Relative motion flexion splinting for the rehabilitation of flexor tendon repairs: A systematic review

Abstract

Introduction

Relative motion splinting has been used successfully in the treatment of extensor tendon repairs and has recently been applied in flexor tendon rehabilitation. The purpose of this systematic review was to identify articles reporting use of relative motion flexion (RMF) splinting following flexor tendon repair and to examine indications for use and clinical outcomes.

Methods

Seven medical databases, four trials registries and three grey literature sources were systematically searched and screened against pre-specified eligibility criteria. Screening, data extraction and quality appraisal were independently performed by two reviewers.

Results

A total of 12 studies were identified, of which three met the review eligibility criteria: one retrospective case series; one cadaveric proof of concept study; and one ongoing prospective case series. The type of splint (including metacarpophalangeal joint position and available movement), exercise programme, and zone of tendon injury varied between studies. Both case series presented acceptable range of movement and grip strength outcomes. The prospective series reported one tendon rupture and two tenolysis procedures; the retrospective series reported no tendon ruptures or secondary surgeries.

Discussion

We found limited evidence supporting the use of RMF splinting in the rehabilitation of zones I-III flexor tendon repairs. Further prospective research with larger patient cohorts is required to assess the clinical outcomes, patient reported outcomes and safety of RMF splinting in comparison to other regimes. Application of the relative motion principles to flexor tendon splinting varied across the included studies, and we suggest an operational definition of relative motion in this context.

Introduction

Flexor tendon injuries have inter-related impacts on both the patient’s quality of life and hand function. ¹ Rehabilitation programmes for individuals who have undergone flexor tendon repairs must balance the need to protect the repaired tendon from rupture or attenuation, with the need to promote tendon gliding in order to reduce the risk of secondary complications. Advancement in surgical repair techniques and materials have enabled early active motion following flexor tendon repair, yet there is a lack of high-quality evidence to inform best practice for post-operative rehabilitation.^2–5

Existing systematic reviews exploring flexor tendon rehabilitation have assessed different approaches to splinting, for example including or excluding the wrist, ⁵ or different exercise programmes, for example passive or active mobilisation.^3,4,6 However, none of these reviews were able to draw firm conclusions. A recent Cochrane review of flexor tendon rehabilitation also highlighted a lack of high-quality evidence to support any specific rehabilitation strategies. ⁷ Current British Society for Surgery of the Hand (BSSH) and British Association of Hand Therapists (BAHT) flexor tendon standards recommend the use of a controlled early active mobilisation regime. ⁸

It has been proposed that the combination of a functional relative motion flexion (RMF) splint and wrist splint ⁹ could be used as an alternative to the more traditional Belfast splint regime ¹⁰ or Manchester short splint. ¹¹ Unlike the Belfast and Manchester splints, the RMF splint is finger-based permitting protected finger motion and functional use of all digits immediately. ⁹ The concept of relative motion applies to muscles that control multiple tendons, such as the extensor digitorum communis (EDC) or flexor digitorum profundus (FDP). The premise is that tendon excursion can be affected by changing the position of one or more metacarpophalangeal joints(s) (MCPJ) relative to those of the other fingers. ¹² This permits active movement of all fingers and all joints, including the MCPJs, through a wide range of motion. Excursion of the repaired FDP tendon is influenced by placing the MCPJ of the affected finger(s) in relatively more flexion than the other fingers. In the splinted position the more flexed MCPJ of the injured digit(s) offloads the repaired FDP tendon(s) levying the ‘quadriga effect’. ⁹ Application of the relative motion concept for management of extensor tendon repairs has been successful using the relative motion extension (RME) splint, whereby the MCPJ of the affected finger(s) is positioned in relatively more extension than the non-injured fingers.^13–18

The aims of this systematic review were to explore the published and unpublished (grey) literature and examine: (i) indications for use of a RMF splint following flexor tendon repair; (ii) reported splint designs; (iii) accompanying rehabilitation exercise programmes (including hand function); and (iv) the clinical and patient reported outcomes.

