

Upper limb
Hand rehabilitation and splinting: refining and deepening your practice
Jean Lichtle
Updated · published
Distal radius fracture: what the tilt really predicts, why the distal radioulnar joint decides rotation, and when rehabilitation should begin.

An isolated radiographic angle should not alarm the patient: it is the combination of several defects, in a young, demanding patient, that matters.
96%of the range of motion and strength of the healthy side regained at eleven years
What follows covers each of these points in detail, with sources. It is there if you need it.
This topic is taught in a course: Hand rehabilitation and splinting: refining and deepening your practice, on-site with Jean Lichtle.
1 / 12
11 article chapters · 44 min in total
Understanding◔ 5 min
The radius bears most of the load of the wrist, the ulna a far smaller fraction.
Population◔ 2 min
Two very different populations lie behind the same diagnosis: the postmenopausal woman and the young person.
Radiograph◔ 4 min
Of 84 correlations tested, only 2 linked radiographic anatomy to function.
Rotation◔ 3 min
Supination is the movement recovered least well, because of the distal radioulnar joint.
Surgery◔ 4 min
In people aged 60 and over, no surgical technique is superior, nor is plaster.
Immobilisation◔ 3 min
Immobilising the wrist does not immobilise the limb: fingers, elbow and shoulder remain to be worked.
After plaster removal◔ 9 min
Supervision does not always change the outcome; a randomised trial shows the opposite.
Complications◔ 3 min
The physiotherapist sees these patients more often than the surgeon, and spots the complication as it sets in.
Clinical case◔ 2 min
Untreated osteoporosis preceded a complex regional pain syndrome at three months.
In practice◔ 5 min
Do not dramatise an isolated angle: rotation and the functional score matter more.
Questions◔ 4 min
Recovery lasts far longer than bone healing, sometimes years.
Train on this


Upper limb
Jean Lichtle
Next comes practice: the course that teaches this topic, with Jean Lichtle.
Our courses on this topic may be covered by DPC or FIFPL, the French continuing-education funding schemes, depending on your profession and situation.
This is the fracture everyone thinks they know: a fall on the hand, a dinner-fork deformity, six weeks in plaster. The detail that changes rehabilitation lies elsewhere. Dorsal tilt, the parameter everyone measures, predicts far less than its radiographic obviousness suggests, and it becomes truly troublesome only when combined with shortening. What decides the lost rotation is the distal radioulnar joint, at the other end of the same fracture line. And the physiotherapist's work does not begin on the day the plaster comes off: it begins during immobilisation, on segments that the plaster does not immobilise.
Synthesis written from primary sources each verified on PubMed through the E-utilities interface: identifier resolved, journal, year and author list checked, abstract read before citation. Full bibliography at the end of the article.
Distal radius fracture in three figures
Three independent results that shift the question from “how do we straighten this wrist?” to “what, here, really changes the patient's life?”
Sources: Soerensen S et al., Hand (N Y) 2024, population study, 5,426 fractures in 522,607 inhabitants (PMID 35856325); Chung KC et al., JAMA Netw Open 2021, WRIST randomised trial, 182 patients assessed at 24 months (PMID 34137830); Reid SA et al., J Physiother 2020, multicentre randomised trial, 67 patients, difference 12° 95 % CI 5 to 20 (PMID 32291223).
What to keep in mind
Four situations that call for a medical opinion without waiting for the next session
A fractured wrist is judged by what the patient can do, not by the angle it forms. Eleven years on, range of motion and strength return to 96 % of the healthy side, whatever the radiograph looks like.

X-rayAnteroposterior wrist X-ray: distal radius fracture, arrow on the metaphyseal fracture line.
Source : Ozaki et al., Medicine, 2016, figure 3 · CC BY
To understand the tilt is to understand why it recurs in plaster, and why it is paid for in rotation more than in flexion.
The distal radius is the widened end of the radius, the one that articulates with the carpus. It carries three articular surfaces: the scaphoid fossa and the lunate fossa, which receive the first row of the carpus, and the sigmoid notch, on the medial border, which receives the head of the ulna and forms the radial side of the distal radioulnar joint. This third surface is the one we forget, and it is the one that decides pronation and supination.
The radius bears most of the load transmitted by the hand to the forearm. The ulna, more slender and articulated with the carpus through the triangular fibrocartilage complex, bears a far smaller fraction. This distribution explains two things: why the distal radius is the part that breaks when one falls on the hand, and why its shortening mechanically transfers load to the ulna, with the consequences that the chapter on the distal radioulnar joint details.
The dominant mechanism is stereotyped: a fall landing on the open hand, wrist in extension. In the Danish population study of 5,426 fractures of the distal end of the forearm, the fall from standing height accounted for 76 % of mechanisms, and the most frequent type of fracture was extra-articular, AO type 2R3A, in 69 % of cases1. In other words: the typical fracture is neither articular, nor comminuted, nor the result of violent trauma. It is an ordinary fall in someone whose bone is fragile.
