A thumb sprain turns on one question, and one only: has the ruptured ligament stayed in contact with its insertion, or has the adductor aponeurosis slipped between the two? In the second case, the Stener lesion, no orthosis will make it heal, and six weeks of immobilisation will only delay surgery. Here is how not to miss it.

Review written from three diagnostic accuracy meta-analyses (2021), two systematic reviews of interventions, two randomised trials and the reference anatomical and clinical series. Every figure carries its source at the point where it is written; the 40 references are listed at the end of the article with their PMID or DOI.

In brief: clinical summary

What the literature says, in one page, before the detail of the chapters.

Ulnar collateral ligament (UCL) sprain of the metacarpophalangeal (MCP) joint of the thumb is an injury that is commonplace in its frequency and formidable in its capacity to be misclassified. The mechanism is stereotyped: forced abduction-extension of the thumb, most often during a fall in which an object held in the hand, a ski pole, a handlebar, a ball, acts as a fulcrum and forces the thumb column to open. The popular name, "skier's thumb", comes from that.

The difficulty is not recognising the sprain: it is separating the injuries that will heal in an orthosis from those that will never heal. That separation rests on an anatomical fact described by Bertil Stener in 1962: when the UCL ruptures completely at its distal insertion, its free end can flip outwards and above the adductor pollicis aponeurosis, which then interposes itself between the ligament and the bony bed where it should reattach (DOI 10.1302/0301-620X.44B4.869). The ligament is there, it is alive, it is simply on the wrong side of a fibrous sheet. No immobilisation crosses that sheet.

The course of action that follows from this reasoning is simple to state. A sprain with no laxity, or with laxity comparable to the uninjured side, calls for immobilisation in an orthosis, then graded rehabilitation. Marked laxity with a Stener effect calls for surgery: reattaching the ligament has no conservative alternative. Between the two, complete rupture without displacement remains, in the words of a recent French-language review, a "controversial" treatment (PMID 28636299).

What follows sets out this reasoning in the order in which it unfolds in the consulting room: what the injury is, why displacement changes everything, how to test it without error, what each investigation really adds, what becomes of a thumb left unstable, and what rehabilitation can and cannot do.

What are the fundamentals to know about UCL sprain of the thumb?

A two-tier anatomy, a mechanism with a single movement, and an epidemiology that depends entirely on where the counting is done.

A joint entitled to only one useful degree of freedom

The thumb MCP joint is a condylar joint whose main function is stability, not mobility. It transmits to the pulp most of the loads of the thumb-index pinch: every time we turn a key, close a jar or hold a pen, the thumb column resists a force that tends to open it outwards. It is the ulnar structures that absorb that opening.

This ulnar layer is made of two bundles, and that duality is not an anatomist's subtlety: it explains the whole of the diagnostic reasoning set out below.

• The proper collateral ligament, running from the head of the first metacarpal to the base of the proximal phalanx, dorsal to the flexion axis. It tightens in flexion and slackens in extension.
• The accessory collateral ligament, more volar, continuous with the volar plate and the ulnar sesamoid. It tightens in extension.

This distribution has been measured directly. Palmer and Louis examined 750 normal thumbs to establish the physiological radioulnar mobility of the MCP joint in full extension, at 15° of flexion and in full flexion: it is in full flexion that the joint is most stable. In 25 cadaveric thumbs, sectioning the adductor aponeurosis and the capsule produced only minimal instability; it was the additional sectioning of the ulnar collateral ligament that produced marked instability, and it showed up best in full flexion. Sectioning the whole set (aponeurosis, dorsal capsule, proper ligament, accessory ligament and volar plate) produced complete instability in every position tested (PMID 722028).

Running over this ligamentous layer is the adductor pollicis aponeurosis, a tendinous expansion that covers the ulnar side of the joint before joining the extensor apparatus. Its arrangement as a superficial sheet, while the ligament lies deep, is exactly what makes the Stener lesion possible. A recent anatomical study has spelled out the detail: in 37 cadaveric thumbs, the intramuscular tendon of adductor pollicis has three components, dorsal, volar and distal, arranged in a lambda; the dorsal component inserts not only on the aponeurosis but also on the joint capsule, dorsal to the ulnar sesamoid, and the thickened portion of the capsule usually called the proper ulnar collateral ligament has a distinct bony insertion, on the proximal aspect of the lateral tubercle of the proximal phalanx, separate from the adductor insertion (PMID 29395587). This anatomical separation is what allows the ligament to detach and retract under then above a structure that itself stays in place.

One mechanism, and an object held in the hand

The mechanism is a forced abduction and/or hyperextension of the MCP joint. In the surgical series of Chuter and colleagues, covering 127 patients operated on over ten years, at least 66 % of the injuries arose from a hyperextension-abduction mechanism (PMID 19389670).

