Rehabilitation following extra-articular proximal phalangeal fractures of the fingers in adults: a scoping review

Abstract

Introduction

Proximal phalangeal fractures are common and can have a significant impact on hand function. Therefore, it is important to optimise post-operative rehabilitation. A scoping review was undertaken to map the existing evidence on rehabilitation of proximal phalangeal fractures of the fingers in adults.

Methods

A comprehensive search was conducted which included database searching, reference searching, hand searching of journals, and searching for grey literature. Eight articles were included after screening for eligibility.

Results

Three studies researched surgical interventions and five studies conservative management. The immobilisation period varied between 5 days to 3 weeks in the surgical studies, and between 3 to 7 weeks in the conservative studies. Active exercise therapy was started immediately with conservative management, while in the surgical studies time to commence exercises varied between 5 days and 3 weeks. All studies reported good results in mobility with a mean total active motion ranging from 240° to 258.9°. Patients reported little pain at final follow-up and grip strength recovered to 96% compared to the unaffected side. Studies reporting on function and patient satisfaction lacked transparency.

Conclusions

All studies had a moderate to high risk of bias and the results of the included studies should therefore be interpreted with caution. More high-quality randomised controlled studies with an a priori research protocol and a standard set of outcome measures are necessary to research whether early motion, an intrinsic plus splint leaving the wrist free, and the inclusion of additional treatment modalities can result in a better and/or faster recovery.

Introduction

Phalangeal fractures are the second most common fractures of the upper extremity after fractures of the radius and ulna^1–3 and account for approximately 59% of all hand fractures.^1,2,4,5 Anakwe et al. (2011) reported an annual incidence of 1.07 phalangeal fractures per 1000 persons in a single trauma unit in Scotland and similar incidence rates are reported in North America with 12.5 phalangeal fractures per 10,000 persons in the US, ² and 1.81 per 1000 (95% CI: 1.31–1.73) in British Columbia. ⁶ About 9.1–22.5% of hand fractures are proximal phalangeal fractures^5,7,8 of which the vast majority, approximately 75–85%, are extra-articular fractures. ¹ Sex and gender distribution show that males are at a 2.08 greater relative risk for sustaining a hand fracture than females.^1,6,9 Typical injury mechanisms include accidental falls, sports-related injuries, and direct blows to the hand.^9,10

In general, uncomplicated, stable extra-articular fractures without malalignment can be treated conservatively.^4,11,12 Most authors suggest placing the hand in MCP flexion and IP extension as this intrinsic plus position reduces the displacing force of the interosseus muscles that contributes to the apex volar angulation often seen in proximal phalangeal fractures.^13,14 In the case of an unstable fracture, considerable malalignment and/or involvement of the joint surface, surgery is often indicated.^4,15 However, there are currently insufficient randomised controlled trials (RCTs), systematic reviews and meta-analyses comparing the different surgical treatment options to draw definite conclusions on the superiority of one technique over the other.^4,16–19

In addition to that, rehabilitation should be optimized as these fractures are common injuries in the working age group resulting in high costs for the health care system as well as secondary costs to society due to absenteeism and a reduction in productivity.^15,20 Rehabilitation aims to restore the functionality of the hand by increasing the mobility of the fingers and regaining strength so that the activities of daily life can be resumed.^1,21 Active and passive mobilizations are a frequently used treatment modality to restore the range of motion (ROM).^13,21–23 It is suggested in the literature to start mobilization exercises as soon as fracture stability can be assured to prevent scarring, adhesion formation, and joint stiffness.^23–26 The systematic review of Feehan and Bassett ²⁷ (2004) concluded that early motion facilitated a faster recovery of metacarpal fractures, which can lead to a quicker resumption of activities of daily living and an earlier return to work. However, no articles on proximal phalangeal fracture management met the eligibility criteria and due to this, the benefit of early motion in this patient population could not be demonstrated at that time.

An initial literature search on the rehabilitation of extra-articular proximal phalangeal fractures did not yield high-quality RCTs, systematic reviews, and meta-analyses. The aim of this scoping review is therefore to present an overview of the existing body of evidence on rehabilitation strategies for the treatment of proximal phalangeal fractures, identify gaps in the existing literature and identify what future research is needed to develop evidence-based rehabilitation guidelines.

Methods

The PRISMA extension for scoping reviews (PRISMA-ScR) was used to guide the reporting of this scoping review (Supplementary file: Appendix A)^28,29 and follows the framework of Arskey and O’Malley.^30,31 To reduce the risk of bias and to increase the overall credibility of the results, a research protocol was developed a priori (Supplementary file: Appendix B), ³² however, it is currently not possible to register scoping reviews at PROSPERO.^33,34

Step 1: Identifying the research question

“What is known from the existing literature about the therapeutic management of proximal phalangeal fractures of the digits in adults?”

