Conservative interventions for treating distal radial fractures in adults

  • Review
  • Intervention

Authors

  • Helen HG Handoll,

    Corresponding author
    1. University of Teesside, Centre for Rehabilitation Sciences (CRS), Research Institute for Health Sciences and Social Care, Middlesborough, Tees Valley, UK
    • Helen HG Handoll, Centre for Rehabilitation Sciences (CRS), Research Institute for Health Sciences and Social Care, University of Teesside, School of Health and Social Care, Middlesborough, Tees Valley, TS1 3BA, UK. h.handoll@tees.ac.uk. H.Handoll@ed.ac.uk.

    Search for more papers by this author
  • Rajan Madhok

    1. University of Manchester, Cochrane Bone, Joint and Muscle Trauma Group, Manchester, UK
    Search for more papers by this author

Abstract

Background

Fracture of the distal radius is a common clinical problem particularly in elderly white women with osteoporosis.

Objectives

To determine the most appropriate conservative treatment for fractures of the distal radius in adults.

Search methods

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (June 2005), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 2, 2005), MEDLINE, EMBASE, CINAHL, PEDro, the National Research Register (UK), conference proceedings and reference lists of articles.

Selection criteria

Randomised or quasi-randomised controlled clinical trials involving skeletally mature people with a fracture of the distal radius, which compared commonly applied conservative interventions for fracture fixation. These included the application of an external support (plaster cast or brace) and fracture manipulation.

Data collection and analysis

Both review authors performed independent trial selection and quality assessment. Data were extracted for anatomical, functional and clinical, including complications, outcomes. The trials were grouped into categories relating to manipulation of displaced fractures; use and extent, including forearm position, of immobilisation; use of braces; different casting materials and techniques; and duration of immobilisation. Although quantitative data from some trials are presented, the lack of good quality trials and trial heterogeneity inhibited pooling of results.

Main results

One trial was newly included in this update. In all, there are 37 trials, involving a total of 4215 mainly female and older patients, meeting the inclusion criteria for this review. Comprehensive details of the individual trials are provided in tabular form, and their results, grouped as indicated above, have been presented in text and graphs. The poor quality and heterogeneity in terms of patient characteristics, interventions compared and outcome measurement, of the included trials meant that no meta-analyses were undertaken.

Authors' conclusions

There remains insufficient evidence from randomised controlled trials to determine which methods of conservative treatment are the most appropriate for the more common types of distal radial fractures in adults. Therefore, at present, practitioners applying conservative management should use an accepted technique with which they are familiar, and which is cost-effective from the perspective of their provider unit. Patient preferences and circumstances, and the risk of complications should also be considered.

Prioritising research questions to clarify the most appropriate conservative treatment for this common fracture is warranted. Researchers should differentiate between extra-articular and intra-articular, and non-displaced and displaced fractures, ascertain patient preferences, and agree a core outcome data set.

摘要

背景

成人遠端橈骨骨折之保守性治療

遠端橈骨骨折是常見的臨床疾病,尤其常發生於患有骨質酥鬆症的老年白種女性

目標

確定對於成人遠端橈骨骨折最適當的保守性治療

搜尋策略

我們搜尋the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (2005年 六月), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 2, 2005), MEDLINE, EMBASE, CINAHL, PEDro, the National Research Register (UK), 各醫學會的記錄及文章後的參考文獻

選擇標準

包含對骨骼已成熟的成人之遠端橈骨骨折,比較各種常用的保守性固定治療的隨機對照研究或有類似水準之研究. 治療包括外固定(石膏或支架)及骨折復位的手法.

資料收集與分析

所有負責文獻回顧的作者皆獨自選擇試驗及評估其研究品質. 資料收集包括解剖, 功能, 臨床方面的結果及併發症. 所有試驗被歸類為幾個範疇, 有對於移位骨折的復位手法; 骨折固定的使用與範圍, 包括前臂的固定姿勢; 支架的使用; 不同的石膏材料及技術; 及固定時間的長短. 雖然有些試驗提供了數量化的資料, 缺乏高品質的試驗及試驗的差異性大限制了試驗結果的統合.

主要結論

本次更新增加納入一個試驗. 包含37個試驗, 共有4215位多為女性,年齡較大的病患, 符合本次文獻回顧收案的條件. 各試驗完整的細節以表格的方式呈現. 而這些試驗的結果則如上述方式分類, 並呈現於內文及圖表中. 由於試驗的品質不佳, 病患群特徵, 治療及結果的差異性大, 無法做綜合分析.

作者結論

對於成人常見的遠端橈骨骨折類型, 目前沒有足夠從隨機對照研究得到的證據, 決定何種保守性治療最為合適. 因此, 目前臨床工作者欲使用保守性治療時, 應使用自己熟悉, 符合經濟效益, 且被人廣為接受的方法. 同時也要考慮病患的偏好及產生併發症的風險. 對於這類常見的骨折, 決定最合適的保守性治療的研究是刻不容緩的. 研究者應區分關節外及關節內的骨折, 無位移及有位移的骨折, 了解病患的偏好, 並建立一的評估結果的核心資料組.

翻譯人

本摘要由臺灣大學附設醫院王政為翻譯。

此翻譯計畫由臺灣國家衛生研究院(National Health Research Institutes, Taiwan)統籌。

總結

對於手腕骨折, 沒有足夠的證據決定最佳的非手術治療. 年長女性會因為跌倒並用手撐地而造成手腕骨折(前臂兩根骨頭, 其中之ㄧ的遠端骨折). 治療方式通常不需手術, 而是將骨頭復位, 並以石膏或支架固定. 目前從臨床試驗沒有得到足夠證據, 決定如何及何時中度位移的骨折應接受徒手復位. 也沒有足夠的證據決定最好的固定方式及固定時間的長短.

Plain language summary

Not enough evidence to tell what type of non-surgical treatment is best for treating a broken wrist

In older women, a broken wrist (from a fracture at the lower end of one of the two forearm bones) can result from a fall onto an outstretched hand. Treatment is usually non-surgical and may include putting the broken bone back into position and immobilising the wrist in a plaster or brace. There was not enough evidence from trials to determine whether, and if so when, moderately displaced fractures should be manipulated back into position. Nor was there enough evidence to determine the best method and duration of immobilisation.

எளியமொழிச் சுருக்கம்

முறிந்த மணிக்கட்டுக்கு அறுவை சிகிச்சை அல்லாத எந்த சிகிச்சை சிறந்தது என்று கூற போதுமான ஆதாரங்கள் இல்லை.

வயதான பெண்களில், ஒரு உடைந்த மணிக்கட்டானது (இரண்டு முழங்கை எலும்புகளில் ஒன்றின் கீழ் இறுதியில் ஒரு எலும்பு முறிவு கொண்ட) கையை நீட்டிய படி வீழ்வதால் விளையலாம். வழக்கமாக இதற்கான சிகிச்சை, அறுவை சிகிச்சை அல்லாத சிகிச்சையாகும் .இது எலும்புத் துண்டுகளை மீண்டும் அதே இடத்தில் வைப்பதையும், மற்றும் மோசமாக இடம்பெயர்ந்தது என்றால் மணிக்கட்டை ஒரு சுண்ணச்சாந்து கட்டில் முடக்குதலையும் உள்ளடக்கும். மிதமான இடம்பெயர்ந்த முறிவுகளை சரியான நிலைக்கு பொருத்தும் முறை என்ன? மற்றும் அதனை எப்பொழுது செய்வது என்பதனை பற்றிய ஆதாரங்கள் இல்லை. மேலும், சிறந்த முறை எது? எவ்வளவு காலம் முடக்கி (immobilisation) வைக்க வேண்டும் என்று தீர்மானிக்க போதுமான ஆதாரம் இல்லை.

மொழிபெயர்ப்பு குறிப்புகள்

மொழிபெயர்ப்பு: சி.இ.ப.ஏன்.அர். குழு

Background

Fractures of the distal radius are a common clinical problem affecting skeletally mature people. The young sustain this injury as a result of significant local trauma. The elderly have predisposing risks of disuse or post menopausal osteoporosis. The lifetime risk of sustaining a distal radius fracture in a white person, aged 50 years, based in the USA or Northern Europe has been estimated to be 15% in women and 2% in men (Cummings 1985). The true costs associated with the injury have been difficult to quantify as this injury is frequently treated on an outpatient basis. In a recent prospective one-year study of patients, aged 35 years and above, with Colles' fracture (a common type of distal radius fracture) treated in six centres in the UK, one in five patients were admitted to hospital, the proportion increasing with age (O'Neill 2001).

These fractures are generally closed and usually involve displacement of fracture fragments. They may be either extra-articular (leaving the articular or joint surface of the distal radius intact) or intra-articular (where the articular surface is disrupted). Numerous classifications have been devised to define and group different fracture patterns (Chitnavis 1999). Simple classifications based on clinical appearance, and often named after those who described them, remain in common use. In particular, 'Colles' fracture' is still the terminology used for a fracture in which there is an obvious and typical clinical deformity - of dorsal displacement, dorsal angulation, dorsal comminution (small fragments of bone), and radial shortening.

The most suitable method of management of this injury is not clearly defined. The final choice of treatment method will be influenced by many considerations including the nature of the fracture, bone stock and fragility, the presence of local complications (compound injury, nerve injury) or other injuries, the patient's general medical condition, the expected functional loading (activity demands), and patient motivation.

The basic treatment options available for these fractures are:

  • closed or open (including arthroscopically assisted) reduction;

  • external splintage: immobilisation or support, or both (plaster of Paris cast, brace, bandage);

  • external fixation using either pins and plaster or an external fixator;

  • percutaneous pinning;

  • internal fixation with pins, nails, screws and plates;

  • replacement of lost bone stock (metaphyseal defect) by temporary bone scaffold (bone graft) material or any other suitable substance (bone cement or substitute).

The complications from this injury are diverse and unexpectedly frequent (Altissimi 1984; Atkins 1989; Cooney 1980). Some complications are associated with the injury itself. As well as concomitant injuries to soft tissues, fracture displacement can further compromise blood vessels, tendons and nerves, with median nerve dysfunction being most common complication (Belsole 1993). Late complications include midcarpal instability and post-traumatic arthritis, which can occur several months or years after injury (Knirk 1986; Taleisnik 1984).

Complications can also result from treatment interventions and include residual finger stiffness (Gartland 1951), and pin track infection and soft tissue injury from external fixation and percutaneous pinning. Reflex sympathetic dystrophy (RSD), also referred to as algodystrophy, Sudeck's atrophy and sometimes shoulder-hand syndrome (Fernandez 1996) is a major complication requiring many months of physiotherapy to alleviate symptoms (pain and tenderness, impairment of joint mobility, swelling, dystrophy, and vasomotor instability) in serious cases. The etiology of RSD is often unclear.

The evaluation of treatment interventions is generally based on anatomical, functional and clinical outcomes and presence of complications. Anatomical restoration and residual deformity are usually based on radiological measurements such as radial length or shortening, volar tilt or dorsal angulation, radial angle or radial inclination, and the relative position of the distal ends of the radius and ulna, ulnar variance (see Table 1). Functional and clinical outcomes usually include wrist and forearm mobility (range of movement), pain, grip strength, residual soft tissue swelling, local complications, cosmetic appearance, patient satisfaction with treatment modality and activities of daily living.

Table 1. Some definitions of radiological parameters
ParameterDefinitionNormal value
Dorsal angulation (dorsal or volar or palmar tilt)Angle between a) the line which connects the most distal points of the dorsal and volar cortical rims of the radius and b) the line drawn perpendicular to the longitudinal axis of the radius. Side view of wrist.Palmar or volar tilt: approximately 11-12 degrees.
Radial lengthDistance between a) a line drawn at the tip of the radial styloid process, perpendicular to the longitudinal axis of the radius and b) a second perpendicular line at the level of the distal articular surface of the ulnar head. Frontal view.Approximately 11-12 mm.
Radial angle or radial inclinationAngle between a) the line drawn from the tip of the radial styloid process to the ulnar corner of the articular surface of the distal end of the radius and b) the line drawn perpendicular to the longitudinal axis of the radius. Frontal view.Approximately 22-23 degrees.
Ulnar varianceVertical distance between a) a line drawn parallel to the proximal surface of the lunate facet of the distal radius and b) a line parallel to the articular surface of the ulnar head.Usually negative variance (e.g. -1 mm) or neutral variance.

This review, which looks at conservative interventions, is one of five Cochrane reviews of interventions used in the management of these fractures. Another review is of surgical interventions, and also addresses the conservative versus surgical management question (Handoll 2005a). The other three reviews are anaesthesia interventions during manipulation of displaced fractures and surgery (Handoll 2005b), methods of closed reduction (Handoll 2005c), and rehabilitation interventions such as mobility exercises (Handoll 2005d).

Objectives

The aim of this systematic review is to determine the most appropriate conservative treatment for fractures of the distal radius in skeletally mature people. The order and, in part, the emphasis of the review comparisons (and associated null hypotheses) set out below reflect the decision points, and associated timing, for the management of these fractures.

We aimed to test the following null hypotheses:

(1) No differences exist between outcomes from reduction, delayed reduction or no reduction of displaced fractures.

The following comparisons were made:
(a) Manipulation versus no manipulation
(b) Delayed manipulation versus immediate manipulation

(2) No differences exist between outcomes from different methods or durations (including none) of immobilisation involving plaster or brace use.

The following comparisons were made:
(a) No immobilisation (minimal support) versus immobilisation
(b) Forearm held in different positions by plaster
(c) Plaster or synthetic cast type A versus type B
(d) Forearm held in different positions by brace
(e) Brace versus plaster cast
(f) Different casting materials or techniques, or both
(g) Different durations of immobilisation

We also planned to study the outcome of different age groups and of comparable fracture groups. Neither, even distinguishing between extra-articular and intra-articular fractures, and non-displaced and displaced fractures, was possible.

Methods

Criteria for considering studies for this review

Types of studies

Any randomised or quasi-randomised (method of allocating participants to a treatment which is not strictly random e.g. by date of birth, hospital record number, alternation) controlled clinical trials of interventions listed below were considered.

Types of participants

Patients of either sex who had completed skeletal growth, with a fracture of the distal radius.

The characteristics of the participants included in the trials were noted with an emphasis on age, gender and fracture type.

Types of interventions

All randomised comparisons of conservative interventions involving plaster cast or brace use with placebo, no intervention or an alternative cast or brace intervention, in the treatment of fractures of the distal radius. Also included are randomised comparisons evaluating the reduction of displaced fractures. Trials evaluating different techniques of reduction including the choice of anaesthesia/analgesia were excluded.

The main treatment options covered in this review were various methods of support and immobilisation such as plaster casts, wrist braces and bandages applied after closed reduction of the fracture, if done.

Types of outcome measures

(1) Anatomical outcome (anatomical restoration and residual deformity)
Radiological parameters include radial length or shortening and shift, dorsal angulation, radial inclination or angle, ulnar variance. Composite measures include malunion and total radiological deformity. Definitions of four of the most commonly reported radiological parameters are presented in Table 1.

(2) Functional outcome
Range of movement (wrist and forearm mobility), pain, grip strength, activities of daily living. Also, patient functional assessment instruments such as Short Form-36 (SF-36), the Disability of the Arm, Shoulder, and Hand questionnaire (DASH) and the Patient-Rated Wrist Evaluation (PRWE) (MacDermid 2000).

(3) Clinical outcome
Residual soft tissue swelling; early and late complications associated with distal radial fractures or their treatment, including reflex sympathetic dystrophy (RSD) and osteoarthrosis; cosmetic appearance; and patient satisfaction with treatment.

(4) Resource use
Number of outpatient attendances and other costs.

Search methods for identification of studies

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (June 2005), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 2, 2005) (see Appendix 1), MEDLINE (1966 to June week 1 2005), EMBASE (1988 to 2005 week 24), CINAHL (1982 to June week 2 2005), PEDro - physiotherapy evidence database (http://www.pedro.fhs.usyd.edu.au/index.html accessed 14 June 2005), OTseeker - The Occupational Therapy Systematic Evaluation of Evidence Database (http://www.otseeker.com accessed 2 June 2005) and reference lists of articles. We also searched Current Controlled Trials at http://www.controlled-trials.com (accessed June 2005) and the UK National Research Register at http://www.update-software.com/national/ (up to Issue 2, 2005) for ongoing and recently completed trials. We hand searched the British Volume of the Journal of Bone and Joint Surgery supplements (1996 onwards), and abstracts of the American Society for Surgery of the Hand annual meetings (2000 to 2004: http:/www.assh.org/), the American Orthopaedic Trauma Association annual meetings (1996 to 2004: http://www.ota.org/education/amabstracts.htm) and American Academy of Orthopaedic Surgeons annual meeting (2004 and 2005: http://www.aaos.org/wordhtml/libscip.htm). We hand searched final programmes of SICOT (1996 & 1999) and SICOT/SIROT (2003), and the British Orthopaedic Association Congress (2000, 2001, 2002 and 2003). We hand searched Orthopaedic Transactions and various supplements of Acta Orthopaedica Scandinavica. We also scrutinised weekly downloads of "Fracture" articles in new issues of 17 journals (Acta Orthop Scand; Am J Orthop; Arch Orthop Trauma Surg; Clin J Sport Med; Clin Orthop; Emerg Med Clin North Am; Foot Ankle Int; Injury; J Accid Emerg Med; J Am Acad Orthop Surg; J Arthroplasty; J Bone Joint Surg Am; J Bone Joint Surg Br; J Foot Ankle Surg; J Orthop Trauma; J Trauma; Orthopedics) from AMEDEO (http://www.amedeo.com). No language restrictions were applied.

In MEDLINE (OVID-WEB) the search strategy was combined with all three sections of the optimal MEDLINE search strategy for randomised trials (Higgins 2005) (see Appendix 2).

Similar search strategies used for EMBASE (OVID-WEB) and CINAHL (OVID-WEB) are shown in Appendix 3 and Appendix 4.

Data collection and analysis

We independently assessed potentially eligible trials for inclusion; any disagreement was resolved by discussion. Titles of journals, names of authors or supporting institutions were not masked at any stage. Both review authors (HH and RM) independently assessed methodological quality of included studies and any disagreement was resolved by discussion. Data were extracted by one reviewer (HH) and checked by a second reviewer (RM) for all studies.

For the first version of the review, there was no systematic return to trialists for additional details of trial methodology or data. The few instances where extra information was obtained are noted in the 'Characteristics of included studies' table. In preparation for the first update, we contacted trialists of studies only published as conference abstracts and contact authors of ongoing trials that were indicated as being completed in the UK National Research Register. For the second update, we contacted trialists of two newly identified studies for further information on their studies. For the third update, we contacted trialists of the newly identified studies and of the studies listed as ongoing in the previous update for further information on their studies. For the fourth update, we contacted trialists of the newly identified study, of the study listed as ongoing in the previous update and of the two studies in 'Studies awaiting assessment' for further information on their studies. For the fifth update, we contacted trialists of the two newly identified studies and of the newly included trial, previously in 'Studies awaiting assessment', for further information on their studies.

Quality assessment
A modification of the Cochrane Bone, Joint and Muscle Trauma Group quality assessment tool (see Group details) was used in the evaluation of the included studies. The scoring scheme for 11 aspects of trial validity, plus brief notes of coding guidelines for some items, is shown in Table 2. The scheme was slightly modified for the first update to be compatible with that used in the surgical treatments review, as well as abiding by the Cochrane Group's requirement to rate active follow up. From the fifth update (Issue 4, 2005) of the review, the scores of the individual items were no longer summed.

Table 2. Methodological quality assessment scheme
ItemsScoresNotes
(1) Was the assigned treatment adequately concealed prior to allocation?3 = method did not allow disclosure of assignment.
1 = small but possible chance of disclosure of assignment or unclear.
0 = quasi-randomised, or open list or tables.
Cochrane code (see Handbook): Clearly Yes = A; Not sure = B; Clearly No = C.
(2) Were the outcomes of participants who withdrew described and included in the analysis (intention to treat)?3 = withdrawals well described and accounted for in analysis.
1 = withdrawals described and analysis not possible, or probably no withdrawals.
0 = no mention, inadequate mention, or obvious differences and no adjustment.
 
(3) Were the outcome assessors blinded to treatment status?3 = effective action taken to blind assessors.
1 = small or moderate chance of unblinding of assessors, or some blinding of outcomes attempted.
0 = not mentioned or not possible.
 
(4) Were important baseline characteristics reported and comparable?3 = good comparability of groups, or confounding adjusted for in analysis.
1 = confounding small, mentioned but not adjusted for, or comparability reported in text without confirmatory data.
0 = large potential for confounding, or not discussed.
Although many characteristics including hand dominance are important, the principal confounders were considered to be age, gender, type of fracture.
(5) Were the trial participants blind to assignment status after allocation?3 = effective action taken to blind participants.
1 = small or moderate chance of unblinding of participants.
0 = not possible, or not mentioned (unless double-blind), or possible but not done.
 
(6) Were the treatment providers blind to assignment status?3 = effective action taken to blind treatment providers.
1 = small or moderate chance of unblinding of treatment providers.
0 = not possible, or not mentioned (unless double-blind), or possible but not done.
 