The role of the ski pole has been studied with rare precision. Engkvist and colleagues compared injured skiers with a control population of 1,619 uninjured skiers: no type of grip in use eliminated the risk of thumb injury, and the frequency was even higher with a grip carrying a large upper platform. The way the skier held the pole in relation to the strap, by contrast, made no difference at all. The explanation the authors propose is mechanical: during the fall, the skier keeps hold of the pole until the last moment before the hand hits the ground; the grip then stays in the palm and acts as a hypomochlion, that is, as a fulcrum around which the thumb is forced into abduction and extension (PMID 7068299).

This observation has a practical consequence that comes as a surprise: prevention does not run through equipment. A review devoted to the subject puts it bluntly: poles without straps do not reduce the incidence of skier's thumb, but teaching the skier to let go of the pole during the fall might reduce the risk (PMID 7740248). Thirty years later, a two-season study in the Dolomites observed that none of the patients with distal upper limb injuries was wearing wrist or thumb protection (PMID 41155774).

An epidemiology that depends on where the counting is done

This is where general figures deserve mistrust, and the literature itself supplies the counter-example that prevents generalisation.

Seen from a ski resort, skier's thumb is a skiing condition: in Engkvist's series, thumb injuries came just after knee injuries among alpine skiing injuries, at 17 % of all injuries, and three quarters of them were UCL injuries of the MCP joint (PMID 7068299). A prospective study run over two consecutive seasons in the Dolomites, between December 2023 and March 2025, finds a comparable profile: of 195 patients attending the emergency department for injury to the forearm, wrist, hand or fingers, 96 (49.2 %) had a fracture and 33 (16.9 %) an isolated UCL injury. Fractures occurred more often on blue runs (56.2 % against 33.3 %, p < 0.001), while UCL injuries were concentrated on red runs (54.5 %) and were frequently linked to the pole being trapped during the fall (PMID 41155774).

Seen from a city trauma department, the picture is reversed. In the British series of Chuter and colleagues, 127 patients operated on for acute UCL rupture over ten years, mean age 40 years (12 to 81 years), male to female ratio 3 to 2, the commonest cause was a simple fall (49 %), followed by sports accidents; skiing accounted for only 2.4 % of cases. The authors say so explicitly in their conclusion: in their population, UCL injuries of the thumb are rarely caused by skiing (PMID 19389670).

The practical lesson is twofold. First, do not require a history of skiing to consider the diagnosis : most of the thumbs operated on in the British series had never seen a ski slope. Second, do not carry rates from one population to another: a practice in a mountain area and an urban practice do not see the same proportions, and there is no single incidence figure valid everywhere.

What the thumb loses when the ulnar layer gives way

The functional consequence of UCL insufficiency is well identified. A systematic review devoted to outcomes after thumb UCL injury states it in its aim: untreated, this injury can lead to reduced pinch strength, pain, instability and osteoarthritis (PMID 23615487). An orthopaedic review confirms this mechanism: UCL injuries are painful and unstable especially during lateral pinch and grip (PMID 36548149).

In other words, the patient does not complain about the thumb when opening the hand: the complaint comes when the thumb is used as a pillar. This is an important practical point for the history, and for the choice of functional follow-up tests.

Why does the Stener lesion change the whole management plan?

Because it changes neither the pain, nor the appearance, nor the mechanism: only the fact that healing has become impossible.

What Stener described

In 1962, Bertil Stener published in the Journal of Bone and Joint Surgery an observation that has organised the management of this sprain for sixty years: when the ulnar collateral ligament ruptures completely at its distal insertion on the proximal phalanx, its free end can retract proximally and flip above the proximal border of the adductor aponeurosis. The aponeurosis, left in place, is then found interposed between the ligament and the bony surface on which it should reattach (DOI 10.1302/0301-620X.44B4.869).

This configuration bears his name. It has one unique consequence, and therein lies all its clinical importance: the ligament can no longer reach its insertion bed, whatever the position of immobilisation and however long it lasts. A review devoted to this lesion puts it thus: interposition of the aponeurosis distinguishes the Stener lesion from other UCL injuries and prevents healing, which makes surgery necessary (PMID 30865860).

What makes this flip possible

The stump flips only under two conditions. First a complete distal rupture : as long as one bundle stays in continuity, the ligament cannot retract enough. Then an abduction wide enough for the free end to cross the proximal border of the aponeurosis, which behaves like a ledge: easy to cross one way, hard to cross back the other.

The anatomical study by Sato and colleagues sheds light on this mechanism by showing that the ligament and the adductor do have distinct bony insertions on the lateral tubercle of the proximal phalanx: the ligament can therefore detach independently, while the adductor stays anchored and keeps its sheet stretched across the field (PMID 29395587). The same study notes that the dorsal component of the intramuscular tendon of the adductor also inserts on the joint capsule, which gives this sheet a stiffness that is underestimated when it is described as a mere "aponeurosis".

How often does it occur? A badly framed question

The question "what percentage of complete ruptures are Stener lesions?" comes up constantly, and it has no single answer, for a methodological reason that must be named: a Stener lesion is only known to be present by operating, and only thumbs judged unstable are operated on. Every available series is therefore a series of patients already selected by a clinical threshold, and the rate they report is that of their threshold, not that of the condition.