Methods

The review protocol was pre-registered with PROSPERO (CRD42020197169) ¹⁹ and completed using the PRISMA guidelines. ²⁰ Literature searching was undertaken between August-September 2020, and eligible studies were those detailing the use of RMF splints following flexor tendon repair. Full eligibility criteria are listed in Table 1.

Table 1.

Review eligibility criteria.

Inclusion	Exclusion
1. Studies reporting the use of a relative motion flexion splint following flexor tendon repair in the hand (all zones of injury and full or partial repairs).2. Any study design. 3. Report available in English language or translation available without charge.	1. Studies reporting flexor tendon rehabilitation methods other than those described as a relative motion flexion splint. 2. Systematic reviews or other literature reviews. 3. Studies reporting the use of the relative motion extension splint.

Search strategy

Seven medical databases were searched (Medline, Scopus, AMED, PEDro, EMCARE, EMBASE, CINAHL), using the search terms provided in Table 2. Title, abstract and keyword search settings were used, with subject headings, if available. The search strategy was piloted and refined in Medline. Hand Therapy and Journal of Hand Therapy were hand searched for relevant articles, as were the reference lists of identified papers. Grey literature sources were also searched to identify relevant unpublished and non-peer reviewed material, using: TRIPdatabase; OpenGrey; a naïve Google search for the first 25 hits; and clinical trials databases (clinicaltrials.gov, clinicaltrialsregister.eu, lsrctn.com, Cochrane Central Register of Controlled Trials). Grey literature search terms were ‘relative motion flexion’ and ‘flexor tendon’. All searches were conducted in the English language.

Table 2.

Review search terms.

Population	Intervention (surgical)a	Intervention (rehabilitation)a
Flexor AND tendon*	Injur*OR repair*OR surger*	Relative AND motion AND flexion AND (splint OR orthos*) ORYoke

^aThe * symbol denotes the search function that includes the specified word ± additional letters.

Duplicate manuscripts were removed before screening. Title, abstract, and full text screening were independently performed by two reviewers (RR and LN) using the eligibility criteria presented in Table 1. Cases of disagreement were resolved by discussion. Authors were contacted by email where additional information was required to determine eligibility.

Data extraction and methodological quality assessment

Data were extracted using a pre-piloted form that had been developed by LN and refined by RR (online Appendix 1). Methodological quality assessment was completed using the Joanna Briggs Institute (JBI) critical appraisal tool for case series, ²¹ with an additional question to ensure appropriate ethics approval (online Appendix 2). Data extraction and methodological quality assessment were completed independently by RR and LN, and any disagreements were resolved by discussion.

Results

Study characteristics

The PRISMA flow diagram is illustrated in Figure 1. During screening, four authors were contacted with a request for additional information, of these, two replied. One respondent did not have any data regarding RMF splinting following flexor tendon repair, ²² while the other was able to provide preliminary data from an ongoing randomised controlled trial (RCT). ²³ In total, 12 unique studies were identified, of which three met the review inclusion criteria: a retrospective case series, ⁹ a cadaveric proof of concept study, ²⁴ and a prospective case series from the RCT discussed above. ²³

Figure 1.

PRISMA flow diagram.

All studies were undertaken in the USA, published between 2019 and 2020 and had a single-site design. A combined total of 19 flexor tendon patients (20 digits) and four cadaver models were involved. Participants in the Henry and Howell case series were predominantly male (7/10), with a median age of 20 years, range 12–49 years. ⁹ Ethnicity details were not reported. Age, sex and ethnicity information was not available in the data provided by O’Connell et al. ²³ Study details are provided in Table 3.

Table 3.

Studies reporting use of relative motion flexion (RMF) splinting following flexor tendon repair.

