Rehabilitation for distal radial fractures in adults

Helen HG Handoll; Joanne Elliott

doi:10.1002/14651858.CD003324.pub3

. 2015 Sep 25;2015(9):CD003324. doi: 10.1002/14651858.CD003324.pub3

Rehabilitation for distal radial fractures in adults

Helen HG Handoll ^1,^✉, Joanne Elliott ²

Editor: Cochrane Bone, Joint and Muscle Trauma Group

PMCID: PMC9250132 PMID: 26403335

Abstract

Background

Fracture of the distal radius is a common clinical problem, particularly in older people with osteoporosis. There is considerable variation in the management, including rehabilitation, of these fractures. This is an update of a Cochrane review first published in 2002 and last updated in 2006.

Objectives

To examine the effects of rehabilitation interventions in adults with conservatively or surgically treated distal radial fractures.

Search methods

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL 2014; Issue 12), MEDLINE, EMBASE, CINAHL, AMED, PEDro, OTseeker and other databases, trial registers, conference proceedings and reference lists of articles. We did not apply any language restrictions. The date of the last search was 12 January 2015.

Selection criteria

Randomised controlled trials (RCTs) or quasi‐RCTs evaluating rehabilitation as part of the management of fractures of the distal radius sustained by adults. Rehabilitation interventions such as active and passive mobilisation exercises, and training for activities of daily living, could be used on their own or in combination, and be applied in various ways by various clinicians.

Data collection and analysis

The review authors independently screened and selected trials, and reviewed eligible trials. We contacted study authors for additional information. We did not pool data.

Main results

We included 26 trials, involving 1269 mainly female and older patients. With few exceptions, these studies did not include people with serious fracture or treatment‐related complications, or older people with comorbidities and poor overall function that would have precluded trial participation or required more intensive treatment. Only four of the 23 comparisons covered by these 26 trials were evaluated by more than one trial. Participants of 15 trials were initially treated conservatively, involving plaster cast immobilisation. Initial treatment was surgery (external fixation or internal fixation) for all participants in five trials. Initial treatment was either surgery or plaster cast alone in six trials. Rehabilitation started during immobilisation in seven trials and after post‐immobilisation in the other 19 trials. As well as being small, the majority of the included trials had methodological shortcomings and were at high risk of bias, usually related to lack of blinding, that could affect the validity of their findings. Based on GRADE criteria for assessment quality, we rated the evidence for each of the 23 comparisons as either low or very low quality; both ratings indicate considerable uncertainty in the findings.

For interventions started during immobilisation, there was very low quality evidence of improved hand function for hand therapy compared with instructions only at four days after plaster cast removal, with some beneficial effects continuing one month later (one trial, 17 participants). There was very low quality evidence of improved hand function in the short‐term, but not in the longer‐term (three months), for early occupational therapy (one trial, 40 participants), and of a lack of differences in outcome between supervised and unsupervised exercises (one trial, 96 participants).

Four trials separately provided very low quality evidence of clinically marginal benefits of specific interventions applied in addition to standard care (therapist‐applied programme of digit mobilisation during external fixation (22 participants); pulsed electromagnetic field (PEMF) during cast immobilisation (60 participants); cyclic pneumatic soft tissue compression using an inflatable cuff placed under the plaster cast (19 participants); and cross‐education involving strength training of the non‐fractured hand during cast immobilisation with or without surgical repair (39 participants)).

For interventions started post‐immobilisation, there was very low quality evidence from one study (47 participants) of improved function for a single session of physiotherapy, primarily advice and instructions for a home exercise programme, compared with 'no intervention' after cast removal. There was low quality evidence from four heterogeneous trials (30, 33, 66 and 75 participants) of a lack of clinically important differences in outcome in patients receiving routine physiotherapy or occupational therapy in addition to instructions for home exercises versus instructions for home exercises from a therapist. There was very low quality evidence of better short‐term hand function in participants given physiotherapy than in those given either instructions for home exercises by a surgeon (16 participants, one trial) or a progressive home exercise programme (20 participants, one trial). Both trials (46 and 76 participants) comparing physiotherapy or occupational therapy versus a progressive home exercise programme after volar plate fixation provided low quality evidence in favour of a structured programme of home exercises preceded by instructions or coaching. One trial (63 participants) provided very low quality evidence of a short‐term, but not persisting, benefit of accelerated compared with usual rehabilitation after volar plate fixation.

For trials testing single interventions applied post‐immobilisation, there was very low quality evidence of no clinically significant differences in outcome in patients receiving passive mobilisation (69 participants, two trials), ice (83 participants, one trial), PEMF (83 participants, one trial), PEMF plus ice (39 participants, one trial), whirlpool immersion (24 participants, one trial), and dynamic extension splint for patients with wrist contracture (40 participants, one trial), compared with no intervention. This finding applied also to the trial (44 participants) comparing PEMF versus ice, and the trial (29 participants) comparing manual oedema mobilisation versus traditional oedema treatment. There was very low quality evidence from single trials of a short‐term benefit of continuous passive motion post‐external fixation (seven participants), intermittent pneumatic compression (31 participants) and ultrasound (38 participants).

Authors' conclusions

The available evidence from RCTs is insufficient to establish the relative effectiveness of the various interventions used in the rehabilitation of adults with fractures of the distal radius. Further randomised trials are warranted. However, in order to optimise research effort and engender the large multicentre randomised trials that are required to inform practice, these should be preceded by research that aims to identify priority questions.

Keywords: Adult; Aged; Female; Humans; Male; Fractures, Bone; Fractures, Bone/rehabilitation; Physical Therapy Modalities; Radius Fractures; Radius Fractures/rehabilitation; Randomized Controlled Trials as Topic; Wrist Injuries; Wrist Injuries/rehabilitation

Plain language summary

Rehabilitation as part of treatment for adults with a broken wrist

Background and aim

Particularly in older women, a broken wrist (comprising a fracture at the lower end of the radius, one of the two forearm bones) can result from a fall onto an outstretched hand. Treatment usually includes putting the bone fragments back in place, if badly displaced, and immobilising the wrist in a plaster cast. Surgery may be considered for more seriously displaced fractures. Rehabilitation with interventions such as exercises and other physical interventions is used to help prevent complications, such as stiffness and aching, restore function and speed up recovery.

This review set out to evaluate the effects, primarily on wrist function, of different rehabilitation interventions for treating these injuries.

Search results

We searched the scientific literature up to January 2015 and found 26 randomised controlled studies, involving 1269 mainly female and older patients. Only four of the 23 treatment comparisons covered by these 26 studies were tested by more than one study. Participants of 15 studies were initially treated with plaster cast immobilisation. Some or all participants in the other 11 studies were treated with surgery. In seven studies, the rehabilitation intervention being tested started during wrist immobilisation. In the other 19 studies, rehabilitation started when the cast had been removed.

All studies were small and were designed in a way that may affect the reliability of their findings. Most studies did not report on patient‐reported outcome measures of function and did not follow up patients for long enough. We judged the quality of the reported evidence as either low or very low and thus we are not confident that the results described below are true.

Key results

Interventions started during immobilisation   Two studies provided very low quality evidence that rehabilitation (hand therapy or task‐orientated therapy) improved hand function after the plaster cast was removed, but not in the longer‐term. One study provided very low quality evidence that outcome after supervised exercises did not differ from outcome after unsupervised exercises. Four studies provided very low quality evidence of some slight benefits of four different single methods of rehabilitation that were given with standard care.

Interventions started post‐immobilisation, mainly after removal of the plaster cast   There was very low quality evidence from one study of improved function for a single session of physiotherapy, primarily advice and instructions for a home exercise programme, compared with 'no intervention' after cast removal. There was low quality evidence from four very different studies of no clinically important differences in outcome in patients receiving routine physiotherapy or occupational therapy in addition to instructions for home exercises versus instructions for home exercises from a therapist. There was very low quality evidence of better short‐term hand function in the participants given physiotherapy than in those given either instructions for home exercises by a surgeon (one study) or a progressive home exercise programme (one study). Both studies comparing physiotherapy or occupational therapy versus a progressive home exercise programme after surgery involving plate fixation (a metal plate and screws are used to hold the broken bone in place) found low quality evidence in favour of a structured programme of home exercises preceded by instructions or coaching. One study provided very low quality evidence of a short‐term, but not persisting, benefit of accelerated compared with usual rehabilitation after surgery involving plate fixation.

For studies testing single interventions applied post‐immobilisation, there was very low quality evidence of no clinically significant differences in outcome in patients receiving passive mobilisation, ice, pulsed electromagnetic field (PEMF), PEMF plus ice, whirlpool immersion, and a dynamic extension splint for patients with a stiff wrist, compared with no intervention. This finding applied also to single studies comparing PEMF versus ice, and a new type of massage treatment for swelling when compared with the traditional approach. There was very low quality evidence from single studies of a short‐term benefit of continuous passive motion immediately after removal of external fixation, intermittent pneumatic compression and ultrasound therapy.

Conclusions

We concluded that there was not enough evidence available to determine the best form of rehabilitation for people with wrist fractures. Priority questions need to be identified before further studies are done.

Background

Description of the condition

Fracture of the distal radius is one of the most common fractures in many predominantly white and older populations (Sahlin 1990; Singer 1998). It has been estimated that a 50‐year‐old white woman in the USA or Northern Europe has a 15% lifetime risk of a distal radius fracture; whereas a white man of the same age has a lifetime risk of a little over 2% (Cummings 1985). A prospective survey, conducted in six centres in the UK, of Colles' fracture in patients aged 35 years and above, reported the overall annual incidence of this fracture to be 9/10,000 in men and 37/10,000 in women (O'Neill 2001). In 2000, an incidence of 195 distal radius fractures per 100,000 adults was reported for the population (534,715 adults) served by the Edinburgh Royal Infirmary (Court‐Brown 2006). Court‐Brown 2006 reported that the average age of patients with these fractures was 56 years (46% were over 65 years) and that the male:female ratio was 31:69. However, this hides the bimodal distribution of the age data, with a cross‐over of the two peaks of incidence at around 50 years of age; the second and higher peak being populated mainly by older women (see Figure 5 within Costa 2015). Distal radius fractures are usually treated on an outpatient basis, with around 20% of patients (mainly older people) requiring hospital admission (Cummings 1985; O'Neill 2001). This percentage is likely to be underestimated given the increasing use of surgery and the ageing population (Nellans 2012).

Most fractures of the distal radius in older people result from low‐energy trauma, such as a fall from standing height or less. In younger adults, these injuries are usually sustained through high‐energy trauma, such as a traffic accident. The pattern of incidence reflects the bone loss from osteoporosis in older people, as well as an increased number of falls by older women (Nguyen 2001).

These fractures are generally closed and usually involve displacement of fracture fragments. They may be either extra‐articular (leaving the joint surface of the distal radius intact) or intra‐articular (the joint 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 (commonly referred to as a 'dinner fork' deformity) of dorsal displacement, dorsal angulation, dorsal comminution (fragmentation), and radial shortening.

The majority of distal radial fractures are treated conservatively (non‐operatively). This usually involves the reduction of the fracture if displaced, and forearm immobilisation in a plaster cast or brace for around six weeks. Surgical treatment usually involves either closed or open reduction, followed by external or internal fixation and a similar period of immobilisation. Since the last version of the review, there has been a marked increase in surgical intervention, particularly in the use of internal fixation using locking plates. For example, Mattila 2011 reported that the number of plate fixations for these fractures had more than doubled between 2006 and 2008 in Finland. Such clinical interventions (cast immobilisation and surgery followed by immobilisation) are often referred to as 'first definitive' treatments (HSCIC 2015). Although in the following, we refer to these as 'definitive' treatments, this descriptor should not be taken to diminish the status of rehabilitation.

The variety of these 'definitive' treatment options is shown in the Cochrane reviews of conservative (Handoll 2003c) and surgical management of these fractures (Handoll 2003a). The latter review is now replaced by reviews on different surgical methods, such as percutaneous pinning (Handoll 2007), with the exception of internal fixation for which a review is underway (Hoare 2014).

These injuries can result in increased morbidity, with long‐term functional impairment, pain and deformity (Edwards 2010; Handoll 2003a; Handoll 2003c). They are also associated with a high incidence of varied complications (McKay 2001); for example, serious complications, such as persistent neuropathies of the median, ulnar or radial nerves, have been reported in one in three patients (Cooney 1980). One major complication is complex regional pain syndrome type 1 (CRPS‐1), also referred to as reflex sympathetic dystrophy, algodystrophy, Sudeck's atrophy and shoulder‐hand syndrome. Serious cases of CRPS‐1 require many months of therapy to alleviate symptoms (pain, tenderness, impairment of joint mobility, swelling, dystrophy, vasomotor instability) (Atkins 1996).

Description of the intervention

Rehabilitation refers to the overall process of helping people to make the best possible recovery from their injury. The issues surrounding the rehabilitation of patients with a distal radial fracture can be expressed in terms of four basic questions:

What sort of intervention(s) should be used?
Who should provide them?
When and for how long?
Why?

A variety of interventions are available for use. Advice, patient education and supervision for active and passive mobilisation exercises, therapist‐applied mobilisation techniques, continuous passive motion, strengthening exercises, supportive splints, physical methods of pain management such as transcutaneous electrical nerve stimulation (TENS), heat treatment, massage, wound care, manual aids and occupational/home assessment are some of the more common therapeutic methods used to maximise the patient's functional recovery (Collins 1993). A small selection of these, commonly advice and mobility exercises, may be employed on a general basis for all patients. Usually though, interventions are selected and adapted by clinicians to meet the specific rehabilitation challenges presented by individual patients. Specific rehabilitation programmes of usually physical interventions (primarily exercises) based on a standardised protocol may also be applied; but, because it is not always possible to standardise to the last detail, some flexibility is common. Although drugs may be prescribed, for instance for pain relief, these are not reviewed here.

As well as the doctors, commonly orthopaedic surgeons, providing definitive treatment, other clinicians are often involved in the rehabilitation of patients with these injuries. These other clinicians may be physiotherapists, occupational therapists or nurses, many of whom are specialised in hand and/or upper limb therapy. The distinctions between the activities and roles of these clinicians often overlap and also vary geographically. Generally, physiotherapists aim to help the patient restore or achieve optimal movement and physical function. Occupational therapists share this aim but focus on helping patients to achieve independence in activities of daily living. Nurses may play a varied role, including that of rehabilitation, but plaster cast management and care of surgical wounds would be typical activities. These latter activities are viewed as part of clinical treatment for the purposes of this review.