Step 2: Identifying relevant studies

A comprehensive search strategy was pursued by one reviewer (NV) in March 2021 to identify both published and unpublished evidence.^29,35 First, ten healthcare databases were searched electronically: CINAHL complete, AMED, MEDLINE, Embase, Cochrane Library, PEDro, Web of Science, Scopus, NICE evidence, and PubMed.^32,36 Additionally, hand searching of the Journal of Hand Therapy, Hand Therapy, the Journal of Hand Surgery (EU) and Hand Clinics was performed. ³² Moreover, hand searching through reference lists as well as forward citation chaining through Google Scholar of all the eligible articles was completed to find additional relevant research that may have been missed. ³² It was attempted to locate completed but unpublished literature by searching clinicaltrials.gov and the World Health Organization International Clinical Trial Registry Platform portal. ³³ Complementary, online available conference programmes of the Federation of European Societies for Surgery of the Hand, the European Federation of Societies for Hand Therapy, the International Federation of Societies for Surgery of the Hand, and the International Federation of Societies for Hand Therapy were searched to identify relevant conference abstracts and posters. Lastly, the Open Access Theses and Dissertation database and the ProQuest Dissertations and Thesis database were searched for relevant dissertations. A rerun of this search strategy was performed by one reviewer (NV) in June 2021 and September 2021. A final search of all the electronic databases was performed in January 2023.

Search terms used were: ‘proximal phalan* fracture’, ‘extra-articular’, ‘extraarticular’, ‘extra articular’, ‘physical therapy’, ‘physiotherapy’, ‘hand therapy’, ‘conservative treatment’, ‘conservative management’, ‘conservative care’, ‘post-operative’, ‘postoperative’, ‘post operative’, ‘dynamic treatment’, ‘dynamic management’, ‘functional brace’, ‘functional treatment’, ‘functional management’, ‘early (active) motion’ ‘early (active) mobili?ation’, ‘immediate motion’, ‘immediate mobili?ation’, ‘splint’, ‘brace’, ‘orthos*’, ‘Lucerne cast’. The search strategy was adapted to the search engine of each specific database but consisted of a combination of the identified search terms with Boolean language operators.^32,36 The exact search strategy used for each database can be found in Appendix C (Supplementary file).

Step 3: Study selection

After completing the literature search, search results were merged using the QCRI Rayyan Platform systematic review management software and duplicate records of the same report were removed.^33,34 The eligibility criteria used to select relevant studies were developed in correspondence with the PCC mnemonic (Population - Concept - Context) as described in the Joanna Briggs Institute (JBI) guidelines for scoping reviews ²⁹ and are presented in Table 1. Paediatric fractures were excluded because the remodelling potential in children differs from that in adults and a different approach is often recommended. ³⁷ Thumb fractures were excluded as the (postoperative) immobilisation, recovery period and rehabilitation exercises can differ significantly from those for other fingers due to the unique anatomy and biomechanics of the thumb.^38,39 Intra-articular fractures were also excluded as these fractures can differ significantly due to the additional challenge of obtaining an accurate reduction and secure fixation of the articular surface. ⁴⁰

Table 1.

Inclusion and exclusion criteria.

Inclusion criteria	Exclusion criteria
Participants • 18 years of age or older • Diagnosed with an extra-articular phalangeal fracture of the fingers	Participants • 17 years old or younger • Intra-articular fractures • Fractures of the thumb • Studies conducted on animals
Concepts • Articles reporting on therapeutic management of extra-articular phalangeal fractures, in particular: The duration and position of the immobilisation and the content of the (post-operative) physiotherapy	Concepts • Articles solely reporting on the surgical technique • Articles that do not report about standardized therapeutic/physiotherapy interventions for the (post-operative) management of extra-articular phalangeal fractures.   • Articles that only present mixed results (for example intra- and extra articular combined, metacarpal, and proximal fractures combined)
AND • Articles that included at least one outcome measure related to active or passive range of motion, TAM of the finger, grip and/or pinch strength, function, pain, self-reported disability, return to work or complications/adverse events
Context • Any geographic location • Any healthcare setting	Context • Cadaver research
Study Type • Articles in English, Dutch, French, German • randomised controlled trials, pragmatic controlled trials, randomised crossover trials • Prospective and retrospective cohort studies • Single case experimental design studies, case-(control) studies, case series, cross • Systematic reviews with or without meta-analysis • Abstracts, conference papers, posters • Dissertations and theses	Study Type • Guidelines • Expert opinions • Narrative reviews or other review articles that do not contain any data/results reported elsewhere

All titles and abstracts were screened for eligibility by one reviewer (NV) and those not fulfilling the inclusion criteria were removed.^33,36 If there were any doubts about whether the reference should be included, it was carried through to full-text review. ³² All full-text papers were obtained to screen them for eligibility. ³² As this study was conducted to obtain an academic degree, all decisions about the inclusion or exclusion of the articles were taken by one reviewer (NV), except for three articles where the reasons for inclusion were discussed during supervision (NG and AK) which aided decision-making.