What can honestly be written are conditional rates, with their condition:

• Conditional on laxity greater than 35° tested in extension : a Stener lesion was found in 15 of the 17 cases concerned, i.e. 87 %, in the prospective clinical study of Heyman and colleagues (PMID 8519106).
• Conditional on a bone fleck displaced proximal to the joint line on the plain radiograph, the displaced fleck sign: of 228 patients reviewed, 17 (7.5 %) had this sign; all 17 had frank instability with no endpoint and a grade III rupture, and 14 of them (94.1 %) a Stener lesion confirmed intraoperatively. Interobserver reproducibility of the radiographic reading was excellent (κ = 0.94) (PMID 39831878).

These two figures, 87 % and 94.1 %, are positive predictive values of given clinical or radiographic signs, and not the prevalence of the Stener lesion among thumb sprains. Confusing them leads to a common error of reasoning: concluding that "about 9 complete ruptures in 10 are Stener lesions", which none of these data establishes.

What displacement changes for the patient

The difference in prognosis is not played out in the weeks that follow, but in those that will have been lost. A thumb whose ligament is displaced can improve perfectly well in an orthosis: pain decreases, swelling settles, the patient resumes activity. This is not healing, it is a lull. The instability persists and shows itself later, at the pinch, and that is when the patient comes back, several months on, with what will now be called chronic instability.

The French series of Agout and colleagues on chronic instability of the thumb MCP joint opens on exactly this point: the functional result of a severe sprain is good when repair is carried out at the acute stage, but the diagnosis is often missed, leading to chronic instability (PMID 28576699).

How can laxity be tested without error and without making the injury worse?

The valgus stress test is easy to describe and hard to make say anything. Here is the order of operations, the positions, the published thresholds, and why they are not the same from one article to another.

Step one: the radiograph, before touching the thumb

This is the point on which the French-language literature is most explicit. In the review by Poujade and Chick, bilateral comparative clinical examination is the central element of the diagnosis, but it must be preceded by a radiograph (PMID 28636299). The reason is twofold: a displaced avulsion fracture changes management straight away without any test being needed, and forcing a thumb carrying an unstable bone fragment brings nothing but risk.

The radiograph is not a confirmatory investigation, it is a triage investigation. It looks for:

• an avulsion fracture of the base of the proximal phalanx on the ulnar side, present in 21 % of the operated thumbs in Chuter's series (PMID 19389670) ;
• the displacement of that fleck proximal to the joint line, which is the high positive predictive value sign described above (PMID 39831878) ;
• a dislocation, a fracture of the metacarpal head, or any injury that contraindicates testing.

Step two: the position, which decides what the test explores

The thumb MCP joint is not tested in one position but in two, and the two do not inform about the same structure. This follows directly from the two-bundle anatomy described above.

This division is not a convention: it has been measured. On cadaveric specimens, sectioning the proper collateral ligament alone significantly increased valgus instability of the flexed MCP joint, with much less laxity when the joint was tested in extension. When the accessory ligament/volar plate complex was sectioned in turn, instability in extension increased to the point of no longer differing significantly from the values obtained at 30° of flexion (PMID 8519106). Palmer and Louis had established the same hierarchy by a different route, showing in 750 normal thumbs that full flexion is the position of greatest stability, and on cadaveric specimens that sectioning the UCL reveals instability there most clearly (PMID 722028).

Step three: how to hold the thumb, and why rotation matters

The fear of "creating" a Stener lesion by testing a thumb is an old one. It has been tested directly on six fresh cadaveric specimens, with sequential sectioning of the accessory ligament, the proper ligament and then the ulnar sagittal band, and measurement of radial deviation in neutral rotation, in pronation and in supination, at 0° and at 30° of flexion. The result: no Stener lesion was created in any position as long as the fascial origin of the ulnar sagittal band remained intact. After a defect had been created in that band, a Stener lesion was produced in two specimens, but only with the thumb flexed and in supination. Pronation added stability, supination took it away (PMID 28774254).

The practical conclusion applies directly: test in neutral rotation, without supinating the thumb, with a progressive and controlled stress. This is not a precaution of principle, it is the condition under which the test was shown not to be harmful.

Step four: the thresholds, and why there are several

This is where the literature calls for most caution. Three values circulate; they do not come from the same study, they are not measured in the same position and they do not answer the same question.

Three clarifications are worth putting in black and white, because they are what separates quoting a threshold from understanding it.

Heyman's 35° threshold is measured in extension. This is not a detail: in the same study, testing in flexion gives different values for the same injuries. Applying 35° to a test performed in flexion means using a figure outside its measurement setting.

Fricker's two thresholds are cumulative. Their text couples radial deviation under stress greater than 30° and a difference greater than 20° from the uninjured side, not one or the other (PMID 7740248). A thumb that is naturally lax on both sides can exceed 30° without exceeding a 20° difference.