The issue of when to commence rehabilitation is controversial. Rehabilitation could start as soon as possible after the injury and continue throughout, or rehabilitation could be seen as a subsequent stage in patient management and undertaken after the initial clinical treatment (e.g. cast immobilisation) is over. Therefore, the two key phases for management of these injuries are during initial clinical treatment, which usually involves immobilisation, and post‐immobilisation (after plaster cast or external fixator removal). Upon receiving initial treatment, such as fracture reduction and application of a plaster cast, patients are usually given instructions to carry out straightforward exercises. These typically include elevation of the injured arm in the first few days post‐injury and exercising of the non‐immobilised joints in order to alleviate and/or counter swelling and stiffness. More extensive and intensive rehabilitation intervention is more frequent post‐immobilisation; at this stage, limited range and quality of movement, reduced grip strength, and pain are typical reasons for initiating rehabilitation interventions.

The 'why' question mainly concerns the clinical indication for the intervention(s). Our main focus is on studying the effects of rehabilitation interventions on preventing complications associated with the fracture and/or treatment and on optimising functional recovery and achievement of activities required for daily living. Rehabilitation interventions may also be prescribed to treat complications, such as CRPS‐1, of these fractures. We acknowledge the difficulties in distinguishing the two situations since there will be overlap but, given our main aim, we noted the reasons for starting or providing the interventions in individual trials. The aims, including intended trial populations, and the primary outcome(s) of individual trials helped us to distinguish between those trials evaluating interventions to resolve or prevent 'problems' and those investigating treatment options for complications. The latter are not included in this review. Similarly excluded are trials primarily investigating interventions for pain relief, acceleration of bone healing, osteoporosis or secondary prevention of fractures.

Editorial feedback for this version of the review prompted consideration of a possible fifth question: 'where'? Although pertinent, we decided against an additional category at this time because 'where should treatment take place?' is generally subsidiary to other questions when set in the context of the care pathway for these fractures. For example, the prime consideration of supervised rehabilitation versus home exercises resolves around the question of who provides rehabilitation (a clinician versus the patient) rather than care in a [hospital] clinic versus exercises at home. While we describe the locations of the interventions tested in the review, so far none of the comparisons is predominantly linked with locality. Nonetheless, this item will be reconsidered before the next version of the review.

How the intervention might work

The central rationale for rehabilitation has been discussed above. A comprehensive summary of rehabilitation interventions, several of which are usually used in combination, is not provided here. Instead Table 1 provides some illustrative examples of putative mechanisms of a few single rehabilitation interventions, only two of which (advice and instructions, joint mobilisation) are in common use.

1. Descriptions and putative mechanisms of some rehabilitation interventions.

Intervention	Description	Aim	Putative mechanism
Advice and instructions	These usually include patient education on the ways to cope with the consequences of the injury (such as pain) and avoid some of the common problems, such as stiff joints, associated with immobilisation and instruction on exercising uninvolved joints during immobilisation, with the wrist joint added subsequently These are delivered in various ways	To empower the patient in their own recovery	Facilitating patient participation in their own recovery by providing information on the means. One key intervention is joint mobilisation (see below)
Cross‐education	This involves strength training (exercises) of the opposite non‐injured hand (cross‐education)	To improve function of the involved hand	Strength training in one limb produces neural adaptations that increase strength in the contralateral limb
Dynamic wrist splint	A splint applied to the forearm that provides a continuous low‐load stretch while enabling hand movement	To improve hand function by increasing range of wrist motion	Stretching the wrist in a steady way will improve physiological parameters such as blood flow to the muscles
Ice	Ice or iced water or other cooling devices	To relieve pain and reduce swelling	Ice reduces metabolic activity within the tissues and thus should prevent secondary damage. It also reduces pain signals to the central nervous system
Joint mobilisation (active and passive)	This involves the movement of a joint and can take two basic forms, which are not mutually exclusive: Active: under the control of the patient’s own musculature  Passive: via an externally applied force, such as by a therapist Passive mobilisation can also be applied by machines that repeatedly move the joint(s) through a prespecified range of motion	To restore or maintain range of movement and avoid complications	Movement is part of keeping active and fit. Not moving results in loss of mobility (stiffness and joint contracture) and reduced muscle strength The range of movement under active control is usually less than that available passively; but a key advantage is that active movement emulates physiological function Passive mobilisation may be undertaken by the patient, but is more usually applied by a therapist Automation of the intervention via a machine can free up clinician time but involves additional technology and restricts the patient to a specific location
Soft tissue compression (cyclic or intermittent)	This is delivered in various ways, generally involving machines providing an intermittent pneumatic compression delivered via an inflatable device round the hand or wrist	To reduce or prevent swelling	Pulsed compression of surrounding tissues helps in circulation and prevents build up of fluid that would impair movements and recovery

Study ID	Number	Gender & age (years)	Fracture  (Ex = exclusion)	Definitive  treatment  and timing	Complications  (Ex = exclusion)	Comorbidities or  prior function  (Ex = exclusion)	Comments
Bache 2001	98	F: 84%  Age: median 69  range 50 to 92	DRF  Ex: bilateral	Plaster cast Post‐immobilisation	Ex: early signs CRPS‐1 or CTS	Ex: unable to follow exercise programme independently, impaired mental function, prior wrist or joint problems, physiotherapy required for other reason	Population was without serious complications or comorbidities
Basso 1998	38	F: 84%  Age: median 57 and 63  range 15 to 69	Colles'  Ex: intra‐articular, severe comminution, open	Plaster cast Post‐immobilisation	Ex: very poor hand function, severe injuries including to ligaments	Ex: unable to cope with assessments	Population was without serious complications and restricted by fracture type
Bighea 2013	20	F: unknown% Age: unknown	Wrist (osteoporotic)	Plaster cast (probably) Post‐immobilisation	N/A	N/A	Abstract report only
Brehmer 2014	81	F: 73%  Age: mean 53  range 21 to 83	DRF  Ex: bilateral, prior fracture, fracture not stable enough	Surgery ‐ internal fixation Post‐immobilisation (strictly: post‐definitive treatment)	Not stated	Not stated	Population limited by suitability for surgery. Professional athletes excluded
Challis 2007	21	F: 76%  Age: mean 51	DRF  Ex: open, surgical fixation required	Plaster cast Definitive treatment period	Ex: nerve or tendon damage	Ex: pathology in intact arm	Population limited by severity of injury
Cheing 2005	83	F: 66%  Age: mean 63  range 17 to 80	DRF  Ex: unstable, previous fracture	Plaster cast Post‐immobilisation	Ex: CRPS‐1	Ex: inflammatory arthritis, prior neurovascular injuries of hand, various circulatory diseases etc	Population was without serious complications or comorbidities, including contraindications for intervention (PEMF)
Christensen 2001	32	F: 90%  Age: mean 66  range 42 to 82	Colles'	Plaster cast Post‐immobilisation	N/A	N/A	Population limitations not known (no report of complications)
Cooper 2001	17	F: 94%  Age: mean 66  range 41 to 81	DRF  Ex: open	Plaster cast Definitive treatment period	Ex: extensive soft tissue injuries	Ex: frail elderly with limited mobility, impaired mental function, neurological conditions	Population was without serious comorbidities
Gronlund 1990	40	F: 88%  Age: median 74.5  range 47 to 93	Colles'  Ex: bilateral, unstable	Plaster cast Definitive treatment period	Ex: neuromuscular pain in limb	Ex: wrist arthritis, dementia, other condition preventing participation	Population was without serious complications or comorbidities
Jongs 2012	40	F: 70%  Age: median 66 and 58	DRF  Ex: bilateral, stable and healed/healing	Conservative or surgical treatment Post‐immobilisation (10 weeks from injury)	Wrist contracture	Ex: unlikely to co‐operate	Surgeon referral Population limited to those with wrist contracture (thus a complication) and expected compliance
Kay 2000	40	F: 68%  Age: mean 53	DRF  Ex: prior fracture	Plaster cast or pins and plaster cast Post‐immobilisation	Ex: only residual impairment from a prior fracture	Ex: inability to understand English	Study population includes those with complications (as reported)
Kay 2008	56	F: 70%  Age: mean 55	DRF  Ex: bilateral	Plaster cast or pins and plaster cast Post‐immobilisation	Ex: only residual impairment from a prior fracture	Ex: pre‐existing inflammatory condition, inability to understand English, unable to participate	Study population includes those with complications (as reported)
Knygsand‐Roenhoej 2011	30	F: 72%  Age: mean 78	DRF  Ex: bilateral	Plaster cast or internal or external fixation Post‐immobilisation	Subacute oedema	Ex: mental impairment, infection, lymphedema, disease of internal organs	Population limited to those with subacute oedema (thus a complication) but without serious comorbidities
Krischak 2009	48	F: 65%  Age: mean 55  range 18 to 76	DRF  Ex: previous fracture	Internal fixation with locking plate Post‐immobilisation (strictly: post‐definitive treatment)	Ex: CTS, CRPS‐1	Ex: mental impairment, functionally dependent, inflammatory joint disease	Population was without serious complications or comorbidities and limited by suitability for surgery
Kuo 2013	22	F: 65%  Age: mean 62	DRF  Ex: open	External fixation Definitive treatment period	Ex: neurological defects or severe soft tissue damage	Ex: rheumatism, arthritis, neurological deficits, previous hand or forearm injury	Population was without serious complications or comorbidities and limited by suitability for surgery
Lazovic 2012	60	F: 100%  Age: mean 66  range 55 to 78	DRF  Ex: bilateral, intra‐articular, previous fracture	Plaster cast Definitive treatment period	None stated	Ex: inflammatory osteoarthritis, peripheral vascular disease, contra‐indications to PEMF	Population was without serious comorbidities, including contraindications for intervention (PEMF)
Maciel 2005	45	F: 76%  Age: mean 56	DRF	Plaster cast or pins and plaster cast  Post‐immobilisation	Ex: signs of CRPS‐1	Ex: inability to understand English, psychiatric disorder, inflammatory joint condition	Population was without serious complications or some comorbidities
Magnus 2013	51	F: 100%  Age: mean 63	DRF Ex: bilateral	Plaster cast or surgery and plaster cast Definitive treatment period	None stated	Ex: history of upper limb neurological problems, cognitive impairment	Population was without serious comorbidities
Pasila 1974	135	F: 93% Age: range 16 to 65 (criteria)	Colles'	Plaster cast Definitive treatment period	N/A	N/A	Study population not described in terms of exclusions
Rozencwaig 1996	7	F: unknown% Age: unknown	DRF	External fixation Post‐immobilisation	N/A	N/A	Abstract report only
Souer 2011	94	F: 65%  Age: mean 50	DRF  Ex: stable, complex fracture	Internal fixation with locking plate Post‐immobilisation (strictly: post‐definitive treatment)	None stated	Ex: dependency in basic functioning	Population limited by independency in function and suitability for internal fixation
Svensson 1993	43	F: 100%  Age: median 72  range 55 to 90	Colles'  Ex: previous fracture	Plaster cast Post‐immobilisation	Ex: CRPS‐1	Ex: neurological disease, disfiguring rheumatic disease	Referral for rehabilitation Population was without serious complications or comorbidities
Taylor 1994	30	F: 80%  Age: mean 63  range 39 to 78	Colles'	Plaster cast Post‐immobilisation	None stated	Ex: other upper limb fracture	Study population probably includes those with complications
Toomey 1986	24	F: 83%  Age: mean 60  range 40 to 80	Colles'	Plaster cast Post‐immobilisation	Ex: CRPS‐1 (shoulder‐hand syndrome) and other conditions, occupational therapy for involved hand	Ex: associated fractures, rheumatoid arthritis	Referral for rehabilitation Population was without serious complications
Wakefield 2000	96	F: 91%  Age: mean 73  range 55 to 90	DRF  Ex: bilateral, previous fracture	Plaster cast Post‐immobilisation	Ex: CRPS‐1	Ex: impaired mental function	Population was without serious complications and cognitive impairment
Watt 2000	18	F: 94%  Age: mean 76	Colles'	Plaster cast Post‐immobilisation	None stated	Ex: no significant past history	Attending outpatients Study population may have included complications. It is not clear what "past history" covers

