Taskoparan 2010

  1. a

    CT: Computed tomography
    MRI: Magnetic resonance imaging
    SD: Standard deviation
    WOSI: Western Ontario Shoulder Instability Index

Methods

Aim: "To compare the functional and stability outcomes of the patients with acute anterior shoulder dislocation, who were stabilised at external versus internal rotation"

Method of randomisation: alternation based on Emergency Department numbers

Assessor blinding: no

Intention-to-treat analysis: no

Loss to follow-up: Rowe and Constant-Murley scoring was planned at six months, and control examinations (radiography and MRI) at 6, 12 and 24 months, thus a two-year follow-up. Presumably the intent was to follow up re-dislocations for the same period. However, individual data in Table 1 indicate that only 12/16 (75%) participants in Group A and 13/17 (76.5%) in Group B were followed up to one year (or were already accounted for as re-dislocators). Only 6/16 (37.5%) participants in Group A and 11/17 (64.7%) in Group B were followed up to two years (or were already accounted for as re-dislocators). Thus, while there was no loss to follow up at the short-term (six-month time point), there was potentially important loss at one year (25% in Group A, 23.5% in Group B) and especially at two years (62.5% in Group A, 35.3% in Group B)

Participants

Setting: Emergency Department in Turkey

Period of data collection: 2004 to 2008

Participants: 33

Inclusion criteria: acute primary traumatic anterior dislocation; admission on the first day of reduction

Exclusion criteria: hyperlaxity; [implicitly] neural damage

Sex: 31 male

Age: mean 34.9 years (SD 17.9 years) in Group A, 28.9 years (14.8 years) in Group B

Subgroups: age < 20 years, n = 1; age 21 to 30 years, n = 21; age 31 to 40 years, n = 4; age > 40 years, n = 7

Pre-injury status: 7/16 (43.8%) in Group A and 9/17 (52.9%) in Group B sustained their injury in sport

Assigned: 16/17 (Group A/B)

Assessed: 16/17 (at 6 months); 12/13 (at 1 year); 6/11 (final follow-up at two years)

Interventions

Group A: immobilisation in 10 degrees of external rotation in a locally-manufactured, hard polyethylene splint, removed only for showering. The immobiliser incorporated the wrist. No information was given on who applied the immobilisation or gave the accompanying instructions

Group B: immobilisation in Velpeau bandaging on the first day, switching to a "waist-assisted sling on the second day", removed only for showering. No information was given on who applied the immobilisation or gave the accompanying instructions

Both groups: immobilisation commenced on the first day after reduction and continued for three weeks. Immediately afterwards, rehabilitation commenced, initially with isometric and pendular exercises. When isometric exercises were painless, isotonic exercises were commenced. Avoidance of sporting activities for three months was recommended

Outcomes

Follow-up points were six months (Constant-Murley score, Rowe score), 1 year and 2 years (radiographs and MRI) and unspecified (re-dislocation, adverse events, apprehension test)

(Outcomes were not prespecified as primary or secondary)

Re-dislocation: was not prespecified as an outcome but was reported in the results. The mechanism for gathering these data was not reported

Constant-Murley score (not used in review)

Rowe score (not used in review)

Adverse events: were not prespecified as an outcome, but severe functional limitation was reported in one participant

Apprehension test: this was interchangeably termed the "intimidation test," and no further information was given

Notes

"Both groups were similar in basic clinical parameters"

Reduction was by the Hippocratic method in 20 participants (11 in Group A), the Kocher manoeuvre in 12 participants (five in Group A) and the Stimson manoeuvre in one participant (Group B)

Adverse events: "There was no severe functional limitation of any patient during follow up except one patient in the internal rotation group with 30° limitation in abduction and 10° in internal rotation in the sixth and twelfth months. This patient was 75 years old and had additional rotator cuff problems"

Adherence: "All patients adapted quite well to the fixation methods, and fixation was performed for three weeks"

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High risk"Patients with odd emergency department administration numbers received stabilization at internal rotation ... and patients with even numbers received stabilization at external rotation"
Allocation concealment (selection bias)High riskAllocation was based on case record number (see above)
Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskIt is unclear who applied the splints and gave initial instructions, and whether they were otherwise independent of the study, raising the possibility of performance bias
Blinding of outcome assessment (detection bias)
All outcomes
High riskThere is no mention of blinding and some outcomes may have been influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBeyond six months there were considerable, unexplained losses to follow up
Selective reporting (reporting bias)High riskNo study protocol is available. Two of the key outcomes reported (re-dislocation and the apprehension test) were not prespecified in the Methods section, and it is unclear how data on re-dislocation were obtained. No specific data on adherence are presented
Other biasUnclear riskAd hoc external rotation splints were used, and no mention is made of how the desired position was determined. This has implications for consistency and adherence, as well as positioning. Adherence is not specifically reported