Bilateral comparison is not optional. Palmer and Louis examined 750 normal thumbs precisely because physiological radioulnar mobility varies a great deal between individuals (PMID 722028). This is what the French-language review recalls when it makes "bilateral and comparative" examination the central element of the diagnosis (PMID 28636299).

The endpoint, a qualitative sign worth a figure

Measuring an angle in the consulting room is approximate, and admitting it is more useful than pretending otherwise. The quality of the end of range provides distinct information: in the study of the displaced fleck sign, all 17 patients with this sign had frank instability with no endpoint in the consulting room, and all had a grade III rupture confirmed intraoperatively (PMID 39831878).

A wide opening that stops firmly and a moderate opening with no stop do not tell the same story. The report sent to the doctor therefore gains from describing both: the estimated range, in flexion and in extension, compared with the opposite side, and the quality of the endpoint.

What clinical examination is really worth

The meta-analysis by Rashidi and colleagues, covering 17 studies and 519 subjects, assessed clinical examination, ultrasound and MRI separately. To rule out a UCL rupture, clinical examination reaches a pooled sensitivity of 97 % (95 % CI: 93-99), statistically comparable to that of ultrasound (96 %) and of MRI (99 %), p = 0.3. But to confirm a rupture, its specificity falls to 85 % (95 % CI: 78-91), significantly lower than that of MRI (100 %, p = 0.04) (PMID 33459856).

Translated into practice: a normal test, well conducted and comparative, is a good argument for going no further. An abnormal test, on the other hand, is not enough to decide between the situations that call for an orthosis and those that call for theatre, which is exactly what imaging is there for.

What is each imaging investigation really worth?

Three meta-analyses were published in 2021 on exactly this question. They do not give the same figures, and that is instructive.

The three 2021 meta-analyses, side by side

The subject has the rare distinction of having been synthesised three times in the same year, by three independent teams, with slightly different questions. Putting them side by side is more honest than quoting only one.

Why the figures do not coincide

An attentive reader will notice that the specificity of ultrasound for the Stener lesion is 90 % in Raheman, 94 % in Qamhawi, and that the ultrasound specificity for displaced rupture falls to 72 % in Rashidi. These three figures describe three different things, and it should be said rather than choosing the most flattering one:

• The target is not the same. "Stener lesion" and "displaced rupture" are not exactly synonymous in the primary studies: some count any displacement of the stump, others require aponeurotic interposition seen intraoperatively.
• The reference standard is not the same. Qamhawi accepts as reference either surgical exploration or clinical stability (PMID 33596684) while others include only operated thumbs, which introduces verification bias.
• The primary studies are the same, or nearly so. Three meta-analyses drawing on a common pool of small series do not constitute three independent proofs.
• The underlying level of evidence is low. Qamhawi states it bluntly: the nine ultrasound studies and the six MRI studies included were all at high risk of bias.

Ultrasound: operator-dependent, and not only in principle

Ultrasound has two advantages of its own that neither radiography nor MRI possesses: it is dynamic and it is comparative in real time. A reported case illustrates exactly this use: in a patient attending for conservative management of thumb pain after a fall on the outstretched hand, the radiograph showed an avulsion fracture at the ulnar base of the proximal phalanx, and ultrasound revealed a UCL that was ruptured and displaced proximal to the adductor aponeurosis, the dynamic imaging confirming the displacement. The diagnosis led directly to surgical repair, with no further imaging (PMID 27298646).

Its ultrasound criteria have been formalised. Melville and colleagues reviewed 26 thumbs with surgically proven rupture, 17 complete displaced ruptures, 7 complete non-displaced and 2 partial, and identified two criteria present in all the displaced cases: the non-visualisation of the ligament in its normal position, and the presence of a heterogeneous mass proximal to the MCP joint. Applied at a second reading, these two criteria reached 100 % sensitivity, specificity and accuracy (PMID 23001117).

This 100 % deserves to be read with the caveats the authors themselves give: it is a retrospective re-reading, by the same readers, of criteria derived from the same series. It is a demonstration of internal consistency, not a prospective validation, a difference of the same order as the one separating the fitting of a model from its verification on new data. The closest prospective series included only 10 patients, of whom 2 displaced ruptures were correctly identified (PMID 32930402).

The main ultrasound pitfall is named in a review devoted to interpretation errors in hand and wrist ultrasound: the confusion between the adductor aponeurosis and a displaced rupture (PMID 25148155). That is, precisely, the error that would produce a false positive Stener lesion, and it explains part of the imperfect specificities in the table above.

And what about stress radiographs?

Stress radiographs have long been proposed, with two limitations. The first is that they require a stress applied during the exposure, which raises exactly the questions of safety and reproducibility dealt with in the previous chapter. The second is that the indirect signs proposed have not lived up to their promise: a French anatomical study set out to verify the Rotella and Urpi criterion, the loss of parallelism between the sesamoids and the metacarpal head as a sign of complete rupture, and to establish its relationship with the Stener lesion (PMID 17897862).