Study ID	What	Who	When	Where	How long	Comments
	Intervention	Provider	Started	Location	Duration	Treatment for all or control group participants
Bache 2001	Physiotherapy	Physiotherapist	Within a week of plaster cast removal; 5‐6 weeks immobilisation	Outpatients clinic	Contents and discharge at discretion of physiotherapists	All participants taught home exercises by physiotherapist at outpatients clinic
Basso 1998	Ultrasound	Not reported	After plaster cast removal; average 4 weeks immobilisation	Fracture clinic?	5 minutes	Sham control  All participants given instructions. Physiotherapy only if poor hand function
Bighea 2013	Physiotherapy	Physiotherapist	Probably after 4 weeks immobilisation	Not stated	Four weeks; not sure if continued after examination	Instructions for a home exercise programme
Brehmer 2014	'Accelerated' post‐op rehabilitation started at 2 weeks	Hand therapist	At 2 weeks post‐surgery	Ambulatory surgery centre	Over 5 weeks.	Both groups began with gentle active range of motion at 3 to 5 days after surgery; removable custom splint applied. Same rehabilitation programme applied in control group but started at 6 weeks and condensed to 3 weeks
Challis 2007	Cyclic pneumatic soft tissue compression during cast immobilisation (inflatable cuff placed under the plaster cast)	Probably physiotherapist for instructions	After week in a split cast (mean 9.4 days after fracture), upon replacement with a full forearm plaster	At home after instructions at fracture clinic	Pump applied for 10 minutes each morning and evening during 5 weeks full forearm cast immobilisation	Device also incorporated into control group cast. Both groups received instructions for hand exercises during 6‐week immobilisation period and were given a 4‐week exercise programme after cast removal
Cheing 2005	Pulsed electromagnetic field (PEMF) or ice, or both	Physiotherapist	After plaster cast removal; average 6 weeks immobilisation	Outpatients clinic?	30 minutes of each for 5 consecutive days	Sham PEMF. All participants taught home exercises by physiotherapist at outpatients clinic
Christensen 2001	Occupational therapy	Occupational therapist	After plaster cast removal; 5 weeks immobilisation	Rheumatology department	Twice weekly sessions until progress plateau	All had instructions for home exercises by occupational therapist
Cooper 2001	'Early' therapy ‐ oedema management, exercises, monitoring, information	Hand therapist	After initial treatment; within 4 days of fracture	Clinic?	During immobilisation, routinely 4 weeks but up to 6 weeks as required. Weekly contact	All had instructions for home programme: skin care, exercises, oedema control at fracture clinic. Post‐immobilisation care comprised an individualised home programme and, if required, attendance of hand therapy group
Gronlund 1990	Occupational therapy	Occupational therapist	After initial treatment; 1 to 3 days after plaster cast application	Rheumatoid disorders outpatients clinic	Single session with referral as required	All had instructions for home exercises by occupational therapist at casualty ward. Referral for occupational therapy if required after plaster cast removal
Jongs 2012	A dynamic splint fitted, which stretched the wrist into extension but allowed intermittent movement	Physiotherapist	After 10 weeks from fracture; post‐immobilisation	Physiotherapy and then home	8 weeks. Participants were instructed to wear the splint for as long as possible during the day, aiming for at least six hours overall per day	Both groups received routine care consisting of exercises and advice for 8 weeks, with fortnightly monitoring. Participants continued with the exercises and advice unsupervised until 12 weeks
Kay 2000	Passive mobilisation	Physiotherapist	After plaster cast or plaster & pins removal; about 6 weeks immobilisation	Physiotherapy	6‐week course: 2 x week for 3 weeks, 1 x week for 3 weeks	All attended physiotherapy for initial advice and instructions for home exercises given by physiotherapists
Kay 2008	One session of physiotherapy; primarily advice and instructions	Physiotherapist	After plaster cast or plaster & pins removal; about 6 weeks immobilisation	Physiotherapy	Single session	Control group had no physiotherapy input. Both groups had initial assessment and review at 3 and 6 weeks by an experienced hand therapist
Knygsand‐Roenhoej 2011	Modified "manual edema mobilization" (MEM). Massage then exercise in segment just massaged.	Occupational therapists	Post‐cast immobilisation, or surgery (internal or external fixation): between 4 to 10 weeks	Department of occupational therapy	6‐week course	'Traditional' oedema treatment In all, oedema treatment was performed three times a week for four weeks and then twice a week for two weeks. Subsequently, until acceptable function reached. All patients had ROM and strengthening therapy at hospital and home exercises
Krischak 2009	Physiotherapy (at discretion of therapist; but instructed to implement exercises that could be done by the patients at home)	Physiotherapist (participant choice)	One week post‐surgery; 2 weeks of splint from surgery (removed for therapy)	Physiotherapy	12 sessions lasting 20 to 30 minutes each, over a 6‐week period	Instruction for home exercises in person with an illustrated exercise guidance booklet and training diary provided. Progressive week‐based exercise plan. Instructions for performing groups of exercises, requiring approximately 20 minutes, in morning and evening
Kuo 2013	Early mobilisation of digits; including massage, passive and active range of motion of finger joints, activity of daily living training	Occupational therapist	In the week after surgery	Department of occupational therapy?	3 x 45‐minute sessions each week for 6 weeks until removal of external fixator	Active ROM of shoulder and elbow joints; and hand joints after week 3 onwards. All had standard progressive rehabilitation programme, including scar management, active and passive wrist range of motion and strengthening for six weeks after fixator removal
Lazovic 2012	Pulsed electromagnetic field (PEMF)	Not stated	After clinical and radiographic check at 7 days (mean 8 days post‐fracture)	Rehabilitation clinic	Provided 5 days a week for 2 weeks, 30 minutes daily	No other therapy control Instructions for home exercise programme (uninvolved joint mobilisation) during cast immobilisation (mean 28 days). Then instructed to perform light activities in the pain‐free range of movement until the follow‐up (2 to 3 days). Wrist elevation for increased oedema
Maciel 2005	"Activity‐focussed" physiotherapy	Physiotherapist	After plaster cast or plaster & pins (K‐wire) removal; about 6 weeks immobilisation	Physiotherapy?	Up to 6 weeks, until return to regular wrist activity	All participants taught routine exercises by physiotherapist at fracture clinic on day of cast removal. Control group had single session of advice on exercises from physiotherapist within one week of trial entry
Magnus 2013	Strength training of opposite hand (cross‐education)	Orthopaedic surgeon	Immediately after first clinic visit	Instructions at fracture clinic Unsupervised home exercises	26 weeks	All participants had standard rehabilitation, with three successful paper‐based exercise protocols for the injured arm: during cast use (approximately 6 weeks) 6 ‐8 weeks and 9 weeks onwards. Patients encouraged to continue exercise to 26 weeks)
Pasila 1974	Advice & instructions provided by physiotherapist; supervised exercises	Physiotherapist	After initial treatment; day after plaster cast application	Physical medicine department	Until physiotherapist considered patient able to carry on unsupervised	Control group had same oral & written instructions provided by physician/surgeon
Rozencwaig 1996	Continuous passive motion	"Therapist", probably occupational therapist	After external fixation for 6 to 8 weeks	Not reported. Clinic?	Not reported. Until criteria for recovery met?	All participants had occupational therapy
Souer 2011	Occupational therapy	"Hand therapist"	At the first post‐operative visit to the surgeon's office after volar plate fixation	Not reported.  Clinic?	Frequency and duration of  the rehabilitation programme were at the discretion of the treating hand therapist	Control group had surgeon‐directed home  exercise programme; also supplied with wrist splint
Svensson 1993	Intermittent pneumatic compression	Occupational therapist	After plaster cast removal; treatment started 1‐46 days afterwards	Rheumatological ward	20 minutes before each occupational therapy sessions: thrice weekly for 3 weeks	All participants had occupational therapy. Instructions for exercises given on first session
Taylor 1994	Passive mobilisation	Physiotherapist	Within 3 working days after plaster cast removal; 6 weeks immobilisation	Physiotherapy department	5 minutes during twice weekly physiotherapy sessions; sessions ended when acceptable function or function plateau	Sham control: soft tissue massage. All participants had physiotherapy
Toomey 1986	Whirlpool	Physiotherapist	Within 6 days after plaster cast removal; average 6 weeks immobilisation	Physical medicine department	First 15 minutes scheduled twice weekly physiotherapy over 6 weeks	Control group: towel wrap for 15 minutes
Wakefield 2000	Physiotherapy	Physiotherapist	After plaster cast removal; average 37 days immobilisation	Local hospital or clinic	Contents and discharge at discretion of physiotherapists	All participants taught home exercises by physiotherapist at fracture clinic
Watt 2000	Physiotherapy	Therapist; probably physiotherapist	After plaster cast removal; average 43 days immobilisation	Physiotherapy department	Contents and discharge at discretion of therapists	Control group had oral and written instructions provided by surgeon/registrar

	Clearly defined study population?	Interventions sufficiently described?	Main outcomes sufficiently   described?	Relevant outcomes reported?	Appropriate timing of outcome measurement?   (Yes = ≥ 12 months)
Bache 2001	Yes	Partial: "semi‐standardised" with agreed discharge criteria. (Pragmatic trial)	Yes	Yes	No: 12 weeks
Basso 1998	Yes	Partial: care providers not identified. (Yes for intervention)	Yes	Partial: no PROMs, limited other outcomes but evidence of systematic approach.	No: 8 to 16 weeks
Bighea 2013	No: osteoporotic 'wrist' fracture	Partial: Not clear when started (probably post‐immobilisation); insufficient details of providers	Partial: limited information	Yes: although insufficient details	No: 4 weeks post‐cast removal
Brehmer 2014	Partial: no detail on decision for non‐surgical Table 9 treatment	Yes	Yes	Yes	Partial: 6 months
Challis 2007	Partial: no detail on decision for non‐surgical treatment	Yes	Yes	Partial: no PROMs	No: 10 weeks
Cheing 2005	Yes	Yes	Yes	Partial: no PROMs except pain	No: 5 days
Christensen 2001	Partial: Colles treated by below‐elbow cast. No exclusion criteria	Partial: insufficient details of occupational therapy	Partial: insufficient detail	Partial: no PROMs, reliance on Gartland and Werley outcome measure	Partial: 9 months
Cooper 2001	Yes	Yes	Yes	Yes	No: 4 weeks post‐cast removal
Gronlund 1990	Yes	Partial: insufficient details of occupational therapy	Yes	Partial: no PROMs, reliance on modified Gartland and Werley outcome measure	No: 13 weeks
Jongs 2012	Yes	Yes	Yes	Yes	No: 12 weeks
Kay 2000	Yes	Yes	Yes	Partial: subjective functional scale was new in the study	No: 6 weeks
Kay 2008	Yes	Yes	Yes	Yes	No: 6 weeks
Knygsand‐Roenhoej 2011	Yes	Yes	Yes	Yes	Partial: 6 months
Krischak 2009	Yes	No: choice of physiotherapist by patient; the contents of the 12 sessions of therapy was at the discretion of the physiotherapist, although with an instruction to implement exercises that could be done by the patients unassisted at home    Home exercise group intervention was well described and illustrated	Yes	Yes	No: 6 weeks
Kuo 2013	Yes: although no detail on indication for surgery	Yes	Yes	Yes: the Manual Ability Measure‐36 is an acceptable measure of function	No: 12 weeks
Lazovic 2012	Yes	Partial: care providers not identified	Yes	Yes	No: 2 to 3 days after cast removal
Maciel 2005	Yes	Yes	Yes	Yes	Partial: 24 weeks
Magnus 2013	Partial: no details on fracture type or detail on indication for surgery	Yes	Yes	Yes	Partial: 26 weeks
Pasila 1974	Partial: not fully defined ‐ "typical" Colles fracture and cast and < 65 years	Partial: exercise protocol not described	Yes	Partial: no PROMs, "subjective" attitude not described	No: 12 weeks
Rozencwaig 1996	Partial: no information on participations other than unstable fractures and external fixation	No: insufficient information	No: very limited information	No: no PROMs, inadequate outcome assessment	No: just up to initial recovery (a few weeks)
Souer 2011	Yes	No: while formal occupational therapy included supervised exercises, the content, frequency and duration of occupational therapy was at the discretion of the hand therapist Home exercise group intervention was well described	Yes	Yes	Partial: 6 months
Svensson 1993	Yes	Yes	Yes	Partial: visual analogue scale for pain and use of hand in daily activities but no PROMs. Otherwise, clearly a systematic approach to data collection	No: 3 months
Taylor 1994	Yes	Yes: note that therapists decided upon the details of the specific passive mobilisation treatment	Yes	Partial: no PROMs, but clearly a systematic approach to data collection for the few outcomes	No: until discharge (mean 26 days)
Toomey 1986	Yes	Yes	Yes	Partial: no PROMs, but clearly a systematic approach to data collection.	No: 6 weeks
Wakefield 2000	Yes	No: while this was a pragmatic trial there were insufficient details on the interventions and reported variation in physiotherapy between the trial centres	Yes	Partial: no PROMs, but clearly systematic and otherwise comprehensive	Partial: 6 months
Watt 2000	Partial: not fully defined ‐ Colles' fracture, treated with plaster cast, no "significant past history"	Partial: treatment, while recorded, was at the discretion of the hospital therapists with the content of the treatment recorded but not controlled	Yes	Partial: no PROMs, but clearly a systematic approach to data collection for the few outcomes	No: 6 weeks

Occupational or other hand therapy
PICO	Population: 2 trials, (Cooper 2001), 17 participants; and (Gronlund 1990), 40 participants. All with distal radius fracture treated with plaster cast immobilisation  Intervention: occupational or other hand therapy ‐ Cooper 2001: routine therapy for cast wear for 4 weeks; Gronlund 1990: occupational therapy single session Control: instructions for home care at fracture clinic/casualty ward Outcomes: DASH, modified Gartland and Werley composite outcome score, grip strength, pinch grip, range of motion, dexterity, finger movement, complications and cast problems, referral to hand therapy, use of appliances and home help, oedema, participant satisfaction; follow‐up 4 weeks (Cooper 2001); 13 weeks (Gronlund 1990)
Findings	No pooling undertaken due to lack of common outcomes and heterogeneity. Cooper 2001 found better functional and clinical results in the intervention group at four days after removal of plaster cast, with some beneficial effects continuing one month later. Fewer people in the early therapy group met the criteria for attending the post‐immobilisation hand therapy classes. Gronlund 1990 found that early "occupational therapy" resulted in significantly better hand function at cast removal but not at 13 weeks for older people with stable fractures treated with plaster casts. Assessment at occupational therapy revealed a need for manual aids, home help and plaster cast adjustment as well as a lack of understanding of the instructions for exercises and advice initially provided to all trial participants
GRADE	Very low quality evidence. Reasons for downgrading according to source of data: ‐1 level for study limitations (mainly, performance bias; with confounding from a major imbalance in age in Gronlund 1990), ‐2 levels for serious imprecision (small trials, no pooling)
Comments	Applicability reduced by inadequate follow‐up in both trials, and no PROMs in Gronlund 1990 (Table 4) As well as the small sample size, there are potential problems with confounding in Cooper 2001 due to differences in baseline characteristics and care programmes (there was no information on the numbers requiring longer plaster cast immobilisation). Thus, these promising results need confirmation in a larger sample size, with a longer duration of follow‐up. Small sample size and incomplete assessment of outcome, for instance in the numbers referred for occupational therapy after plaster cast removal, in Gronlund 1990 mean that there was insufficient evidence to confirm either a lack of longer‐term difference in outcome or a short‐term advantage of early occupational therapy
Cyclic pneumatic soft tissue compression
PICO	Population: 1 trial (Challis 2007), 19 followed up of 21 participants with distal radius fracture treated with plaster cast immobilisation Intervention: cyclic pneumatic soft tissue compression using an inflatable cuff placed under the plaster cast ‐ applied at home (Table 3) Control: usual care during immobilisation Outcomes: grip and pinch strength and range of motion; follow‐up 10 weeks
Findings	Intervention group had better and earlier recovery in grip and pinch strength and range of motion at end of treatment (6 weeks from fracture) and 10 weeks (4 weeks post‐immobilisation)
GRADE	Very low quality evidence. Downgraded: ‐1 study limitations (major baseline imbalance in gender, age and fracture severity); ‐2 serious imprecision (very small pilot trial)
Comments	Applicability reduced by incomplete description of study population; no PROMs; inadequate length of follow‐up (Table 4).  The major differences in the baseline characteristics of the two groups (e.g. 4/10 versus 0/9 males) may have contributed to the favourable results for the intervention group
Early digit mobilisation programme during external fixation
PICO	Population: 1 trial (Kuo 2013), 22 participants with distal radius fracture treated with external fixation  Intervention: early digit mobilisation, including massage and passive stretching (18 sessions) ‐ clinic (Table 3) Control: usual care during external fixation Outcomes: Taiwanese version of the MAM‐36, grip and pinch strength, range of motion (workspace), finger dexterity, fracture displacement; follow‐up 12 weeks
Findings	There were no significant between‐group differences in manual ability (measured using the Taiwanese version of the MAM‐36 at 3, 7 or 12 weeks post‐fracture, or in grip and pinch strengths and finger dexterity or fracture displacement at 12 weeks. Although range of digit motion, expressed as finger "workspace" and thumb "workspace", was greater in the intervention group, the clinical importance of this is unclear
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (high risk of performance and detection biases), ‐1 level for imprecision (small trial); ‐1 level for indirectness (non‐validated version of MAM‐36, uncertain clinical relevance of workspace outcome measure)
Comments	Applicability reduced by insufficient follow‐up (Table 4), but also uncertainty over the clinical relevance of outcome measures  These preliminary findings, if true, do not show an important effect that could warrant implementation of this potentially inconvenient and costly intervention: participants needed to attend 3 times a week for 6 weeks; 45 minutes each session
Pulsed electromagnetic field (PEMF)
PICO	Population: 1 trial (Lazovic 2012), 60 women with distal radius fracture treated with plaster cast immobilisation Intervention: pulsed electromagnetic field ‐ 10 sessions in clinic (Table 3) Control: no PEMF Outcomes: PRWE, range of motion, complications, oedema; follow‐up 2 to 3 days after cast removal
Findings	At 2 to 3 days after cast removal, the between‐group differences in PRWE pain and activity/function scores in favour of PEMF are unlikely to have represented clinically important differences. Range of motion was better in the PEMF group with clinically borderline differences in supination, flexion and extension. Although fewer participants of the PEMF group had complications, the diagnosis of these was provisional. The clinical relevance of the results of the hand oedema measure, which favoured the PEMF group, is unclear
GRADE	Very low quality evidence. Downgraded: ‐2 levels for serious study limitations (quasi‐randomised and non‐blinded trial ‐ high risk of selection and detection biases); ‐1 level for imprecision (small trial)
Comments	Applicability reduced by incomplete description of intervention and very inadequate length of follow‐up (Table 4)  As well as concerns about validity of the findings, the inadequate length of follow‐up means the consequences of the preliminary findings are unknown. Potentially inconvenient and costly: participants needed to attend each weekday for 2 weeks; 30 minutes each session
Cross‐education (strength training of the non‐fractured hand)
PICO	Population: 1 trial (Magnus 2013), 39 followed up 51 participants with distal radius fracture treated with cast immobilisation or surgical repair and cast immobilisation Intervention: standard rehabilitation plus strength training of opposite hand (cross‐education) ‐ at home (Table 3) Control: standard rehabilitation Outcomes: PRWE, grip strength and range of motion; follow‐up 26 weeks
Findings	Between‐group differences in patient‐reported function (PWRE) were not significant at the three follow‐ups (9, 12 and 26 weeks). The intervention group had better grip strength results at all three follow‐ups but the between‐group difference was only statistically and clinically significant at 12 weeks. A similar finding in favour of cross‐education at 12 weeks applied to range of motion. More people in the cross‐education group withdrew from the trial
GRADE	Very low quality evidence. Downgraded: ‐2 levels for serious study limitations (high risks of performance bias and attrition bias ‐ imbalance in participants lost to follow‐up (9/27 versus 3/24) and in missing data points; imbalance in numbers operated); ‐1 level for imprecision (small trial)
Comments	Applicability reduced by incomplete description of study population; insufficient length of follow‐up (Table 4) Potential confounding from imbalances in numbers who had surgery and numbers who withdrew from the trial casts some doubts on the validity of the trial findings of a potential and interim benefit of the intervention on grip strength and range of motion at 12 weeks. These findings were not confirmed by the PWRE scores and as acknowledged by the trial authors: "More investigations are warranted before changes to clinical practice can be recommended"