(7) Were care programmes, other than the trial options, identical?3 = care programmes clearly identical.
1 = clear but trivial differences, or some evidence of comparability.
0 = not mentioned or clear and important differences in care programmes.
Examples of clinically important differences in other interventions were: time of intervention, duration of intervention, anaesthetic used within broad categories, operator experience, difference in rehabilitation. Note, operator experience was not listed in the original review, though it was considered.
(8) Were the inclusion and exclusion criteria for entry clearly defined?3 = clearly defined.
1 = inadequately defined.
0 = not defined.
In contrast to the surgical treatment review, a clear description of fracture type or displacement was not essential, although still preferable, for the top score.
(9) Were the outcome measures used clearly defined?3 = clearly defined.
1 = inadequately defined.
0 = not defined.
 
(10) Were the accuracy and precision, with consideration of observer variation, of the outcome measures adequate; and were these clinically useful and did they include active follow up?3 = optimal.
1 = adequate.
0 = not defined, not adequate.
Active follow up was not considered explicitly in the original review.
(11) Was the timing (e.g. duration of surveillance) clinically appropriate?3 = optimal. (> 1 year)
1 = adequate. (6 months - 1 year)
0 = not defined, not adequate. (< 6 months)
In the original version of this review, the criterion for an optimal score was "1 year and above", rather than above one year.

Data analysis
Where available, quantitative data, both dichotomous and continuous, that were reported in individual trial reports for outcomes listed in the inclusion criteria have been presented in the graphs. Relative risks and 95% confidence intervals have been presented for dichotomous outcomes and mean differences and 95% confidence intervals for continuous outcomes. In the light of the disparate nature and general poor quality of the included studies, there was no pooling of results in this review. The statistics presented in the graphs are for reference only and should not be considered as an estimate of treatment effect.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies.

The two newly identified trials (Liebman 2004; Sahin 2005) considered for this, the fifth, update, have been left in 'Studies awaiting assessment'; Liebman 2004 awaits publication and Sahin 2005, located since the last search in June 2005, awaits translation from Turkish. Additional abstracts were located for both trials (Jackson 2002; Shah 2002) previously in 'Studies awaiting assessment': further information from the trialists from these two trials has not been forthcoming. We have, however, now included the former trial (Jackson 2002) but excluded the latter study (Shah 2002), which is unlikely to be anything more than the self-declared prospective comparison. As before, we note that many more study reports were checked in the process of collecting the studies for the original review and updates, but found not to meet the inclusion criteria of this review; most of the 15 excluded studies are retained primarily for illustrative purposes, mainly in terms of the scope of this review.

Summaries of the past and present trial populations of past and the present versions of this review as well as the changes between updates are presented in Table 3.

In all, 37 trials involving comparisons of conservative interventions were included. Two trials respectively evaluated whether displaced fractures should be reduced and the timing of any closed reduction before cast application. It should be noted that the comparison of different techniques of closed reduction is covered in a separate review (Handoll 2005c). The other 35 trials examined the need for forearm immobilisation and the different methods and durations of immobilisation involving plaster cast or brace use.

Table 3. Numbers and status of studies in the published versions of the review
VersionTrial statusChanges
1st version
Issue 2, 1999
The original review had 29 included trials and 1 trial awaiting assessment. 
2nd version
Issue 3, 2001
The first update had 31 included trials, 5 excluded studies and 3 studies listed as ongoing.Three new trials were included (Cornwall 2001; Kongsholm 1981; Stoffelen 1998); one previously included trial (Milliez 1992), investigating the duration of immobilisation after surgical treatment, was excluded; three newly identified trials were placed in the ongoing trials category and four others were excluded. Additional findings were included from newly identified reports of five already included trials (Bunger 1984; Gibson 1983; Moir 1995; Sarmiento 1980; Stewart 1984).The results of our contact of trialists of the three studies only published as conference abstracts (Gibson 1983; McMillan 1996; Ross 1984) are summarised in the Notes section of the Characteristics of Included Studies Table.
3rd version
Issue 2, 2002
The second update had 33 included trials, 6 excluded studies, 3 listed as ongoing studies and 1 study awaiting assessment.Two studies (Aladin 2001; Ho 1986) were newly included. Of the three other newly identified trials, one (Harris 2002) was ongoing (see Characteristics of Ongoing Studies Table), one (Cohen 2001), pending further response from the lead trialist, was placed in Studies Awaiting Assessment and one (Davison 2001) was excluded (see Characteristics of Excluded Studies Table). Additional trial reports were identified for three already included trials (Bunger 1984; Cornwall 2001; Stewart 1984), one (Aladin 2001) formerly listed as an ongoing trial (Moran 1999) and one already excluded trial (Smith 1999). The new report of Cornwall 2001 showed that the previous report (Cornwall 2000) was an interim analysis of 70 out of the 127 patients in the trial.
4th version
Issue 2, 2003
The third update had 36 included trials, 11 excluded studies, 1 listed as ongoing and 3 studies awaiting assessment.Of seven newly identified studies, two were included (O'Connor 2002; Tumia 2003), one excluded (Moran 2002), one was ongoing (Hearne 2003) and three (Jackson 2002; Kowalski 2002; Shah 2002) were placed in Studies Awaiting Assessment; pending further information from trialists. Upon obtaining separate data for patients with distal radial fracture, one trial (Cohen 2001) previously in Studies Awaiting Assessment was included. Inquiries on the progress and status of three trials (Harris 2002; Singhania 2001; Shanker 2000), formerly listed in Ongoing studies, resulted in their exclusion. One trial (Pool 1973), often mistakenly referred to as a randomised trial, was also added to the excluded trials list.
5th version
Issue 2, 2004
The fourth update had 36 included trials, 14 excluded studies and 2 studies awaiting assessment.The only newly identified study (Shah 2003) was excluded. A study (Hearne 2003) previously listed as ongoing was excluded as was another study (Kowalski 2002), previously listed in Studies Awaiting Assessment. The two other studies (Jackson 2002; Shah 2002) in Studies Awaiting Assessment were retained, again pending further information from trial investigators.
6th version
Issue 4, 2005
The fifth update had 38 included trials, 15 excluded studies and 2 studies awaiting assessment.The two newly identified trials (Liebman 2004; Sahin 2005) considered for this, the fifth, update, have been left in 'Studies awaiting assessment': Liebman 2004 awaits publication and Sahin 2005, located only after the month of last search (June 2005), awaits translation from Turkish. Further abstracts were located for both trials (Jackson 2002; Shah 2002) previously in 'Studies awaiting assessment'. further information from the trialists from these two trials has not been forthcoming. We have, however, now included the former trial (Jackson 2002) but excluded the latter study (Shah 2002), which is unlikely to be anything more than the self-declared prospective comparison.

The majority of the included trials were fully reported in medical journals. Trial reports of eight trials (Aladin 2001; Cornwall 2001; Gibson 1983; Ho 1986; Jackson 2002; Kongsholm 1981; McMillan 1996; Ross 1984) were only available as abstracts. De Bruijn 1987, which consisted of two studies reported as one, was a published thesis as well as a paper in Dutch.

The publication dates of the trials span 23 years; Sarmiento 1980 being the earliest. Translations were obtained for one study (Blatter 1994) published in German and one in Danish (Sorensen 1986). The trials were mostly single centre studies conducted in 11 countries, with the majority being in the UK (16 trials) and Scandinavian countries (12 trials). Notable exceptions were Jackson 2002, which was conducted in four hospitals in Canada, and Tumia 2003, which was conducted at five trauma centres in three European countries.

The 37 included trials involved a total of 4215 mainly female and older participants. Details of individual trials are provided in the 'Characteristics of included studies' table. All of the trials providing information had more female participants, mostly at least three quarters of the trial populations. Three trials (McAuliffe 1987; Millett 1995; Wahlstrom 1982) exclusively recruited female patients. Ten trials, including five of the eight trials only reported as abstracts, provided no information on age; the other three trials with abstract reports only referred to adults and a lower age limit of 18 and 40 years respectively. Fifteen trials explicitly excluded children by setting a lower age limit, and information provided for the lowest age of participants of a further 10 trials indicates that the vast majority of participants were skeletally mature. No upper age limits were explicitly applied.

The variety of descriptions used for fracture type such as Colles', displaced, undisplaced, intra-articular, extra-articular, simple, comminuted and unstable, plus the differences in the classification systems, such as Frykman or Older, when used, or radiological parameters applied (extent of radial shortening or dorsal angulation, or both) meant that a summary of the fracture types based on displacement and articular involvement could not be done. Thus, basic divisions of fractures into a) undisplaced or minimally displaced fractures and displaced fractures or b) extra-articular and intra-articular fractures could not be performed for the whole group of trials. Some trials (Aladin 2001; Ferris 1989; Gibson 1983; Ho 1986; Kongsholm 1981; McMillan 1996; Nielsen 1981) provided no information about fracture type, whilst insufficient information was provided by those others that did not either explicitly and exclusively include displaced fractures. However undisplaced fractures or minimally displaced fractures were solely included in eight trials (Abbaszadegan 1989; Christensen 1995; Cohen 2001; Davis 1987; Jensen 1997; O'Connor 2002; Stoffelen 1998; Vang Hansen 1998). Two further trials (De Bruijn 1987; Dias 1987) altered their control intervention for undisplaced fractures.

The comparisons in the 37 included trials have been grouped according to the main question, or questions, addressed by each trial. Nine trials (De Bruijn 1987; Dias 1987; Gibson 1983; Gupta 1991; Jackson 2002; Kongsholm 1981; Stewart 1984; Van der Linden 1981; Wahlstrom 1982) had more than two intervention groups. These fell inside the same comparison group for Gupta 1991, Jackson 2002, Kongsholm 1981 and Wahlstrom 1982 (all had three intervention groups); but were in different comparison groups for De Bruijn 1987 (four), Dias 1987 (three), Gibson 1983 (three), Stewart 1984 (three) and Van der Linden 1981 (five). Dias 1987 and De Bruijn 1987 allocated undisplaced fractures to a different control intervention (bandage) than displaced fractures (modified plaster cast and brace respectively). De Bruijn 1987 consisted of two trials but failed to provide separate data for the groups by individual trial. Kongsholm 1981 compared two methods of reduction, with or without anaesthesia, as well as two types of plaster cast; only the latter comparison is reviewed here. Missing baseline data for some trials prevents exact enumeration of patients in some comparison groups.

In the first version of our review, two trials in which some or all participants had surgical treatment were included as these compared conservative interventions. In the first update, one of these (Milliez 1992) was excluded as it is reviewed in our review of surgical interventions (Handoll 2005a). However, the other trial (Cohen 1997) in which 10 of the 30 participants had percutaneous pinning, has been retained even though we consider that people who had surgical treatment form a distinct group in that surgery changes the question addressed, the intervention and outcome (e.g. extra surgical complications). These reservations should be borne in mind when viewing this trial.

The following comparisons based on the two null hypotheses listed in the objectives were made.

(1) Reduction of displaced fractures

Two comparisons covering the use of manipulation were included.

(a) Manipulation versus no manipulation
Kelly 1997 tested the result of manipulation under Bier's block anaesthesia versus no manipulation in 30 people with "moderately" displaced Colles' fracture. All subsequently had plaster cast immobilisation for five weeks.

(b) Delayed manipulation versus immediate manipulation
McMillan 1996 tested the effect of delaying manipulation for one week compared with immediate manipulation in 80 people. No details of the extent of displacement of the fractures were given.

(2) Plaster or brace management

(a) No immobilisation (minimal support) versus immobilisation
Two trials (Abbaszadegan 1989; Dias 1987) compared the immediate application of elastic or crepe bandage versus plaster cast immobilisation in 177 people with undisplaced or minimally displaced fractures.

(b) Forearm held in different positions by plaster
There were six trials (Blatter 1994; Gibson 1983; Gupta 1991; Van der Linden 1981; Wahlstrom 1982: Wilson 1984) involving around 650 participants in this category.

(c) Plaster or synthetic cast type A versus type B
Some of the above set of trials also belong to this category as different plaster casts (below-elbow, above-elbow) were used to maintain wrist position. Associated with these were differences in arm mobility. Trials examining different casting materials or techniques, or both are examined in category 2f. Seven trials are reviewed in the present category. Wrist mobility was the basis of the comparison involving 90 people with displaced fractures allocated either to a standard or a modified cast allowing wrist flexibility in Dias 1987. Forearm rotation was blocked but some flexion and extension at the elbow allowed with a modified sugar-tong cast which was compared with a below-elbow cast in the 126 participants of Aladin 2001. Similar but unconfirmed restrictions to mobility are likely with the modified sugar-tong fibreglass splints that were compared with volar-dorsal fibreglass splints or cylindrical plaster casts in the 101 participants of Jackson 2002. Cornwall 2001 and Sorensen 1986 compared above-elbow with below-elbow plaster immobilisation in 227 people. Lastly, Kongsholm 1981 and Van der Linden 1981 included a comparison of a dorsal splint (back slab) with a full (circular) below-elbow plaster cast in approximately 300 people.

(d) Forearm held in different positions by brace
There were two trials (Sarmiento 1980; Stewart 1984) involving over 290 people in this group.

(e) Brace versus plaster cast
There were 11 trials (Bunger 1984; De Bruijn 1987; Ferris 1989; Gibson 1983; Ho 1986; Ledingham 1991; Moir 1995; O'Connor 2002; Ross 1984; Stewart 1984; Tumia 2003) with around 1785 participants.

(f) Different casting materials or techniques, or both
Four trials (Cohen 1997; Cohen 2001; Nielsen 1981; Rosetzsky 1982), involving 170 people, investigated the use of different materials and associated techniques for below-elbow casts. Only minimally displaced and stable fractures were included in Cohen 2001. As stated above, 10 participants of Cohen 1997 had percutaneous pinning.

(g) Different durations of immobilisation
Seven trials compared the use of plaster cast for different durations. One other (De Bruijn 1987) compared the use of a bandage versus an above-elbow functional brace for three weeks after one week of immobilisation for both groups. Four (Christensen 1995; McAuliffe 1987; Millett 1995; Vang Hansen 1998) of the seven plaster cast trials compared immobilisation for three or four versus five or six weeks. These involved 331 people. Christensen 1995 only included undisplaced fractures. One trial (Davis 1987) compared one up to two weeks versus four up to five weeks plaster immobilisation and two trials (Jensen 1997; Stoffelen 1998) compared one-week versus three-weeks plaster immobilisation. All 169 participants of these three trials had undisplaced or minimally displaced fractures.

Risk of bias in included studies

The quality of trial methodology based on trial reports was disappointing. The results for individual trials are presented below. Information specific to the first three items of the quality score are given in the methods section of the 'Characteristics of included studies' table.

No trial explicitly reported using a blinded method of randomisation, but concealment of allocation was considered highly likely in Cohen 2001 (use of sealed numbered and consecutively opened envelopes), Rosetzsky 1982 (randomisation was by closed envelopes, allocated according to predetermined randomisation list) and Tumia 2003 (telephone randomisation). Two other trials (Kelly 1997; Sorensen 1986) used envelopes and three others (Millett 1995; Moir 1995; Van der Linden 1981) gave some details of the method used, without reporting adequate safeguards. Twenty-one trials did not describe their method of randomisation. The remainder used quasi-randomised methods based on either dates of birth, record numbers, dates of attendance or injury, admission sequence or alternation (Christensen 1995; De Bruijn 1987; Dias 1987; McAuliffe 1987; Nielsen 1981; O'Connor 2002; Sarmiento 1980; Stewart 1984).

Clear statements or evidence of intention-to-treat analysis (item 2) were rarely presented in trial reports. Allied with this but not graded was loss to follow up. Nine trials recorded 15% or above loss to follow up, with a third and nearly two thirds of trial participants being omitted from the final analyses of Sarmiento 1980 and Ho 1986 respectively.

Blinding of outcome assessors (item 3) was only rarely reported and nowhere performed for all outcomes at all times. Total blinding of outcome assessment is unrealistic for trials testing these types of interventions but it is possible and was reported for some outcomes in nine trials. Blinded clinical assessment was done at three and six months in Stewart 1984. Functional outcomes were blindly assessed in five trials (Dias 1987; Ledingham 1991; Moir 1995; O'Connor 2002; Tumia 2003) and radiological outcomes in four trials (Cornwall 2001; Cohen 1997; Jackson 2002; Millett 1995). In the later report of Cohen 2001, assessment of movement restriction and muscle strength after cast removal was performed by a clinician who was not informed of the cast type used; however, the earlier report of the trial emphasised that blinding was not practical (Petty 1998). Independent functional assessment was reported in McAuliffe 1987.

As implied in the 'Description of included studies' section, the description of baseline characteristics (item 4) was often sparse or even non-existent. Where quantitative data were provided for all those recruited, rather than for those analysed, there often remained insufficient evidence to confirm comparability of groups. Where differences, often of fracture type, were evident, no adjustments were made.

Blinding of patients and treatment providers (items 5 and 6) is unlikely in these trials and none was claimed.

Comparability of care programmes (item 7) comprising interventions other than the trial interventions proved hard to confirm. In particular, clear differences in the type of anaesthesia used during manipulation, the variation in the timing of the interventions (both start and duration), as well as the lack of general information on the comparability of the experience of the health professional applying the interventions (operator bias), meant that most studies scored low for this item.

Linked with description of participants' characteristics were the trial inclusion and exclusion criteria (item 8) which together help to define a study population. Many trials provided sufficient criteria to define the study population. In contrast to our review of surgical treatments, an explicit description of fracture type was not necessarily required to obtain the top score for this item.

The definition (item 9), adequacy (item 10), including active follow up, as well as the appropriateness of the timing (item 11) of outcome measurement were also poor. Rather than providing data for individual functional outcomes, many trials provided overall functional assessments using often modified schemes of scoring systems which often included anatomical and clinical outcomes. The variety of schemes used is evident from inspection of the 'Characteristics of included studies' table. Length of overall follow up, when defined, ranged from five weeks (Aladin 2001; Wahlstrom 1982) to three years (Millett 1995). Seventeen studies had less than six months overall follow up and some participants of two studies (Gupta 1991; Sarmiento 1980) were followed up less than six months. Though follow up in Jackson 2002 was claimed for six months, the results available so far only pertained to follow up at 28 days.

Trial quality assessment table (Items 1-11 described in Table 2)

(1) Reduction of displaced fractures

(a) Manipulation versus no manipulation
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 1 0 1 0 0 3 3 3 1 0 Kelly 1997

(b) Delayed manipulation versus immediate manipulation
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 0 0 0 0 0 0 0 0 0 0 McMillan 1996

(2) Plaster or brace management

(a) No immobilisation (minimal support) versus immobilisation
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 0 0 0 0 0 1 3 1 1 1 Abbaszadegan 1989
0 0 1 1 0 0 0 3 1 1 0 Dias 1987

(b) Forearm held in different positions by plaster
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 1 0 1 0 0 0 1 1 1 3 Blatter 1994
1 0 0 0 0 0 0 0 0 0 0 Gibson 1983
1 0 0 0 0 0 0 1 1 1 1 Gupta 1991
1 0 0 1 0 0 0 1 1 0 1 Van der Linden 1981
1 0 0 0 0 0 0 1 1 0 0 Wahlstrom 1982
1 3 0 3 0 0 0 3 1 1 0 Wilson 1984

(c) Plaster or synthetic cast type A versus type B
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 0 0 1 0 0 0 0 1 0 0 Aladin 2001
1 0 1 0 0 0 0 1 1 0 0 Cornwall 2001
0 0 1 1 0 0 0 3 1 1 0 Dias 1987 (repeat)
1 0 1 0 0 0 0 1 1 1 0 Jackson 2002
1 0 0 0 0 0 0 0 0 0 1 Kongsholm 1981
1 1 0 0 0 0 0 3 1 1 1 Sorensen 1986
1 0 0 1 0 0 0 1 1 0 1 Van der Linden 1981 (repeat)

(d) Forearm held in different positions by brace
1 2 3 4 5 6 7 8 9 10 11 Study ID
0 0 0 0 0 0 0 0 1 0 0 Sarmiento 1980
0 1 1 1 0 0 0 1 3 1 1 Stewart 1984

(e) Brace versus plaster cast
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 1 0 0 0 0 0 3 3 1 1 Bunger 1984
0 0 0 0 0 0 3 3 3 3 1 De Bruijn 1987
1 0 0 0 0 0 0 3 1 0 0 Ferris 1989
1 0 0 0 0 0 0 0 0 0 0 Gibson 1983 (repeat)
1 0 0 0 0 0 0 1 1 0 1 Ho 1986
1 1 1 1 0 0 1 3 1 1 1 Ledingham 1991
1 1 1 1 0 0 0 3 3 1 1 Moir 1995
0 1 1 1 0 0 3 3 3 3 0 O'Connor 2002
0 0 0 0 0 0 0 0 1 0 0 Ross 1984
0 1 1 1 0 0 0 1 3 1 1 Stewart 1984 (repeat)
3 1 1 3 0 0 0 3 3 1 1 Tumia 2002

(f) Different casting materials or techniques, or both
1 2 3 4 5 6 7 8 9 10 11 Study ID
1 0 1 1 0 0 0 3 1 1 0 Cohen 1997
3 1 0 0 0 0 0 1 3 1 0 Cohen 2001
0 0 0 0 0 0 0 1 1 0 0 Nielsen 1981
3 1 0 0 0 0 0 3 1 1 0 Rosetzsky 1982

(g) Different durations of immobilisation
1 2 3 4 5 6 7 8 9 10 11 Study ID
0 0 0 3 0 0 0 3 1 1 1 Christensen 1995
1 1 0 1 0 0 0 3 1 1 0 Davis 1987
0 0 0 0 0 0 3 3 3 3 1 De Bruijn 1987 (repeat)
1 1 0 1 0 0 1 3 3 1 1 Jensen 1997
0 1 0 1 0 0 0 1 1 1 1 McAuliffe 1987
1 1 1 1 0 0 1 3 3 1 3 Millett 1995
1 0 0 0 0 0 0 1 1 1 1 Stoffelen 1998
1 1 0 0 0 0 1 3 1 1 1 Vang Hansen 1998

Effects of interventions

The outcomes reported in the included trials' reports are listed in the 'Characteristics of included studies' table. These are grouped by anatomical, functional and clinical outcome. Where available, data by intervention group have been presented for complications in this table as well as graphically. In general, most continuous data, commonly presented for anatomical outcomes, were incomplete: standard deviations and, less often, denominators were missing. As noted before, functional outcomes were often combined and presented within an overall scoring system. Aside from complications, other clinical outcomes were rarely reported separately.

