Prospective Multicenter Randomized Controlled Trial Comparing Early Protected Movement and Splinting for Fifth Metacarpal Neck Fracture

Abstract

Background:

Fifth metacarpal neck fractures account for 20% of all hand fractures, yet there remains debate with respect to management, particularly when conservative treatment is initiated. The objective of this study is to compare functional and patient-reported outcomes (PROs) in patients treated with early protected movement or splint immobilization.

Methods:

This national multicenter prospective randomized controlled trial compared 2 groups; elastic bandage with early protected movement versus immobilization with splinting. Demographic characteristics were collected at baseline. Functional outcomes (grip strength testing) and PROs (Brief Michigan Hand Questionnaire [bMHQ]) were collected at 4, 8, and 12 weeks post-intervention. Grip strength values of the injured hand were normalized to both the non-injured hand (at baseline), and the Canadian reference values.

Results:

Thirty-seven participants from 5 Canadian centers were randomized into the splint (n = 21) or elastic bandage group (n = 16). There were no significant differences in the bMHQ score between the splint (52.1 ± 27.2) or the elastic bandage (46.6 ± 20.4) groups (P = .51). There were no differences in baseline grip strength between the splint (15.3 ± 8.9 kg) and elastic bandage (19.9 ± 7.5 kg) groups. At 8 weeks, the elastic bandage group had a significantly higher grip strength than the splint group (93% vs 64%, respectively: P < .05), when standardized as a percentage of the Canadian reference values.

Conclusion:

Patients with Boxer’s fractures treated with early protected movement had better functional outcomes by 8 weeks post-treatment as compared to the Canadian reference values of those treated with immobilization and splinting. Providers should manage Boxer’s fractures with early protected movement.

Introduction

Fifth metacarpal neck fractures, commonly referred to as Boxer’s fractures, account for nearly 20% of all hand fractures. ¹ Despite the high incidence of Boxer’s fractures, the optimal treatment is subject to debate. ^2,3 Currently, a variety of operative techniques are used, including percutaneous wire fixation, ⁴ plate fixation, ^{5 -8} and intramedullary fixation. ^{9 -15} Although these surgical techniques offer a stable reduction, they are limited by their cost and the need for extensive dissection, which potentially introduces complications. At present, it remains unclear whether they provide superior functional outcomes compared to conservative management. For this reason, Boxer’s fractures are commonly managed non-operatively. ^2,3,16

Boxer’s fractures are predominately diagnosed in young men ³ who typically obtain the injury by striking a hard, rigid object with a closed fist. ¹⁷ Boxer’s fractures can be managed conservatively from splint immobilization with or without reduction ^18,19 to short-term buddy taping and bracing. ²⁰ Although the duration of immobilization varies among surgeons, ¹⁹ nonetheless, a recent prospective study observed that a single week of pressure dressing followed by immediate mobilization resulted in similar functional outcomes as compared to 3 weeks of immobilization. ¹⁹ In addition, a prospective randomized trial comparing elastic bandage with early movement to reduction and splinting demonstrated that early mobilization resulted in a satisfactory range of motion and grip strength after 4weeks. ²¹ By circumventing the need for prolonged immobilization, patients can begin active motion protocols that may expedite their return to work and mitigate productivity losses. ^3,16

Previous investigations comparing elastic bandage with early movement versus immobilization for treatment of Boxer’s fracture failed to assess patient-reported outcomes (PROs). ^19,21 Patient-reported outcomes are tools used to assess health-related quality of life, utility, and satisfaction ^22,23 and can inform how patients perceive their condition, symptoms and the delivery of care. ^{22 -24} Unfortunately, only a few prospective investigations have assessed PROs in Boxer’s fracture patients treated with conservative techniques. ^25,26

Although conservative management of Boxer’s fracture has been well established, poor patient compliance with postoperative rehabilitation remains a significant issue. This has led to a difficulty in evaluating the efficacy of various treatment protocols, ^3,21,27 which was reflected in a recent Cochrane review that suggested there is a lack of robust level-1 evidence available to help guide conservative treatment of Boxer’s fractures. ²⁸ Further, there are few prospective investigations examining early protected movement for treatment of Boxer’s fractures, ²¹ as well as a significant paucity of demographic data, such as socioeconomic class or employment status, in patients with Boxer’s fractures. Thus, the objective of this study was to prospectively compare functional and PROs between supportive splinting and early protected movement for treatment of Boxer’s fractures. Secondary objectives were to characterize the relationship between Boxer’s fractures patient demographics and PROs and functional outcomes. We hypothesized was that no is significant difference in short-term function and pain between treatment with elastic bandage bandage or removable splint. The secondary hypothesis was that patients with more labor-intensive occupations would have lower overall functional outcomes.