The radiographic sign that does have a documented value is static and requires no stress: it is the bone fleck displaced proximal to the joint line on a plain film, with a positive predictive value of 100 % for grade III rupture and of 94.1 % for the Stener lesion, and excellent interobserver reproducibility (κ = 0.94), at the cost of a low frequency, since it was present in only 17 of the 228 patients in the series, i.e. 7.5 % (PMID 39831878). A rare but near-certain sign: when it is there, it saves going further; when it is absent, it says nothing.

Skier's thumb or player's thumb: how to tell acute from chronic?

Two pictures that everyday vocabulary confuses, and whose treatment has almost nothing in common.

The historical term denotes the chronic form

The history of the vocabulary is worth recalling, because it explains the confusion. In 1955, C. S. Campbell described in the Journal of Bone and Joint Surgery an injury of the ulnar layer of the thumb MCP joint in Scottish gamekeepers, in whom the repeated occupational movement progressively stretched the ligament: the gamekeeper's thumb, in French the pouce du garde-chasse (DOI 10.1302/0301-620X.37B1.148). It is therefore a chronic attrition injury that gave the condition its name, and a recent orthopaedic review recalls this 1955 precedence (PMID 36548149).

The term "skier's thumb" came later to denote the acute traumatic form, and usage has ended up applying the two interchangeably. One also comes across "player's thumb" for the forms seen in athletes exposed to repeated valgus loading. This imprecision is not merely lexical: it leads to applying to a chronically lax thumb the reasoning built for a freshly ruptured one, and to immobilising for six weeks an injury that is months old.

What becomes of a thumb whose instability went unnoticed

The French series of Agout and colleagues is the most useful reference on this point, because it follows the three surgical options for chronic instability over seven years. Of 67 patients operated on between 2000 and 2012 for chronic post-traumatic instability of the thumb MCP joint, 55 were reassessed after a mean follow-up of 84 months (24 to 164 months). Of these, 48 (87.3 %) described themselves as satisfied or very satisfied (PMID 28576699).

The detailed results draw a hierarchy that was not necessarily expected:

• Mean QuickDASH : 17.4 (0 to 89.5) after primary repair; 25.7 (0 to 58.3) after ligament reconstruction; 17.8 (0 to 50.0) after arthrodesis.
• Pinch strength relative to the uninjured side : 89 % after repair, 84 % after reconstruction, 94 % after arthrodesis.
• Pain relief : significantly better in the arthrodesis group.
• Persistent instability : 6 of the 10 patients in the reconstruction group remained unstable at final follow-up.

The authors conclude that primary repair should be preferred whenever it is possible and that, contrary to earlier publications, the results of ligament reconstruction were no better than those of arthrodesis (PMID 28576699). In other words: the chronic thumb has no solution as clean as the acute thumb, and the option that best restores strength is also the one that abolishes joint motion.

One nuance: delay is not always a disaster

It would be excessive to conclude that a thumb not repaired within days is lost, and two sets of data forbid it.

The systematic review by Samora and colleagues, covering 14 studies and 293 thumbs (32 treated without surgery, 261 operated on, including 200 acute and 93 chronic injuries, with a mean follow-up of 42.8 months), concludes that there are excellent clinical results after surgical treatment, of acute as well as chronic injuries, with no significant difference between repair and reconstruction respectively. It even specifies that after a long delay or after a failed non-operative treatment, excellent results remain achievable (PMID 23615487).

In professional athletes, delay is sometimes a deliberate strategy. In the NFL series of Werner and colleagues, covering 36 thumbs in 32 players followed from 1991 to 2014, repair, when it was needed, was deferred to the end of the season, and all the players, including those with combined ulnar and radial injuries, played again the following season (PMID 27566241).

These two results do not contradict the urgency of the diagnosis: they shift it. The diagnosis is urgent; surgery is not always. What costs is not operating late, it is never knowing that an operation was needed.

What rehabilitation for a UCL sprain of the thumb?

A short literature, two useful randomised trials, and a table of evidence levels where the empty space takes up more room than the full. Saying so is more useful than filling it in.

The real state of the evidence, before the protocols

The reference systematic review on therapeutic interventions in acute complete rupture of the thumb UCL searched MEDLINE, EMBASE, CINAHL and SPORTDiscus from inception to January 2018. It retained six studies, and describes them all as at high risk of bias. Three were retrospective comparative series of surgical techniques; three were randomised trials, of which two compared rehabilitation regimens in operated patients, cast against early mobilisation, and a new splint against a standard splint (PMID 30057756).

This is little, and the methodological consequence must be drawn: on this subject, a week-by-week rehabilitation recommendation is not supported by trials. What follows therefore distinguishes explicitly between what rests on data and what rests on biomechanical reasoning.