Single session of physiotherapy
PICO	Population: 1 trial (Kay 2008), 47 followed up of 56 participants with distal radius fracture, after plaster cast immobilisation (15 had pins and cast) Intervention: single session of physiotherapy, primarily advice and instructions for a home exercise programme (Table 3) Control: no intervention Outcomes: PRWE (pain and function), QuickDASH, grip strength, range of motion, thumb motion, web space, complications, participant satisfaction with intervention, compliance (just physiotherapy group), request for physiotherapy; follow‐up 6 weeks
Findings	Intervention group had better function (clinically important between‐group differences favouring the intervention in PRWE scores for pain and function, and QuickDASH scores for general, sports and work functioning). There were no significant between‐group differences at six weeks in grip strength, range of motion or complications. Slightly fewer participants requested more physiotherapy in the physiotherapy group
GRADE	Very low quality evidence. Downgraded: ‐2 levels for serious study limitations (performance and detection biases; attrition bias ‐ differences in lost to follow‐up); ‐1 level for imprecision (small trial)
Comments	Applicability reduced by inadequate length of follow‐up (Table 4) The assessment at baseline and at 3 and 6 weeks follow‐up by an independent physiotherapist, who noted complications, could be considered as part the care package provided to both groups and thus 'no intervention' is not as stark as it seems. Prior care before and during cast immobilisation may also have varied and impacted on the results. These are promising yet preliminary results given the small, short‐term follow‐up, and single centre setting
Physiotherapy or occupational therapy
PICO	Population: 4 trials, (Bache 2001), 75 followed up of 98 participants; (Christensen 2001), 30 followed up of 32 participants; (Maciel 2005), 33 followed up of 45 participants; (Wakefield 2000), 66 followed up of 96 participants. All with distal radius fracture, after plaster cast immobilisation (7 participants had K‐wires and cast in Maciel 2005)  Intervention: Physiotherapy or occupational therapy ‐ Bache 2001: physiotherapy (contents/schedule at discretion of therapists); Christensen 2001: occupational therapy (2 x weekly until no more progress); Maciel 2005: 'activity focussed physiotherapy' up to 6 weeks; Wakefield 2000: physiotherapy (contents/schedule at discretion of therapists). All participants received instructions for home exercises Control: instructions for home exercises from therapist; (single session of advice on exercises from physiotherapist in Maciel 2005) Outcomes: PWRE, Levine functional analysis score, activities of daily living, modified Gartland and Werley composite outcome score, grip strength, range of motion, pain, SF‐36 (quality of life), complications, referral to occupational therapy, number of sessions, duration of therapy; follow‐up 12 weeks (Bache 2001); 9 months (Christensen 2001); 24 weeks (Maciel 2005); 6 months (Wakefield 2000)
Findings	No pooling undertaken due to lack of common outcomes and heterogeneity None of the trials found a clinically significant effect of the routine provision of either occupational therapy (Christensen 2001), physiotherapy (Bache 2001; Wakefield 2000) or 'activity‐focussed' physiotherapy (Maciel 2005). Participants attended on average 37 therapy sessions in Christensen 2001, whereas the median number of sessions or contacts was 3 in both Bache 2001 and Wakefield 2000, and averaged 4.4 sessions in Maciel 2005. Of the participants in the control groups, none received additional physiotherapy or occupational therapy in Bache 2001 and Christensen 2001, and two received remedial physiotherapy in Wakefield 2000
GRADE	Low quality evidence (overall). Reasons for downgrading according to source of data: ‐1 level for study limitations (mainly, performance bias; with attrition bias for Wakefield 2000), ‐1 level for imprecision (small trials, no pooling)
Comments	Applicability reduced by insufficient description of intervention and insufficient follow‐up in Bache 2001 (Table 4) Applicability considerably reduced by incomplete description of population, intervention and outcomes, no PROMs and insufficient follow‐up in Christensen 2001 (Table 4) Applicability much reduced by insufficient follow‐up in Maciel 2005 (Table 4) Applicability much reduced by incomplete description of intervention, no PROMs and insufficient follow‐up in Wakefield 2000 (Table 4) Although four trials addressed essentially the same issue, namely routine or formal provision of therapy in addition to instructions for home exercises after plaster cast removal, the form of the therapy varied. In particular, the therapy tested by Maciel 2005 was specifically focussed on restoring optimal motor performance of activities that were limited in the individual participants. Three trials (Bache 2001; Christensen 2001; Wakefield 2000) focussed on older people and all four trials excluded those with serious complaints already manifest at cast removal, such as pain (Christensen 2001) or CRPS‐1 (Bache 2001; Maciel 2005; Wakefield 2000). Three trials (Bache 2001; Maciel 2005; Wakefield 2000 explicitly selected patients who were able to understand instructions Individually, none of these trials provide sufficiently robust evidence to confirm the finding of a lack of clinically important differences. All are prey to a type 2 error (false conclusion of no difference). Baseline differences hampered the analysis of the results of Bache 2001, but there was some evidence that the physiotherapy group tended to improve more from a less favourable starting position. The loss to follow up of eight (19.5%) of the 41 participants who started the trial interventions in Maciel 2005 could also have given rise to important bias. Ultimately, no pooling of data was possible and at best the general agreement in these four studies can only be viewed as weak evidence
Continuous passive motion
PICO	Population: 1 trial (Rozencwaig 1996), 7 participants with distal radius fracture, after external fixation Intervention: occupational therapy plus continuous passive motion ‐ at home (4 to 6 hours/day for 1 month) (Table 3) Control: occupational therapy Outcomes: time to achieve independent status; follow‐up to initial recovery (a few weeks)
Findings	The 3 participants in the intervention group took less time (mean 12 days) to achieve a completely independent status than the 4 control group participants
GRADE	Very low quality evidence. Downgraded: ‐2 levels for serious study limitations (quasi‐randomised and non‐blinded trial ‐ high risk of selection and detection biases); ‐1 level for (serious) imprecision (very small trial)
Comments	Applicability substantially reduced by incomplete description of study population, intervention, outcome measures, no PROMs and insufficient length of follow‐up (Table 4). This is a totally inadequate trial from which no conclusions relating to practice should be drawn
Pulsed electromagnetic field (PEMF)
PICO	Population: 1 trial (Cheing 2005), 83 participants with distal radius fracture, after plaster cast immobilisation  Intervention: pulsed electromagnetic field (PEMF) (with or without ice) ‐ 5 sessions in clinic (Table 3) Control: sham PEMF (with or without ice) Outcomes: pain, range of motion, adverse events, oedema; follow‐up 5 days
Findings	None of the differences in pain, oedema (volume) and range of motion between the combined PEMF groups and combined sham PEMF groups were clinically important or statistically significant. There were no adverse effects recorded
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (mainly relating to unclear risk of selection, performance and detection biases), ‐1 level for imprecision (small trial); ‐1 level for indirectness (very short follow‐up, measured before last session of intervention)
Comments	Applicability considerably reduced by lack of PROMs and very inadequate follow‐up (Table 4) The clinical relevance of these results, if true, is uncertain given the premature and very short‐term follow‐up. Additionally, confounding could have resulted from baseline imbalances in some outcome measures. Potentially inconvenient and costly: participants needed to attend each weekday for 1 week; 30 minutes each session
Ice
PICO	Population: 1 trial (Cheing 2005), 83 participants with distal radius fracture, after plaster cast immobilisation  Intervention: ice (with real or sham PEMF) ‐ 5 sessions in clinic (Table 3) Control: no ice (with real or sham PEMF) Outcomes: pain, range of motion, adverse events, oedema; follow‐up 5 days
Findings	Statistically significant but clinically unimportant findings of less pain and worse extension results in the ice treatment group(s). None of the differences in oedema (volume) and other range of motion variables between the combined ice groups and combined control groups were clinically important or statistically significant. There were no adverse effects recorded
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (mainly relating to unclear risk of selection, performance and detection biases), ‐1 level for imprecision (small trial); ‐1 level for indirectness (very short follow‐up, measured before last session of intervention)
Comments	Applicability considerably reduced by lack of PROMs and very inadequate follow‐up (Table 4) The clinical relevance of these results, if true, is uncertain given the premature and very short‐term follow‐up. Additionally, confounding could have resulted from baseline imbalances in some outcome measures. Potentially inconvenient: participants needed to attend each weekday for 1 week
Pulsed electromagnetic field (PEMF) plus ice
PICO	Population: 1 trial (Cheing 2005), 39 participants with distal radius fracture, after plaster cast immobilisation  Intervention: PEMF plus ice ‐ 5 sessions in clinic (Table 3) Control: sham PEMF and no ice Outcomes: pain, range of motion, adverse events, oedema; follow‐up 5 days
Findings	Statistically significant but clinically unimportant findings of worse extension but better ulnar deviation results in the intervention group. None of the between‐group differences in pain, oedema (volume) and other range of motion variables were clinically important or statistically significant. There were no adverse effects recorded
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (mainly relating to unclear risk of selection, performance and detection biases), ‐1 level for imprecision (small trial); ‐1 level for indirectness (very short follow‐up, measured before last session of intervention)
Comments	Applicability considerably reduced by lack of PROMs and very inadequate follow‐up (Table 4) The clinical relevance of these results, if true, is uncertain given the premature and very short‐term follow‐up. Additionally, confounding could have resulted from baseline imbalances in some outcome measures. Potentially inconvenient and costly: participants needed to attend each weekday for 1 week; 30 minutes each session
Passive mobilisation
PICO	Population: 2 trials (Kay 2000), 39 followed up of 40 participants; and (Taylor 1994), 30 participants. All with distal radius fracture, post‐plaster cast immobilisation  Intervention: passive mobilisation ‐ Kay 2000: 9 sessions over 6 weeks; Taylor 1994: 5 minutes during twice weekly physiotherapy (Table 3) Control:Kay 2000: all attended physiotherapy session for advice & instructions; and assessment at 3 and 6 weeks, more at physiotherapist's discretion; Taylor 1994: sham (soft tissue massage) and physiotherapy Outcomes: pain & functional disability, grip strength, range of motion, web space angle, complications, number of physiotherapy sessions, time to discharge, costs; follow‐up 6 weeks (Kay 2000); until discharge (mean 26 days) (Taylor 1994)
Findings	No pooling undertaken due to lack of common outcomes and heterogeneity. Both trials found no significant between‐group differences in short‐term outcome; this was primarily active wrist extension in Taylor 1994. There were no between‐group differences in complications. Participants allocated passive mobilisation received on average six more treatments than those in the control group in Kay 2000, which estimated additional passive mobilisation to be nearly three times as expensive as a regimen of advice and exercises alone. There was little difference (< 1 session) in numbers of sessions between the two groups in Taylor 1994
GRADE	Very low quality evidence. Reasons for downgrading according to source of data: ‐1 level for study limitations (mainly, performance bias; or selection and performance biases), ‐2 levels for serious imprecision (small trials, no pooling)
Comments	Applicability much reduced by lack of PROMs and inadequate follow‐up (Table 4) Neither trial was sufficient to take their lack of significant differences as evidence of no effect. Passive mobilisation delivered as in Kay 2000 was much more costly. The soft tissue massage control of Taylor 1994 cannot be considered a 'placebo'; it may still have a therapeutic role, for instance in redistributing tissue fluid, and thus potentially diminish any effects of passive mobilisation
Intermittent pneumatic compression
PICO	Population: 1 trial (Svensson 1993), 31 followed up of 43 participants with distal radius fracture, referred for rehabilitation after plaster cast immobilisation  Intervention: intermittent pneumatic compression (20 minutes) before occupational therapy ‐ minimum 3 x weekly for 3 weeks (Table 3) Control: occupational therapy only Outcomes: use of hand (VAS), pain, range of motion, complications, oedema, number of sessions
Findings	There were no data to examine the report of improved wrist extension and tendencies for improvements in other outcomes favouring the intervention group. Only a few participants in each group (numbers not stated) were considered to require further occupational therapy after three weeks
GRADE	Very low quality evidence. Downgraded: ‐2 levels for serious study limitations (performance and detection biases, attrition bias ‐ missing results for 12 participants), ‐1 level for imprecision (small trial)
Comments	Applicability reduced by no PROMs and insufficient follow‐up (Table 4) The evidence from this small, inadequately reported and flawed trial is insufficient to inform practice. Potentially inconvenient and costly: participants needed to stay 20 minutes extra each of 9 sessions
Ultrasound
PICO	Population: 1 trial (Basso 1998), 38 participants with distal radius fracture, after plaster cast immobilisation  Intervention: ultrasound ‐ 5 minutes (Table 3) Control: sham ultrasound Outcomes: range of motion, referral for physiotherapy
Findings	There was no between‐group difference in wrist motion but fewer ultrasound participants were referred for physiotherapy at eight weeks (2/19 versus 8/19; P = 0.05).
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (quasi‐randomised, thus high risk of selection bias), ‐1 level for imprecision (small trial); ‐1 level for indirectness (relating to outcome)
Comments	Applicability considerably reduced by incomplete details on intervention, limited outcomes, no PROMs and insufficient follow‐up (Table 4) Baseline and care programme comparability were not confirmed in this small quasi‐randomised trial; the very limited evidence is insufficient to inform practice. The underlying mechanism for the intervention is not established
Whirlpool
PICO	Population: 1 trial ( Toomey 1986), 24 participants with distal radius fracture, referred to Physical Medicine Department after plaster cast immobilisation  Intervention: whirlpool bath immersion for 15 minutes prior to exercises ‐ 12 sessions, twice weekly (Table 3) Control: towel only Outcomes: pain, grip strength, range of motion, oedema; follow‐up 6 weeks
Findings	The intervention resulted in interim oedema without clinically significant differences in other outcomes by the end of treatment
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (performance bias, unclear risk of several other biases), ‐2 levels for serious imprecision (very small trial)
Comments	Applicability reduced by no PROMs and insufficient follow‐up (Table 4) Whirlpool bath immersion prior to exercises was reported as common in Canadian physiotherapy departments in the early 1980s. The inadequate sample size, unresolved questions on participant numbers, and a variable and short‐term follow up amount to potentially flawed and insufficient evidence for this modality. Potentially inconvenient and costly: participants who need to stay 15 minutes extra each of 12 sessions
Dynamic wrist extension splint
PICO	Population: 1 trial (Jongs 2012), 40 participants with wrist contracture who had been referred to physiotherapy at least 10 weeks from their conservatively or surgically treated fracture  Intervention: routine care plus dynamic wrist extension splints, worn up to six hours at home on a daily basis for eight weeks (Table 3) Control: routine care (exercises and advice) Outcomes: PRWE, range of motion, Canadian Occupational Performance Measure (COPM), adverse events, splint repair, adherence; follow‐up 12 weeks
Findings	There were no clinically or statistically between‐group differences in PRWE, range of motion and COPM at 8 or 12 weeks. There were two adverse events, which resolved after modification of the splint, and two other splints required repair. Thus problems were encountered in a quarter of the group allocated splints. Compliance with the splint regimen was poor
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (performance and detection biases), ‐1 level for imprecision (small trial), ‐1 level for indirectness (population is subgroup of distal radius patients)
Comments	Applicability reduced by insufficient follow‐up (Table 4) There was poor adherence to the intended six hour daily application for the splint, with complaints by some participants of the limitations that the splints imposed on day‐to‐day activities. Thus this is an inconvenient and demanding intervention. The trial authors' conclusion that dynamic splints are unlikely to be therapeutic are consistent with the findings of a lack of effectiveness for 'stretch', including that administered via splints, in Katalinic 2010