The force that travels up from the heel of the hand compresses the dorsal cortex, which collapses, while the volar cortex fails in tension. The distal fragment then goes backwards: this is the dorsal tilt, described by Pouteau and then by Colles, which gives the dinner-fork deformity seen from the side. The volar-tilt variant, known as Goyrand-Smith, results from the opposite mechanism, wrist in flexion, and is far rarer.
This anatomy of displacement has a direct practical consequence: the tilt is not only produced by the impact, it is maintained by collapse of the dorsal cortex. Once the cancellous bone is impacted, there is no longer a posterior column to hold the reduction. That is the mechanical reason why a reduction in plaster is so often lost again, and why secondary redisplacement is a frequent event and not a technical accident.
Why the distal fragment tilts backwards, and why it holds poorly
The dorsal cortex collapses under compression; once impacted, it no longer offers any support to maintain the reduction
Schematic diagram. The redisplacement figure comes from Karantana A, Handoll HH, Sabouni A, Cochrane Database Syst Rev 2020, review of 26 trials, mean of six trials reporting this outcome, spread 3.3 to 75 % (PMID 32032439). Physiological volar tilt and its reference value are discussed in the chapter on measurements.
Four key figures to keep in mind
What the large series establish about the shape of the fracture, its healing and its aftermath
Sources: Soerensen S et al., Hand (N Y) 2024, 5,426 fractures (PMID 35856325) for shape and mechanism; Karantana A et al., Cochrane Database Syst Rev 2020, mean of six trials (PMID 32032439) for redisplacement; Schmidt V et al., J Hand Surg Eur Vol 2024, 242 patients examined at 11-13 years (PMID 37684021) for recovery.
Key points
Two populations, two prognoses, two rehabilitation plans. Confusing them is the first error of orientation.
The most robust population study available followed an average-risk population of 522,607 inhabitants and manually validated the radiographs and records of 5,426 fractures of the distal end of the forearm. The overall incidence stands at 207.7 per 100,000 per year, broken down into 323.4 in women and 93.3 in men. Women accounted for 77 % of fractures, and incidence rose markedly in them after 50. The incidence of distal radius fractures proper was 203.0 per 100,000 per year, that of isolated ulna fractures only 3.81.
Lifetime cumulative risk gives the measure of the public health problem: in high-income countries, 6 % of women will have had a wrist fracture by 80, and 9 % by 902.
Incidence of fractures of the distal end of the forearm
A ratio of 3.5 between women and men, in a manually validated population
Single source: Soerensen S, Larsen P, Korup LR et al., Epidemiology of Distal Forearm Fracture: A Population-Based Study of 5426 Fractures, Hand (N Y) 2024;19(1):24-29 (PMID 35856325). Average-risk population of 522,607 inhabitants, manual review of radiographs and records.
Behind this overall figure lie two distinct clinical realities, and the authoritative trials do not concern the same one.
The first is that of the postmenopausal woman with fragile bone, who falls from standing height. This is the population of the WRIST trial, whose mean age was 70.1 years with 87.9 % women3. In her, a wrist fracture is a sentinel fracture : it heralds the later bone risk, and its detection should trigger an assessment of osteoporosis and of fall risk. Our article on osteoporosis and the prevention of fragility fractures details this cascade and what it requires in practice.
The second is that of the young person in high-energy trauma, a bicycle fall, a sports accident, a fall from height. The functional demand there is entirely different, associated injuries are more frequent, and it is in this population that anatomical restoration retains the most meaning, as what follows shows.
The trap of generalisation
Most of the large trials of this condition, WRIST in particular, recruited patients aged 60 and over. Their reassuring conclusions about the low prognostic value of radiographic parameters do not transpose mechanically to the young patient. The Brogren cohort, whose age range is wider and which finds an effect of combined malunion at 2 years5, is a reminder that tolerance of deformity depends on what the patient asks of the wrist.

X-rayLateral and anteroposterior wrist X-rays: Colles fracture, arrows on the metaphyseal line of the distal radius with dorsal tilt of the fragment.
Source : Wong et al., International journal of emergency medicine, 2015, figure 5 · CC BY
Four measurements, of which only one appears linked to function in older people, and a combination that weighs more than each of its parts.