The orthosis: the only question settled by a randomised trial

Sollerman and colleagues randomised 63 consecutive thumbs with a UCL injury of the MCP joint, operated or not, between plaster immobilisation and functional treatment with a splint. The splint allowed flexion-extension of the MCP joint while preventing ulnar and radial deviation of the thumb. Of the 40 thumbs treated without surgery, 21 received a cast and 19 a splint; of the 23 operated on, 10 received a cast and 13 a splint. At a mean follow-up of 15 months (11 to 41 months), there was no difference in stability, range of motion, thumb strength or time off work between the groups. Patients did, on the other hand, judge the splint markedly more comfortable than the cast (PMID 1767639).

This result is old, 1991, but it remains the only randomised trial directly comparing the two forms of immobilisation, and its conclusion is consistent with what is seen elsewhere in upper limb trauma: at equal effectiveness, the removable orthosis wins over rigid immobilisation on comfort and adherence.

The second support is biomechanical. Gil and colleagues tested, on 10 fresh cadaveric hands, a radial-based orthosis stabilising the MCP joint, made by a certified hand therapist, under increasing valgus loads of 20, 40, 60, 80 and 100 N applied 15 mm distal to the joint. The orthosis significantly reduced the mean abduction angles at every load applied, even though it leaves the interphalangeal and trapeziometacarpal joints free (PMID 28711411).

The published time frames, and where they come from

Two sources give an explicit timetable, and it is useful to quote them as they stand rather than produce an average of them.

Fricker and Hintermann propose: controlled active range-of-motion exercises from the 3^rd-4^th week (after the injury for conservative treatment, after the operation for those operated on), the protective splint continued until the 6^th week, and unrestricted use from the 12^th week (PMID 7740248).

Poujade and Chick indicate, whatever treatment is chosen, a rehabilitation started as early as the 4^th week to limit stiffness, with return to sport depending on the stage of the sprain and on the possibility of wearing a rigid splint (PMID 28636299).

These two timetables are authors' proposals, consistent with each other, published in peer-reviewed journals but not derived from a comparative trial. Presenting them as a validated protocol would credit them with more than they carry.

After surgical repair: early mobilisation has been tested

This is one of the rare points where a randomised trial, small though it is, exists. Crowley and colleagues compared, after UCL repair with a Mitek bone anchor, an early active mobilisation regimen with standard immobilisation in a thumb spica orthosis for 4 to 6 weeks. The results of this pilot trial: return to full hand function on average at 6 weeks against 8, return to work at 7 weeks against 11, and no difference in final range of motion (PMID 23970193).

The reach of this result is real but bounded: a pilot trial, a small sample, and conditional on repair with a bone anchor, the authors relying explicitly on biomechanical work suggesting that this type of construct is strong enough to tolerate early controlled active mobilisation. It extrapolates neither to other constructs nor to thumbs treated without surgery.

The caveat deserves to be stated in both directions. An orthopaedic review recalls that no repair or reconstruction technique using native tissue restores strength equivalent to that of the ligament before injury and that augmentation with suture tape gives excellent short-term results with an earlier return to function, though the literature is lacking in the long term (PMID 36548149). The physiotherapist who receives an operated thumb therefore has every interest in knowing the construct used before choosing a progression.

Conservative treatment of a complete rupture: the grey zone

This is the point on which the literature is most uncomfortable, and where honesty means not deciding in its place.

On one side, the systematic review by Samora and colleagues writes that non-operative treatment often failed, requiring surgery, a finding based on 32 thumbs treated without surgery against 261 operated on, an imbalance that itself reflects prevailing practice (PMID 23615487). On the other, the French-language review explicitly describes as "controversial" the therapeutic choice in cases of laxity without a Stener effect, and reserves the mandatory surgical indication for laxity with a Stener effect (PMID 28636299).

What the physiotherapist can do with this, in concrete terms: do not promise a patient with a complete non-displaced rupture that the orthosis will be enough, and set a dated reassessment criterion from the outset. Laxity that persists on comparative testing at the end of the immobilisation period is not a failure of rehabilitation, it is information that belongs in the record and that justifies going back for a surgical opinion.

What rehabilitation does, in order

For want of dedicated trials, the progression below is built from the published time frames above and from the anatomical reasoning of the earlier chapters. It is presented as such, and not as a validated protocol.

Modalities and level of evidence

The table below grades each modality using GRADE logic, not by how reasonable it seems, but by what the retrieved literature actually supports.

When and how to return to sport and to work?

A very high return to sport rate, a spread of time frames that is just as high, and a decisive criterion that is not elapsed time.

What the return-to-sport data show

The most complete systematic review on the subject collected 23 articles, 11 reporting patients and 12 giving expert opinion, for a total of 311 patients operated on for an injury of the ulnar ligament complex of the thumb. Return to sport reached 98.1 % in aggregate, no drop in performance was found after surgery, and 32 patients (10.3 %) had a postoperative complication. The methodological quality of the included studies was modest, with a mean MINORS score of 9.4. Above all, the authors note that the recommended return times vary by sport and by author, while all recommend initial immobilisation of the thumb (PMID 37323971).