The poor methodological quality, the heterogeneous nature of the interventions, patient populations, and incomplete results of trials within each comparison group hinder quantitative analysis. Care must be taken in the interpretation of the results presented below, ordered by the comparisons given in the 'Description of studies' section.

(1) Reduction for displaced fractures
The use or delay in use of manipulation was evaluated in two studies (Kelly 1997; McMillan 1996) involving 110 people with displaced Colles' fracture.

(a) Manipulation versus no manipulation
Kelly 1997 compared manipulation under Bier's block anaesthesia with no manipulation in 30 people, aged 65 years or above, with moderately displaced Colles' fracture. Both groups had plaster cast immobilisation for five weeks. Initial fracture displacement, defined by radial shortening, dorsal angulation and radial tilt, was reported as comparable in the two groups, although radial tilt was statistically significantly lower in the manipulation group (mean 4.0 versus 9.5 degrees). Final displacement at plaster removal at five weeks was similar; notable was the mean loss in radial length of about two millimetres from the fracture position in both groups. Approximate values for mean values and standard deviations of these three parameters abstracted from graphs confirmed that none of the differences were statistically significant (not shown) at five weeks despite the clear post-manipulation reduction in the manipulated group. No action was taken to correct the two fractures in each group which displaced beyond the limits of the study as they were considered to have united. There were no significant differences in functional outcome, either in stiffness, grip strength, pain or overall (see Graph 01.01) at 13 weeks. Likewise, complications (see Graph 01.02) and cosmetic deformity (see Graph 01.03) whether assessed subjectively or objectively were similar in the two groups. Kelly 1997 concluded that manipulation was unnecessary in selected elderly patients with only moderately displaced fractures.

(b) Delayed versus immediate manipulation
McMillan 1996 compared the effect of delaying manipulation for one week to allow initial swelling to subside with immediate manipulation in 80 people aged over 50 years, with displaced Colles' fracture. No quantitative data were available for this trial which was only published as an abstract. McMillan reported that the anatomical results in the two groups were similar, with no increase in complications nor discomfort in the delayed manipulation group. Contact with one of the trialists resulted in no new information about this trial.

(2) Plaster or brace management

(a) No immobilisation (minimal support) versus immobilisation
Two trials (Abbaszadegan 1989; Dias 1987) tested this comparison in 177 people with minimally or undisplaced fractures. The immediate application of elastic (Abbaszadegan 1989) or crepe bandage (Dias 1987) was compared with plaster cast immobilisation for four and five weeks respectively.

Bandage use did not appear to be detrimental to anatomical outcome in either study. Abbaszadegan 1989 reported a statistically significantly lower mean radial shortening in the bandage group at eight weeks (0.15 mm versus 0.55 mm). Dias 1987 found no difference in 'bony deformity', nor was there a difference in anatomical grades between the two groups (see Graph 02.01).

Recovery of function (grip strength, range of motion) was better with less pain in the bandage group in Abbaszadegan 1989 at all times (incomplete data). Dias 1987 also reported a more rapid and sustained recovery of function by three months. Overall functional grades were better in the bandage group (see Graph 02.02) in Dias 1987. There were no significant differences in the complications occurring in either group (see Graph 02.03).

Both studies concluded that bandage use enabling earlier mobilisation was safe with a potential for a faster and better functional outcome for patients with undisplaced and uncomplicated Colles' fracture.

(b) Forearm held in different positions by plaster
There were six studies (Blatter 1994; Gibson 1983; Gupta 1991; Van der Linden 1981; Wahlstrom 1982; Wilson 1984) involving around 650 people in this category. These compared three aspects of wrist position: supination or neutral versus pronation; palmar or neutral flexion versus dorsiflexion; and no ulnar deviation versus ulnar deviation.

Supination (or neutral) versus pronation
Three trials (Gibson 1983; Wahlstrom 1982; Wilson 1984) involving around 150 people with, probably all displaced, Colles' fractures compared the wrist held by plaster in supination or neutral with that held in pronation. The extension of the plaster above the elbow in Gibson 1983 and Wilson 1984 for participants of the supinated group would have more effectively held this position than in Wahlstrom 1982 where a below-elbow plaster was used for all three groups. There were no qualitative data available for Gibson 1983 which was only published as an abstract. Gibson 1983 reported that there was no significant difference in functional outcome at 20 weeks between the two groups. There was no report of anatomical or clinical outcomes. The follow-up period in Wahlstrom 1982 was only five weeks, which probably covered the time of plaster immobilisation. Only anatomical outcome (see Graphs 03.01 and 04.01) and rereduction (see Graphs 03.03 and 04.02) were recorded for the three groups with wrists held in supination, neutral ("mid-way") and pronation respectively. Wrist anatomy after reduction held better with fewer displacements (10 degrees or above) and need for further reduction in the pronated group compared with either the neutral or supinated groups. At plaster removal, trial participants immobilised in an above-elbow plaster in supination had less loss of radial deviation and volar tilt from the reduced position (see Graph 03.01) than those in the pronated group in Wilson 1984; these differences were not statistically significant. Neither were the results of the functional assessment (see Graph 03.02). Given the lack of significant difference in either anatomical or functional outcome, Wilson 1984 who mentioned the extra effort in applying an above-elbow splint, the extra weight and cumbersome nature of this and associated finger stiffness, concluded that the above-elbow plaster with the forearm in supination could not be recommended.

Palmar or neutral flexion versus dorsiflexion
Two trials (Blatter 1994; Gupta 1991) involving 254 people with displaced fractures compared wrist positioning in either palmar flexion or neutral with dorsiflexion. Anatomical results for both trials were better in the dorsiflexion group. Overall functional outcome was also better in the dorsiflexion group as shown by the lower proportion of patients with a poor or only fair overall outcome or significant disability (see Graphs 05.01 and 05.02). Gupta 1991 reported without data, that the palmar flexion group had the most redisplacements, and more hand joint stiffness. Aside from a statistically non-significant higher proportion of problems associated with plaster wear (see Graph 05.03), mainly pressure pain encountered in the dorsiflexion group, all other outcomes in Blatter 1994 favoured the dorsiflexion group.

No ulnar deviation versus ulnar deviation
Five variations of plaster management were compared in 250 people with undisplaced and displaced fractures in Van der Linden 1981. Here ulnar deviation was compared with no ulnar deviation and a dorsal splint with a full plaster cast. Insufficient data were presented to allow confirmation of the lack of significant differences between treatment groups for both anatomical and functional results at eight-months follow up. Van der Linden 1981 concluded that the method of immobilisation (ulnar deviation or full plaster cast) was of only minor importance and that final outcome depended more on the original displacement of the fracture and success of reduction.

(c) Plaster or synthetic cast type A versus type B
Some of the above set of trials such as Wilson 1984 also belong to this category as different casts were used to maintain specific wrist positions with associated differences in arm mobility. Seven trials (Aladin 2001; Cornwall 2001; Dias 1987; Jackson 2002; Kongsholm 1981; Sorensen 1986; Van der Linden 1981) are reviewed for this category as these compared different cast types independent of wrist position.

In Dias 1987, 90 people with displaced Colles fractures requiring reduction were allocated to either a modified cast allowing greater wrist movement or to a standard plaster cast for five weeks. Dias 1987 observed an early phase of rapid deterioration in wrist anatomy, followed by a slower deterioration which continued after cast removal. 'Bony deformity' and anatomical grades (see Graph 06.01) were similar in the two groups. Overall functional grades (see Graph 06.02) were better in the modified cast group, and a faster recovery of wrist movement and resolution of wrist swelling, better grip strength and less pain were reported for this group. There was no specific mention of complications. Dias 1987 concluded that the early mobilisation of the wrist using the modified cast hastened functional recovery without adversely affecting anatomical outcome in patients with displaced fractures.

Aladin 2001 assessed radiological outcome, up to five weeks, of 126 people with 128 fractures that were immobilised in either a modified sugar-tong cast that blocked forearm rotation or in a below-elbow cast. Though there was a statistically significant difference in the mean values for dorsal angulation of the two groups, the actual difference (mean difference = 2.3 degrees) is not clinically significant (see Graph 07.01). There was no difference between the two groups in the incidence of unacceptable anatomical outcome or redisplacement at five weeks (see Graphs 07.02 and 07.03).

Jackson 2002 compared modified sugar-tong fibreglass splints versus volar-dorsal fibreglass splints versus cylindrical plaster casts in 101 adults with displaced extra-articular fractures that had been successfully reduced. Only data for radiological outcomes were presented in the two abstract reports of this trial. Though there were some discrepancies between the two abstracts, it seems likely that the results for redisplacement applied at one-week follow up. There was no statistically significant difference between the two groups in either of these outcomes for any of the three comparisons (see Graphs 08.01, 09.01 and 10.01).

Cornwall 2001 evaluated the radiological outcome, up to fracture healing, of immobilisation in an above-elbow plaster cast with a below-elbow plaster cast in 127 adults. There were insufficient data available from the conference abstract report of the trial to confirm the lack of statistically significant differences between the two groups. Sorensen 1986 compared immobilisation in a full above-elbow plaster cast with a below-elbow dorsal slab in 100 people with displaced distal radial fractures. Anatomical outcome at six weeks (see Graph 11.01) was better in the dorsal slab group, but otherwise there were no significant differences in either functional outcomes or complications (see Graphs 11.02, 11.03 and 11.04). Like Wilson 1984 (see above), Sorensen 1986 concluded that the greater ease of application of a dorsal slab made it preferable.

Kongsholm 1981 and Van der Linden 1981 compared application of a dorsal splint with a full below-elbow plaster cast. As stated above, insufficient data are available to confirm the lack of significant difference in either anatomical or functional outcomes between these two cast types in Van der Linden 1981. Neither were there data available in the Kongsholm 1981, reported only in a conference abstract, to test if the higher proportion of recurrent fracture dislocation in the dorsal splint group and the reported better end results for the circular (full plaster cast) plaster group were statistically significant.

(d) Forearm held in different positions by brace
Two trials (Sarmiento 1980; Stewart 1984) involving people with 298 fractures compared above-elbow brace which held the forearm in supination (hand faces upwards) with another brace either above-elbow holding another position (Sarmiento 1980) or a less restrictive below-elbow brace (Stewart 1984). As these comparisons were dissimilar, and even more so in the initial plaster management (above-elbow for Sarmiento 1980; back slab for Stewart 1984), these two trials are treated separately below.

Sarmiento 1980 varied the time of immobilisation in plaster and then brace according to fracture type. Anatomical and functional outcome were represented, without data, as being not significantly different between the two groups for undisplaced fractures. Of the 93 displaced fractures, only data for 71 were given for trial follow up; which varied between four to 60 weeks. Fewer fractures in the supinated group had excessive loss in anatomical position (see Graph 12.01), resulting in fewer with only fair or poor overall anatomical grades (see Graph 12.02). Overall functional outcome was also better in the supinated group (see Graph 12.03). Complications were not presented separately by treatment group but of note is that the only major complication, that of Sudeck's atrophy, occurred in a trial participant with an undisplaced fracture. The favourable results for an above-elbow supinated brace prompted others such as Stewart 1984 to test the so-called 'Sarmiento brace'.

Stewart 1984 compared the Sarmiento brace with a below-elbow brace which did not fix the hand position, for displaced fractures. Anatomically there was no difference in outcome between the two groups (see Graph 13.01). Similarly, no statistically significant differences between the two groups in functional outcomes, including poor finger flexion (see Graph 13.03), were reported. Separate data for the two groups for complications and problems of brace use were not presented.

(e) Brace versus plaster cast
No quantitative data were available for the three trials published as abstracts (Gibson 1983; Ho 1986; Ross 1984), nor for De Bruijn 1987 where the results for two separate trials involving different comparisons were lumped together. The other seven trials (Bunger 1984; Ferris 1989; Ledingham 1991; Moir 1995; O'Connor 2002; Stewart 1984; Tumia 2003) involved 1000 people, which is just over one half of the total for this general comparison. Two pairs of trials (De Bruijn 1987 and Ho 1986; Moir 1995 and Tumia 2003) tested comparable interventions. None of the other trials compared the same brace or brace application including hand position and restriction of movement. Due to the variation of the interventions under comparison and the lack of compatible outcome data, no summary data are presented for this section.

Bunger 1984 compared an above-elbow functional brace with forearm in supination versus a below-elbow plaster in 150 people. Early treatment, particularly wrist position varied according to whether the fracture was displaced and unstable. Anatomically, the brace group had better results with significantly less displacement and less subsequent reduction at two weeks (see Graph 14.09) and a better overall end result at six weeks, in particular regarding dorsal angulation (see Graphs 14.01). Improvement in functional outcomes with time was noted in both groups, a significant difference in overall functional outcome only evident at six-months follow up when more participants in the brace group had a good or excellent outcome. Early complications (see Graph 14.01), mainly neural, were similar in both cases. Without supporting data, Bunger 1984 reported that both displaced and unstable fractures showed better functional results at six months in the brace group. They proposed that the benefit arose primarily from a reduction in fracture redisplacement, a complication that occurred early in the treatment. The good results of both groups were attributed to participation in the trial.

De Bruijn 1987 and Ho 1986 compared an above-elbow backslab replaced at one week (in De Bruijn 1987), or one to two weeks (Ho 1986), by an above-elbow functional brace in supination with a below-elbow backslab in approximately 280 people with Colles' fracture. The duration of immobilisation was four weeks in De Bruijn 1987 and six weeks in Ho 1986. As explained above there are no data available to check the stated conclusion in De Bruijn 1987 that there was no place for treatment with the above-elbow functional brace. Nor are there data to examine the counter claim in Ho 1986, based respectively on the results of two thirds and one half of the trial participants, that functional bracing gave better anatomical results and improved short term functional results.

Ferris 1989 compared a four-part below-elbow brace with forearm cast in 47 people with a Colles' fracture. At splintage removal at five weeks the dorsal displacement from reduced position was less in the brace group (see Graph 14.01). The significantly enhanced functional outcome (range of movement, grip and torque strength) at five weeks in the brace groups was reported to be not significant at nine weeks. Ferris 1989 considered that the brace was well tolerated but noted the extra medical supervision required for the brace group such as for the hand swelling encountered in all brace-group participants in the first week that had required adjustment to the brace (see Graph 14.09).

Gibson 1983 compared a functional brace in supination with either a supinated above-elbow cast or pronated-elbow cast in 105 people with a Colles' fracture. They observed an earlier recovery of function in the brace group at nine weeks, which had disappeared at 20 weeks. There was no report of complications or other outcomes.

Ledingham 1991 compared a functional plaster cast brace with a standard plaster cast in 60 people with a displaced Colles' fracture. Both anatomically and functionally (see Graphs 14.02 and 12.08) the brace group was better before twelve weeks. By six months, five plaster-cast group participants compared with no brace-group participants were graded as poor or only fair in function (see Graph 14.08). Ledingham 1991 considered that there was no significant difference in function by this time. Six cases of superficial radial nerve paraesthesia (numbness) occurred in the brace group, a likely consequence of the brace used. These were transient and all resolved. Ledingham 1991 considered the brace had advantages but that it required a degree of skill and experience in its application. As a consequence, a prefabricated brace that was easier for inexperienced staff to apply was under development (see Moir 1995).

Both Moir 1995 and Tumia 2003 compared the "Aberdeen Colles' fracture brace", a prefabricated functional brace applying three-point loading, with a standard plaster cast. Despite similarities in the measurement of outcome, lack of full and compatible data precluded pooling. Separate reports of the findings of the two trials are given below.

All 85 participants of Moir 1995 had manipulation of their displaced Colles' fracture. Anatomical and functional outcomes were generally summarised as medians in the comprehensive report of this trial. Moir 1995 noted that whilst the brace group was initially less severely displaced, there were more intra-articular fractures. Both groups were reported as losing equivalent positions during splintage. The brace group was indicated as having a superior functional result at six-months follow up as well as better function (grip, pinch strength, pain, finger motion) during splintage. Complications of brace use included three cases of superficial radial nerve paraesthesia, three cases of discomfort requiring brace removal and one incorrect application (see Graph 14.09). In all, Moir 1995 concluded that this brace, which was adjustable and was intended for use by less experienced staff, was preferable. These conclusions were questioned by Moiz 1996 who pointed out inconsistency in the claims of statistical significance for pinch strength at three and six-months follow up as well as pointing out that the criterion for assessing finger movement was inappropriate for patients with plaster casts. Moiz 1996 also queried the cost implications of the brace.

Tumia 2003 extended the examination of the Aberdeen brace through a multicentre and international trial involving 339 people with minimally displaced (non-manipulated) or displaced (manipulated) fractures. The pre-publication report of this trial presented separate results for non-manipulated and manipulated fractures; these data were, however, incomplete together with a few internal inconsistencies in the report. No significant difference in anatomical scores at seven weeks was reported. Relative grip strength was reported as significantly better during brace use but not subsequently; the grip strength results for the two treatment groups were very similar at six months. Pain scores data were inconsistent between text and table in the report but appeared similar in the two treatment groups. No statistically significant differences between the two groups were reported for overall functional scores at three follow-up times. Given the similarities in anatomical and longer term functional results of the brace and plaster cast groups, Tumia 2003 highlighted the alleged ease of application of the Aberdeen brace, the superior grip strength during brace use and the avoidance of cast changes. There were insufficient data, including no cost or procedural results and no information on complications, to confirm these claims. Tumia 2003 indicated that a further development of the brace was in progress.

O'Connor 2002 compared a lightweight and removable wrist splint with a below-elbow plaster cast in 76 people with minimally displaced Colles' fracture. One person in each group required manipulation for secondary displacement (see Graph 14.09). No difference in radiological outcome at 12 weeks was reported. Participants of the splint group had greater mobility by the time of splint removal at six weeks: the differences between the two groups in extension, flexion and pronation were statistically significant (see Graph 14.03). Differences in range of movement between the two groups had disappeared by 12-week follow up (see Graph 14.04). No statistically significant differences between the two groups were reported for grip strength or in the results of a subjective assessment rating each trial participant's ability to perform 15 activities of daily living. Overall functional scores were better in the splint group at six weeks; on average the splint group had a good score whereas the plaster cast group had a fair score according to the rating system of Stewart et al (Stewart 1985). Again, there was no statistically significant difference in the functional scores by 12 weeks (see Graph 14.06). There were similar numbers of people in the two groups with specific complications. Though the clinical implications of the actual differences in visual analogue scores are not clear, the participants of the splint group tended to be less dissatisfied and found fewer problems with their splintage than those in the cast group (see Graph 14.10). Splint group participants registered slightly more pain at two weeks but slightly less at six weeks. The splint, which cost almost the same as a plaster cast, took less time to apply and could be adjusted as required: nearly half (14 out of 32) of the participants of the cast group had a change of cast during the six weeks treatment period.

Ross 1984 compared an Orthoplast functional brace in supination versus a standard plaster cast in 405 people. The anatomical difference between injured and uninjured arms was reported to be less in the brace group. As well as being younger, participants of the brace group were reported as having less pain, less disability, less weakness and better range of movement at three to four months follow up. Ross 1984 recommended the brace, but without providing supporting data.