Materials and Methods

A multicenter, prospective parallel group randomized controlled trial was employed. The study was performed in conjunction with the Canadian Plastic Surgery Research Collaborative, a national trainee-led research group. ²⁹ The participating Canadian centers (Sunnybrook Health Sciences Center, Toronto, Ontario; Halifax Infirmary, Halifax, Nova Scotia; The Ottawa Hospital, Ottawa, Ontario; McGill University Health Center, Montreal, Quebec; and Centre Hospitalier de l’Université de Montréal, Montreal, Quebec) were selected due to the high volume of cases and experience of the surgeon. Ethics was obtained at each center and all participants provided written informed consent. The study is presented in adherence to the Consolidated Standards of Reporting Trials guidelines. ³⁰

Participants

All patients >18 years of age who presented to the emergency department (ED) and were referred to the site principal investigator (PI) for treatment of Boxer’s fractures between December 2016 and June 2018 were assessed for eligibility for enrolment.

Eligibility

To be eligible for participation in the study, individuals had to be ≥18 years of age, have a radiographically diagnosed Boxer’s fractures, and have been seen in consultation by a plastic surgeon within 10 days of injury. All participants were also required to speak, read, and write in English.

Exclusion

Participants were excluded from the study if they had prior surgical treatment for a Boxer’s fracture or had a co-existing hand/wrist condition or injury. Participants were also excluded if their Boxer’s fracture had more than 70° angulation, had clinically significant malrotation, had undergone reduction in the ED or were surgical candidates.

Randomization

Participants were seen within 3 to 5 days of referral. Following consultation, patients were invited to participate in the study once a radiographic diagnosis of Boxer’s fractures was made, and there was confirmation that closed reduction was not performed or required. Eligible patients completed informed consent and were randomized via a computer-generated random-numbers table (random.org) to either group. Randomization was performed using an allocation ratio of 1:1. Allocation was assigned by the PI to patients using a prefabricated datasheet. Due to the nature of the intervention, surgeons, study personnel, and patients were not blinded to the treatment.

Sample Size

The sample size was calculated to detect a significant association for the study primary objective, specifically the comparison of functional and PROs between supportive splinting and early protected movement for treatment of Boxer’s fractures. The power calculation was not aimed at secondary study objectives which were to characterize the relationship between Boxer’s fractures patient demographics and PROs and functional outcomes.

The target sample size (n = 43 per group) was calculated based on a superior trial design. A sample size calculation using a sampling ratio of 1, a standard deviation (SD) of 10, and group means differing by 7, ³¹ determined that 32 participants per group was sufficient to achieve adequate power (power = 0.80, α = .05). This number was inflated to account for an estimated 25% attrition rate. The PI (P.B.) estimated that a minimal clinically important difference of 7 points was reasonable for our cohort; a value estimated by using a sample of patients with Dupuytren’s contracture in which 7 points on the Brief Michigan Hand Questionnaire (bMHQ) was clinically signficant. ³²

Interventions

Elastic bandage

Participants randomized to the elastic bandage group were treated with an elastic bandage for the palm and wrist worn for 4 weeks without a splint. A 2″ elastic bandage was specifically used (Figure 1). Participants were provided standardized written and verbal instructions for behaviour modification (Table 1).

Figure 1.

Elastic bandage.

Table 1.

Standardized Instructions Provided to Participants in the Elastic Bandage Group at Baseline and the Splint Group at Week 4.

Type of activity	Behavior modifications
Encouraged	• You are encouraged to flex and extend your finger when it is not in use
Permitted	• You can use your hand for activities of self-care (eating meals, brushing your teeth, getting dressed) • You can use your hand for other minimal activities (reading a book or newspaper, using a keyboard) • You can carry light objects with your hand (a glass of water or other objects less than one pound)
Discouraged	• You should not be using your hand for strenuous activities (lifting a full grocery bag, performing household maintenance)

Supportive splinting

Participants randomized to the splint group were fitted with a custom plaster/fibreglass splint. A traditional plaster splint or Dynacast Prelude fiberglass splint was applied and secured using a 2″ elastic bandage (Figure 2). The splint immobilized the fifth metacarpophalangeal joint at 60° flexion and did not involve the proximal interphalangeal joint, leaving the joint free for movement (Figure 2). ²¹ Patients were encouraged to wear the splint most of the time but were allowed to use their hand for light activities. At 4 weeks follow-up, all splints were discontinued, and participants were provided the same behavioural modifications as the elastic bandage group (Online Appendix 1).

Figure 2.

Supportive splint.

Outcomes Measures

After consultation with the PI and allocation to their group, participants completed the bMHQ (Online Appendix 1) and the Demographics on Socioeconomic Class and Current Employment Questionnaire (DSCCEQ).

Brief Michigan Hand Questionnaire

The bMHQ is a validated, ³³ patient-reported, 12-item questionnaire pertaining to several aspects of hand function, and is measured via a 1 through 5 Likert scale. The items are totaled and normalized to yield a summary score on a scale of 0 to 100, in which higher scores indicate better overall function and satisfaction.

Demographics on Socioeconomic Class and Current Employment

The DSCCEQ is a patient-reported 7-item assessment of socioeconomic class and current employment. This assessment is a mixture of scaled questions and open-ended written questions that examines self-reported race, education status, employment status, family income, and type of work.