This spread is found again in the professional series. Of 23 American football players operated on for a thumb UCL injury, 22 (95.7 %) played again in the NFL, after a mean delay of 132.2 days, with a standard deviation of 126.1 days. One-year career survival was 87.0 %, with no significant difference in games per season, career length or performance compared with matched control players (PMID 29480741).

This standard deviation, almost as large as the mean, deserves emphasis: it means that the time to return is essentially individual, and that no mean value amounts to a prescription.

The criterion that decides, and those that do not

Three elements condition the return, and their hierarchy is not the one people expect.

Stability, first. A thumb that stays lax on comparative testing is not ready, whatever the number of weeks that have passed. It is also the criterion that decides whether to go back for a surgical opinion rather than continue rehabilitation.

The possibility of protecting, next. The French-language review is explicit: return to sport depends on the stage of the sprain and on the possibility of wearing a rigid splint (PMID 28636299). That is why a rugby or American football player sometimes returns before a skier or a climber: in some sports a rigid orthosis is tolerated by the rules and by the movement; in others it is not.

Pinch strength, finally, as the criterion for resuming maximal loading. It is the parameter that the surgical series measure when they assess their results: 89 %, 84 % and 94 % of the uninjured side depending on the technique in Agout's series (PMID 28576699). Measuring it with a dynamometer, comparatively, gives the record an objective criterion where the patient's own sense discriminates poorly.

Return to work, which does not obey the same rules

Return to work is the most concrete criterion for most patients, and it was measured in two of the available trials. Sollerman and colleagues found no difference in time off work between cast and functional splint (PMID 1767639). Crowley and colleagues, after repair with a bone anchor, observed a return to work at 7 weeks with early active mobilisation against 11 weeks with standard immobilisation (PMID 23970193).

These figures describe populations, not jobs. A thumb does not recover at the same pace depending on whether it turns a screwdriver all day, types on a keyboard or holds a motorbike handlebar. The occupational history, which movements, how often, what scope for adjustment, is more informative here than the calendar.

What does a growing skeleton change in children and adolescents?

In children, load is directed preferentially towards bone rather than ligament, and the management that follows is more permissive than one might imagine.

An injury that changes site

In adults, forced abduction ruptures the ligament or avulses a small bone fragment at its insertion. In children, the same load finds a different weak point, and a purely ligamentous injury is unusual enough to have been the subject of a case report entitled, precisely, " True skier's thumb in childhood ", a true skier's thumb in a child (PMID 11890924). The title says the essential: in this age group, the expected injury is bony, and the ligamentous injury is the exception worth reporting.

This difference in site has an immediate practical consequence: in children the radiograph has a diagnostic value that it does not have in adults, where it mainly serves to rule out. The avulsion fracture of the base of the proximal phalanx is the usual presentation there.

What the only available comparative series shows

Huynh, Tang and Cheung retrospectively reviewed all patients under 18 managed at their centre for an avulsion fracture of the thumb UCL, that is, 47 children, grouping them by initial treatment and measuring the time to return to full activity (PMID 30857433). The results are counter-intuitive and deserve to be read in full:

• 10 children operated on at the outset : all united without complication, with a mean recovery time of 6.2 ± 1.8 weeks.
• 37 children treated conservatively : most united, but with a longer and far more variable time, 9.2 ± 8.3 weeks, adjusted hazard ratio 2.3 (95 % CI: 1.0-5.5).
• 4 cases (11 %) required conversion to surgery; all then united without complication.
• Above all: 19 fractures that met the usual criteria for a surgical indication were treated with immobilisation and united successfully in 7.1 ± 2.9 weeks.
• The characteristics of the fracture, articular surface involvement, displacement, rotation, did not predict which children would fail conservative treatment or which would have a prolonged recovery.

The authors conclude that, while surgery provides definitive treatment without delaying recovery, immobilisation is not an unreasonable choice even in the presence of a large or displaced fracture, and that in the event of failure, second-line surgery gives good results (PMID 30857433).

What that changes for the physiotherapist

• A far greater variability in time frame : a standard deviation of 8.3 weeks around a mean of 9.2 means that some children take several months. Announcing a fixed time frame to parents is preparing to be contradicted.
• Active monitoring of union : 11 % of conversions to surgery, and no initial characteristic allows them to be anticipated. Clinical and radiographic follow-up beats prediction.
• Rehabilitation aimed at use, not performance : in children, the return of the hand to play and to school is a better indicator than a strength measurement.
• Particular vigilance for the persistently painful thumb : the true ligamentous injury does exist in children, even if it is rare (PMID 11890924).

What do real clinical cases teach us?

Four published observations, each chosen because it defeats a different certainty. All are genuine indexed case reports, cited with their identifier.