Physiotherapy versus instructions for home exercises by an orthopaedic surgeon
PICO	Population: 1 trial (Watt 2000), 16 followed up of 18 participants with distal radius fracture, after plaster cast immobilisation Intervention: routine physiotherapy (Table 3) Control: home exercise sheet and simple home instructions given at outpatients by orthopaedic surgeon/registrar Outcomes: grip strength and range of motion (wrist extension), number of physiotherapy sessions; follow‐up 6 weeks
Findings	Trial found significantly better grip strength and wrist extension at six weeks in participants given physiotherapy. Physiotherapy group participants attended an average of five sessions
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (performance and detection biases); ‐2 levels for serious imprecision (very small trial, little data)
Comments	Applicability much reduced by incomplete descriptions of study population and physiotherapy; limited outcomes and no PROMs, and inadequate length of follow‐up (Table 4) These promising yet preliminary results need confirmation with larger numbers, longer‐term follow up, and a more comprehensive appraisal of outcome, and replication in different settings
Pulsed electromagnetic field (PEMF) versus ice
PICO	Population: 1 trial (Cheing 2005), 44 participants with distal radius fracture, after plaster cast immobilisation  Intervention: PEMF plus ice ‐ 5 sessions in clinic (Table 3) Control: ice Outcomes: pain, range of motion, adverse events, oedema; follow‐up 5 days
Findings	Statistically significant but clinically marginal findings of worse pain but better extension results in the PEMF group. However, these are consistent with similar differences in baseline values between the two group. None of the between‐group differences were clinically important or statistically significant. There were no adverse effects recorded
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (mainly relating to unclear risk of selection, performance and detection biases), ‐1 level for imprecision (small trial); ‐1 level for indirectness (very short follow‐up, measured before last session of intervention)
Comments	Applicability considerably reduced by lack of PROMs and very inadequate follow‐up (Table 4) Ideally, the effects of the two interventions should be established before a comparison of their relative effects. This is not the case so far. Serious deficiencies in the measurement of outcome and baseline imbalances in the trial testing this comparison mean that no conclusions can be drawn on the relative effectiveness of PEMF versus ice in treating pain, swelling and stiffness after plaster cast removal
Modified "manual edema mobilization" (MEM) versus 'traditional' oedema treatment
PICO	Population: 1 trial (Knygsand‐Roenhoej 2011), 29 followed up of 30 participants with distal radius fracture who had subacute oedema 4 to 10 weeks post‐trauma/surgery (plaster cast (11); surgery (18: external or internal fixation)) and a 60 mL or more in volume difference between the upper extremities Intervention: Modified "manual edema mobilization" (MEM) (Table 3) Control: 'Traditional oedema treatment': elevation, compression, and functional training Outcomes: Activities of daily living, Canadian Occupational Performance Measure, complication, oedema, number of occupational therapy sessions; follow‐up 6 months
Findings	The trial found no clinically important difference in measures of function up to nine weeks or pain at 9 or 26 weeks between these two methods of treating subacute oedema after treatment for distal radius fracture. Two complications were reported in the MEM group
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (high risk of performance and detection biases); ‐1 level for imprecision (small trial); ‐1 level for indirectness (population is subgroup of distal radius patients; also could depend on clinicians providing treatment)
Comments	Applicability reduced by insufficient length of follow‐up (Table 4). Additionally, the results apply to the patient subgroup with subacute oedema  Although the traditional therapy group tended to receive more treatment sessions, the reasons for this may be other than the treatment itself.These include differences in the two clinicians providing treatment, a slightly higher level of oedema in the traditional group at baseline and other differences in the interventions. Oedema is often associated with hand and wrist trauma and so any future research to test the use of MEM could be expanded to include other injuries

Exercise therapy supervised by a physiotherapist versus instructions for the same exercises given by an orthopaedic surgeon
PICO	Population: 1 trial (Pasila 1974), 96 followed up of 135 participants with distal radius fracture, after plaster cast immobilisation Intervention: exercise therapy supervised by a physiotherapist (Table 3) Control: instructions for the same exercises given by an orthopaedic surgeon Outcomes: grip strength and range of motion, time to return to work, visits to physiotherapy department; follow‐up 12 weeks
Findings	Trial found no significant between group differences in strength or range of motion, time to return to work. The physiotherapy group participants visited the physical medicine department an average of four times
GRADE	Very low quality evidence. Downgraded: ‐2 levels for serious study limitations (performance and detection biases, attrition bias (29% losses), selective outcome reporting bias); ‐1 level for imprecision (small trial)
Comments	Applicability much reduced by incomplete descriptions of study population and physiotherapy; limited outcomes and no PROMs, and inadequate length of follow‐up (Table 4). The serious methodological flaws, including a large loss to follow up, and inadequate sample size of Pasila 1974 mean that the lack of significant differences in various measures of recovery between the two participant groups cannot be considered as reliable evidence. Noteworthy is that this was a comparatively young population, over two‐thirds of whom were under 40 years old, and thus not generally representative

Physiotherapy versus a home exercise programme
PICO	Population: 1 trial (Bighea 2013), 20 participants with osteoporotic wrist fracture, after plaster cast immobilisation Intervention: physiotherapy galvanic bath, 30 minutes exercise each session programme; 5 x week for 4 weeks (Table 3) Control: home exercise programme, instructions for 15 minutes 2 x daily sheet Outcomes: PRWE, range of motion; follow‐up 8 weeks
Findings	The trial reported better PRWE scores and greater increases in extension and flexion in the physiotherapy group after four weeks
GRADE	Very low quality evidence. Downgraded: ‐1 level for study limitations (performance and detection biases, selective outcome reporting bias); ‐2 levels for serious imprecision (very small trial, little data)
Comments	Applicability much reduced by incomplete descriptions of study population, interventions and outcomes, and inadequate length of follow‐up (Table 4) These preliminary results need confirmation with larger numbers, longer‐term follow up, and a more comprehensive appraisal of outcome, and replication in different settings
Physiotherapy or occupational therapy versus progressive home exercise programme after volar plate fixation
PICO	Population: 2 trials (Krischak 2009), 46 followed up of 48 participants; and (Souer 2011), 76 followed up of 94 participants. All with distal radius fracture, after volar plate fixation Intervention: physiotherapy or occupational therapy: 12 sessions 20‐30 minutes (Krischak 2009); at discretion of therapist (Souer 2011) (Table 3) Control: progressive home exercise programme Outcomes: PRWE, DASH, Mayo wrist scores, pain, grip and pinch strength and range of motion, complications, change of treatment, compliance, costs; follow‐up 6 weeks (Krischak 2009), 6 months (Souer 2011)
Findings	Both trials in this category found some evidence in favour of a structured programme of home exercises preceded by instructions or coaching compared with supervised therapy after surgery. At the end of the six‐weeks treatment programme, Krischak 2009 found physiotherapy resulted in significantly poorer self reported function as well as objective measures of function. Although finding no difference between the two groups in self reported function at three or six months follow‐up, Souer 2011 reported poorer grip strength at three months and poorer wrist mobility at six months in the occupational therapy group compared with independent exercise group
GRADE	Low quality evidence. Downgraded: ‐1 level for study limitations (performance and detection biases); ‐1 level for imprecision (small trials, no pooling)
Comments	Applicability reduced by incomplete descriptions of physiotherapy/occupational therapy; and insufficient length of follow‐up (Table 4) A marked difference in the duration of exercises was explicit in Krischak 2009, supervised therapy amounted to 40 to 60 minutes per week in contrast to 280 minutes of home exercises. It is unclear but seems unlikely from the accounts of both trials that instructions and encouragement for home exercises were provided in the supervised therapy group. If so, this reduces the applicability given that instructions for home exercises are typically part of usual care. It should be noted that even if independent exercises using these rigorous schemes is enough for most patients after volar plate fixation, this low quality evidence does not rule out a role for supervised therapy. The latter though is potentially inconvenient and more costly: in Krischak 2009, participants needed to attend 12 sessions in 6 weeks
Accelerated versus usual rehabilitation after volar plate fixation
PICO	Population: 1 trial (Brehmer 2014), 63 of 81 participants with distal radius fracture after volar plate fixation Intervention: accelerated (started at two weeks) rehabilitation (Table 3) Control: usual rehabilitation (started at six weeks) Outcomes: DASH scores, grip strength, palmar pinch strength, range of motion, adverse events and splint repair, fracture alignment and healing; follow‐up 6 months
Findings	Although DASH scores favoured accelerated rehabilitation at 8 and 12 weeks and six months, the effect size dropped over time and the confidence intervals did not include the MCID in the two later follow‐up times. Grip strength at three and six months was also greater in the accelerated group compared with the usual group, with a significant difference showing at six months. The results for range of motion results favoured the home exercises group but were not clinically important. There were few complications in both groups, and no loss of alignment
GRADE	Very low quality evidence. Downgraded: ‐2 levels for study limitations (high risk of performance and detection biases; and attrition bias); ‐1 level for imprecision (small trial)
Comments	Applicability reduced by incomplete description of population and insufficient length of follow‐up (Table 4).  The early clinical benefit shown by better DASH results of accelerated rehabilitation did not persist at three and six months. There were few complications and none, such as loss of alignment, that conflicted with either intervention. Although the focus was on rehabilitation for this review, participants in the usual group were scheduled to stop using their removable splints three weeks after those in the accelerated group. It is also not clear if any advice was given to the participants in the usual group to engage in at least range of motion exercises before the start of the strengthening protocol.

Date	Event	Description
23 September 2015	New search has been performed	We updated the search to January 2015. We included 11 new trials. These tested 10 new comparisons. Risk of bias assessment replaced the previous assessment of methodological quality. We assessed the quality of evidence using GRADE and compiled the evidence in summary tables.
23 September 2015	New citation required and conclusions have changed	Although the conclusions reflecting the fundamental insufficiency of the evidence are unchanged, we have made some secondary revisions to the implications for practice, in terms of basic standards of care, and research. There has been a change to the byline.

Date	Event	Description
25 July 2008	Amended	Converted to new review format.
9 May 2006	New citation required and conclusions have changed	In this substantive update (Issue 3, 2006) the search for trials was updated to December 2005. We identified six new studies and one full report of a trial formerly listed as an ongoing study. Of these, we included three studies, excluded three studies and placed one study into 'Studies awaiting assessment'. We further excluded one study previously awaiting assessment. There were several format changes made to comply with the Cochrane Style Guide (October 2005). Graphical presentations of the results were modified with totals removed in all cases. There were no substantive changes made to the conclusions.    For details of previous updates, please see 'Notes'.
25 July 2004	Amended	In the third, a minor update published in Issue 4, 2004, all changes resulted from copy‐editing. There were no changes made to the conclusions.
23 February 2004	New search has been performed	In the second, a minor update published in Issue 3, 2004, the search for trials was extended to January 2004. We identified no new studies nor publications of studies listed as ongoing or pending. There were no changes made to the conclusions.
22 March 2003	New search has been performed	In the first, a minor update published in Issue 2, 2003, the search for trials was extended to January 2003. We identified five new studies, three of which were ongoing, one of which was excluded and one of which was placed in studies awaiting assessment. There were no changes made to the conclusions.