The radiographic assessment of a distal radius fracture rests on simple measurements, made on the frontal and lateral views. The physiotherapist does not have to take them, but reading them in a report changes the way rehabilitation is conducted, and above all the way it is discussed with the patient.
| Parameter | What it measures | Reference | What the evidence says |
|---|---|---|---|
| Dorsal tilt (sagittal inclination) |
Orientation of the articular surface on the lateral view. Physiologically inclined towards the palm | Physiological volar tilt | No effect on clinical outcome at 11-13 years6. Matters at 2 years if combined with shortening5 |
| Ulnar variance (radial shortening) |
Relative height of the ulnar head with respect to the radius, on the frontal view | 0 mm | In those aged 70 and over, each millimetre towards normal was associated with 10.4 points better on the daily activities score (95 % CI 3.86 to 16.84)4 |
| Radial inclination | Obliquity of the articular surface on the frontal view | 22° | In those aged 70 and over, each degree of deviation was associated with a loss of strength of 1.1 kg (95 % CI 0.38 to 1.76)4, with no effect on the global functional score |
| Joint involvement (step-off, gap) |
Congruence of the radiocarpal articular surface | Continuous surface | Intra-articular extension influenced neither clinical outcome nor the risk of osteoarthritis at 11-13 years6 |
The secondary analysis of the WRIST trial put this question to the test systematically. In 166 participants aged 60 and over (86.7 % women, mean age 70.9 years), the authors set the radiographic measurements against seven functional outcomes at 12 months. Of 84 correlation coefficients calculated, only 2 were statistically significant, both in the subgroup aged 70 and over, and none concerned the global hand score or the function score4. The authors' conclusion is explicit: precise restoration of the anatomy of the wrist is not associated with better results in older people at 12 months.
The longest prospective follow-up available points the same way and extends it to a wider population. Eleven to thirteen years after the fracture, 292 patients answered the questionnaires and 242 were examined. The median QuickDASH was 5, the EuroQol score 1.0, and the range of motion and grip strength both reached 96 % of the contralateral side. Neither radiocarpal osteoarthritis (6 %) nor nonunion of the ulnar styloid (30 %) affected the results. Above all, neither dorsal tilt, nor radial inclination, nor ulnar variance, nor intra-articular extension influenced the long-term clinical outcome or the risk of osteoarthritis6.
One must nevertheless beware of concluding from this that reduction has no importance at all. Three reservations are in order, and they are clinically decisive.
First reservation: it is the combination that matters. The Brogren prospective cohort followed 123 patients with a DASH repeated at baseline, 3 months, 6 months, 1 year and 2 years. By defining malunion as an ulnar variance of at least 1 mm combined with a dorsal tilt greater than 10°, the authors formed three groups: no malunion (n = 35), with one or the other defect (n = 65), with both (n = 23). Only the group with both defects kept a significantly higher disability: 13 more DASH points than the group without malunion (95 % CI 1.4 to 25), and 13 points more than the group with a single defect (95 % CI 2.2 to 24)5. A single defect did not behave any differently from a correctly reduced wrist.
Second reservation: age modulates everything. The most reassuring results come from older populations with moderate functional demand. The 35-year-old who works with their hands is not in that frame, and no data of this level allow the same tolerance thresholds to be applied to them.
Third reservation: these studies measure global function, not rotation. Hand scores such as the DASH or the MHQ integrate activities that pronation and supination influence, but do not determine on their own. That is precisely the subject of the next chapter.
Key points
Supination is the movement these patients recover least well. This is no accident, and it is not the fault of the radiocarpal fracture line.
The distal radioulnar joint unites the sigmoid notch of the radius with the head of the ulna. Together with the proximal radioulnar joint at the elbow, it forms the pair that carries pronation and supination : it is the radius that rotates around the ulna, and not the reverse. Its stability depends on bony congruence, weak in itself, and above all on the triangular fibrocartilage complex, whose insertions are on the ulnar fovea and styloid.
Now the sigmoid notch of the radius lies a few millimetres from the usual fracture line. The review by Nypaver and Bozentka devoted to this joint in the context of distal radius fracture details its anatomy, its biomechanics, the ways in which the injury presents, at the time of trauma as after treatment, the appropriate assessment and the therapeutic options7. The clinical point to keep is simple: a distal radius fracture is almost always also a matter of the distal radioulnar joint, whether or not the report mentions it.
The mechanical reasoning is direct. The radius tilts dorsally and shortens; its sigmoid notch then changes orientation and height relative to the ulnar head, which has not moved. Three consequences follow:
This reading sheds light on a clinical fact that long follow-up confirms: in the Schmidt cohort, the nonunion of the ulnar styloid affected 30 % of patients at eleven years without affecting the clinical outcome6. In other words, an ununited styloid is not in itself a problem; it is instability of the joint that it stabilises that would be one, and the two should not be confused.
What the distal radioulnar joint explains in a wrist that does not recover
Three complaints, three distinct mechanisms, three different approaches
Source of the anatomical and injury framework: Nypaver C, Bozentka DJ, Distal Radius Fracture and the Distal Radioulnar Joint, Hand Clin 2021;37(2):293-307 (PMID 33892882). The associated rehabilitation approaches are a proposal of clinical reasoning, explicitly not derived from a comparative trial.