Step 4: Charting the data

Data extraction was done by one reviewer (NV). If the available data were insufficient or missing, an attempt was made to contact the author(s) to request the provision of data. Three authors were contacted,^41–43 but to date no response has been received. The following categories of information for each study were extracted: study characteristics (year of publication, location, language, study design, population, sample size, eligibility criteria), patient characteristics (age, sex, fracture site, type of fracture, hand dominance), characteristics of the intervention (medical treatment, hand therapy, compliance), assessments characteristics (outcome measures, time points), and outcomes.

Step 5: Collecting, summarising, and reporting results

As the data of this scoping review did not lend itself to performing a meta-analysis,^33,44 a descriptive synthesis of the data is provided which follows the above-mentioned structure of data extraction: study characteristics, patient characteristics, characteristics of the intervention, assessments characteristics, and outcomes.

Although usually no formal assessment of methodological quality is included in a scoping review,^28,29,35 it was decided to integrate critical appraisal to identify strengths and limitations of the included studies as this helps to determine the reliability of the evidence.^45–47 The level of evidence was determined according to the Oxford Centre for Evidence-Based Medicine (OCEBM) scale ⁴⁸ and the JBI critical appraisal tools were used to evaluate the methodology of the study.^49–51

Results

Search results

A total of 1623 articles were identified through the search of 10 databases, and an additional 105 records were identified through complementary search strategies. After removing 740 duplicates, 988 titles and abstracts were screened for eligibility with 842 records being excluded. Subsequently, 146 full texts were screened, of which 7 articles met the eligibility criteria. In addition, one more article was identified during the rerun of the database search in June 2021 and was also included in this review. No additional records were identified during the rerun searches in September 2021 or January 2023. A detailed overview of the literature review process, including the reasons for exclusion at each step, can be found in the PRISMA-flowchart (Figure 1).

Figure 1.

PRISMA-flowchart of the literature review process.

Study characteristics

Six of the included studies were case series of which five were conducted prospectively^42,52–55 and one retrospectively. ⁴¹ In addition, one RCT ⁵⁶ and one case report ⁴³ were included. The general characteristics of the included studies can be found in Table 2.

Table 2.

General characteristics of the included studies.

	Journal	Year of publication	Location	Language	Design	OCEBM level of evidence
Gangatharam, 2021	Hand surgery and rehabilitation	2021	Qatar	English	Prospective case-series	4
Jun et al. 2021	Archives of plastic surgery	2021	Seoul, Korea	English	Retrospective case-series	4
Byrne, Jacques and Gurfinkel, 2019	Journal of hand surgery (EU)	2019	Perth, Australia	English	Prospective case-series	4
Ataker et al. 2017	Journal of hand surgery (EU)	2017	Turkey	English	Prospective case-series	4
Held et al. 2016	Hand surgery	2013	Cape Town, South Africa	English	Prospective case-series	4
Franz et al. 2012	Journal of hand surgery	2012	Switzerland	English	Prospective, randomised multicenter study	2
Franz, von Wartburg and Hug, 2010	Handchirurgie · Mikrochirurgie · Plastische Chirurgie	2010	Bern, Switzerland	German	Prospective case-series	4
Loosli and Garrick, 1987	The American journal of sports Medicine	1987	San Francisco, USA	English	Case report	5

Patient characteristics

A total of 267 proximal phalangeal fractures were included in this review and an overview of the participants’ characteristics is presented in Table 3. Additional information on the demographics of the participants such as race, occupation, education level and mechanism of injury was often lacking except for the study of Held et al. (2013) which included information on the mechanism of injury and employment status, the study by Ataker et al. (2017) which included information on the mechanism of injury, and the study by Franz, von Wartburg, and Hug (2010) which included information on occupation.

Table 3.

Patient’s characteristics.