Primary author, year, location	Study Design	Number of participants	Zone of injury	Type of injury	Surgical details	Splint type	Mediana days between injury and surgery (range)	Mediana days between surgery and therapy (range)
Henry, 2020, USA	Retrospective case series	10 (10 digits) 2 lost to follow-up	Zone I/IIFDPSingle digit	Sharp [n = 6] Ragged [n = 2] Staged reconstruction [n = 1] Type IV FDP avulsion [n = 1]	General or regional anesthetic Pulleys vented, FDS not repaired4 strand repair using FiberWire™ with epitendinous suture [n = 9] FDP to distal phalanx tie-over button [n = 1]	Two splints: i) Wrist splint for 3 weeks full time, either prefabricated (wrist 0–20° extension, n = 7) or forearm dorsal blocking splint (wrist 0-20° flexion, n = 3) ii) RMF splint for 8–10 weeks, affected finger in 30-40° relative MCPJ flexion	6 (0-23)b [n = 9]	6 (5-20)
Chung, 2019, USA	Cadaver study	4	Zone III FDPCadaver	Sharp [n = 4]	Single, simple interrupted 6-0 nylon suture	Two splints: i) Forearm dorsal blocking splint, all MCPJs 50° flexion, wrist fixed in 30° extension ii) RMF splint, affected digit in 15-25° relative MCPJ flexion	Not applicable	Not applicable
O’Connell, 2020, USA	Prospective case series (ongoing)	9 (10 digits) 1 lost to follow-up	Zone II/IIIFDP/FDS1–2 digits ± digital nerve injury	Not reported	General or regional anesthetic	Three splints: i) Day: wrist splint for 3 weeks (wrist 0° extension) ii) Day: PEAMO splint, MCPJ extension of the affected digit(s) restricted to 50-70° flexion. Full time for 3 weeks, discarded at 7 weeks. iii) Night: dorsal blocking splint for 6 weeks, wrist 0° extension, MCPJs 30-50° flexion, IPJ 0° extension	Not reported	3 (2-5)

FDP: flexor digitorum profundus; FDS: flexor digitorum superficialis; IPJ: interphalangeal joint; MCPJ: metacarpophalangeal joint; PEAMO: protected early active motion orthosis; RMF: relative motion flexion.

^aCalculated by the review authors.

^bAn additional participant underwent a staged tendon graft, time from their initial injury to repair = 217 days.

Methodological quality assessment

The retrospective case series by Henry and Howell ⁹ largely met the JBI quality criteria, however, the timing of range of motion and grip strength measurements were not standardised and there were no functional or patient reported outcomes. The biomechanical cadaveric research performed by Chung et al. ²⁴ was a proof of concept study and therefore not all of the JBI criteria were applicable. A standardised testing procedure was reported, although there was a lack of information regarding tendon quality. The preliminary data provided by O’Connell et al. ²³ was rated as ‘unclear’ for seven of the 11 quality assessment criteria, including participant selection and demographics, and consecutive inclusion. However, the research is ongoing and data were provided in a summary format for the purposes of this review. Methodological quality assessment scores are provided in Table 4.

Table 4.

Methodological quality assessment scores using the Joanna Briggs Institute (JBI) case series critical appraisal tool.

JBI case series critical appraisala	Henry 2019	Chung 2019	O’Connell 2020b
1. Ethics approval	Yes	Yes	Yes
2. Clear criteria	Yes	Unclear	Unclear
3. Standard measurements	Yes	Yes	Yes
4. Valid methods used for identification of condition	Unclear	Not applicable	Unclear
5. Consecutive inclusion	Yes	Not applicable	Unclear
6. Complete inclusion of participants	Yes	Not applicable	Yes
7. Reporting of demographics	Yes	Not applicable	Unclear
8. Clear reporting of clinical information	No	Unclear	Unclear
9. Outcomes clearly reported	Unclear	Unclear	Yes
10. Clear reporting of sites/clinics demographics	Yes	Not applicable	Unclear
11. Appropriate statistical analysis	Yes	Yes	Unclear

^aThe JBI tool was modified to include a question on ethics approval.