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Meeting criteria for attendance of post‐immobilisation hand therapy group	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
2 Grip strength (kg) at 4 weeks (post‐immobilisation)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3 Range of motion at 4 weeks (post‐immobilisation)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.5 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.6 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Oedema (mL) at 4 weeks (post‐immobilisation)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5 Any pain at rest at 4 weeks (post‐immobilisation)	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6 Finger mobility at 4 weeks (post‐immobilisation)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
6.1 Kapandji score (thumb opposition, 1 to 10 locations of increasing difficulty)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.2 Index finger TAM (total active motion) (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.3 Middle finger TAM (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.4 Ring finger TAM (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.5 Little finger TAM (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7 Complications	2	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
7.1 CRPS‐1	2	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.2 Median nerve compression	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.3 Ulnar nerve compression	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.4 Tendon rupture	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Grip strength (kg)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 10 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Pinch strength (kg)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 10 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Range of motion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Flexion/extension at 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Flexion/extension at 10 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.3 Supination/pronation at 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.4 Supination/pronation at 10 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Manual Ability Measure‐36 ‐ Taiwan version 45 questions (36 to 180: best result)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 At 3 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 At 7 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.3 At 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Grip, pinch and 'three jaw chuck' pinch strengths (% of uninvolved hand)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Grip power at 7 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Grip power at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Pinch strength at 7 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 Pinch strength at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 Three jaw chuck pinch strength at 7 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 Three jaw chuck pinch strength at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Range of motion (% of other hand)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 'Finger workspace' at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Thumb 'workspace' at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 PRWE scores at 2 to 3 days after cast removal	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 Pain score (0 to 50: worst pain)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 Activity/function score (0 to 50: worst function)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Range of motion at day 2 to 3 after cast removal	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Hand oedema: difference between hands in circumference (mm)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4 Complications	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.1 Total complications	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 CRPS 1 (symptoms)	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.3 Median nerve compression	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.4 Finger stiffness	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.5 Adverse effect of PEMF	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Patient‐Rated Wrist Evaluation (PRWE) (0 to 150: worst results)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 At 9 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 At 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.3 At 26 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Grip strength of fractured hand (kg)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 At 9 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 At 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 At 26 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Range of motion: supination/pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 At 9 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 At 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.3 At 26 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Range of motion: flexion/extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 At 9 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 At 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.3 At 26 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 PRWE scores at 3 and 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 PRWE (pain) at 3 weeks (0 to 100: worst pain)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 PRWE (function) at 3 weeks (0 to 100: worst function)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.3 PRWE (pain) at 6 weeks (0 to 100: worst pain)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.4 PRWE (function) at 6 weeks (0 to 100: worst function)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 QuickDASH scores at 3 and 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 QuickDASH (general) at 3 weeks (0 to 100: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 QuickDASH (work) at 3 weeks (0 to 100: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 QuickDASH (sports) at 3 weeks (0 to 100: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 QuickDASH (general) at 6 weeks (0 to 100: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 QuickDASH (work) at 6 weeks (0 to 100: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 QuickDASH (sports) at 6 weeks (0 to 100: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Grip strength (kg) at 3 and 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 At 3 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 At 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Range of motion at 6 weeks (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 Extension	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 Flexion	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.3 Radial deviation	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.4 Ulnar deviation	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.5 Pronation	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.6 Supination	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Change in range of motion at 6 weeks (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5.1 Extension	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 Flexion	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.3 Radial deviation	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.4 Ulnar deviation	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.5 Pronation	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.6 Supination	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Complications	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
7 Request for more physiotherapy	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Pain and volume at day 5	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 Pain (scale 0: no pain to 10 cm: worst imaginable) during active wrist movements	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 Volume (ml)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Range of motion at day 5	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Grip strength (kg) at 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2 Range of motion at 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Web space angle (degrees) at 6 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4 Wrist extension at discharge (4 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5 Number of treatments	2	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
6 Time to discharge (days)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
7 Complications at 6 weeks	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
7.1 Carpal tunnel syndrome	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.2 Finger stiffness (continuing)	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.3 Complex regional pain syndrome (continuing)	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.4 Malunion	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Greater than 30% loss of wrist motion (flexion‐extension) at 8 weeks	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
2 Referral for physiotherapy	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Grip strength at end of treatment (kg)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2 Pain (scale 0: no pain to 5: excruciating) at end of treatment	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3 Range of motion at end of treatment	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.5 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.6 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Finger flexion at end of treatment	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 Thumb MCP (metacarpal phalange) flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 Index finger MCP flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.3 Long finger MCP flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.4 Ring finger MCL flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.5 Little finger MCP flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Oedema (ml)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5.1 Short term: at end of each session	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 At end of treatment	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Canadian Occupational Performance Measure at 9 weeks (clinically important improvement)	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
1.1 Ability to perform the key activity (change 2+ in score)	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 Satisfaction with ability to perform key activity (change 2+ in score)	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Pain (VAS: 0 to 100: worst pain)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Pain at rest (9 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Pain at rest (26 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Pain when active (9 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 Pain when active (26 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Complications	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4 Number of occupational therapy sessions	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5 Receiving oedema treatment after scheduled period	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
5.1 Receiving oedema treatment after 6 weeks	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 Receiving oedema treatment after 9 weeks	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Oedema: volume difference between injured and non‐injured side (mL)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
6.1 Oedema (9 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.2 Oedema (26 weeks)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Strength and power at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 "Grip strength" (kg/cm2)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 Hand pumping power (mmHg)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Range of motion at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 PRWE scores at 6 weeks (0: normal to 150: worst outcome)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2 DASH score (0 to 100: higher scores = worse upper‐extremity function)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 DASH scores at 3 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 DASH scores at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Mayo wrist score (0 to 100: higher scores = best functional outcome)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Mayo scores at 3 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Mayo scores at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Pain at rest (VAS: 0 to 10: worst pain)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 Pain at 3 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 Pain at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Grip strength (kg)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5.1 Grip strength at 3 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 Grip strength at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Pinch strength (kg)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
6.1 Pinch strength at 3 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.2 Pinch strength at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7 Range of motion at 3 months	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
7.1 Extension‐flexion arc (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.2 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.3 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.4 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.5 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.6 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.7 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8 Range of motion at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
8.1 Extension‐flexion arc (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.2 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.3 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.4 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.5 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.6 Radial deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.7 Ulnar deviation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
9 Complications	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
9.1 Carpal tunnel release 2‐3 months post‐initial treatment	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
9.2 Loss of alignment of lunar facet fragment	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
9.3 Extensor pollicis longus tendon rupture	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
9.4 Implant removal for tendon irritation	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 DASH score (0 to 100: higher scores = worse upper‐extremity function)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1 At 8 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 At 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.3 At 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Grip strength (lb)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 Grip strength at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 Grip strength at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Pinch strength (lb)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 Pinch strength at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 Pinch strength at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Range of motion at 12 weeks	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Range of motion at 6 months	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5.1 Pronation (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 Supination (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.3 Flexion (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.4 Extension (degrees)	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Complications	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.1 Carpal tunnel syndrome	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.2 Loss of alignment or non‐union	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.3 Extensor pollicis longus tendon rupture	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.4 Re‐operation to remove screw	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

PERMALINK

Rehabilitation for distal radial fractures in adults

Helen HG Handoll

Joanne Elliott

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Description of the condition

Description of the intervention

How the intervention might work

1. Descriptions and putative mechanisms of some rehabilitation interventions.

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

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

Results

Description of studies

Results of the search

1.

Included studies

Design

Sample sizes

Setting

Participants

2. Summary of study populations.

Interventions

3. Summary of rehabilitation interventions.

Comparisons

(1) Rehabilitation intervention versus no intervention

(1a) Rehabilitation started during the definitive treatment period

(1b) Rehabilitation started post‐immobilisation

(2) One rehabilitation intervention versus another rehabilitation intervention

(2a) Rehabilitation started during the definitive treatment period

(2b) Rehabilitation started post‐immobilisation

(3) Any method (context) of delivering or providing rehabilitation interventions versus any other method of delivering or providing rehabilitation interventions

(3a) Rehabilitation started during the definitive treatment period

(3b) Rehabilitation started post‐immobilisation

Outcome measurement

4. Assessment of items relating to applicability of trial findings.

Excluded studies

Ongoing studies

Studies awaiting classification

Risk of bias in included studies

2.

3.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

(1) Rehabilitation intervention versus no intervention

Methods	Method of randomisation: sealed envelopes contained in a box  Assessor blinding: yes for objective measures; some participants revealed their treatment despite requests not to do so beforehand  Intention‐to‐treat analysis: likely, but post‐randomisation exclusions: 4 developed CRPS‐1 and 1 developed carpal tunnel syndrome  Loss to follow‐up: 18 (+ above 5 exclusions) (at 12 weeks)
Participants	Selly Oak Hospital, Birmingham, UK  98 participants  Inclusion criteria: distal radius fracture, treated by plaster cast immobilisation, living at home, age over 50 years, participants able to follow an exercise programme independently, informed consent.  Exclusion criteria: medical history of dementia, Alzheimer's or psychiatric or confused state, multiple limb fractures or bilateral fracture, requiring physiotherapy for other reasons, pre‐existing inflammatory joint disorder. Past medical history of wrist problems or operations on affected side. Early manifestation of CRPS‐1 or carpal tunnel syndrome.  Classification: AO and Frykman  Sex: 82 female (84%)  Age: median 69 years; range 50 to 92 years  Assigned: 43/55 [physiotherapy/control]  Assessed: 36/45 (at 4 weeks); 35/40 (at 12 weeks)
Interventions	Timing of intervention: following plaster cast removal (5 to 6 weeks immobilisation). All participants given explanation of home care based on standardised advice and exercise sheet by physiotherapist  1. Referral for routine physiotherapy at outpatients clinic. Contents of treatment at discretion of physiotherapists; these involved different combinations of physiological mobilisation, progressive active exercise, passive stretching, accessory movements of wrist and radioulnar joints. Discharge criteria: functional ROM, full function, plateau of improvement  2. Home exercises alone
Outcomes	Length of follow‐up: (median) 12 weeks; also (median) 4 weeks.  1. Functional: grip strength, ROM (pronation; supination; flexion; extension; radial deviation; ulnar deviation); functional analysis scale (Levine 1993), pain (VAS). Referral to occupational therapy 2. Number of contacts with physiotherapist; duration of physiotherapy, reasons for discharge  3. Complications: carpal tunnel syndrome & CRPS‐1 (mainly excluded from follow‐up)
Funding, ethics and patient consent	No mention of funding or ethics  Informed patient consent
Notes	Draft trial report received from Mrs Sarah Bache, now based in Australia, on 30 August 2001, and further details on 5 September. Further discussion on outcome measures on 12 September with feedback from trial statistician Louise Hiller
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Randomised" but statistician involved
Allocation concealment (selection bias)	Unclear risk	Draft report: "Consented patients were randomised individually by staff with sealed envelopes contained in a box". However, the trial investigator confirmed that there was "a possible chance of disclosure of assignment"
Blinding of participants and personnel (performance bias)   Subjective outcomes	High risk	Neither participants nor personnel providing care were blinded
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Neither participants nor personnel providing care were blinded
Blinding of outcome assessment (detection bias)   Subjective outcomes	Unclear risk	Abstract: "Blind assessment were performed at baseline (within one week of cast removal), and at four and 12 weeks" Draft report: "some patients discussed their treatment at review even though it was stated in the patient information sheet that this should not be done"
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	Unclear risk	As above
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Low risk	Outcomes unlikely to be affected by lack of blinding
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	There were 5 post‐randomisation exclusions and imbalance in the numbers lost to follow‐up (16% versus 27% at 12 weeks)
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No longer‐term follow‐up
Selective reporting (reporting bias)	Unclear risk	No protocol available. However, consistencies noted with draft report and trial registration document. Area under the curve analyses in response to detected baseline differences
Major baseline imbalance bias?	Unclear risk	Some trends towards patients randomised to physiotherapy and advice and exercise being more symptomatic at baseline. (Adjustments made in analyses)
Other performance bias (e.g. differential expertise bias)	Low risk	No problems detected None of the control group patients received physiotherapy or occupational therapy

Methods	Quasi‐randomised: by year of birth  Assessor blinding: yes  Intention‐to‐treat analysis: likely  Loss to follow‐up: none probably
Participants	Edgeware General Hospital, UK  38 participants  Inclusion criteria: Colles' fracture, manipulated and treated with plaster cast. Complete transverse extra‐articular break with minimal degree of dorsal displacement and comminution Exclusion criteria: (often by example of the 13 excluded participants) age < 15 years, intra‐articular involvement, palmar/no displacement, severe dorsal comminution, damage to ulnar styloid, severe disruption of DRUJ (> 25 degrees dorsal displacement or > 6 mm radial shortening) and triangular fibrocartilage, carpal injury, inadequate reduction, more than one manipulation, open fracture, multiple trauma, history of injury to the contralateral wrist, inability to cope with measuring technique, very poor hand function following POP removal Classification: none  Sex: 22 female (58%)  Age: median 57 years [ultrasound] and 63 years [control]; range 15 to 69 years  Assigned: 19/19 [ultrasound / control]  Assessed: 19/19 (at 8 weeks)
Interventions	Timing of intervention: following plaster cast removal (on average after 4 weeks immobilisation; range 3 to 8 weeks)  All participants were given instructions to use hands as much as possible. No physiotherapy "unless hand function was poor"  1. Ultrasound: 46.39 kHz at intensity 74 W/cm2 applied for 5 minutes to back of wrist. Joint actively mobilised during treatment  2. Sham ultrasound. Joint actively mobilised for 5 minutes but machine not active (generator still switched on)
Outcomes	Length of follow‐up: 8 weeks; also 2 weeks and to end of treatment for those prescribed physiotherapy after 8 weeks  1. Functional: ROM (extension‐flexion) loss  2. Referral for physiotherapy at 8 weeks, length of follow‐up
Funding, ethics and patient consent	No mention of funding or ethics. (The equipment was provided already calibrated, and reviewed routinely by the manufacturer") All participants gave informed consent
Notes	Request for further information sent 8 August 2001
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	"Patients were randomly allocated by year‐of‐birth.."
Allocation concealment (selection bias)	High risk	"Patients were randomly allocated by year‐of‐birth.."
Blinding of participants and personnel (performance bias)   Subjective outcomes	Unclear risk	No subjective outcomes
Blinding of participants and personnel (performance bias)   Objective outcomes	Low risk	"In the control group, the procedure was the same as in the treatment group but the generator, although switched on, was not active"
Blinding of outcome assessment (detection bias)   Subjective outcomes	Unclear risk	No subjective outcomes
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	Low risk	"When reviewing a patient at 8 weeks, the observer (OB) had no access to the records and was unaware of the treatment given until after assessment of hand function and pain"
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Low risk	Outcomes unlikely to be affected by lack of blinding
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	There appears to be no loss to follow‐up. However, it is stated that "It was considered ethically unacceptable to retain patients in the study who had very poor hand function following removal of the plaster"
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No longer‐term follow‐up
Selective reporting (reporting bias)	Unclear risk	No protocol available
Major baseline imbalance bias?	Unclear risk	Incomplete information including on duration of prior immobilisation
Other performance bias (e.g. differential expertise bias)	Unclear risk	Incomplete information on care programmes. Possible provision of physiotherapy for poor hand function