Stewart 1984 compared a below-elbow brace or above-elbow brace in supination with a standard plaster cast in 243 people with displaced Colles' fracture. For this comparison the results of the two brace groups have been lumped together. Anatomically there was no significant difference reported for the three groups, though there were proportionally fewer people with only fair or poor overall anatomical grades in the combined brace groups (see Graph 14.02). Similarly, functional outcome including poor finger flexion (see Graph 14.05) was reported as being uninfluenced by the method used. Stewart 1984 noted the disproportionate number of return visits (38) of braced group participants compared with only three in the plaster cast group because of problems with the splintage. The report concluded that there was no reason to change from the use of standard plaster casts for uncomplicated Colles' fracture.

(f) Different casting materials or techniques, or both
Four trials evaluated materials or techniques used for below-elbow casts. Three (Cohen 1997; Nielsen 1981; Rosetzsky 1982) compared a new material applied in different way with a standard plaster or fibreglass cast. The fourth trial (Cohen 2001) compared a new casting technique, focused rigidity casting, with standard casting. The four trials emphasised different aspects of cast function or focused on different inadequacies of standard materials or techniques, or both. These latter included the heavy weight, water solubility, limited strength, poor X-ray transmission and rigidity of plaster casts and rigidity and need for cast changes for fibreglass and other standard casts. All four trials actively obtained feedback on the problems and satisfaction of cast use from trial participants, as well as discussing the relative costs of the casts involved at the time of the trial. However, there were only scant details on functional outcomes with follow up extending only up to the time of cast removal.

Polyurethane cast with "zipper" versus cylindrical plaster cast.
Rosetzsky 1982 compared immobilisation for six weeks using a below-elbow polyurethane cast with "zipper" with a below-elbow plaster of Paris cylinder cast, in 50 people with undisplaced or displaced distal radial fractures. The potential advantages of the polyurethane cast were that it was lighter, water repellent, hardened faster and enabled, as was done in the trial, reduction of displaced fractures after cast application while the cast set. The zipper also allowed some adjustment and easy removal. No specific anatomical data were reported. Fewer participants in the polyurethane group failed to retain reduced fracture position (see Graph 15.01) but there was no need in either group for rereduction of the displaced fracture. Secondary adjustment of casts resulting from problems such as restricted finger movement was similar in both groups as were the number of trial participants who noted problems, often local pressure points, during cast use (see Graph 15.01). No incidences of skin damage occurred. Rosetzsky 1982 reported that the polyurethane cast was twice as costly as a standard plaster cast and moreover required considerable practice to make and apply. Given these considerations, Rosetzsky 1982 concluded that the polyurethane cast should only be used in special cases specifically where low weight and water insolubility were prime considerations.

Hexelite (thermoplastic) bandage versus plaster cast
Nielsen 1981 compared immobilisation in a thermoplastic polyester bandage applied after heating in hot water with a standard plaster cast in 183 people, of which 55 had Colles' fractures. There was no report of anatomical or functional outcomes. Subjective complications such as pain, skin pressure, poor hand function and poor stability during cast wear, were similar in both groups with the exception of water intolerance which was, not surprisingly, worse in the plaster group (see Graph 16.01). There were more cases of skin damage in the thermoplastic group. As the thermoplastic material was eight times more expensive and more difficult to handle, without clear advantages, Nielsen 1981 abandoned its use after the trial.

QuickCast (shrinkable polymer) versus fibreglass tape cast
Cohen 1997 compared immobilisation for up to 6.5 weeks using a polymer-coated casting material which was shrunk using hot air to fit the arm against a short-arm fibreglass tape cast, in 30 people with undisplaced or displaced distal radial fractures. Percutaneous pins were used to stabilise fractures in 10 people. As indicated above, the surgical treatment of these people places them in a different category where additional considerations such as the complications of pin track infection come into play. Cohen 1997 recognised this but although randomisation was stratified by the severity of the fracture, most of the results were presented for the whole group. No significant difference between groups was observed in the final redisplacement from the reduced position (see Graph 17.01: note denominators used are for the whole group). In all, three fractures were considered to have redisplaced, one, belonging to the QuickCast group, of which required further treatment (see Graph 17.02). Similar numbers of complications occurred in both groups, although only the QuickCast group had burns resulting from the process of cast application. Trial participants considered they had marginally more problems in the fibreglass group but rated their satisfaction similarly. The key difference was that on average there was one fewer cast change in the QuickCast group, mainly as further heat could be used to reshape the cast as required. Cohen 1997 noted that QuickCast was over twice the cost of a fibreglass cast but that this would be probably outweighed by the reduction in time and labour resulting from fewer cast changes.

Focused rigidity casting versus standard casting
Cohen 2001 compared focused rigidity casting (FRC), where elasticated synthetic tape was applied to confer relative rigidity at the fracture site but more flexibility elsewhere, with standard casting in 200 people, 35 of whom had undisplaced fractures of the distal radius. Length of follow up, up to cast removal, averaged four weeks, ranging from 14 to 42 days overall. There were no cases of fracture displacement, delayed or non-union or adverse effects recorded in the limited follow-up period (see Graph 18.02). There were no statistically significant differences between the two groups in movement limitations or loss of muscle power at cast removal (see Graph 18.01). One person in each group found their cast excessively restricting and one person in the FRC group found their cast very uncomfortable but no one expressed dissatisfaction overall. Based on the results of a specially developed scoring scheme, the Bradford Plaster Index, aimed at assessing activities of daily living, satisfaction, overall functional impairment and comfort during cast use, Cohen 2001 reported that FRC-group participants perceived they were more able to perform activities of daily living and had a greater satisfaction with their ability to perform these. Separate data for distal radial fracture patients were unavailable for cast changes and adjustments (only an option for FRC). A saving in the amount of casting tape used and the facility for re-application of the same cast after examination was noted for the FRC method.

(g) Different durations of immobilisation
Seven trials compared the use of plaster cast for different durations. Four trial studies (Christensen 1995; McAuliffe 1987; Millett 1995; Vang Hansen 1998) compared three to four weeks versus five to six weeks immobilisation, one trial (Davis 1987) compared one to two weeks versus four to five weeks immobilisation and two trials (Jensen 1997; Stoffelen 1998) compared one versus three weeks immobilisation. De Bruijn 1987 compared bandage with an above-elbow functional brace for three weeks, after a week-long application of a below-elbow backslab and above-elbow backslab respectively. As explained above, De Bruijn 1987 failed to provide supporting data for the conclusions given in this trial and is not considered further in the following.

Christensen 1995 and Vang Hansen 1998 compared three-weeks with five-weeks immobilisation in a plaster back slab, in 133 people. All 33 participants of Christensen 1995 and over half in Vang Hansen 1998 had minimally displaced distal radial fractures. Both trials reported that there were no statistically significant differences in either anatomical or functional outcomes, but only data for long-term pain were available (see Graph 19.01). Aside from tendon rupture (see Graph 19.04), separate data for complications were not given.

McAuliffe 1987 and Millett 1995 included both undisplaced and displaced fracture patients, and involved 108 and 90 participants respectively. In McAuliffe 1987, fractures were remanipulated if necessary, after an initial application of a cast for one week. Immobilisation in a plaster cast was then continued for a further three weeks or five weeks. There were no significant differences reported between groups for anatomical outcomes at either three or twelve months. Overall assessments by trial participants of pain, disability and wrist mobility were comparable (see Graph 19.02). However significantly less pain and greater grip strength were reported in the early group when assessed by an independent physiotherapist. An improvement in range of movement was also noted in the early group. There was no mention of complications. Millett 1995 compared three-weeks immobilisation followed by two weeks in a flexible cast with five-weeks plaster immobilisation. Anatomical and clinical outcomes were presented at various time intervals up to three years but, as the numbers of trial participants assessed at these times were not given, only exploratory analyses could be done to test the claims in the report. Radiological examinations at three years were reported to show a statistically significant increase in the average radial tilt (2 degrees) and decreased radioulnar joint space (0.3 mm) in the five-week group, but no statistically significant difference in numbers with osteoarthritis. The regain of grip strength and wrist movement was more rapid in the early-mobilisation group. The higher joint mobility and mean grip scores in the early -mobilisation group were reported as reaching statistical significance at three and six months respectively. An exploratory analysis, based on intention-to-treat analysis, for grip strength at three and six months shows the difference between the two groups reaching statistical significance at three months (see Graph 19.03). By three years, no statistical differences between the two groups were reported in pain, grip strength, residual deformity, hand swelling, disability and overall complication rate. Millett 1995 concluded that the short-term improvement in functional recovery, lack of detrimental evidence and patient preference supported early mobilisation.

Davis 1987 compared tubigrip bandage versus plaster cast for three weeks after the application of a back slab for 7 to 13 days. Though Davis 1987 did not report radiological outcomes, there was no difference in the number of secondary fracture displacements (2 versus 3; see Graph 20.05) with the two cases of cosmetic deformity occurring in the plaster group. Short-term functional outcome was similar for pain and mean grip strength (see Graphs 20.01 and 20.02) in both groups. Overall functional outcome was superior in the bandage group, with an earlier recovery of domestic skills (incomplete data). Overall functional grades were better in the bandage group (see Graph 20.03). There was no major difference in the complications occurring in either group though more occurred in the plaster group (see Graph 20.04). Finally, Davis 1987 reported greater patient satisfaction with bandage use (see Graph 20.05) and concluded that bandage use enabling earlier mobilisation was safe with a potential for a faster and better functional outcome for people with undisplaced and uncomplicated Colles' fracture.

Based on the favourable conclusions cited in trials testing early mobilisation (Christensen 1995; McAuliffe 1987), or no immobilisation (Abbaszadegan 1989; Dias 1987), Jensen 1997 and Stoffelen 1998 went further by comparing one-week with three-weeks immobilisation in plaster in 114 people with undisplaced or minimally displaced Colles' fractures. Radiological examination of 48 trial participants at six months in Jensen 1997 showed no cases of non-union, and no statistically significant difference in excess angulation or axial radial shortening (see Graph 21.01). Functional outcome in only one participant of the three-weeks group was adversely affected by excessive angular displacement. Stoffelen 1998 reported no further displacement of fractures in either group. Final functional outcome assessed within an overall functional scoring system was similar in both groups for both trials (see Graphs 21.02 and 21.03). Additional pain was experienced by trial participants at cast removal at one week, and the need for a mild pain reliever was advised in Jensen 1997. Pain preventing the removal of plaster was not experienced by any participants of Stoffelen 1998. This trial was initially intended to compare the use of removable splints with plaster immobilisation for three weeks, but inhibiting pain experienced by patients in the first days after a fracture, prompted use of plaster immobilisation for the first week. Separate results for complications were not provided by Jensen 1997. Though more cases of algoneurodystrophy (reflex sympathetic dystrophy) occurred in the group immobilised for three weeks in Stoffelen 1998, this was not statistically significant (see Graph 21.04). Both Jensen 1997 and Stoffelen 1998 considered that undisplaced or minimally displaced Colles' fractures could be safely treated with one week immobilisation in a dorsal plaster cast, but emphasised that care had to be taken in selecting patients.

Discussion

Given the frequent occurrence of distal radial fracture in adults and the general use of conservative interventions in the management of the more common types of these fractures, the overall number of randomised trials (37) evaluating various conservative treatment options and associated number of participants (just over 4200) are comparatively few. Very few trials were comparable in terms of inclusion criteria and patient characteristics, primarily fracture type, the interventions compared, overall care programs or outcome measurement. The majority of trials were methodologically flawed such that there was a real prospect of serious bias. Indeed, perhaps through under reporting, the possibility of selection, performance, exclusion or ascertainment bias, or a combination of these could not be ruled out for any trial. Much of the emphasis in this discussion inevitably refers to methodology rather than drawing out conclusions of effectiveness of the interventions under investigation.

While small trials with under 50 participants in each intervention group may provide robust evidence of effect in some areas of health care, it is unlikely that any of these small trials could provide conclusive evidence to establish the superiority of one intervention over another. Furthermore the apparent comparability of results of interventions tested within some trials should not be interpreted as evidence of no effect or no difference. Apart from trial size, the general lack of allocation concealment, the lack of assessor blinding and, less commonly, severe losses to follow-up, all serve to undermine the validity of the trial results.

Other issues which are more specific to this subject involve the severity of injury and fracture type, the execution of the interventions and associated care programmes, and the usefulness and relevance of outcome measurement.

The term 'distal radius fracture' covers a range of injuries which includes the commonly referred 'Colles' fracture. It is commonly perceived that the choice of intervention will depend to a large extent on the severity and type of injury. At a basic level, this is often viewed in terms of whether the, usually closed, fracture is displaced and whether there is articular involvement. Other factors such as patient age, comorbidity, cost, available resources and patient choice will also contribute to this decision. Thus an inadequate description of inclusion criteria, a lack of baseline characteristics presented by intervention group and/or an important but unadjusted difference in patient characteristics, do not assist an evaluation of the applicability of trial results. The variety of fracture classification systems, with associated issues of reliability and validity further complicates this area (Jupiter 1997). However, in general these trials testing conservative interventions involved closed minimally to moderately displaced Colles' fractures in the elderly for which conservative rather than surgical treatment is conventionally applied. Other types of distal radius fracture, such as those associated with major violence which may be compound in nature, or the 'reverse Colles' or Smith's fracture, are far less common and no trial specifically focused on these.

Incomplete description of trial interventions and the measures taken to provide otherwise comparable care programmes also gave rise to concern. Of particular note was the frequent lack of sufficient information to determine if there were any important differences in the timing of interventions in the context of the immediate post traumatic reaction such as swelling, the use of anaesthesia, the experience of health care workers, and general rehabilitation.

The length of overall follow up of some trials was inadequate and no trial provided a completely satisfactory presentation of outcome. In particular, the common reliance on often modified functional scoring systems and subsequent presentation by an overall score or grade, hampered interpretation of trial results. Though understandably an attractive option, the use of such unvalidated scales is questionable. Many such as those based on Gartland and Werley (Gartland 1951) were termed 'functional' despite including anatomical results. Since, in general, these latter have no proven relationship to eventual functional outcome, their inclusion will undermine a so-called summary of 'functional' outcome. Functional scales often included serious complications and some trials failed to report these separately. Finally, patient satisfaction and resource implications were rarely mentioned.

This review raised a number of fundamental questions. For example, when should a distal radius fracture be manipulated? How should the fracture be immobilised, whether with or without prior manipulation? What methods and materials should be used for immobilisation? How long should the immobilisation last? And, though not expressed in terms of a null hypothesis, should remanipulation be undertaken for any redisplacement after closed reduction and when should this take place? So far, the evidence available from randomised controlled trials is insufficient to provide answers to these questions.

Summarising the findings of this review according to the null hypotheses listed in 'Objectives' we find:

There is no conclusive evidence of difference in outcome between reduction or no reduction of displaced fractures, namely:

  • Manipulation versus no manipulation

  • Delayed manipulation versus immediate manipulation

Likewise, there is no conclusive evidence of difference in outcome between different methods of plaster and brace management listed below:

  • No immobilisation (minimal support) versus immobilisation

  • Forearm held in different positions by plaster

  • Plaster or synthetic cast type A versus type B

  • Forearm held in different positions by brace

  • Brace versus plaster cast

  • Different types of cast material

  • Different durations of immobilisation

Authors' conclusions

Implications for practice

Conventionally, practitioners treating distal radius fractures in adult patients aim to restore, as best as they can, the anatomy of the broken wrist, and to control pain, allow healing, and maintain the function of uninjured joints in the limb, by applying some form of immobilisation. Many variations of reduction/immobilisation have been suggested and used accordingly. This review has failed to find evidence for important differences in outcome between these variations.

From a health policy and health care funders' perspective it seems justifiable to ask the question as to why the least costly option for managing this common fracture should not become the norm since there is no evidence available of additional benefits for the patients from the various conservative treatment regimens?

In the meantime, practitioners applying conservative management for distal radial fracture should use an accepted technique with which they are familiar, and which is cost-effective from the perspective of their provider unit. Patient preferences and their circumstances as well as the risk of complications of the fracture and treatment should also be considered.

Implications for research

Given that a distal radius fracture in adults is a common injury and given that there is very limited knowledge about the best method of treatment, either conservative or surgical (Handoll 2005a), further research is called for. However, rather than embarking on further small, single-centre and spasmodic trials which, often with inadequate methodology and reporting, are unlikely to provide the good generalisable evidence required, it is important to develop a more coherent strategy for future research. Some thoughts on this are presented below.

Systematic reviews of the main treatment choices for these fractures are now available. These cover methods of closed reduction (Handoll 2005c), conservative interventions (this review), surgical interventions (Handoll 2005a), anaesthetic technique (Handoll 2005b) and rehabilitation (Handoll 2005d). All conclude that there is a general insufficiency of evidence to inform practice. An overview of the whole treatment process thus seems warranted.

Previously we suggested that one approach to determine the way forward for future randomised trials would be a careful and comprehensive review of the management of patients with this fracture. This could serve to standardise good quality care but also would enable people to distinguish where and on what basis important treatment decisions (such as whether to reduce a fracture) are made and to allow an examination of their underlying uncertainty. We considered that reaching a consensus on the important areas of uncertainty should help to gather momentum for larger, probably multicentre, trials.

We have now completed a project that has taken our systematic review of the evidence some way along these lines (Handoll 2003a). Essentially, we placed the evidence drawn from our reviews into the context of the care decisions made within a 'typical' care programme. Then, through a consultation exercise involving key players, including several with a personal or close familial experience of these fractures, we sought to identify key treatment questions and research priorities. While we were unable to reach a consensus in this complex area, we concluded that we had provided a template for further action and that the wealth of insights gained should help to inform a future research agenda.

In the meantime, the following points should still be considered.
(1) A deeper understanding is required of patient preferences regarding outcomes of treatment, and any trade off between benefits and adverse effects of the various methods of management. This would inform debate about the most cost-effective method of managing this common injury.
(2) Future research would be facilitated by the development and adoption of a standard core data set for classification and outcome reporting for distal radial fracture.
(3) Such a dataset should include, at minimum, a basic fracture classification (non-displaced: displaced; intra-articular: extra-articular), patient preferences, validated functional and activities of daily living outcomes, preferably rated by patients, and resource use measures in addition to traditional measures such as anatomical position and range of movement.
(4) The design and reporting of any future trial should conform to the CONSORT statement (Begg 1996; Moher 2001) or any future development of it.

Acknowledgements

We thank Noelle Murphy for her contributions in the early stages of the original version of this review. We thank Jenny Duthie for secretarial support in the original version of the review. We are grateful for the helpful comments at editorial review of the original version from the following: Bill Gillespie, Geoffrey Hooper, Gordon Murray, John Stothard, Marc Swiontkowski, Joanna Tasker. We are grateful to Bill Gillespie and Marc Swiontkowski for their helpful comments on the third update.

We thank Lesley Gillespie for her help in developing the search strategy and Leeann Morton and Bill Gillespie for their help during editorial review of the first and second updates. We thank Lesley Gillespie for her help during the editorial review of the third update.

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

Data and analyses

Download statistical data

Comparison 1. Manipulation versus control
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Functional grading: poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Redisplacement or displacement beyond study criteria1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Re-manipulation or secondary manipulation1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.3 Algodystrophy (RSD)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.4 Finger stiffness1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Cosmetic deformity1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Cosmetic deformity (subjective)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.2 Cosmetic deformity (objective)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 1.1.

Comparison 1 Manipulation versus control, Outcome 1 Functional grading: poor or fair.

Analysis 1.2.

Comparison 1 Manipulation versus control, Outcome 2 Complications.

Analysis 1.3.

Comparison 1 Manipulation versus control, Outcome 3 Cosmetic deformity.

Comparison 2. No immobilisation versus immobilisation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Functional grading2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Not excellent2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Poor or fair2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Pressure sores1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.2 Problems with bandage / cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.3 Median nerve compression (CTS)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 2.1.

Comparison 2 No immobilisation versus immobilisation, Outcome 1 Anatomical grading.

Analysis 2.2.

Comparison 2 No immobilisation versus immobilisation, Outcome 2 Functional grading.

Analysis 2.3.

Comparison 2 No immobilisation versus immobilisation, Outcome 3 Complications.

Comparison 3. Cast position: supination versus pronation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement2 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 Mean loss of volar tilt (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.2 Mean loss of radial deviation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.3 Mean loss of dorsal angulation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
2 Functional grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Redisplacement (10+ degrees)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.2 Rereduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 3.1.

Comparison 3 Cast position: supination versus pronation, Outcome 1 Anatomical displacement.

Analysis 3.2.

Comparison 3 Cast position: supination versus pronation, Outcome 2 Functional grading.

Analysis 3.3.

Comparison 3 Cast position: supination versus pronation, Outcome 3 Complications.

Comparison 4. Cast position: neutral (mid-way) versus pronation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 Mean loss of dorsal angulation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
2 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Redisplacement (10+ degrees)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Rereduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 4.1.

Comparison 4 Cast position: neutral (mid-way) versus pronation, Outcome 1 Anatomical displacement.

Analysis 4.2.

Comparison 4 Cast position: neutral (mid-way) versus pronation, Outcome 2 Complications.

Comparison 5. Palmar flexion or neutral versus dorsiflexion
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Functional grading: not excellent2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Palmar flexion versus dorsiflexion1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Neutral flexion versus dorsiflexion2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Functional grading: poor or fair2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Palmar flexion versus dorsiflexion1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Neutral flexion versus dorsiflexion2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Discomfort (pressure pain) in plaster1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 5.1.