		Gangatharam 2021	Jun et al. 2021	Byrne, Jacques and Gurfinkel, 2019	Ataker et al. 2017	Held et al. 2016	Franz et al. 2012	Franz, von Wartburg and Hug, 2010	Loosli and Garrick, 1987	Total
Number of patients		13	37	101	22	23	66	15	1	278
Number of proximal phalanx fractures		13	12	101	22	23	75	20	1	267
Age
Mean (SD)		35.08	42 (11)	35 (17)	29.7 (6)	36	49	45.1	18
Range			16–65			18–60	16–93	15–77		15–93
Sex
Male	N (%)	13 (100.00%)	30 (81.08%)	72 (71.29%)	14 (63.64%)	18 (78.26%)	34 (51.52%)	7 (46.67%)	1 (100.00%)	189 (67.99%)
Female	N (%)	0 (0.00%)	7 (18.92%)	29 (28.71%)	8 (36.36%)	5 (21.74%)	32 (48.48%)	8 (53.33%)	0 (0.00%)	89 (32.01%)
Fracture site
Index	N (%)	5 (41.67%)		10 (9.90%)	4 (18.18%)	5 (21.74%)	14 (18.67%)	5 (25.00%)		43 (16.17%)
Middle	N (%)	1 (8.33%)		6 (5.94%)	2 (9.09%)	9 (39.13%)	7 (9.33%)	1 (5.00%)	1 (100.00%)	27 (10.15%)
Ring	N (%)	3 (25.00%)		17 (16.83%)	7 (31.82%)	5 (21.74%)	13 (17.33%)	4 (20.00%)		49 (18.42%)
Little	N (%)	3 (25.00%)		68 (67.33%)	9 (40.91%)	4 (21.74%)	41 (54.67%)	10 (50.00%)		135 (50.75%)
Not specified	N (%)		12 (100.00%)							12 (4.51%)
Type of fracture
Transverse	N (%)			56 (55.45%)		9 (39.13%)	50 (66.67%)	9 (45.00%)		124 (46.62%)
Spiral	N (%)			30 (29.70%)	8 (36.36%)		9 (12.00%)	5 (25.00%)	1 (100.00%)	53 (19.92%)
Oblique	N (%)			15 (14.85%)	6 (27.7%)	9 (39.13%)	13 (17.33%)	6 (30.00%)		49 (18.42%)
Longitudinal	N (%)						3 (4.00%)			3 (1.13%)
Comminuted or Complex	N (%)				8 (36.36%)	5 (21.74%)				13 (4.89%)
Not specified	N (%)	12 (100.00%)	12 (100.00%)							24 (9.02)
Hand dominance
Dominant hand	N (%)				16 (72.73%)		32 (42.67%)	7 (46.67%)		55 (21.07%)
Non-dominant hand	N (%)				6 (27.27%)		43 (57.33%)	8 (53.33%)		57 (21.84%)
Not specified	N (%)	12 (100.00%)	12 (100.00%)	101 (100.00%)		23 (100.00%)			1 (100.00%)	149 (57.09%)

Characteristics of interventions

A detailed overview of the interventions provided in the studies is presented in Table 4. The proximal phalangeal fractures were treated surgically in three studies,^41,42,53 and conservatively in the remaining five studies.^{43,52,54–56} In all studies, the hand was immobilised in the intrinsic plus position, apart from the case report by Loosli and Garrick (1987) where the finger was buddy taped to the adjacent fingers without the provision of an additional splint or cast. In three studies a dorsal hand-based splint or cast was provided,^52,55,56 while in five studies a forearm-based splint was used.^{41,42,53,54,56}

Table 4.

Overview of the interventions.