^bRelates to a subset of data from an ongoing study. Items were marked as unclear, rather than no, where information was missing from this preliminary dataset.

Indications for relative motion flexion splinting after flexor tendon repair

Henry and Howell included patients with single-digit flexor tendon repairs in zones I or II. ⁹ When both the FDP and flexor digitorum superficialis (FDS) were involved, only the FDP tendon was repaired. Eight tendons underwent primary repair, there was one two-stage reconstruction and one type IV FDP avulsion. Surgical repairs were predominantly 4-strand core suture followed by a peripheral suture and pulley venting at the discretion of the surgeon. O’Connell et al. included patients with single or two-digit FDP and/or FDS injuries in zones II and/or III. ²³ Mechanism of injury, type of repair, and other surgical details were not available in the preliminary data. For their biomechanical study, Chung et al. sharply lacerated FDP tendons in zone III and repaired each with a single simple interrupted nylon 6–0 suture. ²⁴ Measurement of tendon elongation and observation of tendon integrity were performed in intact and repaired FDP tendons both in and out of a RMF splint after cyclic loading at ≥11 Newtons (N).

Splint design and schedule for use

While all three studies reported use of the relative motion flexion concept, there were variations in splint design, including wrist and MCPJ position, and the relative difference between repaired and uninjured MCPJ flexion angles. The schedule of splint wear also varied.

Henry and Howell reported the use of two different splints and wrist positions: the initial three (out of ten) patients had their wrists immobilised in 0 to 20° flexion within a custom forearm-based dorsal blocking splint, while subsequent participants were immobilised in a prefabricated wrist splint with the wrist positioned in 0 to 20° extension. ⁹ The MCPJ of the affected digit was placed in 30–40° more flexion than the neighbouring MCPJs using a 4-finger RMF splint (Figure 2). Wrist splint wear was reduced to night and during ‘at risk’ activities at three weeks, and was discontinued at six weeks postoperatively. Discontinuation of the RMF splint occurred between 8–10 weeks postoperatively.

Figure 2.

Representation of the relative motion flexion (RMF) splint described by Henry and Howell.The affected middle finger metacarpophalangeal joint (MCPJ) is positioned in 30-40? more flexion than the other MCPJs and is paired with a prefabricated wrist splint to immobilise the wrist in 0-20? extension. Thermoplastic material provided by Orfit.

O’Connell et al. used a combination of two splints during waking hours. A custom wrist immobilisation splint (wrist neutral) was worn full-time for the initial three weeks postoperatively in conjunction with a thermoplastic tape strap, which blocked the affected finger in approximately 50–70° of MCPJ flexion ²³ (Figure 3). The strap, referred to as a ‘protected early active motion orthosis’ (PEAMO) was worn full time for six weeks postoperatively (apart from being removed for exercises after three weeks). At night, a custom forearm-based dorsal blocking splint was worn for the first six postoperative weeks, positioning the wrist and interphalangeal joints in neutral and MCPJs in 30–50° flexion.

Figure 3.

Protected early active motion orthosis described by O'Connell et al.The Protective Early Active Motion Orthosis (PEAMO) blocks end range metacarpophalangeal extension of the affected middle finger at 50-70? and is worn in combination with a thermoplastic wrist immobilisation splint positioned in neutral.Image courtesy of April O?Connell.