What this changes in the session
Faced with a wrist that plateaus in supination at 6 or 8 weeks, the reflex response is to insist on stretching. The useful reasoning consists first of distinguishing soft-tissue stiffness, which gives way progressively, from a bony block or an incongruence, which will not give way. A limitation that no longer moves after three to four weeks of well-conducted work, especially if it comes with pain at the ulnar border, deserves to be reported to the surgeon rather than forced.
The question is not for the physiotherapist to decide, but answering it correctly avoids letting a patient believe that they were badly treated.
The WRIST trial recruited 304 adults aged 60 and over with an unstable, isolated distal radius fracture, in 24 health systems in the United States, Canada and Singapore. One hundred and eighty-seven were randomised between volar locking plate (n = 65), external fixator with or without supplementary wires (n = 64) and percutaneous pinning (n = 58); 117 chose non-operative treatment with plaster8.
At 24 months, in 182 patients assessed, the hand scores were as follows: 88 for the volar plate (95 % CI 83 to 92), 83 for the external fixator (78 to 88), 85 for pinning (79 to 90) and 85 for plaster (79 to 90), with no clinically relevant difference after adjustment. Pain did not differ either. Between 12 and 24 months, the scores had hardly changed any further3.
It is in this same study that one of the most counter-intuitive results in the literature is found: the malunion was far more frequent after plaster (59.1 %) than after volar plate (8.0 %), external fixator (17.0 %) or pinning (9.8 %), with a highly significant difference. And yet, this malunion was not associated with the difference in outcome at 24 months3. Surgery restores alignment better, without that better alignment translating into better function in this population.
In adults operated on for a fracture displaced backwards, the British DRAFFT trial randomised 461 patients between percutaneous pinning and volar locking plate. More than 90 % were followed to the end. Both groups regained their function at 12 months, and no clinically relevant difference was found on the PRWE at 3, 6 or 12 months (difference at 12 months of −1.3 points, 95 % CI −4.5 to 1.8, p = 0.398). The number of complications did not differ, and pinning proved cheaper by £7272.
The Cochrane review by Karantana, which included 26 trials and 1,946 adults, is more cautious still, and for a reason of method: all the trials included were at high risk of bias, and allocation concealment was secure in only one. Of the eleven trials comparing pinning and plaster after reduction (917 participants), the quality of evidence was very low for all outcomes reported, to the point that the authors declare themselves uncertain of any effect on patient-reported function9.
Two descriptive figures nevertheless emerge, useful in daily practice: the redisplacement leading to secondary treatment affected on average 12 % of patients in plaster (3.3 to 75 % depending on the series), and the pin-track infection warranting antibiotics affected 7.7 % on average of patients with pinning (0 to 15 %). Two studies also reported finger stiffness after plaster removal that was less frequent after pinning (20 % versus 36 %)9.
A randomised trial published after the review, DRAFFT2, clarifies this finding in 500 adults aged 16 and over, in 36 British hospitals, with a dorsally displaced fracture manipulated under local, regional or general anaesthesia (mean age 60 years, 83 % women, 79 % follow-up). It finds no difference in PRWE at twelve months between moulded plaster and pinning (21.2 versus 20.7; adjusted difference −0.34, 95 % CI −4.33 to 3.66), nor at earlier time points. On the other hand, 33 of 255 patients in plaster (13 %) had to be operated on for loss of reduction in the first six weeks, against a single reoperation in the pinned group21. The 12 % in the review is a mean of six widely dispersed trials; the 13 % in DRAFFT2 is the figure from a single large trial.
| Study | Population | Comparison | Main result |
|---|---|---|---|
| WRIST 20213 | 304 adults aged 60 and over, unstable fracture, 24 centres | Volar plate, external fixator, pinning, plaster | Hand scores comparable at 24 months (88 / 83 / 85 / 85), p = 0.70 |
| UK DRAFFT 20152 | 461 adults operated on, fracture displaced backwards | Pinning versus volar plate | No clinically relevant difference in PRWE at 12 months (difference −1.3; 95 % CI −4.5 to 1.8), pinning less costly |
| Cochrane 20209 | 26 trials, 1,946 adults, mostly older | Pinning versus plaster, and technical variants | Very low evidence, uncertainty about function; 12 % redisplacement in plaster, 7.7 % pin infection |
| DRAFFT2 202221 | 500 adults aged 16 and over, dorsally displaced fracture reduced, 36 hospitals | Pinning versus moulded plaster | No significant difference in PRWE at 12 months (21.2 versus 20.7; 95 % CI of the difference −4.33 to 3.66); 13 % reoperations for loss of reduction in plaster, only one with wires |
| WRIST complications 20198 | 296 adults aged 60 and over | Complication profile by technique | Median nerve compression 18.5 % after plate; pin infection 23 to 26 %; plaster: more complications of all severities (RR 1.88) |
It is in this context, with no clear difference between techniques, that the American Academy of Orthopaedic Surgeons and the American Society for Surgery of the Hand published in 2020 updated clinical practice guidelines on the evaluation and treatment of acute distal radius fractures, following a formalised methodology: seven recommendations based on the best available data, intended to guide practitioners who manage these patients19.