		Medical intervention	Splint	Hand therapy intervention	Additional therapy	Compliance
Surgical studies	Gangatharam 2021	Surgical treatment: Closed reduction K-wire fixation	Until the 10-12th postoperative day: Cast. 10-12th postoperative day until 5th postoperative week: Forearm-based thermoplastic splint (wrist 30° extension, MCP 70° flexion, IP extension). 2 splints used for rehabilitation exercises.	10-12th postoperative day - 4th postoperative week: o Progressive ROM exercises using 2 thermoplastic splints, 10 repetitions every 2-3 waking hours.	No	Template splints and target blocks were used to increase compliance, but compliance itself was not assessed
				o Template 1: PIP and DIP joint flexion against the splint
				o Template 2: Blocks MCP and PIP in extension to stimulate DIP movement
			5th postoperative week – 8th postoperative week: Thermoplastic splint only for the night	5th postoperative week:
				o Finger closure without any template splint but with target blocks fabricated to the subject’s needs varying sizes from 1,5 in.-0,25 in.
				o Light functional activities
				o if extensor lag: activities to facilitate extension
				8th-12th postoperative week:
				o Strengthening programme: Light to high resistive exercise depending on the subject’s tolerance level
	Jun et al. 2021	Surgical treatment: ORIF (extensor tendon saving technique) with low-profile 1,2 mm titanium mini plate	2 weeks: Ulnar or radial gutter splint (wrist 30° extension, MCP 70° flexion, IP extension)	3–5 weeks after surgery:	No	Not assessed
				o Active exercise with pain-free active traction in supination, neutral and pronation.
			Additional 1 week: Soft splint	o Easier exercise if necessary: Passive flexion and active extension via a thermoplastic splint with a rubber band
				4 weeks after surgery: Moderate functional activities
	Ataker et al. 2017	Surgical treatment: ORIF (dorsal (7/22) or mid-axial (15/22) incision) with 1,5 mm mini plate and screws (10/22) or screws only (12/22)	Immediately after surgery: Forearm-based cast splint (wrist 45° extension, MCP 70° flexion, PIP extension, DIP free to move). 5th day after surgery: No splint used in postoperative rehabilitation protocol	Early phase (5th day – 4th week):	Neuromuscular electric stimulation for patients with an extension lag at the PIP joint (20° min) in intermediate phase	Not assessed
				o Combined and isolated active assistive ROM (blocking) exercises of DIP, PIP and MCP, TGE, tenodesis exercises, oedema treatment (ice packs, coban elastic bandages)
				o Light activities were gradually introduced
				o Scar tissue massage: From 2nd postop week if needed.
				o First 3 weeks: 3x week, 60 minutes. Home exercises 4x day, 10 repetitions
				Intermediate phase (5th-6th week):
				o Active resistive and passive ROM exercises, stretching of lateral bands and intrinsics
				Late phase (after 6th week):
				o Strengthening and conditioning exercises
				o 3 weeks of therapy sessions
Conservative treatment studies	Byrne, Jacques and Gurfinkel, 2019	Conservative treatment: Closed fracture reduction under local anesthesia	4 weeks: Dorsal hand-based thermoplastic splint in maximum MCP flexion + elastane buddy stall	o Immediately: Full active PIP and DIP flexion and extension, hourly. Light use of the hand was permitted for activities that were pain free and necessary for daily living while avoiding functional use of the injured digit	No	Not assessed
			After 4 weeks: Splint removal for light activities leaving the buddy stall on	o After 4 weeks: Full active tendon gliding exercises
			At 6 weeks: Splint and buddy stall were removed	o At 6 weeks: Progression to full activities
	Held et al. 2016	Conservative treatment: Closed reduction under local anesthesia	3 weeks: Dorsal slab to maintain reduction (MCP in at least 90° flexion) + buddy tape to control for rotation	Immediately: Full active range of motion exercises in dorsal slab	No	Not assessed
				Week 4-6: Active range of motion exercises at the PIP joint
	Franz et al. 2012	Conservative treatment: Closed reduction under digital block anesthesia	After reduction: Static forearm-finger cast in intrinsic-plus position	Immediately: Patients were instructed to fully extend and fully flex the interphalangeal joints of all fingers several times a day	OT was initiated in 37/66 patients at time of cast removal	Not assessed
			No later than day 4 - 4 weeks (up to 7 weeks in case of tenderness at fracture site):
			Group A: Dorsopalmar forearm plaster splint (wrist 30° extension, MCP 70–90° flexion, fingers free on palmar side).
			Group B: LuCa, padded semirigid circular metacarpal brace (MCP 70–90° flexion, wrist flexion crease and distal flexion crease exposed).
			Group A and B: Buddy tape to unaffected radial adjacent finger, or in case of index to middle finger
	Franz, von Wartburg and Hug, 2010	Conservative treatment: (If necessary) closed reduction under metacarpal block	No later than day 4 -4 weeks (up to 6 weeks in case of tenderness at fracture site): LuCa padded semirigid circular metacarpal brace (MCP 70–90° flexion, wrist flexion crease and distal flexion crease exposed)+ buddy loop to unaffected radial adjacent finger, or in case of index to middle finger	Immediately: Patients were instructed to fully extend and fully flex the interphalangeal joints of all fingers several times a day	OT was initiated in 10/15 patients at time of cast removal	Not assessed
	Loosli and Garrick, 1987	Conservative treatment: Reduction of the gross deformity by traction on the field + closed reduction under local anesthesia	First few days: Ace bandage for general compression	Immediately - 3 weeks: Icing programme (cool water soaks for 10 minutes six times a day, doing active range of motion exercises)	No	Not assessed

Abbreviations: MCP: Metacarpophalangeal joint; PIP: Proximal interphalangeal joint; DIP: Distal interphalangeal joint; IP: Interphalangeal joints; ORIF: Open reduction internal fixation; ROM: Range of motion; TGE: Tendon gliding exercises; OT: Occupational therapy; LuCa: Lucerne-cast.