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    ASES: American Shoulder and Elbow Surgeons
    HRQoL: Health-related quality of life
    ISRCTN: International Standard Randomised Controlled Trial Number Register
    WOSI: Western Ontario Shoulder Instability Index

Chutkan 2012Not a trial: a commentary on Liavaag 2011
Harper 2000No response was received to requests for information on this trial, which is listed in the National Research Register UK. We found no other reports of this trial
Hovelius 1983This trial involved 27 centres. At six centres, allocation to two different durations of immobilisation was based on date of shoulder dislocation. At the other 21 centres, treatment was given according to customary practice. Hence, overall this study was neither randomised nor quasi-randomised
Itoi 2003We have received contradictory information on the status of this preliminary quasi-randomised study in relation to Itoi 2007. Excluded to remove the risk of duplicate populations
Kiviluoto 1980Of 99 people younger than 50 years in this study, 53 were immobilised for one week and 46 for three weeks. However, no indication reveals the method of allocation to the two groups, nor are results presented for these participants. We received no response from the study authors
Staply 2002A contact person at the institution given for this trial as listed in the National Research Register UK indicated that this trial did not appear to have been started. We found no other reports of this trial
Wakefield 2001A draft report of this pilot study of 23 participants, comparing Intensive physiotherapy plus home exercises versus home exercises alone, was shared with review authors in June 2005. At that time, a report was expected to be submitted shortly for publication. We are unable to find a published version of the report, and the contact author is no longer available at the postal or email address given in the trial registry
Whelan 2010Not a trial: a commentary on Finestone 2009
Xu 2003This was not a randomised controlled trial but a comparison, with a historical control group, of systematic rehabilitation therapy versus supporting bandage in 43 people with acute shoulder dislocation

Characteristics of studies awaiting assessment [ordered by study ID]

ISRCTN41070054

MethodsReportedly a randomised controlled trial
Participants38 people with anterior shoulder dislocation
Interventions

1. External rotation splint

2. Sling

Outcomes

Follow-up: not stated

Incidence of recurrent dislocation

NotesThe anticipated end date of this UK-based study, which is listed as completed on the ISRCTN registry, was 1 November 2010. We are unable to find a published report, and we received no response to our email and postal correspondence to the contact author, Mr N Nicolaou

ISRCTN48254181

MethodsStated to be a randomised controlled trial
Participants150 people with primary anterior shoulder dislocation
Interventions1. Arm immobilised in external rotation
2. Arm immobilised in internal rotation
OutcomesFollow-up: not stated
Recurrence
NotesThe anticipated end date of this UK-based study, which is listed as completed on the ISRCTN registry, was 1 May 2008. We are unable to find a published report, and we received no response to our email and postal correspondence to the contact author, Mr A Pimpalnerkar

Itoi 2013

MethodsRandomised controlled trial
Participants109 people with primary traumatic shoulder dislocation. Study conducted in Japan
Interventions

3 weeks' immobilisation in external rotation followed by:

1. no shoulder motion restriction band;

2. a shoulder motion restriction band for 3 weeks; or

3. a shoulder motion restriction band for 6 weeks.

OutcomesRecurrent dislocation and return to sports, assessed at 2-year follow-up
Notes

No difference in recurrence rate was reported

'Awaiting classification' as evaluating a supplement to an immobilisation technique whose general effectiveness is not established

Kelly 2011

MethodsMethod of randomisation: randomised multi-centre - no details
Assessor blinding: no
Intention-to-treat analysis: not known
Loss to follow-up: 18 at one year
ParticipantsSeventy-two people (61 male) 16 to 45 years of age, with acute, first-time, traumatic, anterior dislocation of the shoulder (no fractures)
Interventions

1. External rotation brace

2. Internal rotation using a conventional sling

Outcomes

Follow-up: 1 year and 2 years

Recurrence and perception of shoulder instability: assessed via the Oxford Shoulder Instability Questionnaire at 1 year and 2 years after the first dislocation

Adherence/difficulties with wearing the brace/sling and with the rehabilitation protocol at 3 months

NotesMulticentre trial (7 NHS hospitals in England). One-year results reported in a conference abstract; results from a two-year follow-up questionnaire reported as pending in a 2011 report of the Institute of Orthopaedics

Miller 2007

MethodsMethod of randomisation: randomised multi-centre - no details
ParticipantsThirty people (to date) 18 to 30 years of age with acute, first-time anterior dislocation of the shoulder. Study conducted in USA
Interventions

Three weeks' immobilisation in:

1. a 15-degree external rotation brace; or

2. a sling.