The splint design used by Chung et al. included fixation of the cadaver’s wrist in 30°extension, a custom forearm-based dorsal blocking splint positioning all MCPJs in 50° of flexion, combined with a RMF splint positioning the MCPJ of the digit being tested in 15–25° more flexion relative to the other MCPJs. ²⁴

Exercise programme and recommended functional hand use

The exercise programme, frequency and duration of therapy for each patient was not reported by Henry and Howell, however, semi-structured interviews with the treating therapists, described individualised, patient-centred management. ⁹ It was reported that the surgeon instructed all patients to perform passive combined IPJ flexion, passive IPJ extension with the MCPJ in flexion, and active range of movement with all fingers in the RMF splint; exercise dosage was not reported. Patients were advised to use their operated hand for light function during the first three weeks postoperatively and not to lift or grip strenuously. After three weeks, both hands could be used to lift a ‘light’ bag, and ‘at risk’ activities such as jogging were allowed. Bilateral hand use to the equivalent of 3.5 kg was allowed from six weeks postoperatively. All restrictions to hand use ended between weeks 8–10 after surgery.

O’Connell et al. described a detailed postoperative exercise programme which commenced at day 3–5 postoperatively with both splints in situ. ²³ Exercise dosage was 20 repetitions, twice hourly for all fingers and consisted of passive flexion and active extension, plus gentle passive IPJ extension with the MCPJ in flexion. Exercises added from day six, required removal of the wrist splint for active wrist extension with tenodesis flexion of all fingers to half a fist for repetitions, thrice daily. This was increased to two thirds of a fist from 14 days postoperatively. At day 21, wear of the wrist splint was discontinued and the PEAMO was taken off only for exercises, including isolated active IPJ flexion/extension and ‘active tendon gliding’ for three sets of 10 repetitions, five times per day. Very light function (for example gentle use of a mobile phone touchscreen, but not holding the phone) was permitted from day 3–5 with both splints in situ. From day 28, light functional activities of daily living (ADLs) were commenced in the PEAMO. At day 42, the nocturnal wrist and PEAMO splints were discarded and replaced with buddy straps, which were used as a reminder to avoid heavy ADLs, and blocked IPJ flexion exercises were commenced. Light strengthening began from day 49.

In the biomechanical proof of concept cadaver study, each studied digit was subjected to a minimum load of 11 N while all fingers were passively flexed and extended for 25 cycles. ²⁴

Clinical assessments and patient-reported outcomes

Clinical outcomes for all three reviewed studies are presented in Table 5. Henry and Howell reported range of movement, grip strength and duration of work absence, however data were not available for all patients and were measured at different time points. ⁹ Total Active Motion (TAM) grading ²⁵ was Excellent (n = 4), Good (n = 1) and Fair (n = 3) and grip strength ranged between 63–100 percent of the uninjured hand. There were no patient reported outcome data.

Table 5.

Clinical and patient reported outcomes for relative motion flexion (RMF) splinting following flexor tendon repair.

AuthorYear	Median outcome time point (range)	Range of movement	Median days from surgery to return to work (range)	Mediana grip strength in pounds (range)	Number of tendon ruptures	Other complications reported	Function
Henry2020	10.5 months(5–72)	TAMb Excellent = 4Good = 1Fair = 3	80 (7–112) [n = 5/8]	% of uninjured side100 (63-107) [n = 6/8]	None	None [n = 0/8] (no secondary surgeries or PIPJ flexion contractures)	Not reported
Chung2019	Immediate	Not applicable	Not applicable	Not applicable	None, gapping <2mm	Not applicable	Not applicable
O’Connell 2020	12 weeks(12–16)	TAMa,bGood = 4Fair = 2Poor = 3c	Not reported	51 (40-90) [n = 7/9] (measure for affected hand)	1 [n = 1/10]	2 tenolysis1 infection [n = 3/10]	Quick DASHdMediana 6.25, range 0-30

PIPJ: proximal interphalangeal joint.

^aCalculated by the review authors.

^bTAM – Total active motion percentage = [(PIP flexion + DIP flexion) - degrees (PIP + DIP extension loss)/175°]×100. Excellent 85–100, Good 70–84, Fair 50–69, Poor <50. ²⁵

^cTwo poor outcomes related to two digits on the same hand.