The immobilisation period after surgery was generally shorter (range: 5 days to 5 weeks) compared to the immobilisation period with a conservative approach (range: 3–7 weeks) (Table 4), although the study of Gangatharam ⁴² (2021), in which the fractures were stabilized with K-wires, recommended to wear the splint at night until the 8^th postoperative week. Moreover, in all the studies with a conservative approach, active ROM exercises of the interphalangeal joints were immediately allowed in the splint ^52,54–56 or with the buddy tape, ⁴³ while in all surgical studies, the onset of exercise therapy was delayed (range: 5 days ⁵³ to 3 weeks^41,42). One study included passive ROM exercises and stretching of the lateral bands and intrinsic musculature in the 5th-6th week postoperatively, ⁵³ and two studies explicitly mentioned strengthening exercises in their rehabilitation schedule.^42,53 Three studies indicated that some participants received complementary therapy in addition to standard hand therapy interventions (Table 4).

Assessment characteristics

The total follow-up of the included studies ranged between 6 weeks ⁵⁴ and 15.4 months, ⁵³ with a follow-up of 12 weeks in half of the studies.^41,42,55,56 However, the follow-up time in Ataker’s study varied considerably from 6 months to 48 months. All studies, except for the study by Loosli and Garrick (1987), included some form of quantitative ROM assessment. However, different assessments methods were used throughout the studies including Total Active Motion (TAM),^{42,52,53,55,56} Total Passive Motion (TPM), ⁴² total extension deficit,^42,52–56 goniometry per joint,^41,52,53 or the distance between the pulp of the fingertip and the distal palmar crease (DPC). ⁵³ This diversity in assessment methods makes it hard to compare results between studies. Pain was evaluated in two studies with a Visual Analogue Scale (VAS),^52,53 however, the study by Loosli and Garrick (1987) did not report transparently how pain was assessed. Grip strength was only evaluated in one study using the Jamar dynamometer. ⁴² Patient-reported hand function was assessed in the study by Gangatharam (2021) using the Michigan Hand Questionnaire (MHQ) and in the study of Ataker et al. (2017) by the Disability of the Arm, Shoulder, and Hand questionnaire (DASH). Patient satisfaction was assessed in three studies,^42,55,56 although only the study by Franz et al. (2012) transparently reported that this was done by means of a VAS.

Outcomes

The results showed a mean TAM between 240° and 258.9°.^42,55 The RCT of Franz et al. (2012) compared a forearm-based splint with the hand-based Lucerne-cast. No statistically significant difference in TAM-values were observed between both groups, neither at the time of cast removal, nor at 12 weeks, although the exact TAM-values of both groups were not reported. ⁵⁶ Wrist ROM was statistically higher with the Lucerne-cast at the time of cast removal (group A: 107°; group B: 127°; p = 0.001), however, this difference was no longer statistically significant at 12 weeks (group A: 128°; group B: 137°; p = 0.074). Nevertheless, patient satisfaction with treatment is significantly higher if the wrist can move freely (p = 0.022).^55,56 Extension lags were reported in six studies in which the mean extension lag at 12 weeks varied between 2,5° and 5°.^42,55,56

Patients seem to experience little pain over time as Byrne et al. (2019) reported a median pain score of 0 (IQR: 0.0) at 7 weeks and Ataker et al. (2017) a mean pain score of 0.3 (SD: 0.46; range: 0-1) with a mean follow-up of 15.4 months (range 6–48). Gangatharam ⁴² (2021) reported grip strength of 96% compared to the unaffected side at 12 weeks post-op. The RCT by Franz et al. (2012) reported higher satisfaction with treatment in the Lucerne-cast group (VAS = 9.4) compared to the group in which the wrist was included in the splint (VAS = 8.4) (p = 0.022) (Franz et al., 2012). Gangatharam (2021) reported a ‘mean satisfaction score at discharge’ (12 weeks) of 91%, although it was not transparently reported how this was assessed.

Risk of bias assessment

The level of evidence according to the OCEBM scale is reported in Table 2, and the JBI critical appraisal checklists are available in Appendix D (Supplementary file). A summary of the risk of biases identified per study is presented in Table 5. In only one study, participants were randomised, ⁵⁶ however, treatment allocation was not concealed, and therefore a risk of selection bias was identified in all studies. Due to the nature of the intervention, it was not possible to blind the participants. However, in none of the studies were outcome assessors blinded which can lead to confirmation bias.^57,58 Additionally, it was not stated whether training of assessors took place to increase the standardization and reliability of the measurements. ⁵⁷ Nevertheless, most studies did use reliable and validated outcome measures such as goniometry, grip strength measurements, and validated questionnaires. Lastly, it was unclear in three studies whether all participants were treated identically other than the intervention of interest, as patients who received additional therapy were not analyzed separately.^53,55,56 This adjunct intervention may lead to performance bias.^57,59

Table 5.

Risk of bias summary.