No technique proved superior on what the patient feels. The choice therefore comes down to risk profile, cost and preferences, not to a promise of better function.
Six weeks of waiting are not six weeks of nothing. This is the period when finger mobility, the shoulder and the trophic state are built up, or lost.
A below-elbow plaster leaves the metacarpophalangeal joints, the thumb, the elbow and the shoulder free. Yet these are precisely the segments where the complications that are most costly to recover from develop: stiffness of the metacarpophalangeal joints in extension, stiffening of the shoulder in older people, and oedema that sets in for lack of a muscle pump.
The Cochrane review devoted to rehabilitation after this fracture examined 26 trials and 1,269 patients, mostly older women, distributed across 23 comparisons of which only 4 were covered by more than one trial. Seven of these trials concerned interventions started during immobilisation. The quality of evidence was judged low or very low for all comparisons, which rules out any strong claim11.
What this review reports for the period of immobilisation therefore remains a signal, not a demonstration: very low-quality evidence in favour of hand therapy compared with simple instructions, four days after plaster removal and with an effect persisting for a month; very low-quality evidence of short-term improvement with early occupational therapy; and no difference between supervised and unsupervised exercises in a trial of 96 participants11.
Where the evidence becomes clearer is after osteosynthesis with a volar plate, precisely because the construct is stable. A randomised trial of 30 patients compared immediate mobilisation after surgery with 5 weeks of immobilisation, with assessments at 6 weeks, 9 weeks, 3 months, 6 months and 1 year.
Patients mobilised immediately had better range of motion in the sagittal plane and better grip strength up to 6 months, an advantage in the frontal plane up to 9 weeks and in forearm rotation up to 6 weeks. QuickDASH and PRWE scores were better up to 6 weeks, and the Green and O'Brien score remained significantly different up to one year, with 93 % excellent or good results in this group. Above all, no difference was observed in loss of reduction, nor in pain, duration of physiotherapy or time off work10.
This is a pilot study of 30 patients, level Ib: the signal is consistent and reassuring on safety, but the sample size rules out turning it into a universal rule. The transferable message is this: when the construct is stable and the surgeon allows it, waiting brings nothing and costs mobility.
What we work on during immobilisation, and why
Four areas of work opened before plaster removal, on segments that the plaster does not block
Sources: Handoll HH, Elliott J, Cochrane Database Syst Rev 2015, 26 trials, 1,269 patients (PMID 26403335); Karantana A et al., Cochrane Database Syst Rev 2020, finger stiffness 20 % versus 36 % in two trials (PMID 32032439); Gutiérrez-Espinoza H et al., J Man Manip Ther 2022, manual lymphatic drainage, mean difference of 13.97 to 17.96 ml depending on the time of measurement (PMID 34668847); Chung KC et al., Plast Reconstr Surg 2019 (PMID 31348349).
This is where the literature is most uncomfortable for our profession, and most useful to read honestly.
We must begin with the least flattering data, because ignoring it would amount to selling the patient a promise that the facts do not support.
The analysis of the WRIST trial concerning hand therapy followed patients aged 60 and over, referral and protocol being left to the discretion of the surgeon and the therapist. Eighty per cent of participants received therapy, 70 % combining supervised sessions and home exercises, with on average 9.2 supervised sessions over 14.2 weeks. The results are unambiguous: no difference in patient-reported scores between those who had therapy and the others ; patients who had no therapy even recovered more grip strength ; and those whose therapy was shorter reported better function, better work capacity and greater satisfaction15.
This result calls for a methodological reading, not a rejection. It is a secondary observational analysis within a randomised trial: referral to rehabilitation was not drawn by lot. Patients sent to physiotherapy are, by construction, those who were doing less well, which is enough to produce an unfavourable association without the treatment being the cause. This is confounding by indication in its most classic form. The authors put it cautiously: encouraging the resumption of daily activities might be as effective as formal therapy in these patients.
A later randomised trial asked the same question with the right method. Seventy-four patients over 60 with an extra-articular fracture were allocated between six weeks of supervised physiotherapy (n = 37) and six weeks of a home exercise programme (n = 37), with two-year follow-up and no one lost to follow-up.
The results draw a characteristic curve: at 6 weeks as at one year, the PRWE difference in favour of supervision was 18.6 points (95 % CI 12.8 to 24.3) then 18.5 points (95 % CI 12.7 to 24.2), clinically important gaps. At two years, this difference fell to 3.3 points (95 % CI −2.4 to 9.0), not significant. For secondary outcomes, effect sizes ranged from medium to large at 6 weeks and one year; at two years, only grip strength kept a large effect in favour of supervision13.