^dQuick DASH – Disabilities of the Arm, Shoulder and Hand, abbreviated version. ²⁶

O’Connell et al. reported range of movement, grip strength and the abbreviated Disabilities of the Arm, Shoulder and Hand Score (Quick-DASH) ²⁶ measured at 12 weeks postoperatively. TAM grading was Good (n = 4), Fair (n = 2) and Poor (n = 3). Median grip strength was 23 kg (range 14–41kg) and median Quick-DASH score was 6.25 (range 0–30). Return to work timing was not reported. ²³

For those patients included in Henry and Howell’s case series, there were no FDP ruptures, no secondary surgeries and no PIPJ flexion contractures. ⁹ No ruptures were reported when the RMF splint was in situ in the cadaveric study. ²⁴ At the time of this review, one rupture was reported by O’Connell et al. following an infection and two tenolysis procedures had been performed. ²³

Discussion

This review was developed to explore the use of RMF splinting for the rehabilitation of flexor tendon repairs. Specifically, we aimed to identify: (i) indications for using RMF splinting, (ii) splint designs, (iii) exercise programmes (including hand function), and (iv) clinical and patient reported outcomes. Three studies met our inclusion criteria. Two were case series,^9,23 and the other was a biomechanical (cadaveric) proof of concept study. ²⁴

Splint designs varied across the included studies. Following the principles outlined by Merritt and Howell ²⁷ ‘relative motion’ requires three key elements: 1) the splint must be used to deliver early active motion; 2) the MCPJ of the repaired digit(s) must be held in greater flexion (or extension) than the other MCPJs; and 3) this relative MCPJ position must be maintained throughout a wide range of active MCPJ movement, for all fingers. ²⁷ The RMF splint used by Henry and Howell satisfied all of these elements. The PEAMO strap used by O’Connell et al. offset the involved MCPJ in greater flexion, but did not maintain the same degree of relative flexion throughout range. The static nature of the strap, which was anchored in a figure of eight around the wrist, blocked end-range MCPJ extension of the affected digit(s), rather than holding the MCPJ in relatively more flexion throughout range. The PEAMO splint strategy appears to focus on restricting extension of the affected MCPJ past a certain point, while allowing normal composite flexion. A similar dorsal blocking approach was used in the cadaveric study by Chung et al., in addition to RMF splint positioning. ²⁴ Therefore, a key finding of this review is that the terminology for relative motion splinting appears to be used inconsistently. All three studies met the review eligibility criteria because their splints were described as RMF, however O’Connell et al. have since changed the name of their splint to the Protected Early Active Motion Orthosis, as reported in this review. ²³ To avoid ambiguity, we recommended that consensus needs to be reached on the definition of a relative motion splint. We propose that only splints that follow the relative motion principles outlined by Merritt and Howell (and discussed above)^12,16,27 be termed ‘relative motion’, but welcome debate on this issue.

Another important difference between splints described in this review was the position of the wrist in relation to the MCPJs. O’Connell et al. used a neutral wrist position with the MCPJ of the repaired digit(s) blocked at 50–70° flexion by the PEAMO. ²³ Henry and Howell positioned the wrist in either 0–20° flexion or 0–20° extension plus the RMF splint; ⁹ while the wrists of the cadavers were fixed at 30°extension and positioned in a RMF splint with an additional MCPJ extension block. ²⁴ The safest relative difference between the MCPJ(s) of the affected digit(s) compared with the uninjured fingers is unknown. The MCPJ differential for Henry and Howell’s RMF splint was 30–40°, permitting a potential arc (if 0 to 90° MCPJ motion is considered normal) of 50–60° MCPJ motion for the zone I–II FDP repairs. ⁹ There were no reported tendon ruptures by Henry and Howell (zones I–II FDP repairs) or Chung et al. (zone III FDP repair) wearing any of the larger splints in combination with a RMF splint;^9,24 while O’Connell et al. encountered one tendon rupture associated with an infection. ²³ No firm conclusions should be drawn regarding the type of splint required to supplement the RMF splint for control of wrist motion. The different splint strategies reported in this review represent a shift from the traditional forearm-based dorsal blocking splint ⁵ and expand on the work of Peck et al., who reported successful use of the Manchester short splint in a prospective case series of zone II FDP repairs. ¹¹ The Manchester short splint allowed for active wrist extension up to 45° (and unlimited flexion) combined with a potential 60° total arc of active MCPJ motion (MCPJs blocked at 30°flexion), and was reported to support differential tendon glide, reduce the work of flexion and prevent PIPJ flexion contractures. ²⁸