Methods	Randomised: no details of method  Assessor blinding: no report (unlikely) Intention‐to‐treat analysis: not known  Loss to follow‐up: none probably
Participants	University of Medicine and Pharmacy of Craiova, Craiova, Romania  20 participants  Inclusion criteria: osteoporotic wrist fracture  Exclusion criteria: No details  Classification: none  Sex: not reported Age: not reported  Assigned: 10/10 [physiotherapy / control]  Assessed: ?/? (at 8 weeks?; 4 weeks post‐immobilisation )
Interventions	Timing of intervention: probably following plaster cast removal (after 4 weeks immobilisation)  1. Physiotherapy: 20 minutes galvanic bath then 30 minutes exercise programme performed 5 times per week, once daily, involving active range of motion, hand and finger flexion and extension, and ball resistance. Four weeks of therapy before second examination reported  2. Home exercise programme. Detailed instructions given to follow a 15 minutes twice a day home exercise programme
Outcomes	Length of follow‐up: 8 weeks (timing of randomisation not stated); first examination just after 4 weeks immobilisation; second examination after 4 weeks of rehabilitative treatment  1. Functional: PRWE, ROM (extension, flexion)
Funding, ethics and patient consent	No mention of funding or ethics No mention of informed consent
Notes	Request for further information sent 4 May 2015
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"randomly assigned into two groups"
Allocation concealment (selection bias)	Unclear risk	"randomly assigned into two groups"
Blinding of participants and personnel (performance bias)   Subjective outcomes	High risk	Not blinded
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Not blinded
Blinding of outcome assessment (detection bias)   Subjective outcomes	High risk	Not blinded
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	High risk	Not blinded
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Unclear risk	Outcomes not reported
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	No report of participant flow
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	Long‐term outcomes not reported
Selective reporting (reporting bias)	High risk	No protocol. Conference abstract only. Minimal data
Major baseline imbalance bias?	Unclear risk	Balance claimed but no supporting data
Other performance bias (e.g. differential expertise bias)	Unclear risk	No information

Methods	Method of randomisation: use of computerised random number generation; administered by an independent research co‐ordinator  Assessor blinding: surgeons were but not assessors, including hand therapists Intention‐to‐treat analysis: no, 3 exclusions in the standard rehabilitation group  Loss to follow‐up: 15 (at 6 months)
Participants	TRIA Orthopaedic Center, University of Minnesota, Bloomington, Minnesota, USA  81 participants  Inclusion criteria: age 18 to 85 years; isolated distal radial fracture treated with open reduction and internal fixation using a plate (Hand Innovations DVR plate) and informed consent.  Exclusion criteria: previous distal radial fracture on the affected side; professional athlete; bilateral distal radial fracture; another concurrent fracture; and a distal radial fracture for which, in the surgeon’s judgment, fixation of fracture fragments could not be achieved to allow participation in the accelerated rehabilitation protocol.  Classification: AO fracture classification Sex (of 78): female (73%)  Age: mean 53 years (range 21 to 83 years) Assigned: 36/45 [accelerated / standard] Assessed: 33/30 (at 6 months)
Interventions	Timing of intervention: both groups began with gentle active range of motion at three to five days after plate fixation using the same type of plate. Custom splint applied (30 degrees wrist extension) removed for hygiene, dressing, eating, exercises. Scar management began when sutures were removed in both groups 1. Accelerated rehabilitation group. Patients initiated wrist/forearm passive range of motion and strengthening exercises at two weeks. Weaned from splint at week 3, discontinued splint at week 4. Isotonic and medium putty strengthening at week 4; heavy putty strengthening at week 6 2. 'Standard' rehabilitation group. Patients initiated wrist/forearm passive range of motion and strengthening exercises at six weeks. Weaned from splint at week 6, discontinued splint at week 7. Isotonic and medium putty strengthening at week 7; heavy putty strengthening at week 8
Outcomes	Length of follow‐up: 6 months post‐operatively; also 2, 3, 4, 6, 8 and 12 weeks  1. Functional: Disabilities of the Arm, Shoulder and Hand (DASH) scores; grip strength, palmar pinch strength; ROM (wrist flexion, extension, supination, pronation);  2. Clinical: adverse events and splint repair  3. Others: fracture alignment and healing
Funding, ethics and patient consent	Study received "limited financial support (from a research grant) from" a commercial company but no implants were involved. Approval reported by institutional review board  Informed consent was obtained from all participants
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Randomization was performed by computerized random number generation"
Allocation concealment (selection bias)	Low risk	Although no details, there was blinding and independent randomisation: "Enrollment was performed by a research coordinator who was blinded to fracture severity and, if consent had been obtained postoperatively but before initiation of therapy, to the operative result"
Blinding of participants and personnel (performance bias)   Subjective outcomes	High risk	"All participating surgeons were blinded to each patient’s randomization group as the surgeons’ postoperative care was the same for the two groups. Due to the nature of the study, the hand therapists were not blinded." Participants were also not blinded
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	As above
Blinding of outcome assessment (detection bias)   Subjective outcomes	High risk	Participants were not blinded
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	High risk	"All measurements and assessments were performed postoperatively at scheduled outpatient hand therapy appointments by the unblinded treating therapist, which introduced a potential source of measurement bias"
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	High risk	"All measurements and assessments were performed postoperatively at scheduled outpatient hand therapy appointments by the unblinded treating therapist, which introduced a potential source of measurement bias"
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	Similar losses (at 12 weeks: 6/36 (16.7%) versus 8/45 (17.8%)) but small trial
Incomplete outcome data (attrition bias)   Longer term follow‐up	High risk	Over three times as many loses in the usual group at 6 months: 3/36 (8.3%) versus 12/45 (26.7%)
Selective reporting (reporting bias)	Low risk	No protocol available nor trial registration but assessment was straightforward, systematic and covered sufficient key outcomes
Major baseline imbalance bias?	Unclear risk	Some imbalance in fracture distribution of unknown importance. Accelerated group 5 years younger on average
Other performance bias (e.g. differential expertise bias)	Low risk	Since the same programme was applied but at different times the expertise of the hand therapists would have mattered less

Methods	Method of randomisation: opaque envelopes contained in a box  Assessor blinding: yes for objective measures.  Intention‐to‐treat analysis: yes  Loss to follow‐up: 2 (at 6 and 10 weeks)
Participants	General Hospital, Queensland, Australia  21 participants  Inclusion criteria: distal radius fracture, treated by plaster cast immobilisation, age 18 years or above, informed consent.  Exclusion criteria: open skin lesions, nerve or tendon damage associated with the fracture, fracture required surgical fixation or known pathology in their intact arm.  Classification: "Rockwood + Green" (4 categories: extra‐articular undisplaced, extra‐articular displaced; intra‐articular undisplaced; intra‐articular displaced): distribution not reported.  Sex: 15 female (76%)  Age: mean 51 years  Assigned: 11/10 [pneumatic compression / control]  Assessed: 10/9 (at 10 weeks)
Interventions	Timing of intervention: following initial treatment of fracture in a split cast. Allocation was at a mean of 9.4 days after fracture. The split cast was replaced with a full forearm plaster, worn for 5 weeks. All participants had an inflatable cuff positioned around their forearm under their plaster with the valve of the cuff protruding through the cast. Both groups were instructed to actively make a fist 100 times per day during the immobilisation period. All participants had a total of 6 weeks immobilisation.  1. Cyclic pneumatic soft tissue compression for 5 weeks. Participants were provided with a compression pump apparatus to use at home and received instructions on its use. The apparatus consisted of a compression pump connected to an inflatable cuff positioned around the proximal forearm flexor and extensor muscle bulk under the plaster. The compression pump was designed to pump air into a reservoir and, at set time periods, release a pressurised volume of air into the inflatable cuff. One inflation/deflation of the cuff took 10 seconds and 60 compressions were applied per treatment session. The cyclic pneumatic pressure was applied twice per day (morning and evening) taking ten minutes for each session. After piloting on two cases, the pressurising period of 3.5 seconds was chosen on the basis that it did not produce any pain at the fracture site 2. 'Usual care'. (See above) After plaster cast removal, all participants were given a programme of strengthening and stretching exercises for the hand, wrist, and forearm to be carried out twice a day for four weeks
Outcomes	Length of follow‐up: 10 weeks; also 6 weeks  1. Functional: strength (power grip, pinch grip, key grip, supination (isometric), ROM (pronation/ supination; flexion/extension). Absolute and % 'intact side'
Funding, ethics and patient consent	No mention of funding. Ethical approval from hospital ethics committee  Informed patient consent
Notes	Individual patient data provided in an appendix for 'absolute' outcomes reveal that the standard deviations were probably standard errors. Results presented here are derived from individual patient data and thus the results relative to the uninjured side are not presented in this review given the concerns about data discrepancies
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Participants were then randomly allocated to either the experimental group or the control group by the fracture clinic nurse, who drew opaque envelopes containing the concealed group allocation from a box"
Allocation concealment (selection bias)	Low risk	"Participants were then randomly allocated to either the experimental group or the control group by the fracture clinic nurse, who drew opaque envelopes containing the concealed group allocation from a box"
Blinding of participants and personnel (performance bias)   Subjective outcomes	Unclear risk	No subjective outcomes measured
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Participants and personnel not blinded
Blinding of outcome assessment (detection bias)   Subjective outcomes	Unclear risk	No subjective outcomes measured
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	Low risk	"All participants had an inflatable cuff positioned around their forearm under their plaster with the valve of the cuff protruding through the cast so that the independent assessor remained blind to group allocation during all measurements"
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Unclear risk	Outcomes not recorded
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	"Two participants were lost to follow‐up: one participant (experimental group) removed the plaster prematurely and could not receive the intervention, and the other (control group) failed to attend measurement sessions and withdrew within the first three weeks of the trial"
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No longer‐term follow‐up
Selective reporting (reporting bias)	Unclear risk	No protocol and no trial registration available
Major baseline imbalance bias?	High risk	The populations in the two groups differed importantly: intervention group 5/11 were male, mean age 46 years, and "more severe fractures"; control group 0/10 were male, mean age 57 years
Other performance bias (e.g. differential expertise bias)	Low risk	Same treatment other than intervention applied to both groups

Methods	Method of randomisation: by drawing lots (non‐replacement method)  Assessor blinding: no  Intention‐to‐treat analysis: likely  Loss to follow‐up: none
Participants	Queen Elizabeth Hospital, Hong Kong  83 participants  Inclusion criteria: "stable'" distal radial fracture treated by closed reduction and 6 weeks plaster cast immobilisation. Informed consent. Able to communicate independently  Exclusion criteria: CRPS‐1, inflammatory arthritis, perivascular disease, previous fracture or neurovascular injuries in the affected hand, heart disease, use of heart pacemaker or other auxiliary organs, tuberculosis, viral infections, juvenile diabetes, mycosis, internal haemorrhages, or pregnancy. Recently had deep X‐Ray therapy or pulsed electromagnetic treatment during immobilisation period  Classification: None given  Sex: 55 female (66%)  Age: mean 63 years; range 17 to 80 years  Assigned: 23/22/22/16 [PEMF+ice/ sham PEMF+ice/ PEMF / sham PEMF]  Assessed: 23/22/22/16 (at 5 days)
Interventions	Timing of intervention: 3 to 4 days following plaster cast removal (6 weeks immobilisation)  All treatments were 30 minutes for 5 consecutive days. After the first treatment, the participants were taught and given written instructions for a home exercise programme of active wrist and finger mobilisation exercises and advised to do these twice a day for 20 minutes each session. Exercise compliance was checked by the physiotherapist at each treatment session 1. Pulsed electromagnetic field (PEMF) at 50 Hz with a field intensity of 99 gauss, and ice (1 kg pack of flaked ice wrapped in towel and placed dorsally)  2. Sham PEMF and ice  3. PEMF  4. Sham PEMF
Outcomes	Length of follow‐up: 4 days (before the 5th treatment session); also 2 days (before 3rd session)  1. Functional: pain (VAS during mobilisation), ROM (pronation; supination; flexion; extension; radial deviation; ulnar deviation)  2. Clinical: oedema. Adverse events ("none reported")
Funding, ethics and patient consent	No mention of funding or ethics. Informed consent obtained
Notes	Reply received from A/Prof Cheing on 9 December 2005 who provided further details of the methods, including randomisation, and also stated there were no adverse events reported
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Email (09/12/2005): "Group allocation was done by drawing lots (non‐replacement method)"
Allocation concealment (selection bias)	Unclear risk	Email (09/12/2005): "Group allocation was done by drawing lots (non‐replacement method).". No information to judge
Blinding of participants and personnel (performance bias)   Subjective outcomes	Unclear risk	"participants were blinded as to whether they received PEMF or sham PEMF." Personnel may have been aware of circuit being disconnected at the back of the machine." No blinding for ice intervention
Blinding of participants and personnel (performance bias)   Objective outcomes	Unclear risk	As above
Blinding of outcome assessment (detection bias)   Subjective outcomes	Unclear risk	No mention of blinding of outcome assessment, but follow‐up assessment within the treatment period
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	Unclear risk	No mention of blinding of outcome assessment, but follow‐up assessment within the treatment period
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Low risk	Outcomes unlikely to be affected by lack of blinding
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	No lost to follow‐up but trial failed to record outcome after the end of treatment
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No longer‐term follow‐up
Selective reporting (reporting bias)	Unclear risk	Protocol not available
Major baseline imbalance bias?	Unclear risk	Baseline imbalances in some variables, such as extension
Other performance bias (e.g. differential expertise bias)	Low risk	No problems detected