The two studies therefore do not really contradict each other: they describe the same thing from two angles. Supervised rehabilitation speeds up recovery substantially during the first year, and this advantage then fades as the control group catches up. This is honest, usable information: what we sell the patient is time gained and fewer months of discomfort, not a different wrist at ten years.
What supervised physiotherapy brings, and for how long
PRWE difference between supervised physiotherapy and home programme, randomised trial of 74 patients over 60
Single source: Gutiérrez-Espinoza H, Gutiérrez-Monclus R, Román-Veas J et al., Effectiveness of supervised physiotherapy versus a home exercise program in patients with distal radius fracture: a randomized controlled trial with a 2-year follow-up, Physiotherapy 2024;124:93-100 (PMID 38875842). The vertical bars show the 95 % confidence interval. A lower PRWE score means better function: the difference here is in favour of supervised physiotherapy.
The mobilisation with movement after Mulligan is applied in supination and in wrist extension; the 2024 American guideline grades manual therapy B in this indication20. A multicentre randomised trial, with concealed allocation, blinding and intention-to-treat analysis, included 67 adults treated in plaster (76 % women, mean age 60 years). The experimental group received the same exercises and advice as the control group, plus the mobilisations.
The gains are clear in the short term: supination 12° greater (95 % CI 5 to 20) at 4 weeks and 8° greater (95 % CI 1 to 15) at 12 weeks ; extension 14° greater (95 % CI 7 to 20), flexion 9° greater (95 % CI 4 to 15), QuickDASH 11 points lower (95 % CI 3 to 18) and PRWE 13 points lower (95 % CI 4 to 23) at 4 weeks. The benefits persisted at 12 weeks for supination, extension, flexion and QuickDASH, and patients in the experimental group more often reported being improved (risk difference 22 %, 95 % CI 5 to 39). No clear benefit remained on patient-reported measures at 26 and 52 weeks, and no adverse effect was reported12.
The manual therapy meta-analysis that followed confirms this profile. Of eight eligible trials, six of which were retained for quantitative synthesis, adding mobilisation with movement to an exercise programme gave at 12 weeks a PRWE lower by 10.2 points (p = 0.02) and a DASH lower by 9.86 points (p = 0.0001), with no significant effect on grip strength (3.9 %, p = 0.25). On the other hand, the comparison between supervised physiotherapy with joint mobilisation and a home programme showed no significant difference in flexion (7.1°, p = 0.20) or in extension (11.99°, p = 0.16) at 6 weeks. Manual lymphatic drainage added to conventional treatment reduced oedema consistently and significantly at all measurement times, by 13.97 to 17.96 ml depending on the time (p between 0.03 and 0.002). The GRADE level of evidence ranged from very low to high depending on the comparison14.
In 2024, two academies of the American Physical Therapy Association, the orthopaedic academy and the hand and upper extremity academy, published a rehabilitation guideline for distal radius fracture. Each recommendation carries a letter that sets both the strength of the evidence and the strength of recommendation: A, strong evidence; B, moderate evidence; C, weak evidence; D, conflicting evidence, no recommendation; E, theoretical evidence; F, expert opinion of the writing group. These letters are not to be confused with the levels of evidence on the cards below, which judge the certainty of the trials themselves20.
The Cochrane review is a useful reminder of the real state of knowledge. Across 23 comparisons, none reaches a moderate or high level of evidence. It finds no clinically important difference between routine physiotherapy or occupational therapy added to home exercise instructions and those instructions alone, across four heterogeneous trials of 30, 33, 66 and 75 participants (low-quality evidence). It does, however, note, in two trials of 46 and 76 participants after volar plate, an advantage for a structured home exercise programme preceded by instructions or support, also on low-quality evidence11.
For modalities taken in isolation against no intervention, the finding is starker still: passive mobilisation, ice, pulsed electromagnetic fields, whirlpools, dynamic extension splint in case of stiffness, none showed a clinically important difference, on very low-quality evidence11. This is an absence of evidence, not evidence of an absence of effect; the 2024 guideline asks a different question, that of physical agents integrated into multimodal management, which it grades B20.
Weak evidence
Weak evidence
Supervised physiotherapy versus home programme, people over 60. Clinically important advantage at 6 weeks and 1 year, erased at 2 years except for strength. A randomised trial of 74 patients, no one lost to follow-up13. The 2024 guideline calls for at least one supervised session per week in people aged 60 and over or with complications (B)20.
Weak evidence
Structured home exercise programme after volar plate. Advantage over unsupervised exercise in two trials (46 and 76 participants)11. Consistent with the fact that guidance matters more than setting.