This was followed by an identical short course of physiotherapy

Outcomes

Follow-up: one year

WOSI, subjective portion of ASES shoulder assessment form and a study-specific questionnaire

NotesInterim results reported in a 2007 poster presentation. We received no response to our attempt to contact Dr BS Miller

NCT00707018

MethodsMethod of randomisation: initially quasi-randomised using alternation (10 cases), then use of random numbers chart
Assessor blinding: no
Intention-to-treat analysis: no information
Loss to follow-up: not known
ParticipantsFifty people 14 to 30 years of age with acute, first-time, traumatic, isolated anterior dislocation of the shoulder (no significant fractures)
Interventions

1. External rotation shoulder sling.

2. Internal rotation shoulder sling.

Outcomes

Follow-up: two years

Primary: recurrent instability consisting of a documented episode of anterior shoulder dislocation or multiple episodes of shoulder subluxation

Secondary: participant self-reported outcome measures, time to return to work or sport

NotesThis USA-based study is listed as completed on the ISRCTN registry, with an end date of 1 February 2012. We are unable to find a published report, and we received no response to our attempt to contact Prof EC McCarty

Whelan 2008

MethodsMethod of randomisation: randomised multi-centre - computer-generated list
ParticipantsSixty-one people with first-time, traumatic anterior dislocation of the shoulder (no large bony lesions or polytrauma). All were younger than 35 years of age. Study conducted in Canada
Interventions

Four weeks' immobilisation in:

1. an external rotation brace; or

2. a sling.

This was followed by a similar rehabilitation protocol

OutcomesBlinded assessment of functional indicators was undertaken by independent raters for 2 years post dislocation. Outcomes include recurrent instability, adherence and, at final follow-up, disease-specific HRQoL data
NotesPreliminary results reported in a 2008 conference abstract. We received no response to our attempt to contact Dr DB Whelan or Prof RB Litchfield

Characteristics of ongoing studies [ordered by study ID]

ACTRN12611001183976

Trial name or titleA randomised controlled trial on the effect of physiotherapy and an external rotation brace on re-dislocation rate and shoulder function in patients aged 15 to 40 years with a first time anterior shoulder dislocation
MethodsStated to be a randomised controlled trial
ParticipantsTwo hundred people over 15 years of age with a clinical and radiographic diagnosis of an initial anterior shoulder dislocation caused by a traumatic event and presenting within three days of injury
Interventions

1. Immobilisation in an external rotation brace for three weeks and physiotherapy comprising graduated range-of-motion exercises and strengthening exercises.

2. Immobilisation in an external rotation brace for three weeks and physiotherapy comprising graduated range-of-motion exercises, pulsed ultrasound and massage.

Outcomes

Follow-up: six weeks, 3 months, 1 year and 2 years after dislocation

Primary: re-dislocation assessed by questioning at participant follow-up. May also include clinical diagnosis by x-ray

Secondary: DASH, Oxford Shoulder Disability Index

Starting date7 January 2012
Contact information

Timothy Walker

Physiotherapy Department

Gold Coast Hospital

Nerang Street

Southport QLD 4215

Australia

NotesPersonal communication 13 November 2012: recruitment is expected to take a further 10 years

NCT01111500

Trial name or titleFirst time anterior glenohumeral joint dislocation with a Bankart lesion in young patients: which type of immobilization should be chosen? A prospective randomised study
MethodsStated to be a randomised controlled trial
ParticipantsFifty people 18 to 40 years of age with radiologically verified anterior dislocation requiring reduction (no fractures or Hill-Sachs lesions)
Interventions

1. Donjoy ER brace (external rotation)

2. Thoracobrachial brace (internal rotation)

Outcomes

Three months after dislocation

Primary: anatomical healing of the labrum

Secondary: reduction in re-dislocation rate

One year and two years after dislocation

Secondary: reduction in re-dislocation rate

Starting dateJuly 2007
Contact information

Stéphane Pelet, MD, PhD

CHA-Pavillon Enfant Jésus

Quebec

Canada

G1J 1Z4

1-418-649-0252 extension 3165

stephane.pelet.ortho@gmail.com

NotesPersonal communication with Dr Pelet, 21 January 2013: estimated completion 2014-2015

NCT01648335

  1. a

    ACTRN: Australian New Zealand Clinical Trials Registry
    DASH: Disabilities of the Arm, Shoulder, and Hand
    NCT: National Clinical Trial

Trial name or titleA prospective study of the effect of treatment of first time traumatic shoulder anterior dislocation by immobilization in external rotation on the incidence of recurrent dislocation
MethodsParallel assignment
ParticipantsMales 18 to 29 years of age with first-time dislocation of the shoulder
Interventions

1. Immobilisation in internal rotation in a universal shoulder immobiliser

2. Immobilisation in external rotation in a universal shoulder immobiliser

Outcomes

Follow-up: six months after dislocation

Primary: number of recurrent dislocations

Secondary: range of motion; supine apprehension test

Starting dateJune 2012
Contact informationNone given
Notes 

Ancillary