	Selection bias	Performance bias	Confirmation bias	Attrition bias	(Selective) reporting bias
Gangatharam (2021)	- No randomization, no control group.	- Blinding of study participants not possible.	No blinding of assessors		- Outcome measures and follow-up results were not clearly reported (no confidence intervals, no exact p-values)
	- Unclear if only extra-articular fractures were included and how the condition was identified, type of fracture was not specified.	- Additional therapies were given to some patients to facilitate extension
	- Unclear whether participants were included consecutively
Jun et al. (2021)	- No randomization, no control group.	- Blinding of study participants not possible.	No blinding of assessors		- Outcome measures and follow-up results were not clearly reported (no confidence intervals, no exact p-values).
	- Unclear whether participants were included consecutively	- Unclear which exercises were performed			- Retrospective study
Byrne, Jacques, and Gurfinkel (2019)	- No randomization, no control group.	- Blinding of study participants not possible.	No blinding of assessors	21 patients were lost to follow-up, no intention to treat analysis	- Outcome measures and follow-up results were not clearly reported (unclear which statistical test was used, no confidence intervals, no exact p-values).
	- Unclear how the condition was identified	- No uniform time for assessments, no assessment at start			- Intermediate results were not reported
Ataker et al. (2017)	- No randomization, no control group	- Blinding of study participants not possible.	No blinding of assessors		- Outcome measures and follow-up results were not clearly reported (no confidence intervals, no exact p-values)
		- Assessment only took place at final follow-up, which ranged from 6 months to 48 months
		- 3 patients received additional therapy (NMES)
Held et al. (2013)	- No randomization, no control group.	- Blinding of study participants not possible.	No blinding of assessors	3/23 were lost to follow-up, no intention to treat analysis	- Outcome measures and follow-up results were not clearly reported (no confidence intervals, no exact p-values, no standard deviations).
	- Unclear how participants were recruited	- Unclear how and when outcome measures were assessed			- Intermediate results were not reported
Franz et al. (2012)	- Participants were randomised but treatment allocation was not concealed.	- Blinding of study participants not possible.	No blinding of assessors	3/66 lost to follow-up, no intention to treat analysis	- Outcome measures and follow-up results were not clearly reported (not all p-values were reported, no confidence intervals, no standard deviations)
	- Uneven groups	- Additional occupational therapy was given to some participants.
		- Unclear whether assessments were standardized
Franz, von Wartburg, and Hug (2010)	- No randomization, no control group.	- Blinding of study participants not possible.	No blinding of assessors		- Not all results were reported.
	- Unclear how the condition was identified	- Additional occupational therapy was given to some participants.			- Outcome measures and follow-up results were not clearly reported (no statistical tests, no p-values, no standard deviations, no confidence intervals)
		- Unclear whether assessments were standardized
Loosli and Garrick (1987)	- No randomization, no control group (case report)	- Intervention given was not clearly described.	No blinding of assessors		- Outcome measures and follow-up results were not clearly reported
		- Unclear how and when outcome measures were assessed

Discussion

The results of this scoping review show that the current body of evidence on this subject is very limited and of low quality due to the risk of bias. Studies often lacked transparency regarding the (additional) interventions given and whether assessments were performed in a standardized manner. Consequently, it can be argued that there are threats to both the validity and reliability of the included studies.^45,57 It is therefore advisable to interpret the results of this review with caution to inform practice.^36,60 The findings of this review support the need for additional high-quality RCTs in which assessments are standardized, assessors are blinded, and groups are treated identically except for the intervention being studied. A large RCT (n = 400) is currently undertaken in Nottingham in which the surgical treatment of proximal phalanx shaft finger fractures is being compared with non-surgical treatment using a finger splint (POINT trial, https://www.isrctn.com/ISRCTN88266404).

All studies except the case report by Loosli and Garrick (1987) immobilized the hand in an intrinsic plus position often in combination with a buddy loop to control for rotation. This is consistent with the literature findings recommending this position to relax the interosseus muscle, reducing its displacing force.^9,13 Moreover, the Dutch Association for Hand Surgery (NVvH) ⁶¹ recommends immobilisation in the intrinsic plus position in their guidelines. Consequently, it seems appropriate to recommend this immobilisation position for 3-4 weeks with conservative treatment. Based on the results of the RCT of Franz et al. (2012), a weak recommendation can be formulated to exclude the wrist from the splint as this led to better wrist ROM in the short term and higher patient satisfaction with treatment.^54,62

Fractures that were treated conservatively in the included studies started active exercise therapy immediately. Since this rarely resulted in a loss of fracture reduction (N = 8/227 (3.5%)),^52,54,56 early motion seems to be a safe approach in this patient population. Initiation of exercise therapy was delayed by 5 days to 3 weeks in the studies where surgical treatment was performed,^41,42,53 but none of the surgical studies provided a clear reason for this delay. It can be questioned whether this is linked to the fact that conservative treatment can only be considered for less severe fractures that are inherently stable. ¹² However, it is precisely the aim of surgery to ensure that stability to allow early motion and thus a clear reason for postponing early motion in the surgical studies can therefore not be given. The literature indicates that the total ROM declined to 66% of normal if the immobilisation period exceeded 4 weeks and that time to commencement of active exercise is an independent predictor of total active ROM 6 weeks after proximal phalanx fixation.^24,63,64 This does suggest that early motion could result in superior results, but unfortunately, the results of this scoping review do not allow us to define the optimal time to initiate early motion more precisely. Future research should therefore investigate optimal time parameters to introduce different treatment modalities.