Weak evidence
Manual lymphatic drainage for oedema. Significant, consistent reduction, of 13.97 to 17.96 ml depending on the time of measurement14. Effect on an intermediate outcome, not on function.
Very low
Routine physiotherapy added to exercise instructions. No clinically important difference against instructions alone, across four heterogeneous trials11.
Not demonstrated
Isolated passive modalities : passive mobilisation, ice, pulsed electromagnetic fields, whirlpools, dynamic extension splint. No clinically important difference against no intervention, on very low-quality evidence11 ; integrated into multimodal management, the physical agents are graded B by the 2024 guideline20.
Key points
The physiotherapist sees these patients more often than the surgeon. It is often the physiotherapist who spots the complication as it sets in.
The analysis of complications in the WRIST trial, covering 296 participants aged 60 and over, gives a precise map of what to monitor depending on the technique used8 :
This last point deserves emphasis, because it contradicts a widespread intuition: non-operative treatment is not the treatment without risk. It shifts the risk, it does not remove it.
This is the complication that causes the most concern, and the one where the physiotherapist has the clearest role as sentinel. The Meena meta-analysis recalls its order of magnitude: a reported incidence of 10.5 to 37 % after distal radius fracture depending on the series and the diagnostic criteria used17, a dispersion that mainly reflects the heterogeneity of definitions.
On prevention with vitamin C, the literature offers a textbook case of two meta-analyses published in the same year and reaching opposite conclusions.
| Study | Trials included | Result | Heterogeneity | Authors' conclusion |
|---|---|---|---|---|
| Evaniew et al., J Orthop Trauma 201516 | 3 trials, 890 patients | Relative risk 0.45 (95 % CI 0.18 to 1.13) | I² = 70 % | No significant benefit; low-quality evidence |
| Meena et al., Eur J Orthop Surg Traumatol 201517 | 4 articles | Risk difference 0.41 (95 % CI 0.19 to 0.92); p = 0.03 | I² = 63 % | Significant reduction; better-quality trials needed |
What should we make of this contradiction? Read it for what it is. The two teams agree on two points: the heterogeneity is high (63 and 70 %), and the overall quality of the evidence is low. Evaniew specifies that the heterogeneity was explained by the diagnostic criteria used, not by the dosage or by the treatment of the fracture, and that the result held up in sensitivity analyses for missing data16. The prescription of vitamin C is for the doctor; the physiotherapist's role is to present this measure neither as established nor as useless, but as debated.
What should raise suspicion of complex regional pain syndrome
These situations call for referral back to the doctor. The management documented in the case reported by Świta combined medication, physiotherapy and active therapy, with notable improvement18, but the authors stress the absence of an established therapeutic standard.
A published case, with its identifier, rather than an invented vignette: what the literature actually documents, including its grey areas.
A 75-year-old woman, untreated osteoporosis, complex regional pain syndrome at three months
Świta M, Szymonek P, Talarek K et al., J Clin Med 2024;13(4) (PMID 38398434). Case report accompanied by a literature review.
History. A 75-year-old patient with untreated osteoporosis presents with a Pouteau-Colles fracture after a fall. Initial management combines reduction and plaster immobilisation of the forearm, followed by percutaneous pinning with a Kirschner wire. She then develops signs of complex regional pain syndrome and is admitted to a rehabilitation unit three months after the fracture. The affected forearm is oedematous, warm, painful, with severely limited range of motion.
Management and course. Treatment combines medication, physiotherapy and active therapy. The patient shows notable improvement over the course of these interventions18.
What this case teaches, and what it does not establish. It illustrates three useful things: the late onset of this complication, which can appear when the fracture has healed; the background, untreated osteoporosis in an older woman, which is a reminder that the sentinel fracture had not triggered any bone management; and the fact that combined management can clearly improve the condition of a patient admitted at three months. On the other hand, an isolated case demonstrates no therapeutic efficacy, and the authors themselves stress the absence of an established treatment standard and the difficulty of predicting the occurrence of this complication.
Why no second clinical case
The searches conducted for this article did not identify, on PubMed, a published case of rehabilitation of a distal radius fracture offering a lesson distinct from this one and verifiable on its record. Rather than complete it with a vignette composed for the occasion, this chapter sticks to the case actually documented. A fabricated observation, even a plausible one, would be exactly the flaw that this site forbids itself.
What the data concretely change in the assessment, in what is said to the patient and in the conduct of sessions.
A progression framework, and what underpins it
The time milestones come from the protocols of the trials cited; the framework itself belongs to clinical reasoning
Milestones drawn from: Quadlbauer S et al., J Wrist Surg 2017, immediate mobilisation after volar plate (PMID 28428911); Reid SA et al., J Physiother 2020, assessments at 4 and 12 weeks (PMID 32291223); Gutiérrez-Espinoza H et al., Physiotherapy 2024, grip strength at 2 years (PMID 38875842); Schmidt V et al., J Hand Surg Eur Vol 2024, follow-up at 11-13 years (PMID 37684021). The assembly into a single progression is a proposal of clinical reasoning, not a recommendation derived from a trial.