Neither case series reported any issues in positioning the RMF or PEAMO splints over the surgical dressings. However, the standard RMF splint design (Figure 2) may need to be amended, especially for index and little finger injuries in zones II, where the RMF splint is likely to sit directly over the dressings. Alternative strategies include semi-circumferential finger loops with additional Velcro straps that allow the size of the finger loop to be modified to accommodate dressing requirements or changes in oedema.^9,15

Zone of flexor tendon injury varied across the included studies. Chung et al. examined FDP repairs in zone III, ²⁴ Henry and Howell included FDP repairs in zones I-II (with pulley venting and no repair of FDS), ⁹ and O’Connell et al. included FDP and FDS repairs in zones II-III. ²³ To learn more of the indications and contra-indications for RMF splinting, future research should report surgical details, such as suture material and repair method, type of anaesthetic, complexity of injury, and timing of surgery and rehabilitation. In addition, complications should be listed including rupture, infection, tenolysis and any other secondary procedures.

In both patient series included in this review, the authors reported early active motion (with the respective splints in situ), commencing within the first postoperative week. O’Connell et al. used a standardised exercise programme as part of their RCT, ²³ while the hand therapists involved in the study by Henry and Howell reported patient-centred exercise programmes. ⁹ As a result, the most effective exercise programme to use with the RMF splint is unclear. It has been acknowledged by Henry and Howell that the protective function of the RMF splint may be overly effective in levying the quadriga effect on the FDP, such that it may need to be removed to achieve desired FDP excursion during the initiation of the early active motion exercise programme. ⁹ Prescribed early active motion exercises with RMF splinting will need to find the right balance of optimising tendon glide without stressing the healing repair. Future research should also explore the optimum timing to commence flexion exercises out of the RMF splint.

Light functional hand use (no strenuous lifting or gripping) was encouraged by Henry and Howell as soon as RMF splinting commenced (around 3–5 days after surgery). ⁹ This is similar to the approach adopted with the Manchester short splint, albeit without using the affected digit.^11,28 O’Connell et al. delayed hand use for light activities until week four. ²³ Following extensor digitorum communis repair, early functional hand use and relative motion extension splinting (without a wrist splint) has been associated with improved outcomes (Quick-DASH, Sollerman Hand Function Test, and TAM grades) four weeks after surgery, when compared with a forearm-based splint that immobilised the wrist and MCPJs. ¹⁷ It is hypothesised that RMF splinting with early active mobilisation and hand function could offer similar benefits following flexor tendon repair, but this has not yet been assessed in a randomised controlled trial.

There were few complications in the included studies, although the number of participants was small. Henry and Howell reported no ruptures or secondary procedures, suggesting that the RMF splint adequately protected a 4-strand core repair. ⁹ O’Connell et al. reported one rupture (following an infection) and two tenolysis procedures in their preliminary data; ²³ we await the full trial data to compare complication rates between PEAMO and dorsal blocking splint groups.