Comparison 5 Palmar flexion or neutral versus dorsiflexion, Outcome 1 Functional grading: not excellent.

Analysis 5.2.

Comparison 5 Palmar flexion or neutral versus dorsiflexion, Outcome 2 Functional grading: poor or fair.

Analysis 5.3.

Comparison 5 Palmar flexion or neutral versus dorsiflexion, Outcome 3 Complications.

Comparison 6. Modified cast restricted mobilisation versus plaster cast immobilisation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Functional grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 6.1.

Comparison 6 Modified cast restricted mobilisation versus plaster cast immobilisation, Outcome 1 Anatomical grading.

Analysis 6.2.

Comparison 6 Modified cast restricted mobilisation versus plaster cast immobilisation, Outcome 2 Functional grading.

Comparison 7. Modified sugar-tong cast versus below-elbow cast
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 Dorsal angulation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
2 Unacceptable anatomical result (dorsal tilt > 10 degrees, radial shortening > 5 mm)1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Redisplacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 7.1.

Comparison 7 Modified sugar-tong cast versus below-elbow cast, Outcome 1 Anatomical displacement.

Analysis 7.2.

Comparison 7 Modified sugar-tong cast versus below-elbow cast, Outcome 2 Unacceptable anatomical result (dorsal tilt > 10 degrees, radial shortening > 5 mm).

Analysis 7.3.

Comparison 7 Modified sugar-tong cast versus below-elbow cast, Outcome 3 Complications.

Comparison 8. Modified sugar-tong fibreglass splint versus cylindrical plaster cast
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Redisplacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Rereduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 8.1.

Comparison 8 Modified sugar-tong fibreglass splint versus cylindrical plaster cast, Outcome 1 Complications.

Comparison 9. Modified sugar-tong fibreglass splint versus volar-dorsal fibreglass splint
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Redisplacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Rereduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 9.1.

Comparison 9 Modified sugar-tong fibreglass splint versus volar-dorsal fibreglass splint, Outcome 1 Complications.

Comparison 10. Volar-dorsal fibreglass splint versus cylindrical plaster cast
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Redisplacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Rereduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 10.1.

Comparison 10 Volar-dorsal fibreglass splint versus cylindrical plaster cast, Outcome 1 Complications.

Comparison 11. Above-elbow versus below-elbow plaster cast
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Functional grading: not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 at 12 weeks1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Functional grading: poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 at 12 weeks1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.2 at 6 months1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
4.1 Re-application of plaster1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.2 Delayed fracture healing1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.3 Sudeck's atrophy1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.4 Nerve compression1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 11.1.

Comparison 11 Above-elbow versus below-elbow plaster cast, Outcome 1 Anatomical grading.

Analysis 11.2.

Comparison 11 Above-elbow versus below-elbow plaster cast, Outcome 2 Functional grading: not excellent.

Analysis 11.3.

Comparison 11 Above-elbow versus below-elbow plaster cast, Outcome 3 Functional grading: poor or fair.

Analysis 11.4.

Comparison 11 Above-elbow versus below-elbow plaster cast, Outcome 4 Complications.

Comparison 12. Braced position: supination versus pronation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Loss of volar tilt (at least 2 degrees)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Loss of radial length (at least 2 mm)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Anatomical grading: poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3 Functional grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 12.1.

Comparison 12 Braced position: supination versus pronation, Outcome 1 Anatomical displacement.

Analysis 12.2.

Comparison 12 Braced position: supination versus pronation, Outcome 2 Anatomical grading: poor or fair.

Analysis 12.3.

Comparison 12 Braced position: supination versus pronation, Outcome 3 Functional grading.

Comparison 13. Braced position: supination versus other
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Poor finger flexion1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
Analysis 13.1.

Comparison 13 Braced position: supination versus other, Outcome 1 Anatomical grading.

Analysis 13.2.

Comparison 13 Braced position: supination versus other, Outcome 2 Poor finger flexion.

Comparison 14. Functional brace versus plaster cast
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement (at splintage removal)2 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 Dorsal displacement/angulation (degs)2 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.2 Dorsal shift (mm)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.3 Radial displacement (degs)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.4 Radial shift (mm)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.5 Radial shortening (mm)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
2 Anatomical grading3 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Not excellent3 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Poor or fair3 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Range of movement (at splintage removal)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3.1 Extension (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
3.2 Flexion (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
3.3 Radial deviation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
3.4 Ulnar deviation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
3.5 Pronation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
3.6 Supination (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4 Range of movement (at 12 weeks)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4.1 Extension (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4.2 Flexion (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4.3 Radial deviation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4.4 Ulnar deviation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4.5 Pronation (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4.6 Supination (degrees)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
5 Poor finger flexion1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
6 Functional assessment score (0: no problems to 33: maximum)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
6.1 at 6 weeks1 Mean Difference (IV, Fixed, 95% CI)Not estimable
6.2 at 12 weeks1 Mean Difference (IV, Fixed, 95% CI)Not estimable
7 Functional grading: not excellent2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
7.1 up to and including 3 months2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
7.2 6 months2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
8 Functional grading: poor or fair2 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
8.1 up to and including 3 months2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
8.2 6 months2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9 Complications6 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
9.1 Hand swelling during brace or cast use2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.2 Poor finger motion (in brace or cast)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.3 Discomfort necessitating removal (brace or cast)2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.4 Incorrect application of splintage1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.5 Problems with brace or cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.6 Brace or cast change1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.7 Redisplacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.8 Remanipulation, rereduction or manipulation after secondary displacement5 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.9 Tendon rupture1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.10 Sudeck's atrophy4 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.11 Upper limb dystrophy (temporary)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.12 Reflex sympathetic dystrophy1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.13 Median neuropathy1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.14 Median nerve compression (CTS)3 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.15 Superficial radial nerve paraesthesia3 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.16 Peripheral radial/ulnar neuropathy1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
9.17 Ulnar nerve compression2 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
10 Assessment of brace or cast use (visual analogue score: 0 to 10)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
10.1 Pain during splintage: 2 weeks (0 = no pain)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
10.2 Pain during splintage: 6 weeks (0 = no pain)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
10.3 Problems with splintage: 2 weeks (0 = no problems)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
10.4 Problems with splintage: 6 weeks (0 = no problems)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
10.5 Satisfaction with splintage: 2 weeks (0 = satisfaction)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
10.6 Satisfaction with splintage: 6 weeks (0 = satisfaction)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
Analysis 14.1.

Comparison 14 Functional brace versus plaster cast, Outcome 1 Anatomical displacement (at splintage removal).

Analysis 14.2.

Comparison 14 Functional brace versus plaster cast, Outcome 2 Anatomical grading.

Analysis 14.3.

Comparison 14 Functional brace versus plaster cast, Outcome 3 Range of movement (at splintage removal).

Analysis 14.4.

Comparison 14 Functional brace versus plaster cast, Outcome 4 Range of movement (at 12 weeks).

Analysis 14.5.

Comparison 14 Functional brace versus plaster cast, Outcome 5 Poor finger flexion.

Analysis 14.6.

Comparison 14 Functional brace versus plaster cast, Outcome 6 Functional assessment score (0: no problems to 33: maximum).

Analysis 14.7.

Comparison 14 Functional brace versus plaster cast, Outcome 7 Functional grading: not excellent.

Analysis 14.8.

Comparison 14 Functional brace versus plaster cast, Outcome 8 Functional grading: poor or fair.

Analysis 14.9.

Comparison 14 Functional brace versus plaster cast, Outcome 9 Complications.

Analysis 14.10.

Comparison 14 Functional brace versus plaster cast, Outcome 10 Assessment of brace or cast use (visual analogue score: 0 to 10).

Comparison 15. Cast material: Polyurethane versus plaster
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Redisplacement or secondary displacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Rereduction or secondary reduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.3 Secondary cast adjustment1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.4 Skin damage1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 15.1.

Comparison 15 Cast material: Polyurethane versus plaster, Outcome 1 Complications.

Comparison 16. Cast material: Thermoplastic versus plaster
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Bandage repair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Poor stability in cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.3 Pain in cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.4 Poor hand function in cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.5 Skin pressure in cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.6 Water intolerant cast (subjective inconvenience)1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.7 Skin damage1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 16.1.

Comparison 16 Cast material: Thermoplastic versus plaster, Outcome 1 Complications.

Comparison 17. Cast material: Shrinkable polymer (Quickcast) versus fibreglass
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement (at cast removal)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
1.1 Change in radial length (mm)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.2 Change in radial inclination (degs)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.3 Change in ulnar variance (mm)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
1.4 Change in volar tiit (degs)1 Mean Difference (IV, Fixed, 95% CI)Not estimable
2 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Burns or skin irritation from cast application1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Redisplacement or secondary displacement1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.3 Rereduction or secondary reduction1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.4 Early cast change1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.5 Discomfort at thumb base1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.6 Discomfort at cast rim near elbow1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.7 Significant skin problems1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 17.1.

Comparison 17 Cast material: Shrinkable polymer (Quickcast) versus fibreglass, Outcome 1 Anatomical displacement (at cast removal).

Analysis 17.2.

Comparison 17 Cast material: Shrinkable polymer (Quickcast) versus fibreglass, Outcome 2 Complications.

Comparison 18. Casting technique: Focused rigidity cast (FRC) versus standard cast
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Functional impairment1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Unable to function during cast wear1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Moderate or severe movement limitations at cast removal1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.3 Moderate or severe loss of power at cast removal1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Very uncomfortable cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Loss of position1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.3 Non-union1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 18.1.

Comparison 18 Casting technique: Focused rigidity cast (FRC) versus standard cast, Outcome 1 Functional impairment.

Analysis 18.2.

Comparison 18 Casting technique: Focused rigidity cast (FRC) versus standard cast, Outcome 2 Complications.

Comparison 19. Early mobilisation (3-4 weeks) versus 5-6 weeks plaster immobilisation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Pain (mild with stenuous use)1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2 Patient assessment (pain, disability, mobility): poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 at 3 months1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 at 1 year1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Grip strength (% of non-injured hand)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3.1 At 3 months1 Mean Difference (IV, Fixed, 95% CI)Not estimable
3.2 At 6 months1 Mean Difference (IV, Fixed, 95% CI)Not estimable
4 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
4.1 Tendon rupture1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 19.1.

Comparison 19 Early mobilisation (3-4 weeks) versus 5-6 weeks plaster immobilisation, Outcome 1 Pain (mild with stenuous use).

Analysis 19.2.

Comparison 19 Early mobilisation (3-4 weeks) versus 5-6 weeks plaster immobilisation, Outcome 2 Patient assessment (pain, disability, mobility): poor or fair.

Analysis 19.3.

Comparison 19 Early mobilisation (3-4 weeks) versus 5-6 weeks plaster immobilisation, Outcome 3 Grip strength (% of non-injured hand).

Analysis 19.4.

Comparison 19 Early mobilisation (3-4 weeks) versus 5-6 weeks plaster immobilisation, Outcome 4 Complications.

Comparison 20. Early mobilisation (1-2 weeks) versus 4-5 weeks immobilisation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Mean VAS pain scores (none 0 - 20 cm)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
2 Mean grip strength (mmHg)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
3 Functional grading1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
3.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
4.1 Withdrawal from treatment1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.2 Problems with bandage or cast1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.3 Fracture displaced1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.4 Cosmetic deformity1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.5 Tendon rupture1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.6 Paraesthesia1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.7 Physiotherapy needed1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.8 Shoulder stiffness1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
5 Patient dislike of or non-compliance with treatment1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
Analysis 20.1.

Comparison 20 Early mobilisation (1-2 weeks) versus 4-5 weeks immobilisation, Outcome 1 Mean VAS pain scores (none 0 - 20 cm).

Analysis 20.2.

Comparison 20 Early mobilisation (1-2 weeks) versus 4-5 weeks immobilisation, Outcome 2 Mean grip strength (mmHg).

Analysis 20.3.

Comparison 20 Early mobilisation (1-2 weeks) versus 4-5 weeks immobilisation, Outcome 3 Functional grading.

Analysis 20.4.

Comparison 20 Early mobilisation (1-2 weeks) versus 4-5 weeks immobilisation, Outcome 4 Complications.

Analysis 20.5.

Comparison 20 Early mobilisation (1-2 weeks) versus 4-5 weeks immobilisation, Outcome 5 Patient dislike of or non-compliance with treatment.

Comparison 21. Early mobilisation (1 week) versus 3 weeks plaster immobilisation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Anatomical displacement1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
1.1 Angulation up to 20 degrees1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
1.2 Axial radial shortening1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2 Functional grading: not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
2.1 Not excellent1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
2.2 Poor or fair1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
3 Mean functional score (100 is normal)1 Mean Difference (IV, Fixed, 95% CI)Totals not selected
4 Complications1 Risk Ratio (M-H, Fixed, 95% CI)Totals not selected
4.1 Pain preventing plaster removal1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.2 Further dislocation of fracture1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.3 Initial symptoms of algoneurodystrophy1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
4.4 Persistent algoneurodystrophy1 Risk Ratio (M-H, Fixed, 95% CI)Not estimable
Analysis 21.1.

Comparison 21 Early mobilisation (1 week) versus 3 weeks plaster immobilisation, Outcome 1 Anatomical displacement.

Analysis 21.2.

Comparison 21 Early mobilisation (1 week) versus 3 weeks plaster immobilisation, Outcome 2 Functional grading: not excellent.

Analysis 21.3.

Comparison 21 Early mobilisation (1 week) versus 3 weeks plaster immobilisation, Outcome 3 Mean functional score (100 is normal).

Analysis 21.4.

Comparison 21 Early mobilisation (1 week) versus 3 weeks plaster immobilisation, Outcome 4 Complications.

Appendices

Appendix 1. The Cochrane Library search strategy

#1. RADIUS FRACTURES explode all trees (MeSH)
#2. WRIST INJURIES explode all trees (MeSH)
#3. (#1 or #2)
#4. ((distal near radius) or (distal near radial))
#5. (colles:ti or smith:ti or smiths:ti)
#6. (colles:ab or smith:ab or smiths:ab)
#7. wrist*
#8. (#4 or #5 or #6 or #7)
#9. fracture*
#10. (#8 and #9)
#11. (#3 or #10)

Appendix 2. MEDLINE (OVID-WEB) search strategy

1. exp Radius Fractures/
2. Wrist Injuries/
3. (((distal adj3 (radius or radial)) or wrist or colles or smith$2) adj3 fracture$).ti,ab.
4. or/1-3

Appendix 3. EMBASE search strategy

1. (((distal adj3 (radius or radial)) or wrist or colles$2 or smith$2) adj3 fracture$).tw.
2. Colles Fracture/ or Radius Fracture/ or Wrist Fracture/ or Wrist Injury/
3. or/1-2
4. exp Randomized Controlled trial/
5. exp Double Blind Procedure/
6. exp Single Blind Procedure/
7. exp Crossover Procedure/
8. or/4-8
9. ((clinical or controlled or comparative or placebo or prospective$ or randomi#ed) adj3 (trial or study)).tw.
10. (random$ adj7 (allocat$ or allot$ or assign$ or basis$ or divid$ or order$)).tw.
11. ((singl$ or doubl$ or trebl$ or tripl$) adj7 (blind$ or mask$)).tw.
12. (cross?over$ or (cross adj1 over$)).tw.
13. ((allocat$ or allot$ or assign$ or divid$) adj3 (condition$ or experiment$ or intervention$ or treatment$ or therap$ or control$ or group$)).tw.
14. or/9-13
15. or/8,14
16. Animal/ not Human/
17. 15 not 16
18. and/3,17

Appendix 4. CINAHL search strategy

1. Radius Fractures/
2. Wrist Injuries/
3. or/1-2
4. (((distal adj3 (radius or radial)) or wrist or colles or smith$2) adj3 fracture$).ti,ab.
5. or/3-4
6. exp Clinical Trials/
7. exp Evaluation Research/
8. exp Comparative Studies/
9. exp Crossover Design/
10. clinical trial.pt.
11. or/6-10
12. ((clinical or controlled or comparative or placebo or prospective or randomi#ed) adj3 (trial or study)).tw.
13. (random$ adj7 (allocat$ or allot$ or assign$ or basis$ or divid$ or order$)).tw.
14. ((singl$ or doubl$ or trebl$ or tripl$) adj7 (blind$ or mask$)).tw.
15. (cross?over$ or (cross adj1 over$)).tw.
16. ((allocat$ or allot$ or assign$ or divid$) adj3 (condition$ or experiment$ or intervention$ or treatment$ or therap$ or control$ or group$)).tw.
17. or/12-16
18. or/11,17
19. and/5,18

What's new

DateEventDescription
13 August 2008AmendedConverted to new review format.

History

Protocol first published: Issue 2, 1995
Review first published: Issue 2, 1999

DateEventDescription
22 August 2005New search has been performedIn this 'minor' update (Issue 4, 2005) the search for trials was updated to June 2005.

Two newly identified trials were placed in 'Studies awaiting assessment' pending full publication and translation respectively. Of two studies previously listed in 'Studies awaiting assessment', one was included and the other was excluded. Other format changes were undertaken to comply with the Cochrane Style Guide (November 2004). Graphical presentation of the results was compressed to reduce the number of graphs. There were no substantive changes made to the conclusions.

For details of previous updates, please see 'Notes'.

Contributions of authors

Rajan Madhok initiated the review, critically reviewed the studies and all review drafts. Helen Handoll located the review studies, critically reviewed the studies, contacted trialists, compiled the first draft and subsequent revisions in RevMan of the original review and the first, second, third, fourth and fifth updates. Noelle Murphy developed the protocol and was involved in the early stages of the original version, primarily quality assessment, data extraction and initial data entry into the 'Characteristics of included studies' table, but is not a named author. Helen Handoll and Rajan Madhok are guarantors of the review.

Declarations of interest

None known.

Sources of support

Internal sources

  • University of Teesside, Middlesbrough, UK.

External sources

  • No sources of support supplied

Notes

Fourth 'minor' update
For this update, published Issue 2, 2004, the search for trials was updated to November 2003. The only newly identified study was excluded. A study previously listed as ongoing was excluded as was another study previously listed in 'Studies awaiting assessment'. There were no substantive changes made to the conclusions.

Third 'substantive' update
For this update, published Issue 2, 2003, the review comparisons were restructured to reflect the decision points, and associated timing, for the management of these fractures. Thus, the question of whether to reduce a fracture now appears first. In addition, we changed the emphasis from the use of support materials (bandages) to early mobilisation. This resulted in a new category (No immobilisation (minimal support) versus immobilisation) and the loss of an old category ((Bandage (minimal support) versus plaster cast or brace) (Handoll 1999; Handoll 2001; Handoll 2002d). Two trials (Abbaszadegan 1989; Dias 1987) from the old category appear in the new category and two trials (Davis 1987; De Bruijn 1987) were transferred to the renamed and expanded category: Different durations of immobilisation (formerly: Plaster cast for different durations).

The search for trials was extended to January 2003. Three trials were newly included, one trial is ongoing and three trials, awaiting further information from trialists, are pending assessment. Further information obtained for three trials, previously listed as ongoing, resulted in their exclusion. There were no changes made to the conclusions.

Second 'substantive' update
For this update published in Issue 2, 2002, the search for trials was extended to January 2002. Two trials were newly included, one other trial was designated as ongoing and one, awaiting further information from a trialist, was pending assessment. Two additional papers of already included trials yielded limited new results. There were no changes made to the conclusions.

First 'substantive' update
The correct version of the first update of the review was published in Issue 3, 2001. This replaced an earlier and interim version of the review which was put forward for publication in error in Issue 2, 2001. The main difference between these two versions was the extension of the search from July to December 2000 and the reformation of the search strategy.