Further reading on the site
Far longer than bone healing, and that needs saying from the outset. In the WRIST trial, hand scores were still changing between 12 and 24 months, even though the change did not reach the threshold of significance3. In the eleven-year prospective follow-up, the authors explicitly conclude that recovery is a process that lasts years, and that a decade later, range of motion, strength and QuickDASH had returned to population norms6. In practice, most of it is decided in the first six months, but improvement continues well beyond.
On its own, probably less than it looks. At eleven years, neither dorsal tilt, nor radial inclination, nor ulnar variance, nor joint involvement was associated with clinical outcome or with the risk of osteoarthritis6. In people aged 60 and over, precise restoration of the anatomy was not associated with better results at 12 months4. The important caveat is that the combination of shortening and tilt was accompanied by persistent disability at 2 years in a prospective cohort5, and that this work mostly concerns older patients.
Because it does not depend on the same joint. Pronation and supination are carried by the radioulnar joints, the distal one of which lies directly next to the fracture line. Tilt and shortening of the radius change the relationship of its sigmoid notch to the head of the ulna, which limits rotation at end range7. That is also why mobilisation with movement was studied first and foremost on supination, with a gain of 12° at 4 weeks12.
The question is for the surgeon, but the data allow us to reassure without lying. In patients aged 60 and over, hand scores at 24 months were comparable between volar plate, external fixator, pinning and plaster3. In adults operated on for a dorsally displaced fracture, wires and volar plate did not differ in a clinically relevant way at 12 months2. Each option has its own complication profile8, and it is on that profile, the cost and the patient's preferences that the choice is made, not on a promise of better function.
Yes, with a caveat that must be stated honestly. A randomised trial in patients over 60 showed a clinically important advantage of supervised physiotherapy over a home programme at 6 weeks and at one year, with a gap of about 18 PRWE points, a difference that disappears at two years except for grip strength13. Conversely, an observational analysis of the WRIST trial found no difference between patients who had physiotherapy and those who did not15, but that analysis is exposed to confounding by indication, since referral was not randomised. The 2024 American guideline calls for at least one supervised session per week in people aged 60 and over or in the event of complications20. The reasonable message: rehabilitation speeds up recovery, it does not change the long-term end point.
The question is not settled. Two meta-analyses published in the same year reach opposite conclusions: Evaniew finds no significant reduction in risk across three trials and 890 patients (relative risk 0.45; 95 % CI 0.18 to 1.13; I² = 70 %)16, whereas Meena concludes that there is a significant reduction across four articles (risk difference 0.41; 95 % CI 0.19 to 0.92; p = 0.03; I² = 63 %)17. Both stress high heterogeneity and limited quality of evidence. The decision rests with the prescriber.
None of the sources retained for this article provides a validated criterion for return to sport specific to this fracture, and that should be said rather than inventing a threshold. The reasoning therefore rests on healing confirmed by the surgeon, absence of pain during progressive loading, and recovery of grip strength compared with the healthy side, bearing in mind that it remains the slowest parameter to return: it is the only criterion that kept a significant advantage at two years in favour of supervised physiotherapy13.
It deserves to be reported, never played down. In the WRIST trial, median nerve compression affected 12 of the 65 patients operated on with a volar plate, i.e. 18.5 %, making it the most frequent complication of this technique8. Paraesthesiae in the median nerve territory that appear, persist or worsen after the procedure warrant a medical opinion, all the more so as they may be amenable to a simple procedure if caught early.
Twenty-one references, all verified individually on PubMed through the E-utilities interface: identifier resolved, title, journal, year, pagination and author list checked on the record itself, abstract read before citation. The links point to the PubMed record.
Methods note
The bibliographic foundation was built before writing, subject by subject, by querying PubMed through the E-utilities interface: each identifier was resolved on its XML record, each reference checked for authors, journal, year and pagination, and each abstract read before being cited. Several display years differ from those commonly cited elsewhere, the record being authoritative: Evaniew and Meena are dated 2015 and not 2016, Quadlbauer 2017, Soerensen 2024. The figures given carry their source at the exact place where they are written. The assessments of the level of evidence follow GRADE principles and take up, where they exist, the evaluations published by the Cochrane reviews cited. Two points are explicitly flagged as unresolved by the sources consulted rather than filled with a convenient assertion: the absence of a validated return-to-sport criterion specific to this fracture, and the contradiction between the two meta-analyses on vitamin C. The classification codes in the structured data block were checked against the official tables: ICD-11 NC32.5 on the WHO MMS linearisation, ICD-10 S52.5 on the official ICD-10 browser. Article written on 16 August 2026.