The literature search did not yield any articles comparing different rehabilitation modalities. The studies in this review emphasized active flexion and extension exercises. Initially, ROM exercises from a protective intrinsic plus position can be suggested, which can be performed in the splint when a dorsal splint is provided. After the immobilisation period, tendon gliding exercises can be gradually introduced to promote the motion of tendons through their sheaths and to prevent soft tissue adherence.^21,65,66 In addition, 2 studies incorporated strength training into their rehabilitation program at week 6 ⁵³ and week 8. ⁴² Although no statement can be made whether this added treatment modality led to better results or faster recovery, as it was part of a multimodal approach, it seems safe to introduce progressive strength training after 6–8 weeks. ²¹

Most studies had an observation period ranging between 6 weeks and 12 weeks,^{41,42,52,54–56} and although most of the recovery in ROM, pain, strength, hand use and work participation occurs by week 6, smaller gains are still reported by week 12 and week 26. ⁶⁷ Future studies may wish to consider a longer follow-up period to fully estimate the long-term effects of rehabilitation, as is the case in the upcoming POINT trial with a 1 year follow-up. Moreover, only one study included baseline measurements, ⁴² and in several studies the (final) assessments did not coincide with the end of the therapeutic intervention.^41,53 Due to these baseline differences, natural history, and additional but unregistered treatments may also have influenced the results. ⁵⁷ Although it is not possible, for example, to perform strength measurements immediately after sustaining a fracture due to the risk of loss of fracture reduction, it should be considered to include a baseline measurement of other outcome measures such as ROM, pain, and hand function.

The mean TAM varied between 240° and 258.9° in all studies, which can be considered a functional ROM. ⁶⁸ However, considering the wide variety of ROM assessment methods, it can be recommended to work towards a more uniform way to document the ROM in the future as this will allow comparison between the different studies. ⁶⁹ Goniometry has proven to be a valid and reliable method to assess ROM and goniometry values of the individual joints can provide a more detailed picture of the evolution of individual joints, ⁶⁹ which in turn can guide the rehabilitation process.

The average reported extension deficit in the included studies was less than 5° which could be considered an acceptable result. However, when analyzing the results in more detail, it is apparent that there can be significant differences between the participants. Most of the participants did not show an extension deficit and only a few participants had a significant deficit of 15–25°.^53,55,56 Further research into the reasons why some participants develop an extension deficit while others do not may be beneficial to develop a targeted approach to address this problem, which can then be included in treatment protocols.

Although low pain scores were reported in the included studies,^52,53 it may be recommended to assess pain at the start of rehabilitation as pain in the post-operative week predicts pain and hand use 12 weeks after proximal phalangeal fixation. ⁷⁰ By screening routinely for resting pain at baseline, people presenting with high pain scores can be monitored closely and guided accordingly. ⁷⁰ In addition, it can be recommended to repeatedly assess the average pain during the past 7 days, instead of using a one-time point measure as was done in the included studies, given that this has shown to be more reliable. ⁷¹

A strength of this scoping review is that it attempted to identify all relevant literature to the extent that this was possible through an extensive and transparent search strategy. Moreover, attempts were made to reduce error and increase reliability by means of an a priori research protocol and by reporting transparently in accordance with the PRISMA-ScR guidelines.^28,31,62,72 However, only articles in Dutch, English, French and German were considered and therefore it cannot be ruled out that relevant articles were excluded. Additionally, only one reviewer was involved in the search and selection procedure, and this might have led to reviewer bias.^32,36 Lastly, although an attempt was made to minimise the risk of bias through a rigorous literature review process, the methodological limitations of the individual sources of evidence within this review warrant a cautious interpretation of the results.

Conclusions

The studies included in this review provided low-quality evidence that both conservative as well as surgical treatment for proximal phalangeal fractures can result in good ROM at final follow-up. However, additional high-quality RCTs that compare different rehabilitation approaches are needed to develop evidence-based treatment guidelines.

ORCID iD

Nele Vervloesem https://orcid.org/0000-0002-8900-7764