The basic features of the first update were the inclusion of three more included trials, some modifications to quality scores of trials resulting from the slightly amended quality scoring scheme applied to the now published surgical treatment review, and some more specific pointers to future research. One previously included trial was moved to the surgical treatment review, and hence excluded from this review.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Abbaszadegan 1989

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: unknown (baseline data not given)
Lost to follow up: 12 (15%)
ParticipantsTeaching hospital, Sweden
80 participants
Inclusion criteria: minimally displaced Colles' fracture, dorsal angulation < 95 degrees (see notes) or axial shortening < 2 mm
Exclusion criteria: not stated
Classification: Lidstrom (I & II), Frykman (I to VII, not VI), extra-articular and intra-articular
Sex: 57 female (of 68)
Age: of 68: mean 62 years; range 19 - 91 years
Allocation: ?/? [elastic / POP]
Assessed: 34/34
InterventionsTiming of intervention: not stated. No reduction performed.
(1) Elastic bandage (“functional treatment”) for 4 weeks.
(2) Dorsal plaster cast for 4 weeks.
All participants received information about finger and shoulder mobilisation.
OutcomesLength of follow up: 1 year; also 11 days, 4 and 8 weeks
(1) Anatomical: X-ray initially and at 11 days, 8 weeks. Dorsal angulation, radial shortening.
(2) Functional: Pain (VAS), grip strength, range of movement. Combined score of functional outcomes [Lidstrom].
(3) Clinical:
Complications:
Fracture dislocation: no data
Median nerve decompression: 1/0 [elastic / POP]
Pressure sores: 0/4 (all needed cast change)
NotesDorsal angulation was calculated on the "lateral projection as the angle between the joint surface of the radius and the long axis of the radial shaft".
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Aladin 2001

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: unknown, likely unit of analysis problems
Lost to follow up: not known
ParticipantsTeaching Hospital, UK
126 participants with 128 fractures
Inclusion criteria: displaced Colles' fractures
Exclusion criteria: not stated
Classification: not stated
Sex: not stated (both)
Age: no details
Assigned: 65/63 (fractures) [sugar-tong POP / below-elbow POP]
Assessed: ?/?
InterventionsTiming of intervention: not known. Reduction using Chinese finger-trap method.
Duration of treatment not known, probably 5 weeks
(1) Modified sugar-tong plaster cast where forearm rotation is blocked. Some flexion / extension allowed at elbow
(2) Below-elbow cast
OutcomesLength of follow up: 5 weeks; also 1 week.
(1) Anatomical: X-ray immediately post manipulation, and at 1 and 5 weeks. Dorsal tilt, radial shortening, radial angle and carpal malalignment. 'Acceptable' position (dorsal tilt < 10 degrees, radial shortening < 5 mm).
(2) Functional: none
(3) Clinical:
Complications:
Redisplacement: 43 (66%)/ 43 (69%) [sugar-tong POP / below-elbow POP]
NotesAbstract only
Results probably given for fractures rather than patients. Two participants had two wrist fractures. Estimates of numbers from percentages presented in report are based on the assumption of no losses.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Blatter 1994

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: not known
Lost to follow up: 0 (inferred)
ParticipantsGeneral hospital, Switzerland
50 participants
Inclusion criteria: closed displaced distal radial fracture, AO classification A2.2, A3.2, C1.2 & C2.2
Exclusion criteria: not stated
Classification: AO (see above); extra- and intra-articular
Sex: 37 females
Age: not stated (some < 30 years, others > 50 years)
Assigned: 25/25 [neutral / dorsal flexion]
Assessed: 25/25
InterventionsTiming of intervention: not stated.
Closed MUA. All reduction performed under regional analgesia.
(1) POP dorso-radial slab with wrist in neutral position. POP completed 5-9 days later.
(2) POP dorso-radial slab with wrist in dorsal position (20 degrees flexion). POP completed 5-9 days later.
Total immobilisation for 4 weeks for both groups.
OutcomesLength of follow up: 2 years minimum [range 2-7 years]
(1) Anatomical: X-ray pre and post manipulation, at plaster completion (5-9 days) and at 2 years. Radial shortening, radial angle and dorsal angulation
(2) Functional: range of wrist movement and grip strength, subjective assessment of disability.
(3) Clinical: patient satisfaction with POP.
Complications:
Discomfort with plaster (pressure pain) 3/8 [neutral / dorsal flexion]
NotesGerman paper: full translation obtained.
Adequacy of reduction different between groups for radial angle parameter.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Bunger 1984

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: problems; 5 exclusions due to withdrawal and personal circumstances
Lost to follow up: 14 (9%)
ParticipantsRegional hospital, Denmark
150 participants
Inclusion criteria: Colles' fracture - all grades undisplaced and displaced (criteria given).
Exclusion criteria: ipsilateral fractures of arm, multiple trauma, previously impaired neuromotor function of fractured wrist.
Classification: Frykman (I-VIII); extra-articular and intra-articular
Sex: (of 145) 125 female
Age: not stated
Assigned: 68/82 [brace / POP]
Assessed: 64/72
InterventionsTiming of intervention: not stated.
Reduction of displaced fracture under local anaesthesia.
(1) Above elbow POP splint with forearm in supination. Replaced at 1 week by functional brace (PVC / Hexcelite) in supination. Brace removed at 5 weeks.
(2) Below-elbow POP (undisplaced, stable displaced), POP including elbow, with forearm in pronation (unstable displaced). Replaced at 12 days by POP dorsal splints with wrist in neutral flexion. (Mention of post-reduction Cotton-Loder position in 1983 abstract.) Total immobilisation for 6 weeks.
OutcomesLength of follow up: 6 months; also at 1 day, 2 and 7 weeks and 3 months post MUA.
(1) Anatomical: X-ray pre and post manipulation, and at 1, 2 and 6 weeks. Radial length, radial angulation, dorsal angulation, dorsal shift, radial shift. Combined score [Lidstrom]
(2) Functional: 7 weeks, 3 and 6 months. Combined score [modified Gartland and Werley]: including residual deformity, subjective evaluation (pain, limitations), objective evaluation (grip strength etc), complications. Also ROM and grip strength.
(3) Clinical:
Complications: paraesthesia, finger motion, arthritic changes (within functional score). Early:
Redisplacement: 16/38 [brace / POP]
Rereduction at 2 weeks: 4/7
Median neuropathy: 2/2
Temporary upper limb dystrophy: 3/2
Peripheral radial/ulnar neuropathy: 3/3
Radioulnar pain: no difference
NotesAbstract in 1982 reports trial of 81 people: either subgroup analysis or interim report assumed.
Abstract (1983) reported 150 participants in the trial; there had been some ambiguity in the main report as to whether 5 excluded participants were from 145 or 150 in all: thus, resolved.
Copy of report in Danish (Bunger 1985) not obtained. The Solund 1983 reference gives data for the overall population.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Christensen 1995

MethodsRandomisation by date of presentation (quasi-RCT)
Assessor blinding: not reported
Intention to treat: not known
Lost to follow up: not stated
ParticipantsTeaching hospital, Denmark
33 participants
Inclusion criteria: undisplaced distal radial fracture, Older type 1 and 2, dorsal angulation </= 10 degrees, axial radial shortening </= 4 mm
Exclusion criteria: previous radial fracture of same side, neurological disorders.
Classification: Older (1 & 2)
Sex: 28 female
Age: mean 62 years; range 29-84 years
Assigned: 19/14 [3 weeks POP/ 5 weeks POP]
Assessed: 19/14 (assuming no losses)
InterventionsTiming of intervention: not stated.
Dorsal plaster slab with forearm in pronation, wrist in palmar / ulnar flexion.
(1) Immobilisation for 3 weeks
(2) Immobilisation for 5 weeks
OutcomesLength of follow up: 9 months; also at 3 months.
(1) Anatomical: X-ray pre-cast and 3 months. Radial length, radial angulation, dorsal angulation.
(2) Functional: 3 and 9 months. Combined score [modified Gartland and Werley]: including residual deformity, subjective evaluation (pain, limitations), objective evaluation (grip strength etc), complications.
(3) Clinical: complications (in functional score)
NotesNo data available for quantitative analysis.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Cohen 1997

MethodsMethod of randomisation: not stated, stratified by severity of fracture
Assessor blinding: independent evaluation of cast fit and blinded and independent assessment of radiological outcomes
Intention to treat: likely, not known
Lost to follow up: not known, probably none
ParticipantsTeaching hospital, USA
30 people with distal radius fracture
Inclusion criteria: unilateral distal radius fracture from fall from standing height or less. Treatable by cast with or without percutaneous pins. Within 5 days from injury. Undisplaced and displaced fractures.
Exclusion criteria: fractures requiring open reduction or external fixation devices.
Classification: own classification - Grade I , II, II, based on displacement and stability.
Sex: 22 female
Age: mean 57 years, range 19-89 years
Assigned: 15/15 [Quickcast / fibreglass]
Assessed: 15/15
InterventionsTiming for intervention: within 5 days of injury. Reduction under local anaesthesia for 10 displaced fractures. Cast application was during operation for 10 unstable fractures requiring percutaneous pin fixation.
(1) Short-arm Quickcast (TM) - shrinkable polymer-coated casting material. Applied using warm air.
(2) Short-arm fibreglass tape cast
Casts worn 5.5 to 6.5 weeks.
OutcomesLength of follow up: not stated - until cast removal? Seen at 10-14 day intervals
(1) Anatomical: X-ray after reduction, and 10-14 day intervals. Radial length and inclination, ulnar variance, volar tilt. Fracture displacement.
(2) Functional: not stated
(3) Clinical:
Patient satisfaction with cast
Complications:
During last use:
Burns during application: 2/0 [Quickcast / fibreglass]
Discomfort at base of thumb: 4/3
Discomfort near elbow: 1/0
Loosening / discomfort - early cast change: 2/2
Fracture redisplacement: 2/1
Fracture rereduction: 1/0
(4) Resource use:
Cost of casting material
Cast changes
Notes10 participants, classed as having displaced unstable fractures, had percutaneous pin fixation. However, the results for these could not be separated from the overall group.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Cohen 2001

MethodsRandomisation by numbered sealed consecutively opened envelopes
Assessor blinding: maybe (ROM and grip strength post cast removal) but not claimed as such
Intention to treat: some problems with exclusions but additional information received
Lost to follow up: 1 (3%)
ParticipantsDistrict hospital, UK
35 people with distal radius fracture (of 200 patients with arm and leg injuries in trial)
Inclusion criteria: Minimally or undisplaced distal radial fracture, stable configuration, fresh injury, informed consent
Exclusion criteria: < 16 years, psychiatric morbidity, in police custody, surgery indicated
Classification: none
Sex: not stated (108 females in overall group)
Age: not stated (range 16-84 years in overall group)
Assigned: 14/21 [Focused rigidity cast / complete POP or synthetic cast] (see Notes)
Assessed: 14/20 (14/15 in paper Cohen 2001)
InterventionsTiming of intervention: randomised at first attendance at fracture clinic.
(1) Focused rigidity cast (elasticated synthetic tape applied to confer relative rigidity at fracture site with more flexibility elsewhere). Cast removed 4 weeks (21-42 days)
(2) Synthetic cast or, in 6 participants, circular POP (see Notes). POP back slab initially applied for latter 6 participants as they had very swollen wrists. Cast removed 4 weeks (14-40 days)
OutcomesLength of follow up: just after cast removal (mean 4 weeks).
(1) Anatomical: non-union, loss in position
(2) Functional: at 4 weeks. Movement limitations, muscle power. During cast-use ADL subjective scoring scheme (ADL, satisfaction, comfort) [own scheme: Bradford Plaster Index]
(3) Clinical:
Patient satisfaction with cast
Complications:
Discomfort in cast: 1/0 [Focused rigidity cast / complete POP or synthetic cast]
Loss in fracture position: 0/0
Manipulation: 0/0
Non-union: 0/0
Other complications: 0/0
(4) Resource use:
Referral for physiotherapy (no data), cast replacement, cast adjustment
Notes

Extra details on trial and separate data for wrist fracture patients provided by Andrew Cohen (28/02/02). Some information contradicted trial report: in particular whether synthetic or a full POP cast was used for some patients.

Copy of earlier paper (Petty 1998) provided by Mr Shaw. Some discrepancies with later report: wrist injuries: 20 versus 15 instead of 14 versus 21. Request for further clarification sent 26 August 2002.

Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?YesA - Adequate

Cornwall 2001

MethodsMethod of randomisation: not stated
Assessor blinding: yes, radiological measurements by a blinded observer
Intention to treat: not known
Lost to follow up: not known (24 (34%) reported for interim analysis of 70 participants)
ParticipantsHospital, USA
127 participants
Inclusion criteria: adults with acute closed distal radius fractures - not defined. Some Frykman VIII fractures and mention of articular step-off measurements thus some were intra-articular. Also reduction where necessary so at least some were displaced.
Exclusion criteria: not stated
Classification: Frykman, AO and Fernandez performed (no details)
Sex: not stated
Age: no details but adults
Assigned: ?/? [above-elbow / below-elbow]
Assessed: ?/?
InterventionsTiming of intervention: not known, but acute fractures. Reduction where necessary.
Duration of treatment not known.
(1) Above-elbow cast; neutral forearm rotation
(2) Below-elbow cast
OutcomesLength of follow up: until union; mean 41 days.
(1) Anatomical: X-ray pre and post manipulation, 6 and 15 days and at fracture healing. Volar tilt, radial length, radial inclination and articular step-off.
(2) Functional: none
(3) Clinical: none
Complications: not stated
NotesAbstracts only
No discrete data
Abstract report in 2000 presented interim results for 70 people.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Davis 1987

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: unlikely - post randomisation exclusions
Lost to follow up: 3 (5%)
Participants2 district hospitals, UK
55 participants with 56 fractures
Inclusion criteria: minimally displaced closed distal radius fracture with < 10 degrees dorsal angulation
Exclusion criteria: < 18 years, open fracture, multiple injuries, styloid fractures
Classification: not stated: extra- and intra-articular (6/54) fractures
Sex: 43 female (of 54 fractures)
Age: mean 56 years (of 54 fractures)
Assigned: 28/28 fractures [DTG / POP]
Assessed: 27/25 [after exclusions]
InterventionsTiming of intervention: 7-14 days.
Below-elbow back slab for 7-13 days.
(1) Double tubigrip bandage (DTG) during the day for 3 weeks
(2) POP cylinder with immobilisation for 3 weeks
OutcomesLength of follow up: 7 weeks; also at 1 day post MUA and 5 weeks.
(1) Anatomical: X-ray pre-treatment and 5 weeks. Dorsal angulation.
(2) Functional: 2, 5 and 7 weeks. Range of wrist movement, grip strength, pain, domestic tasks. Combined score grades [Gartland and Werley]
(3) Clinical: patient satisfaction.
Complications:
Dorsal displacement: 2/3 [DTG / POP]
Withdrawal from treatment: 0/2
Treatment failure: 1/1
Cosmetic deformity: 0/2
EPL rupture: 0/1
Shoulder stiffness: 1/0
Physiotherapy required: 1/3
Persistent paraesthesia: 7/7
NotesResults given for fractures rather than patients. One participant had bilateral wrist fractures.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

De Bruijn 1987

MethodsTwo sequential trials
Randomisation by date of attendance in first trial. Numbered forms, ratio 3/1 [functional brace or bandage / Sarmiento brace] used for the second trial. Quasi-RCTs.
Assessor blinding: not reported
Intention to treat: no, data merged incorrectly from both trials; also some exclusions (including missed follow up etc) applied after trial entry.
Lost to follow up: 52 for both trials (21%)
ParticipantsTeaching hospital, The Netherlands
248 people participated in 2 trials
Inclusion criteria: Colles' fracture with united epiphyseal plates
Exclusion criteria: open epiphysis, ipsilateral fracture, fracture > 3 days old, bilateral fractures, multiple injuries, previous fracture, pre-existing limb impairment, follow-up elsewhere. (Withdrawal: inability to cooperate, 2 missed follow-up appointments, < 5 degrees volar angle post-reduction.)
Classification: Sarmiento (types 1 to 4); extra-articular and intra-articular fractures
Sex: 152 female of 196
Age: not stated
Assigned: ?/104 [Sarmiento brace / POP]; 56/? [functional brace or bandage / Sarmiento brace]
(Total Sarmiento: 88)
Assessed: ?/82 [Sarmiento brace / POP]; 47/? [functional brace or bandage / Sarmiento brace]
(Total Sarmiento: 67)
InterventionsTiming of intervention: at reduction and 1 week post-reduction.
Displaced fractures manipulated with haematoma block.
Trial 1:
(1) Above-elbow POP backslab with forearm in supination for 1 week, changed to a Sarmiento's brace in supination for 3 weeks
(2) Below-elbow POP backslab for 4 weeks
Trial 2:
(1) Below-elbow POP backslab then functional below-elbow brace with restricted dorso-radial movement for 3 weeks (displaced fractures), or bandage for 3 weeks (undisplaced).
(2) Above-elbow POP backslab with forearm in supination for 1 week, changed to a Sarmiento's brace in supination for 3 weeks
OutcomesLength of follow up: 1 year; also day 1, 7 and 14, and 4, 6, 10, 14 and 26 weeks
(1) Anatomical: X-ray pre and post manipulation and 1 and 2 weeks. Radial length and angle, and volar angle.
(2) Functional: range of movement, pain and grip strength, motor functions of hand (activities of daily living)
(3) Clinical: cosmetic appearance. Signs and symptoms.
Complications: pressure sores, rereduction, nerve & tendon injury, Sudeck's syndrome, post-traumatic arthritis, Dupuytren's contracture, carpal tunnel compression, equipment problems.
NotesComprehensive thesis.
Data incorrectly combined from 2 trials which had one intervention in common (Sarmiento brace). Non concurrent interventions compared.
Baseline data not available.
Discrete data were not given for "bandage" group
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Dias 1987

MethodsRandomisation by dates of birth (Quasi-RCT)
Assessor blinding: independent and blinded observer at 13 weeks for functional assessment
Intention to treat: no (according to trialist)
Lost to follow up: not known (none?)
ParticipantsTeaching hospital, UK
187 participants
Inclusion criteria: closed unilateral Colles' fracture, displaced (reduction required) and undisplaced, > 55 years, no past history of hand/wrist injury, no generalised joint disease
Exclusion criteria: injury to other hand / wrist, unable to perform daily activities
Classification: not stated, displaced/undisplaced
Sex: 161 female
Age: mean 67 years, range 55-98 years
Assigned: 50/47/43/47 [undisplaced crepe / undisplaced POP / displaced modified POP/ displaced POP]
Assessed: 50/47/43/47
InterventionsTiming of intervention: on first visit to A&E department.
Timing of MUA not stated.
Undisplaced fractures allocated to treatment in either:
(1) Crepe bandage and early mobilisation during the day for 3 weeks
(2) POP cylinder for 5 weeks
Displaced fractures (clinically visible deformity) allocated to treatment by:
(3) Modified POP cylinder allowing greater wrist mobility (extension restricted) for 5 weeks. Velcro strap fixed across palm at night to restrict hand movement.
(4) POP cylinder for 5 weeks
OutcomesLength of follow up: 3 months; also at 1 day post MUA, 1, 5 and 9 weeks post MUA
(1) Anatomical: X-ray at each follow-up time. Radial length, deviation and shift, volar tilt measured [Lidstrom]
(2) Functional: 5 weeks and 3 months. Range of wrist movement, grip strength, pain (VAS), analgesia use, night pain and wrist circumference [Stewart et al]. Also functional score [Gartland and Werley]
(3) Clinical:
Complications: No data; wrist swelling, deformity, discomfort
NotesFull report plus abstract. Abstract reports on minimally displaced fracture group only.
Some information on methodology obtained from trialist.
The entry used to state that "All reduction performed under regional analgesia." but subsequently we have failed to find written records of this, though it is likely.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Ferris 1989

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: problems
Lost to follow up: 13 (declared), 8 in analysis (17%)
ParticipantsTeaching hospital, UK
47 participants (see notes)
Inclusion criteria: closed Colles' fracture [not defined]
Exclusion criteria: < 16 years, inability to follow instructions.
Classification: not stated
Sex: 2/3rd female
Age: mean 60 years; range 22-87 years
Assigned: 20/27 [brace / POP]
Assessed: 18/21
InterventionsTiming of intervention: not stated.
All reductions performed under regional analgesia.
(1) Closed MUA and a 4 part brace applied with wrist in supination. Limited wrist motion allowed (dorsi and palmar flexion). Total application for 5 weeks.
(20 Closed MUA and POP dorso-radial slab with forearm in pronation. POP completed within 48 hours. Total immobilisation for 5 weeks.
OutcomesLength of follow up: 9 weeks. Also at 1 day and 5 weeks post MUA.
(1) Anatomical: X-ray pre and post manipulation and 5 weeks. Radial shortening and dorsal angulation
(2) Functional: range of wrist movement and grip strength [Porter & Stockley] at 5 and 9 weeks
(3) Clinical:
Complications:
Swelling: 20/0 [brace /POP]
Remanipulation: 1/4
Sudeck's: 1/0 (responded to therapy)
NotesShort follow up.
Possibility that a few more patients were recruited than stated in report.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Gibson 1983

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: not known
Lost to follow up: not known
ParticipantsTeaching hospital, UK
105 participants
Inclusion criteria: Colles' fracture [not defined in abstract]. Displaced according to long-term report (see Notes).
Exclusion criteria: not stated
Classification: not stated
Sex: not stated
Age: not stated
Assigned: ?/?/?
Assessed: ?/?/?
InterventionsTiming of intervention: not stated. Closed reduction under IVRA.
(1) Functional bracing in supination (thermoplastic)
(2) Supinated above-elbow POP
(3) Pronated elbow POP
All participants were given physiotherapy to the non-immobilised joints. Immobilisation for 5 weeks.
OutcomesLength of follow up: 20 weeks
(1) Anatomical: X-ray pre and post manipulation and 1, 9 and 20 weeks
(2) Functional: at 9 and 20 weeks, no method detailed
(3) Clinical:
Complications: not detailed
NotesAbstract only.
No discrete data given.
Both named trialists contacted in August 2000. Informed of two papers associated with trial - both 10-year follow ups of the trial patients but data split by treatment group not provided. Some extra details of interventions and indication that all fractures were displaced. Also, there were discrepancies: e.g. only 100 participants reported rather than 105; follow up at 20 weeks not acknowledged nor the use of three types of splintage. Thus, no information has been used from these 2 long-term follow up reports in the assessment of this trial. One trialist commented that the supinated brace, advocated in the abstract, was not used subsequently because patients rejected it.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Gupta 1991