Henry and Howell reported no PIPJ flexion contractions (full active extension), and all but one participant achieved full active PIPJ extension in the study by O’Connell et al..^9,23 Anecdotally, PIPJ flexion contractures are a complication after flexor tendon repair with rehabilitation regimes that use a forearm based dorsal blocking splint. ¹¹ Peck et al. noted that patients using the Manchester short splint had greater PIPJ extension than those using a forearm-based splint and hypothesised that this was because the wrist could be flexed, easing passive tension in the long flexors to concurrently support active IPJ extension via the extensor tenodesis effect. ¹¹ The RMF splint is also thought to reduce passive tension to the affected FDP (quadriga effect) during active finger extension, such that the action of the intrinsics encourages full active IPJ extension. ²⁷

Limitations

Use of the RMF splint permitting early active motion is a new approach to flexor tendon rehabilitation, and we designed our search strategy to include the grey literature in an attempt to capture any unpublished data. Despite systematically searching grey literature repositories, trials databases and a naïve internet browser, it is possible that additional unpublished research was not identified. The identification of preliminary data from O’Connell et al. (who initially presented their PEAMO strap as a RMF splint) suggests that there is an ongoing interest in alternative rehabilitation strategies for the management of flexor tendon repairs.

The studies included in our review represent level IV (case series) and foundation level (cadaver studies) evidence, and are insufficient to suggest changes to practice. ²⁹ RMF and PEAMO approaches require further assessment in adequately powered clinical trials, and within multi-centre and multi-national healthcare settings, before their use can be advocated. Factors to consider include healthcare type, for example private or public services; hospital type, for example specialist hand centres or non-specialist units; and hospital setting, for example urban or rural. All of these contextual factors may influence the availability and regularity of hand therapy input and are important in assessing the generalisability of research findings. It is also important for researchers to report demographic information for their participants, particularly age and occupation. This will assist clinicians in identifying the relevance for their patient population.

In the current review, both case series included small numbers of participants and were conducted within a single healthcare organisation. Henry and Howell included two-staged and primary tendon repairs, and an FDP type IV avulsion, but present individual participant outcome data, allowing these cases to be reviewed separately. ⁹ In both patient series participants were lost to follow-up, further reducing the small sample sizes. However, drop out rates of circa 30% have been previously reported in other studies involving populations with tendon injuries.^11,16,30,31

Methodological quality assessment for this review was independently performed by two authors using the JBI critical appraisal tool for case series. ²¹ Importantly, JH, who was an author of one of the included studies, was not involved in the literature search, screening or quality appraisal stages. Quality assessment was not used as a criterion for inclusion in the review narrative summary, rather as a means of exploring potential limitations with the included studies. Several of the appraisal criteria were not applicable for the biomechanical cadaver study, and we acknowledge that an alternative assessment method may have been preferable, for example QUACS (Quality Appraisal for Cadaveric Studies), ³² however we chose to use the tool outlined in our review protocol to allow comparison across the three studies.

We did not specify the age of participants as a criterion for inclusion in this review. In general, flexor tendon repair outcomes for individuals under the age of 18 years old are good,^33,34 and in Henry and Howell’s series, three of the four patients achieving excellent TAM scores were aged under 18 years. This suggests that the RMF and early active motion programme may be a suitable strategy for young adults, as all three patients adhered to the splint regime and follow-up, although larger cohorts are required to assess outcomes across different age groups.

Conclusions

RMF splinting may offer an important alternative early active motion approach in the rehabilitation of patients following flexor tendon repair. The small patient and cadaveric series identified by this review suggested that a RMF splint paired with wrist immobilisation may provide suitable protection for single/two-digit flexor tendon repairs in zones I-III while allowing functional hand use. To validate this, we require larger cohort studies and RCTs involving multiple centres and reporting comprehensive detail of splint design, exercise programmes, parameters for functional hand use, clinical and patient reported outcomes, as well as comparison to existing splint regimes.

An unexpected finding of this review was variation in the interpretation of the ‘relative motion’ principle. We propose that a RMF splint must incorporate the following three key elements: 1) early active motion; 2) MCPJ of the affected digit(s) positioned in greater flexion; and 3) the relative flexion differential is maintained throughout range. ²⁷