Case 1: the Stener lesion found in a patient referred for conservative care

The situation. A patient attends for conservative management of thumb pain that arose after a fall on the outstretched hand. The plain radiograph shows an avulsion fracture on the ulnar side of the base of the proximal phalanx (PMID 27298646, full text open access: PMC4879008).

What was done. Diagnostic ultrasound shows an ulnar collateral ligament ruptured and displaced proximal to the adductor aponeurosis, that is, a Stener lesion. The dynamic imaging confirms the displacement of the completely ruptured ligament. Surgical repair follows the diagnosis.

What this case defeats. The idea that a patient referred for conservative treatment has already been triaged. The authors note that ultrasound made it possible, in this case, to reach an exact diagnosis while avoiding any further imaging, and that early intervention allowed an optimal result. A thumb arriving in rehabilitation has not necessarily been tested, or imaged.

Case 2: the ulnar mass that was not a Stener lesion

The situation. A woman of 63 presents with thumb instability, pain, and a small palpable mass on the ulnar side of the metacarpophalangeal joint, the classic clinical expression of the ligament stump folded above the aponeurosis (PMID 37116280, full text open access: PMC10163635).

What was found. Intraoperatively, the mass proved to be a collection of granulation tissue, and not a displaced ligament. The rupture was indeed present but lay at the proximal insertion of the ligament and not at its usual distal insertion. The authors call this presentation a pseudo-Stener lesion. The ligament was repaired and the patient resumed her daily activities without restriction after six weeks.

What this case defeats. The absolute value given to the palpable mass. It remains a major warning signal, but it is not pathognomonic: what is felt under the finger is not always the ligament. Management does not change, this thumb had to be referred, but what is said to the patient must remain cautious.

Case 3: two-level injuries, and the argument for a systematic approach

The situation. Three rare cases of skier's thumb with a two-level injury, reported by a Swiss hand surgery team (PMID 39697532, full text open access: PMC11650722).

What the authors draw from it. They set out the diagnostic difficulties, the surgical treatments and the anatomical explanations of these forms, and conclude explicitly on the importance of a systematic diagnostic approach in any skier's thumb type injury, comprising clinical examination, radiography and ultrasound.

What this case defeats. The idea that a commonplace mechanism produces an unambiguous injury. The authors' conclusion is also the practical conclusion of this article: it is not the rarity of a form that justifies a systematic approach, it is the impossibility of knowing in advance which one you are dealing with.

Case 4: what imaging sees when the clinical picture hesitates

The situation. An imaging observation published in a physical and rehabilitation medicine journal, under the title "Stener lesion: an uncommon variant of skier's thumb" (PMID 31464751).

What this case adds. It is a reminder that this injury belongs to the catalogue of situations encountered by rehabilitation practitioners, and not only by hand surgeons. The presence of this image in a PRM journal is in itself an indication of the real patient pathway: many of these thumbs pass through rehabilitation before the full diagnosis is made.

How is this applied in practice?

What is checked at the first appointment, what is written to the doctor, and what to think of when it is not a UCL sprain.

The first assessment checklist

What should be written in the report sent to the doctor

A useful report fits in five lines and avoids wording that commits to nothing. Concretely:

• The estimated valgus range, in flexion and in extension, always compared with the uninjured side, since the published thresholds differ by position and one of Fricker's two criteria is a difference, not an absolute value (PMID 7740248).
• The quality of the endpoint, described in words: firm, soft, or absent.
• The presence or absence of a palpable ulnar mass.
• Whether or not the test was interpretable, and why.
• The precise question being asked : "should this be completed with an ultrasound?", rather than "at your disposal".

When it is not a UCL sprain

A painful thumb or hand after injury, or with no clear injury, may be something else entirely. The corpus of this site covers several of these diagnoses in detail.

Two further situations are worth keeping in mind, because they accompany UCL sprain rather than replacing it. The first is the combined injury of the ulnar and the radial collateral ligament : it accounted for 25 % of the thumb MCP ligament injuries in an American football series followed over 23 seasons, and all required surgery; the authors stress that it is probably frequently missed, and recommend testing both sides (PMID 27566241). The second is the associated dorsal capsular injury, found in 57 % of the operated thumbs in a series of 127, with dorsal capsular invagination in 29 % of cases (PMID 19389670).

Three messages to give the patient, and one not to give

• "The orthosis is not judged on pain." The disappearance of pain does not mean healing; the criterion for coming out is stability on comparative testing.
• "The thumb is reassessed at the end of the immobilisation." Setting the reassessment from the outset stops it being experienced as a failure if it leads to a surgical opinion.
• "Strength comes back after stability, not before." The functional complaint concerns the pinch, and it is the last parameter to recover.
• Not to give: a firm, numbered date for return to sport. In operated professionals, the mean time to return to play was 132 days with a standard deviation of 126 days (PMID 29480741) : announcing a date means announcing a disappointment or a premature return.

Frequently asked questions

References

All the references below have been verified at source: an existing PubMed PMID or CrossRef DOI, and the content checked to make sure it does support the statement it is attached to in the text.