Methods	Method of randomisation: use of sealed envelopes (concealment confirmed by trialist)  Assessor blinding: yes  Intention‐to‐treat analysis: likely but for 2 excluded from analyses (1 death and 1 with severe pain after cast removal)  Loss to follow‐up: none (except 2 exclusions)
Participants	University Hospital Gentofte, Denmark  32 participants  Inclusion criteria: Colles' fracture, treated with plaster cast.  Exclusion criteria: none provided.  Classification: Older's classification  Sex: (of 30) 27 female (90%)  Age: (of 30) mean 66 years; range 46 to 82 years  Assigned: 16/16 [occupational therapy / control]  Assessed: 16/14 (at 9 months)
Interventions	Timing of intervention: following plaster cast removal (5 weeks immobilisation)  All participants were given instructions by occupational therapist for shoulder, wrist and fingers exercises to be performed thrice daily at home  1. Occupational therapy involving active joint exercises for wrist, elbow and shoulder; oedema prevention; coordination exercise; coarse and fine motor‐function exercise; strengthening exercise; sensation exercise; ADL training. "Distributed" around twice weekly sessions until therapist considered no further progress was being made.  2. Home exercises only
Outcomes	Length of follow‐up: 9 months; also 3 months  1. Functional: grip strength, Solgaard modified Gartland and Werley score  2. Number of sessions and overall duration of occupational therapy
Funding, ethics and patient consent	No mention of funding No mention of patient consent. Mention of the approval of protocol by local ethics committee
Notes	Replies received 20 and 21 August 2001
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Randomized trial." No information on sequence generation
Allocation concealment (selection bias)	Low risk	"The allocation principle was by the closed envelope method"  Letter dated 19/08/2001: "The envelopes were sealed and allocation was concealed at the time of randomisation"
Blinding of participants and personnel (performance bias)   Subjective outcomes	Unclear risk	No subjective outcomes except within Gartland and Werley score
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Neither participants nor personnel providing care were blinded
Blinding of outcome assessment (detection bias)   Subjective outcomes	Unclear risk	No subjective outcomes except within Gartland and Werley score
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	Low risk	Letter dated 19/08/2001: "The outcome assessments were made by myself and T.C. Christiansen. At the time of the assessments we did not know to which group the patients were allocated"
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Low risk	Outcomes unlikely to be affected by lack of blinding
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	Letter dated 19/08/2001: "No patients were lost to follow up. Two patients were excluded from the study. One patient died after the 3 months follow up and the other had severe pain at cast removal ‐ had further immobilisation and developed reflex sympathetic dystrophy." Further email (21/08/2001) explained that both patients were in the "non‐occupational group." Incomplete data provided on Gartland and Werley scores. (However, grip strength data provided)
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	As above
Selective reporting (reporting bias)	Unclear risk	No protocol available
Major baseline imbalance bias?	Low risk	No major imbalance
Other performance bias (e.g. differential expertise bias)	Low risk	No problems detected.  Letter dated 19/08/2001: "Both groups received instruction by occupational therapist" "No patients in the non‐occupational group received occupational therapy"

Methods	Method of randomisation: independent person generated sealed numbered opaque envelopes using a random numbers table ‐ researcher had no knowledge of allocation in advance  Assessor blinding: no  Intention‐to‐treat analysis: yes  Loss to follow‐up: none
Participants	Pilgrim Hospital, Boston, UK  17 participants Inclusion criteria: Distal radial fracture treated conservatively with closed reduction and immobilisation, age > 16 years (adult), willing and able to attend the department for assessment and treatment. informed consent Exclusion criteria: frail elderly people with mobility problems preventing attendance, impaired mental or cognitive ability, multiple fractures or extensive soft tissue injuries, surgical treatment or pre‐morbid neurological conditions Classification: None given  Sex: 16 female (94%)  Age: mean 65.5 years; range 41 to 81 years  Assigned: 8/9 [early intervention/control]  Assessed: 8/9 (at 4 weeks post‐removal of plaster cast)
Interventions	Timing of intervention: within 4 days of fracture (routinely 4 weeks immobilisation, or, for some, 6 weeks)  All participants received home treatment programme including written advice about skin care, control of oedema, wrist and forearm exercises at fracture clinic and cast application. Post‐immobilisation care programme for all participants comprised an individualised home programme and, where prespecified criteria were met, attendance of a hand therapy group 1. "Early therapeutic intervention" with oedema management, active range of movement of uninvolved joints (fingers, elbow, shoulder and neck), monitoring of plaster cast, written information and contact number of project team. Weekly contact with member of hand therapy team 2. "Standard intervention" only, started after plaster cast removal
Outcomes	Length of follow‐up: 4 weeks post‐immobilisation; also 4 days 1. Functional: grip strength, ROM (pronation; supination; flexion; extension; radial deviation; ulnar deviation), functional dexterity (9‐hole peg test), pain at rest or during activity (VAS), DASH functional scores, finger movement (total active movement), opposition of thumb (Kapandji scores), pinch grip, and referral to hand therapy 2. Clinical: oedema. Complications: CRPS‐1
Funding, ethics and patient consent	No mention of funding or ethics Written patient consent
Notes	Trial was part of a masters degree in hand therapy
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"A table of random numbers was used"
Allocation concealment (selection bias)	Low risk	"The allocation was then placed into envelopes and sealed with the subject number on the outside. ... When allocating patients to the treatment groups the envelopes were opened in front of the patients if present, or a member of staff if carried our over the telephone, again in order to prevent bias." Trialist confirmed envelopes were "opaque"
Blinding of participants and personnel (performance bias)   Subjective outcomes	High risk	Neither participants nor personnel providing care were blinded
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Neither participants nor personnel providing care were blinded
Blinding of outcome assessment (detection bias)   Subjective outcomes	High risk	There was no blinding
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	High risk	There was no blinding
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Low risk	Outcomes unlikely to be affected by lack of blinding
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Low risk	All participants accounted for. No loss to follow‐up
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No longer‐term outcomes
Selective reporting (reporting bias)	Low risk	No protocol available but comprehensive MSC report indicates no risk
Major baseline imbalance bias?	High risk	Small number of participants with 9 year difference in the two group's mean age (61 versus 70)
Other performance bias (e.g. differential expertise bias)	Unclear risk	Unknown variation in duration of cast immobilisation

Methods	Method of randomisation: involved envelopes ‐ stated to be single‐blind by trialist  Assessor blinding: yes  Intention‐to‐treat analysis: probably  Loss to follow‐up: probably none
Participants	Fredenksberg Hospital, Alsgarde, Denmark  40 participants  Inclusion criteria: Colles' fracture, unilateral fracture, suitable for plaster cast: stable fracture in plaster, attendance at casualty ward within 24 hours of injury, age > 45 years, (implied: resident in hospital catchment area)  Exclusion criteria: unstable fracture (reduced position could not be maintained in plaster), wrist arthritis, other fracture in same limb, neuromuscular pain in limb, dementia or some other condition making participation difficult  Classification: Older  Sex: 35 female (88%)  Age: median 74.5 years; range 47 to 93 years  Assigned: 17/23 [occupational therapy/control]  Assessed: 17/23 (at 13 weeks)
Interventions	Timing of intervention: following reduction and application of plaster cast. (Approximately 5 weeks immobilisation)  All participants given advice about active movement exercises of shoulder and fingers and information on the problems of plaster casts after application of cast (in casualty) 1. Participant attended rheumatoid disorder outpatients clinic 1 to 3 days after initial treatment. Instructions for hand pumping exercises, active finger, elbow, shoulder movements, assessment of the need for appliances (e.g. angled knives), and for home help provided by occupational therapist. Referral to occupational therapist for rehabilitation if required after plaster cast removal 2. Referral to occupational therapist for rehabilitation if required after plaster cast removal
Outcomes	Length of follow‐up: 13 weeks; also 5 and 9 weeks 1. Functional: modified Stewart 1984 (modified Gartland and Werley) functional score (subjective pain, limitations of movement and function; ROM, grip strength, median nerve compression), movement, use of analgesia  2. Clinical: oedema, abnormal sweating, colour, temperature. Complications: CRPS‐1, median & ulnar nerve compression, tendon rupture  3. Use of appliances, home help, plaster cast problems, participant satisfaction, understanding of instructions given at casualty: for intervention group participants only.
Funding, ethics and patient consent	No mention of funding or ethics All participants gave informed consent
Notes	Translation from Danish by Dr Michael Bird  Further details of trial received 20 August 2001. Nine participants previously unaccounted for had unstable fractures, and were re‐admitted, some fractures were fixed with Hoffman external fixation, and were not included in trial
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Completed form 20 August 2001: "Picking an envelope with two possibilities (single blind procedure)"
Allocation concealment (selection bias)	Low risk	In answer to whether allocation was concealed ‐ completed form 20 August 2001: "YES"
Blinding of participants and personnel (performance bias)   Subjective outcomes	High risk	Participants and personnel not blinded
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Participants and personnel not blinded
Blinding of outcome assessment (detection bias)   Subjective outcomes	Unclear risk	From translation: "Five, 9, and 13 weeks after the date of the accident, all patients were examined by the same rheumatoid specialist, who did not know to which group the patients belonged." There was, however, no indication of safeguards such as telling the participants not to tell the assessor which group they were in
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	Unclear risk	As above ‐ blinding possible but not mention of safeguards
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Low risk	Outcomes unlikely to be affected by any deficiencies in blinding
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	Some slight unease concerning recruitment status of 9 participants, accounted for in correspondence, that had unstable fractures, and were re‐admitted, some fractures were fixed with Hoffman external fixation, and were apparently "not included" in the trial
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No longer‐term follow‐up
Selective reporting (reporting bias)	Unclear risk	No protocol available to judge
Major baseline imbalance bias?	Unclear risk	Probably no major imbalance but there were differences in the distribution of fracture types reported
Other performance bias (e.g. differential expertise bias)	Unclear risk	Pragmatic trial but lack of information on occupational therapy referral and mention of two of the treatment group patients already having home help arranged by their general practitioner

Methods	Method of randomisation: off site computerised blocked randomisation and sealed and sequentially numbered envelopes  Assessor blinding: yes at 8 weeks but not at 12 weeks  Intention‐to‐treat analysis: no, post‐randomisation losses, 1 unexplained protocol violation  Loss to follow‐up: 8 (at 12 weeks)
Participants	A Sydney metropolitan hospital (Royal North Shore Hospital), Australia  40 participants Inclusion criteria: distal radius fracture, conservatively or surgically treated patients were referred to physiotherapy by a consultant hand surgeon at least 10 weeks from the time of injury if surgeon was concerned about progress, a stable and united (or uniting) unilateral fracture, wrist contracture evident by a loss of passive extension compared with the unaffected wrist, lived in the Sydney metropolitan region, were willing or likely to co‐operate with the intervention, over the age of 18 years and informed consent Exclusion criteria: patient unlikely to co‐operate  Classification: not stated Sex: female (70%)  Age: median 66 [dynamic splints] versus 58 years [control]  Assigned: 20/20 [dynamic splints / control] (intention‐to‐treat analysis); 19/21 (per‐protocol analysis) Assessed: 15/17 (at 12 weeks)(intention‐to‐treat analysis); 14/18 (per‐protocol analysis)
Interventions	Timing of intervention: recruitment after 10 weeks (median 76 days [dynamic splints] versus 83 days [control]) from fracture. All patients had commenced post‐immobilisation exercises. Acute management had been cast (15 patients), surgery (24 internal fixation, 1 K‐wires) 1. Dynamic splints for 8 weeks plus usual care (see below). Splint was custom‐made from thermoplastic material and "incorporated an axis about the flexion‐extension plane of the wrist. The fingers and thumb were unrestricted. A constant low‐load stretch was applied in the direction of wrist extension via an elastic band, with the stretch set as high as tolerated by each participant. This stretch was adjusted once every two weeks to maintain the wrist at maximal tolerated extension. Participants were instructed to wear the splint for as long as possible during the day, aiming for at least six hours a day of cumulative splint wear.  They were encouraged to actively flex their wrist against the splint intermittently, and were advised to continue activities of daily living whilst wearing the splint wherever possible." (Jongs 2012) 2. 'Usual care'. This consisted of general advice and a home exercise programme, which was monitored but not supervised, provided by a therapist blinded to the allocation. Participants were instructed to perform exercise at least three times throughout the day. Participants were shown the exercises (for range of motion and wrist and grip strength) and given written instructions. Verbal advice given on expected progress, including on pain resolution, and on using hand of the affected wrist as much as possible in day‐to‐day activities All participants continued with the exercises and advice unsupervised until 12 weeks
Outcomes	Length of follow‐up: 8 weeks; also 12 weeks (1 month post‐treatment)  1. Functional: PRHWE, ROM (passive wrist extension; active wrist extension, flexion, radial deviation and ulnar deviation); performance and satisfaction items of the Canadian Occupational Performance Measure (COPM)  2. Clinical: adverse events and splint repair  3. Others: compliance
Funding, ethics and patient consent	No mention of funding but no commercial company with direct involvement; sponsor was the University of Sydney. Ethical approval reported  Informed consent was obtained from all participants
Notes	One participant who was randomly allocated to dynamic splint group ended up in the control group within 10 minutes of allocation because of an unexplained error Per‐protocol and intention‐to‐treat analyses were reported
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"a computerised blocked randomisation sequence was generated prior to the commencement of the trial by an independent offsite person"
Allocation concealment (selection bias)	Low risk	"Participants’ allocations were placed in opaque sealed and sequentially numbered envelopes that were held off‐site. A participant was considered to have entered the trial once his/her envelope was opened"
Blinding of participants and personnel (performance bias)   Subjective outcomes	High risk	Participants and physiotherapists not blind to use or not of dynamic splints
Blinding of participants and personnel (performance bias)   Objective outcomes	High risk	Participants and physiotherapists not blind to use or not of dynamic splints
Blinding of outcome assessment (detection bias)   Subjective outcomes	High risk	Participants not blinded for PRWE score
Blinding of outcome assessment (detection bias)   Objective outcomes (functional impairment)	High risk	Assessor at 12 weeks was not blinded
Blinding of outcome assessment (detection bias)   Complications, number of sessions, return to former activity	Unclear risk	Unblinded therapist contacted patients to monitor and record adherence: patients recorded activity and use of splint in diaries
Incomplete outcome data (attrition bias)   Short term follow‐up (up to 3 months)	Unclear risk	Data provided to check for differences between per‐protocol and intention‐to‐treat analyses (no problem detected); 20% loss to follow‐up but not badly imbalanced between the two groups (5/20 versus 3/20)
Incomplete outcome data (attrition bias)   Longer term follow‐up	Unclear risk	No long‐term follow‐up
Selective reporting (reporting bias)	Low risk	Trial reported according to prior trial registration details
Major baseline imbalance bias?	Unclear risk	There were baseline differences in participant characteristics (e.g. gender: 4/19 versus 8/21 males; surgery: 9/19 versus 16/21) but similar results for baseline measurements of outcome measures
Other performance bias (e.g. differential expertise bias)	Low risk	Comparable