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: likely but not known
Lost to follow up: not stated
ParticipantsTeaching hospital, India
204 participants
Inclusion criteria: closed displaced Colles' fracture [not defined], included comminuted and non-comminuted extra-articular fractures, and intra-articular fractures.
Exclusion criteria: not stated
Classification: Own
Sex: 122 females
Age: mean 46 years; range 18-74 years
Assigned: 60/75/69 [palmar / neutral / dorsal position]
Assessed: 60/75/69
InterventionsTiming of intervention: not stated.
All reduction performed under intravenous sedation
Closed MUA and below-elbow plaster cast applied with:
(1) Wrist in palmar flexion for 6 weeks
(2) Wrist in neutral position for 6 weeks
(3) Wrist in dorsiflexion for 6 weeks
Intra-articular fractures for 5 rather than 6 weeks.
Wrist also held in slight ulnar deviation.
OutcomesLength of follow up: mean 15 months (range 5-24 months); also at 1 day and 10 days post MUA.
(1) Anatomical: X-ray pre and post manipulation, day 10 and final review. Radial length, radial deviation and volar tilt
(2) Functional: Combined outcome: overall grades [Sarmiento criteria] not stated when assessed (at final review?)
(3) Clinical:
Complications: not stated, joint stiffness mentioned.
Notes 
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Ho 1986

MethodsMethod of randomisation: not stated, "randomly distributed"
Assessor blinding: not reported
Intention to treat: problems anticipated due to severe lost to follow-up
Lost to follow up: 70 at 6 months (63%)
ParticipantsTeaching hospital, Hong Kong, China
109 participants with 111 fractures
Inclusion criteria: age > 40 years with Colles' fracture [not defined]
Exclusion criteria: not stated
Classification: not stated
Sex: not stated
Age: not stated
Assigned: 42/69 [brace / POP]
Assessed: ?/?
InterventionsTiming of intervention: not stated, after closed reduction.
(1) Supinated long arm POP for 1-2 weeks, followed by Sarmiento's functional brace in supination for 4 weeks. Wrist palmar flexion and ulnar deviation allowed in brace
(2) Short arm POP for 6 weeks
OutcomesLength of follow-up: 6 months. Also at 6 weeks and 3 months post reduction.
(1) Anatomical: X-ray pre and post reduction, 1-2 weeks and 6 weeks. Radial length and shortening, dorsal angulation, radial angulation, dorsal shift.
(2) Functional: 3 & 6 months. Range of wrist movement, pain, function [modified Gartland and Werley], hand function for activities of daily living [modified Jabsen test]
(3) Clinical:
Complications: deformity
NotesAbstract only.
Severe lost to follow up
Letter sent to trialist on 20 December 2001.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Jackson 2002

MethodsMethod of randomisation: not stated
Assessor blinding: yes for radiological outcomes
Intention to treat: not known
Lost to follow up: not known
ParticipantsFour hospitals, Canada
101 participants
Inclusion criteria: Acute displaced extra-articular distal radius fractures, age > 18 years, intention to treat non-operatively.
Exclusion criteria: failed closed reduction.
Classification: AO 3.1 and 3.2; extra-articular
Sex: not stated
Age: not stated
Assigned: 30/31/40 [modified sugar-tong splint / volar-dorsal splint / cylindrical POP]
Assessed: Assessed: 30/31/40 (at 1 or 4 weeks?)
InterventionsTiming of intervention: not stated. Closed reduction.
(1) Muenster modified sugar-tong fibreglass splint
(2) Volar-dorsal fibreglass splint
(3) Cylindrical POP
No information on length of immobilisation or aftercare.
OutcomesLength of follow up: 6 months; also 7 and 28 days and 2 months
(1) Anatomical: X-ray pre and post manipulation and 1 and 4 weeks
(2) Functional: at 2 and 6 months (no results)
(3) Clinical:
Complications: not detailed
NotesAbstracts only.
No further information forthcoming from trialists. Requests sent to different authors: 8 November 2002, 2 February 2004 and 20 June 2005.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Jensen 1997

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: no - post randomisation exclusions including 7 who did not fulfil inclusion criteria
Lost to follow up: 14 (23%)
ParticipantsTeaching hospital, Denmark
62 participants
Inclusion criteria: minimally displaced or undisplaced Colles' fracture Older I or IIa, dorsal angulation < 5 degrees with no radial shortening or angulation 5+ degrees with up to 2.0 mm radial shortening
Exclusion criteria: previous fractures of either wrist
Classification: Older - modified (I & IIa)
Sex: 32 female (of 48)
Age: mean 57 years, range 19-86 years (of 48)
Assigned: ?/? [1 week / 3 weeks]
Assessed: 22/26
InterventionsTiming of intervention: after admission to A&E department, written consent required for study inclusion.
(1) Dorsal plaster cast for 1 week. Motion within range of comfort after cast removal
(2) Dorsal plaster cast 3 weeks
OutcomesLength of follow up: 6 months; also at cast removal and 12 weeks.
(1) Anatomical: X-ray at cast removal and 6 months. Angulation and axial radial shortening.
(2) Functional: cast removal, 12 weeks and 6 months. Deformity, pain, range of wrist movement, grip strength, (also complications), in functional score [modified Gartland and Werley]. Time off work.
(3) Clinical: times as above.
Complications: within functional score.
Notes 
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Kelly 1997

MethodsRandomisation by pre-numbered envelopes
Assessor blinding: not reported
Intention to treat: likely
Lost to follow up: none
ParticipantsTeaching hospital, UK
30 participants
Inclusion criteria: moderately displaced Colles' fracture, 10 - 30 degrees of dorsal angulation, < 5 mm radial shortening.
Exclusion criteria: < 65 years, previous ipsilateral forearm fractures
Classification: Frykman (1&2, 3&4, 5&6, 7&8); extra- and intra-articular
Sex: 27 female
Age: mean 75 years
Assigned: 15/15 [Manipulation / Control]
Assessed: 15/15
InterventionsTiming of intervention: not stated.
(1) Bier’s block manipulation and immobilisation in forearm plaster. Removed at 5 weeks.
(2) Immobilisation in forearm plaster. Removed at 5 weeks.
All patients assessed for physiotherapy.
OutcomesLength of follow up: 3 months; also 1, 3 and 5 weeks, and 1 month.
(1) Anatomical: X-ray at admission, post reduction and 5 weeks. Radial and dorsal angulation, radial shortening.
(2) Functional: Combined score of subjective and objective outcomes [modified Gartland and Werley]. Grip strength, pain, stiffness.
(3) Clinical: swelling, vasomotor changes, cosmetic deformity, dolorimetry ratio.
Complications:
Redisplacement/displacement: 2/2 [Manipulation / Control]
Rereduction or secondary reduction: 0/0
Algodystrophy (RSD): 1/1
Finger stiffness: 3/3
NotesStresses selection of elderly patients.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Kongsholm 1981

MethodsMethod of randomisation: not stated
Assessor blinding: not known
Intention to treat: not known
Lost to follow up: not known
ParticipantsTeaching hospital, Sweden
106 participants
Inclusion criteria: Colles' fracture [not defined]
Exclusion criteria: not stated
Classification: not stated
Sex: not stated
Age: not stated
Assigned: ?/?/? [3 groups]
Assessed: ?/?/?
InterventionsTiming of intervention: not known, but acute fractures.
Duration of treatment not known.
(1) Manual reduction under local anaesthesia then dorsal plaster splint
(2) Manual reduction under local anaesthesia then circular plaster cast
(3) Mechanically applied continuous traction without anaesthesia then circular plaster cast
OutcomesLength of follow up: 6 months
(1) Anatomical: overall, Lidstrom?
(2) Functional: overall, Lidstrom?
(3) Clinical: pain during reduction (not for this review)
Complications:
Recurrent fracture dislocations: 27% group (1) / 10% groups (2) and (3)
NotesAbstract only No discrete data
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Ledingham 1991

MethodsMethod of randomisation: not stated
Assessor blinding: blind functional assessment
Intention to treat: probable but 3 participants removed from trial before treatment completion
Lost to follow up: 10 (17%) [6 months]
ParticipantsTeaching hospital, UK
60 participants
Inclusion criteria: closed displaced distal radial fracture requiring MUA [not defined]
Exclusion criteria: < 18 years, bilateral fractures, previous history of distal radial fracture or inability to follow instructions
Classification: Frykman
Sex: 50 females
Age: mean 60.5 years [3 point brace / POP]
Assigned: 30/30 [3 point brace / POP]
Assessed: 24/26 [6 months]
InterventionsTiming of intervention: not stated.
All reduction performed under general analgesia.
(1) Closed MUA and a molded functional POP applied with 3 point loading in neutral position. Brace allowed free movement at radiocarpal joint - pronation and supination restricted. Removed 5-6 weeks.
(2) Closed MUA and POP dorso-radial slab with wrist in ulnar deviation and slight palmar flexion. POP completed 1-2 weeks later. Plaster removed 5-6 weeks.
OutcomesLength of follow up: 6 months; also at 1 day post MUA, 7-14 days, 35-42 days and 12 weeks
(1) Anatomical: X-ray pre and post manipulation and 7-14 days, 5 to 7 weeks. Combined score: radial length and angle, dorsal angulation (Van der Linden).
(2) Functional: Combined score: subjective complaints, ROM, Grip strength, complications [Stewart method based on Gartland and Werley]
(3) Clinical:
Complications :
Remanipulation: 0/2 [3 point brace / POP]
Median nerve symptoms (CTS): 14.5% at 6 months for whole study group
Radial nerve neuropraxia: 6/0 - all resolved
EPL rupture: 0/1
Sudeck's atrophy: 0/0
Shoulder-hand syndrome: 0/0
NotesInitial and post-MUA displacements not presented.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

McAuliffe 1987

MethodsRandomisation by alternation (quasi-RCT)
Assessor blinding: not known, but independent assessor of function and pain
Intention to treat: not known
Lost to follow up: 28 (26%) [1 year]
ParticipantsTeaching hospital, UK
108 participants
Inclusion criteria: all women > 60 years with closed Colles' fracture [not defined]
Exclusion criteria: not stated
Classification: Lidstrom (I,IIA-E); extra-articular and intra-articular
Sex: all female
Age: mean 72/73 years
Assigned: 54/54 (3 weeks POP / 5 weeks)
Assessed: 41/39 (1 year)
InterventionsTiming of intervention: not stated but all reduction in casualty department.
Closed MUA and POP dorso-radial slab. POP completed 24 hours later. Radiograph at 1 week and remanipulation if required.
(1) Immobilisation continued for '3 weeks' (19-22 days)
(2) Immobilisation continued for '5 weeks' (34-45 days)
OutcomesLength of follow up: 1 year; also at 1 day post MUA, 1 week and 3 months
(1) Anatomical: X-ray pre and post manipulation, 1 week, 3 months and 1 year. Radial shortening, radial deviation and dorsal angulation, overall result [modified Lidstrom]
(2) Functional: 3 months, 1 year. Range of wrist movement, grip strength, pain. Combined score grading [Gartland & Werley]. Subjective assessment.
(3) Clinical:
Complications: no data presented
NotesBaseline data not given but may be the same as those at 3 months. Groups not matched for degree of displacement or adequacy of initial reduction.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

McMillan 1996

MethodsMethod of randomisation: not stated
Assessor blinding: not known
Intention to treat: not known
Loss to follow up: not known
ParticipantsHospital, UK
80 participants
Inclusion criteria: displaced Colles' fracture, age over 50 years
Exclusion criteria: symptoms of median nerve compression
Classification: not stated
Sex: not stated
Age: not stated
Assigned: ?/?
Assessed: ?/?
InterventionsTiming of intervention: on presentation to Orthopaedic department.
(1) Delayed primary manipulation under regional anaesthesia at 1 week after back slab in displaced position
(2) Immediate primary manipulation
Probably plaster cast was applied after MUA in both groups
OutcomesLength of follow up: not stated
(1) Anatomical: radiological outcomes measured
(2) Functional: discomfort only
(3) Clinical:
Complications:
Median nerve compression: 0/? (1 week)
NotesAbstract only
Mr James, last contacted in August 2000, indicated that he would be contacting Mr Kinninmonth to discuss the trial report and whether to send it on for this review. He seemed to be intending to write up the trial for publication.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Millett 1995

MethodsRandomisation by use of randomisation sheet [information from author]
Assessor blinding: blind and independent assessment of radiographs
Intention to treat: not known
Lost to follow up: 7 (7%); and 10 deaths
ParticipantsTeaching Hospital, UK
90 participants
Inclusion criteria: closed Colles' fracture [not defined] all grades. Normal hand/arm function before injury, females only, normal contralateral forearm and hand.
Exclusion criteria: bilateral fractures, previous history of forearm/hand injury, inability to return for follow up, generalised rheumatic disease.
Classification: Frykman (I to VIII); extra- and intra-articular
Sex: all females
Age: mean 61/60 years; range 22-88 years
Assigned: 45/45 [Viscopaste / POP]
Assessed: ?/?
InterventionsTiming of intervention: not stated.
Displaced fractures (definition?) manipulated under general or local anaesthesia.
(1) Below-elbow POP with wrist in ulnar deviation and supination for 3 weeks, then flexible Viscopaste allowing early joint mobilisation for 2 weeks
(2) Below-elbow POP with wrist in ulnar deviation and supination for 5 weeks
OutcomesLength of follow up: 3 years; also at 1 day post MUA, 3 ,5, 12 and 26 weeks post MUA.
(1) Anatomical: X-ray pre and post manipulation and 1, 3, 5 weeks, 3 and 6 months and 3 years. Radial tilt, radio-ulnar space, dorsal tilt, radial shortening.
(2) Functional: at 3 and 5 weeks, 3 and 6 months and 3 years. Range of wrist movement, grip and pinch strength, finger movement, hand swelling and pain. Subjective assessment.
(3) Clinical:
Cosmetic appearance: 1/3 of participants in both groups had residual deformity. Patient comfort.
Complications: overall rate (not defined), osteoarthritis.
Remanipulation of reduced fractures: 1/4 [Viscopaste / POP] - but no information on numbers reduced.
NotesDifference in numbers lost to follow up in paper and numbers assessed provided by authors.
Five participants of each group had died by 3 years.
Outliers excluded from analyses of radiological and functional outcomes.
The formulation of Viscopaste was altered by the manufacturers after this experiment.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Moir 1995

MethodsRandomisation by computer generated randomisation sheet
Assessor blinding: yes (functional outcomes)
Intention to treat: not known
Lost to follow up: 6 (7%)
ParticipantsTeaching hospital, UK
85 participants
Inclusion criteria: closed displaced distal radial fracture requiring MUA [not defined], fused epiphysis
Exclusion criteria: < 18 years, bilateral fractures, previous distal radial fracture or inability to follow instructions
Classification: Frykman (I-VIII probably); extra- and intra-articular
Sex: 70 females (out of 79)
Age: median 55 / 60 years [3 point brace / POP]; range 21-86 years
Assigned: 49/36 [3 point brace / POP]
Assessed: 44/35
InterventionsTiming of intervention: not stated.
Reduction performed under general or regional anaesthesia.
(1) Closed MUA and a 2 part brace applying 3 point loading. Brace removed 5-6 weeks
(2) Closed MUA and POP dorso-radial slab with wrist in ulnar deviation and slight palmar flexion. POP completed 10-14 days later. Cast removed 5-6 weeks
OutcomesLength of follow up: 6 months. Also at 1 day post MUA, 10-14 days, 5-6 weeks, 8 and 13 weeks post MUA.
(1) Anatomical: X-ray pre and post manipulation and 10-14 days, 5-6 weeks. Radial length, radial angle and dorsal angulation. Overall [modified Lidstrom and Sarmiento demerit scoring system]
(2) Functional: at 8, 13 and 26 weeks. Range of wrist movement, grip and pinch strength, finger movement, hand swelling and pain. Combined score grades: subjective, objective e.g. ROM outcomes and complications [modified Gartland and Werley]
(3) Clinical:
Complications:
Equipment failure (brace discomfort): 3 (excluded participants)/0 [3 point brace / POP]
Incorrect application of treatment: 1(excluded patient)/0
Surgical correction: 1/1 (excluded participants)
Remanipulation: 4/2
Carpal tunnel syndrome: 8/5
Radial nerve neuropraxia: 3/0
Ulnar nerve compression: 1/3
RSD: 2/3
Hand swelling: 11/14
Poor finger movement: 4/11
NotesGroups not comparable for initial displacement - brace group were initially less severely displaced than POP group, nor intra-articular fractures - brace group: 2/3rd, POP group: 1/2th.
Abstracts reported interim data.
Comments on trial from Moiz 1996 questioning some of the conclusions and querying about costs.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Nielsen 1981

MethodsRandomisation by odd and even dates (quasi-RCT)
Assessor blinding: not reported
Intention to treat: not known
Lost to follow up: not known, probably none
ParticipantsTeaching hospital, Denmark
55 people with Colles' fracture
Inclusion criteria: Colles' fracture, also malleolar fracture and finger joint injuries
Exclusion criteria: age < 15 years
Classification: not stated
Sex: not stated
Age: not stated
Assigned: 32/23 [Hexelite / POP]
Assessed: 32/23
InterventionsTiming of intervention: 1 week post trauma after swelling had diminished.
(1) Hexelite (R) - thermoplastic polyester bandage
(2) POP forearm dorsal splint
Duration not given, but similar
OutcomesLength of follow up: not stated, until cast removed?
(1) Anatomical: none given
(2) Functional: patient's opinion of hand function during cast wear
(3) Clinical: pain during cast wear
Complications: (during cast immobilisation)
Poor hand function: 4/1 [Hexelite / POP]
Poor stabilisation: 4/4
Pain: 1/3
Bandage repair: 3/3
Skin pressure: 7/6
Skin damage: 4/1
Inconvenience - water intolerant: 16/21
(4) Resource use: costs of bandage
NotesStudy also enrolled 46 people with ankle fractures and 82 requiring finger bandages
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

O'Connor 2002

MethodsRandomisation by odd and even dates of injury (quasi-RCT)
Assessor blinding: yes, blinded functional assessment by physiotherapist.
Intention to treat: marginal - data for 10 participants not given in baseline characteristics but numbers randomised obtained
Lost to follow up: 7 (+3 requiring further treatment: 2 manipulations and 1 refracture) (13%)
ParticipantsTeaching hospital, Ireland
76 participants
Inclusion criteria: minimally displaced closed Colles' fracture, not requiring manipulation, isolated upper limb injury
Exclusion criteria: previous wrist fracture, impaired cognitive function
Classification: Universal (all type 1: extra-articular, undisplaced)
Sex: 44 females (of 66)
Age: mean 57 years, range 16-81 years (of 66)
Assigned: 37/39 [Splint / POP]
Assessed: 34/32
InterventionsTiming of intervention: at first fracture clinic appointment.
(1) 'Futuro' wrist splint (lightweight and removable) fitted by plaster room technician. Participants were instructed to avoid removing splint for 2 weeks other than to tighten it to adjust as swelling diminished. Then to wear it for a further 4 weeks, only removing it for limited periods for personal hygiene / washing.
(2) Below-elbow POP for 6 weeks.
All participants received the same mobilisation instructions and post-immobilisation physiotherapy.
OutcomesLength of follow up: 12 weeks; also 1, 2 and 6 weeks.
(1) Anatomical: X-ray at presentation, 1, 2 and 6 weeks. Scoring system for displacement (dorsal angulation, radial shortening, radial angulation) [modified Lidstrom]: no results given
(2) Functional: pain, patient problems and satisfaction during splint/cast wear. At 6 and 12 weeks: range of wrist movement, grip strength. Combined functional score [Stewart modification of Gartland & Werley]. Subjective assessment via functional questionnaire.
(3) Clinical:
Complications:
Displacement requiring manipulation: 1/1 [Splint / POP]
Refracture: 0/1
Change of splint/cast: 0/14
Median nerve compression: 1/2
Ulnar nerve compression: 0/2 (1 persisted at 12 weeks)
Sudek's atrophy: 1/1 (resolved by 12 weeks)
(4) Resource use: costs, not quantified
Notes

Lead trialist provided a copy of the uncorrected proofs of article, pending publication in J Hand Surg Br 2003: published report listed in references in updated review (Issue 2, 2004)

Further information from Damian O'Connor received on 14/11/02 regarding numbers randomised, fracture type and care programme.

Abstract (Mullett 2002) involved 71 patients. There were other slight differences, including exclusion criterion: age < 20 years, which clearly didn't apply.

Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Rosetzsky 1982

MethodsRandomisation by closed envelopes, according to predetermined randomisation list
Assessor blinding: not reported
Intention to treat: no, 4 exclusions
Lost to follow up: 1 (2%)
ParticipantsTeaching hospital, Denmark
50 participants
Inclusion criteria: fresh fracture of lower radius and an ulnar avulsion (Colles'). None were open. Displaced and undisplaced.
Exclusion criteria: none given
Classification: none used
Sex: 35 female
Age: mean 45 years, range 13-90 years
Assigned: 25/25 [Polyurethane / POP]
Assessed: 24/22
InterventionsTiming of intervention: not known.
Displaced fractures (dorsal angulation >15 degrees) reduced.
(1) Below-elbow polyurethane cast/brace with zipper. Displaced fractures reduced after cast application, whilst cast set.
(2) Below-elbow POP cylinder. Displaced fractures reduced before cast application.
All casts removed after 6 weeks.
OutcomesLength of follow up: 6 weeks
(1) Anatomical: X-ray at admission and after 6 weeks. Difference in radiological position.
(2) Functional: pain and restricted motion of free joints as reasons for cast adjustment.
(3) Clinical:
Complications:
Secondary adjustment of cast: 8/8 [Polyurethane / POP]
Fracture displacement: 1/5
Secondary reduction: 0/0
Patient inconvenience: 6/4
Skin damage: 0/0
Notes 
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?YesA - Adequate

Ross 1984

MethodsMethod of randomisation: unknown but a quasi-randomised method was used
Assessor blinding: not known
Intention to treat: not known
Lost to follow up: not known
ParticipantsTeaching hospital, UK
405 participants
Inclusion criteria: Colles' fracture [not defined]
Exclusion criteria: not stated
Classification: Own: Type I [extra-articular]; Type II [intra-articular]
Sex: not stated
Age: no details but significant differences between groups noted
Assigned: 121/284 [brace / POP]
Assessed: ?/?
InterventionsTiming of intervention: not known.
Duration of treatment not known.
(1) Orthoplast functional brace, forearm in supination
(2) Below-elbow POP
OutcomesLength of follow up: 3-4 months
(1) Anatomical: radial length, radial deviation and volar tilt
(2) Functional: range of wrist movement, grip strength, disability and pain
(3) Clinical: cosmetic appearance (prominence of ulnar styloid).
Complications: nerve
NotesAbstract only.
No discrete data.
On contact in August 2000, Mr Ross reported that the trial was quasi-randomised, that there were no other publications of the trial and that the trial materials had unfortunately been discarded.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Sarmiento 1980

MethodsRandomisation based on date of fracture (quasi-RCT)
Assessor blinding: not reported
Intention to treat: not known
Lost to follow up: 50 (32%)
ParticipantsTeaching hospital, USA
154 people with 156 distal radius fractures; 93 displaced fractures.
Inclusion criteria: not clarified
Exclusion criteria: not stated
Classification: Sarmiento (I-IV); extra- and intra-articular
Sex: 106 females
Age: mean 49 years; range 15-92 years
Assigned: 156 (?/?)
Assessed: 104 (?/?); subgroup of displaced fractures: 38/33 [Pronation/Supination]
InterventionsTiming of intervention: not stated.
Most reduction performed under regional anaesthesia; some local.
(1) Closed MUA and above-elbow POP with wrist in 20 degrees of volar and ulnar deviation with the forearm in pronation. About 1 week later, Orthoplast brace applied in pronation.
(2) Closed MUA and above-elbow POP with wrist in 20 degrees of volar and ulnar deviation with the forearm in supination. About 1 week later, Orthoplast brace applied in supination.
Cast immobilisation generally between 8-9 days; brace 2-5 weeks.
OutcomesLength of follow up: 4 to 60 weeks; mean 15 weeks
(1) Anatomical: X-ray at pre and post MUA, on removal of POP, on application of brace and at final assessment. Scored using modified Lidstrom. Volar tilt, radial length and deviation
(2) Functional: Grip strength, range of wrist movement, loss of pronation, hand function, elbow motion. Overall score [Gartland and Werley]
(3) Clinical: Complications:
RSD: 1 overall
Carpal tunnel syndrome: 3 in all (mild)
Ulnar nerve injury: 1 in all (resolved)
NotesBaseline data not given by treatment group.
No discrete data presented.
Mixture of un/displaced or extra/intra articular fractures.
Wide range in period of immobilisation/treatment: average time of initial cast application varied according to fracture type.
Data only presented for displaced fracture groups.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Sorensen 1986

MethodsRandomisation by sealed envelopes
Assessor blinding: not known
Intention to treat: some problems, 7 exclusions from baseline
Loss to follow up: 7 (7%)?
ParticipantsHospital, Denmark
100 participants
Inclusion criteria: displaced fractures of the distal 5 cm of radius +/- fracture of ulnar styloid
Exclusion criteria: previous fracture, bilateral fracture, other ipsilateral limb injury, open epiphysis or refracture.
Classification: Van der Linden Type I-IV; extra-articular and intra-articular
Sex: 82 females (of 93)
Age: not stated
Assigned: ?/? [above-elbow POP/ below-elbow POP]
Assessed: 33/60
InterventionsTiming of intervention: not given.
All reductions under local anaesthetic.
Duration of treatment: not stated
(1) Closed manipulation & above-elbow full POP, arm in pronation
(2) Closed manipulation & below-elbow dorsal slab, arm in pronation
OutcomesLength of follow up: 6 months; also at 1, 2, 6, 7 and 12 weeks post MUA.
(1) Anatomical: dorsal angulation and axial compression at post-reduction, 1 and 6 weeks [Bunger, Lidstrom]
(2) Functional: 1, 2, 6, 7 and 12 weeks and 6 months: Range of movement, combined score grades [Gartland & Werley]
(3) Clinical:
Complications:
Sympathetic reflex dystrophy: 1/1 [above-elbow / below-elbow]
Symptoms of nerve compression: 8/11
Prolonged healing time: 0/2
Replastering: 5/7
NotesDanish paper: brief translation obtained.
Mixed intra/extra articular fractures.
Imbalance in numbers at allocation referred to but not fully explained.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Stewart 1984

MethodsRandomisation based on date of birth (quasi-RCT)
Assessor blinding: yes for functional assessment
Intention to treat: problems - 8 excluded before treatment completed
Loss to follow up (6 months): 34 (14%)
ParticipantsDistrict General Hospital, UK
243 participants
Inclusion criteria: patients with displaced Colles' fracture [not defined but included minimally displaced fractures]
Exclusion criteria: not stated
Classification: not stated; extra- and intra-articular fractures
Sex: 207 female
Age: mean 60 years; range 18-86 years
Assigned: 70/72/93 [supination / short brace / POP]
Assessed: 64/67/84 [3 months]; 59/67/83 [6 months]
InterventionsTiming of intervention: within 10 days of injury
Duration of treatment: 6 weeks.
Majority (193 = 79%) of fractures were reduced, mainly under IVRA. Below-elbow back slab applied for 9-10 days. All splintage removed at 6 weeks.
(1) Above-elbow cast bracing in supination
(2) Below-elbow cast brace
(3) Standard forearm POP
OutcomesLength of follow up: 6 months; also assessed at times shown below.
(1) Anatomical: X-ray at presentation, post reduction, 10 days and 6 weeks. Radial angle and length, dorsal angle. Graded modified Sarmiento criteria based on Lidstrom.
(2) Functional: at 3 and 6 months. Combined score grades - subjective, objective including ROM, complications [Gartland and adapted Sarmiento]
(3) Clinical:
Complications:
Carpal tunnel compression: 18/19 [supination & short brace / POP] (3 months); 26 overall (6 months)
Radial nerve injury: 11 in "braced groups" (3 months); 1 at 6 months
Attending physiotherapy for excessive finger / shoulder stiffness after 6 weeks: 70 in all
EPL rupture: 1 in all
Dupuytren's disease: 23 overall (6 months)
Sudeck's osteodystrophy: none
Shoulder stiffness: none
Secondary surgery: 3 (Baldwin's operation) + 9 nerve decompressions overall
Trigger finger: 3 overall (3 months); 1 (6 months)
Ulnar nerve compression: 2 overall (3 months); 1 (6 months)
(4) Resource use:
Equipment problems: 38 visits/142 for both braced groups; 3 participants/93 POP group. Included: Serious sweat rashes: 2 in brace groups (braces removed)
NotesEleven fractures were remanipulated before allocation.
Ratio of intra-articular fracture not identical between groups (reported as not statistically significant).
New data on complications reported in further report included in the second update.
Fractures with "less than 10 degrees of dorsal angulation and scarcely any shortening or radial angulation" were not reduced.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?NoC - Inadequate

Stoffelen 1998

MethodsMethod of randomisation: not stated
Assessor blinding: unlikely, not reported
Intention to treat: not stated, potential problems - see notes
Loss to follow up: not known
ParticipantsTeaching hospital, Belgium
52 participants
Inclusion criteria: minimally displaced, stable distal radius fracture. No dorsal comminution.
Exclusion criteria: dorsal angulation > 10 degrees, radial shortening > 2 mm, radial shift > 2 mm
Classification: Frykman (no data)
Sex: male and female
Age: not stated
Assigned: 25/27 [1 week / 3 weeks]
Assessed: ?/? (25/27 assumed)
InterventionsTiming of intervention: not stated.
(1) Below-elbow cast with wrist in neutral position for 1 week
(2) Same for 3 weeks
A removable Valpeau dressing was applied after plaster removal in some cases.
OutcomesLength of follow up: 1 year; also assessed at 6 weeks, and 3 and 6 months.
(1) Anatomical: no information
(2) Functional: pain (VAS: 0-10) at plaster removal. Overall grading (Cooney score). Grip strength.
(3) Clinical: patient satisfaction
Complications:
Continuation of plaster immobilisation 0/0 [1 week / 3 weeks]
Use of Velpeau dressing: no data
Reflex sympathetic dystrophy: 1/3 (transient); 0/2 (unresolved)
Notes

Reference (no. 9 in paper) to the Cooney scoring system was incorrect.

Original attention was to treat the 1 week group with removable splints but inhibiting pain prompted a change to 1 week of plaster. It is not clear if these patients were initially part of the trial and thence excluded, and whether those in the 3 weeks group who were randomised at the same time were included or excluded.

Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Tumia 2003

MethodsRandomisation by telephone using computer generated randomisation scheme; stratified by severity of fracture (minimally displaced; displaced)
Assessor blinding: blind assessment of function at 8, 12 & 24 weeks
Intention to treat: likely, but case notes lost for 10 people
Lost to follow up: probably none but 10 excluded (3%)
ParticipantsTeaching hospitals: 5 trauma centres based in 3 countries: Sweden, The Netherlands and UK
339 participants
Inclusion criteria: unilateral Colles' fracture. Minimally displaced and displaced fractures. Fused epiphysis.
Exclusion criteria: age < 18 years, previous forearm fracture
Classification: Frykman (intra- and extra-articular).
Sex: 271 female
Age: mean 59 years, range 18-98 years
Assigned: 169/170 [brace / POP]
Assessed: 166/163
InterventionsTiming for intervention: not stated
Reduction for 188 displaced fractures.
(1) Aberdeen Colles-fracture brace: a 2 part plastic brace applying 3 point loading; parts connected by 2 elasticated Velcro strips. Allows wrist movement. Brace removed 5-6 weeks
(2) POP dorso-radial slab completed 10-14 days later to Colles-type cast. Cast removed 5-6 weeks
OutcomesLength of follow up: 24 weeks; also 10 days, and 7, 8 and 12 weeks.
(1) Anatomical: X-ray at 0, 10 and 35 days. Overall score based on dorsal angulation, radial shortening and radial angulation [Bunger].
(2) Functional: 5, 8, 12 & 24 weeks. Pain, grip strength, overall functional scores [modified Gartland & Werley]
(3) Clinical: complications in functional score
Notes

Lead trialist provided a pre-publication draft of the paper and further details of method of randomisation.

Request for further information sent 13/11/02.

Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?YesA - Adequate

Van der Linden 1981

MethodsMethod of randomisation: not clear, “random order operations design”
Assessor blinding: not reported
Intention to treat: unlikely, subgroup analysis only - no denominators
Lost to follow up: not reported
ParticipantsTeaching hospital, Sweden
250 participants
Inclusion criteria: undisplaced and displaced Colles' fracture
Exclusion criteria: not stated
Classification: not stated
Sex: 211 female
Age: range 16-92 years
Assigned: ?/?/?/?/? [5 groups]
Assessed: ?/?/?/?/?
InterventionsTiming of intervention: not stated
Reduction under local anaesthesia except in 48 undisplaced fractures. All below-elbow plaster immobilisation.
Duration of treatment: 10 days + 4 weeks.
(1) Circular plaster cast, hand in palmar flexion and pronation, with ulnar deviation (Cotton-Loder position). Replaced after 10 days by dorsal splint, hand neutral, ulnar deviation preserved.
(2) Dorsal splint, hand neutral, ulnar deviation preserved.
(3) Circular plaster cast, hand neutral, ulnar deviation preserved.
(4) Dorsal splint, hand neutral, without ulnar deviation.
(5) Circular plaster cast, hand neutral, without ulnar deviation.
OutcomesLength of follow up: 8 months; also 10 days, POP removal (+4 weeks), and +4 weeks.
(1) Anatomical: X-ray at all follow-up times. Dorsal angulation, radial shift.
(2) Functional: Pain, grip strength, range of movement.
(3) Clinical: swelling, cosmetic results. Complications:
Redisplacement
Rereduction: 16 in all
NotesBaseline and follow-up denominators not given. Results (mean values) presented without denominators for the 202 displaced fractures only.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Vang Hansen 1998

MethodsMethod of randomisation: not stated
Assessor blinding: not reported
Intention to treat: not known, but exclusions (see notes)
Lost to follow up: 27 (27%)
ParticipantsRegional hospital, Denmark
100 participants
Inclusion criteria: Colles' fracture, Older type 1 and 2, age 18+ years
Exclusion criteria: hemiplegia, pathological fracture, senile dementia, outside hospital area, unwilling to join
Classification: Older (1 & 2)
Sex: 57 female (of 73)
Age: mean 60/61 years (range 18-96 years) (of 73)
Assigned: ?/? [3 weeks POP/ 5 weeks]
Assessed: 37/37 (fractures; 1 year)
InterventionsTiming of intervention: not stated.
If not reduced, wrist immobilised in neutral position. Where reduced (no definition), MUA with local anaesthesia, wrist immobilised with hand in slight pronation, flexion and ulna deviation.
(1) Below-elbow dorsal POP immobilisation for 3 weeks
(2) Below-elbow dorsal POP immobilisation for 5 weeks
OutcomesLength of follow up: 1 year; also at 10 days and at end of immobilisation, and 3 days after that.
(1) Anatomical: X-ray at 10 days, 3-5 weeks, 3-5 weeks + 3 days, 1 year. Appearance, dislocation.
(2) Functional: at 1 year. Pain, range of movement, grip score
(3) Clinical:
Complications:
Dislocation: 6 in all
Tendon rupture: 1/1 [3 weeks POP/ 5 weeks]
NotesSix type 2 fractures excluded from study as they dislocated within 10 days - all had K-wire fixation after remanipulation.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Wahlstrom 1982

MethodsMethod of randomisation: not stated
Assessor blinding: not known
Intention to treat: not known
Loss to follow up: not stated (probably none)
ParticipantsRegional hospital, Sweden
42 participants
Inclusion criteria: people with closed displaced extra-articular Colles' fracture, females > 40 years
Exclusion criteria: not above
Classification: Lidstrom group II a+c
Sex: all female
Age: mean 65 years
Assigned: 16/14/12 [supination / pronation / neutral]
Assessed: 16/14/12
InterventionsTiming of intervention: not stated.
All fractures reduced under haematoma block.
Duration of treatment: not stated.
Forearm (up to MCP joints) POP with forearm in:
(1) Supination
(2) Pronation
(3) Neutral (mid-way)
OutcomesLength of follow up: 5 weeks
(1) Anatomical: X-ray timing injury, post reduction, 10 days, 4-5 weeks. Dorsal angle. Redislocation.
(2) Functional: not given
(3) Clinical:
Complications:
Redisplacement: 8/2/6 [supination / pronation / neutral]
Rereduction: 3/1/1
NotesResults only given for 5-weeks follow up.
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Wilson 1984

  1. a

    Definition: axial compression is radial shortening
    Abbreviations:
    <: less than
    >: greater than
    AO: Arbeitsgemeinschaft fur Osteosynthesefragen / Association for the Study of Internal Fixation (or ASIF)
    A&E: accident and emergency
    CTS: carpal tunnel syndrome
    DTG: double tubigrip bandage
    EPL: extensor pollicus longus
    GA : general anaesthetic
    IVRA: intravenous regional anaesthesia
    MCP: metacarpo-phalangeal
    MUA: manipulation under anaesthesia
    POP: plaster of Paris
    RCT: randomised controlled trial
    ROM: range of movement (wrist / forearm)
    RSD: reflex sympathetic dystrophy
    VAS: visual analogue scale

MethodsMethod of randomisation: not stated
Assessor blinding: not known
Intention to treat: likely
Loss to follow up: none
ParticipantsTeaching hospital, UK
41 participants
Inclusion criteria: people with displaced Colles' fracture [not defined], simple and comminuted, requiring manipulation.
Exclusion criteria: too frail, bilateral fractures or multiple injuries
Classification: not stated
Sex: not stated
Age: not stated
Assigned: 21/20 [supination / pronation]
Assessed: 21/20 [3 months]
Interventions

Timing of intervention: on day of attendance.
Duration of treatment: 4 weeks.
All fractures reduced under regional block by non-specialist junior staff.
(1) Above-elbow POP slab, completed at 24 hours, with forearm in supination. Cut down at 2 weeks to forearm.
(2) Standard forearm POP slab, completed at 24 hours, in pronation

All wrists held in volar flexion and ulnar deviation

OutcomesLength of follow up: 12 weeks
(1) Anatomical: X-ray at presentation, post reduction 1, 2 [above-elbow group] and 4 weeks. Radial deviation and volar tilt [Gartland & Werley]
(2) Functional: Frykman criteria assessment (strength, motion, pain) at 11-12 weeks post-injury
(3) Clinical:
Complications: mention of finger stiffness in group (1)
NotesShort follow up
Risk of bias
ItemAuthors' judgementDescription
Allocation concealment?UnclearB - Unclear

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Davison 2001Children only. Conservative treatment for torus fractures.
Earnshaw 1999Not in scope of review: reduction technique trial.
Harris 2002Randomised trial investigated early mobilisation in people with unilateral Colles' fracture. Four treatment groups: Stable fractures - no manipulation: 1 week versus 5 weeks plaster cast; unstable fractures - manipulation: 2-3 weeks versus 6 weeks plaster cast. Trial was stopped after recruiting 20 patients. Main investigator (Dr T Harris) left post. Information provided by co-investigator, Dr James Metcalfe. No analysis or publication planned.
Hearne 2003

Trial aimed to compare early mobilisation (wrist splint as required and standard physiotherapy) after 10-14 days plaster cast immobilisation versus 4 weeks plaster cast immobilisation in people with stable distal radius fractures with no other relevant or significant pathology, aged 50+ years.

This was aborted with no recruited patients due to difficulties in obtaining a satisfactory definition of a stable fracture and lack of patient consent.

Kongsholm 1987Not in scope of review: anaesthesia and reduction technique trial.
Kowalski 2002Trial compared fibreglass versus plaster of Paris casts in fractures of the arm and leg but did not split by fracture type. No response was received from lead trialist to a request for separate fracture information sent 31/10/2002.
McGeoy 1986Unlikely to be a randomised comparison. Only available as a conference abstract.
Milliez 1992Randomised trial evaluating the duration of post-surgical immobilisation. Now included in our review of surgical treatment.
Moran 2002Randomised trial indicated as having started 01/12/2000 is currently postponed until funding obtained. Message from Mr Moran's secretary on 6 August 2002.
Pool 1973Though sometimes stated to be a randomised trial, allocation into four groups was not random or concurrent: the first 50 patients were placed into Group 1, the next into Group 2 and so on.
Shah 2002Described in two abstracts as a prospective study, where "undisplaced distal radial fractures were divided into two groups": plaster immobilisation versus removable volar splint. There was no indication that this was a randomised trial and no further information was forthcoming.
Shah 2003Non randomised comparison of softcast (combicast) versus plaster of Paris cast in 39 people with distal radial fracture.
Shanker 2000Randomised trial, listed in National Research Register (UK), comparing conservative management versus primary manipulation of minimally displaced Colles' fractures in adults. Trial aborted after recruiting 3 patients. Reason given: "Non runner".
Singhania 2001Trial listed in National Research Register (UK) as a comparison of 3 weeks in plaster cast then "Futura" splint versus conventional treatment (5 to 6 weeks in plaster cast) for patients over 50 years of age with an isolated wrist fracture. Failed to trace Dr A K Singhania, no longer at Hull Royal Infirmary, despite help from Mrs Susan Yu. Study may not have been a randomised trial anyway.
Smith 1999Children only. Trial listed in National Research Register (UK) on the management of torus fractures of distal radius.

Characteristics of ongoing studies [ordered by study ID]

Ancillary