Grip strength

At the baseline assessment, participants were provided standardized instructions, and completed the grip strength test using a JAMAR Grip Dynamometer (Jamal Hydraulic Hand Dynamometer) in the second setting. Grip strength testing was performed on both the injured and non-injured hand and recorded as the average of 3 trials (in kilograms).

Clinical outcomes

Principal investigators monitored for complications and recorded any significant events.

Cost

The cost of each intervention was estimated for the main study site, Sunnybrook Health Sciences Center. These costs are approximate and may vary based on site. As the interventions were provided by the PIs, the costs included in the analysis included only the cost of the elastic bandage and/or splint, with no additional hand therapy costs.

Follow-Up

The DSCCEQ, the bMHQ, and grip strength testing were completed by participants at baseline and at each follow-up visit (weeks 4, 8, 12). In the event that a participant was unable to attend a follow-up visit in-person, the bMHQ was completed over the phone. If a participant failed to respond to the initial contact, 2 further attempts were made. If no follow-up appointment was scheduled, participants were classified as “lost to follow-up.”

Statistical Analysis

Continuous data were reported as means ± SD and categorical data were presented as a frequency and percent. To assess for differences in baseline demographics between the 2groups, a χ² test or Fisher exact test was performed as appropriate. To determine the difference between groups in continuous variables (ie, age, bMHQ, and grip strength), Mann-Whitney U non-parametric tests were performed.

Grip strength values were normalized to both the non-injured hand at baseline and the Canadian reference values. ³⁴ A linear regression was performed to examine the effect of age, family income, employment and education status, and type of work on the baseline bMHQ score and grip strength. An intention-to-treat analysis was performed. A P value of less than .05 was used to denote statistical significance. All analyses were performed using SPSS statistical package version 23 (IBM SPSS Statistics for Windows, Version 23.0).

Results

Participant Demographics

Thirty-seven participants were randomized to the splint (n = 21) or elastic bandage group (n = 16, Figure 3). Randomization at baseline was successful and as there were no statistically significant differences between groups for any study variable (Table 2). Our cohort was predominately male (97%), with an average age of 34 ± 14 years. A large proportion of participants were Caucasian (78%), with 55% earning >$50 000 per year. Most participants (72%) were permanent employees (full time or part time), had completed some degree of college or university (58%), and described their “type of work” as sedentary (47%). There was a significant difference (P = .04) in the number of participants lost to follow-up at week 4 between splint (n = 5) and elastic bandage (n = 10) groups. From baseline to week 12, both groups lost a similar number of participants loss to follow-up (splint n = 17, elastic bandage n = 14).

Figure 3.

Recruitment diagram displaying the number of participants and those lost to follow-up. Brief MHQ indicates Brief Michigan Hand Questionnaire; ED, emergency department; PI, Principal Investigator.

Table 2.

Baseline Participant Characteristics.^a

Participant Characteristics	Total (n = 37)	Splint (n = 21)	Elastic bandage (n = 16)	P
Age, years	34 ± 14	33 ± 15	31 ± 11	.69
Male	26 (96%)	20 (95%)	16 (100%)
Caucasian	29 (81%)	17 (81%)	12 (75%)	.71b
Family income
<$50 000	13 (36%)	7 (35%)	6 (38%)	1.00
>$50 000	23 (64%)	13 (65%)	10 (62%)
Employment
Permanent (full time or part time)	27 (75%)	14 (70%)	13 (81%)	.70
Not employed (unemployed, student, retired)	9 (25%)	6 (30%)	3 (19%)
Type of work
<Medium work (sedentary, light)	19 (53%)	11 (65%)	8 (62%)	1.00
>Medium work (or heavy)	11 (47%)	6 (35%)	5 (38%)
Hand-specific Information
Right-hand dominance	34 (92%)	21 (100%)	13 (81%)	.07
Right-hand injured	29 (78%)	18 (86%)	11 (69%)	.25
MHQ	49.7 ± 24.3	52.1 ± 27.2	46.6 ± 20.4	.51
Grip strength injured hand (kg)	17.3 ± 8.5	15.3 ± 8.9	19.9 ± 7.5	.15
Grip strength non-injured hand (kg)	42.1 ± 10.4	40.4 ± 11.4	44.1 ± 9.4	.20

^a Mann-Whitney U non-parametric test was performed on continuous data

^b A χ² or Fisher t test was used for categorical data. Significance was set at P < .05.

Brief Michigan Hand Questionnaire

At baseline, there were no statistically significant differences (P = .51) in the bMHQ score for the splint or the elastic bandage groups (Table 3). There was a similar increase in bMHQ from baseline to week 4 (P = .07) and week 8 (P = .15) for splint and elastic bandage groups (Table 3). At week 12, there were no statistically significant differences in bMHQ scores between the splint group and elastic bandage group (86.9 ± 9.4 and 77.1 ± 8.8, respectively; P = .16, Table 3, Figure 4).

Table 3.

Brief Michigan Hand Questionnaire Score.^a

MHQ Score	Total		Splint		Elastic bandage		P value
Baseline	N = 37	49.7 ± 24.3	N = 21	52.1 ± 27.2	N = 16	46.6 ± 20.4	.51
Week 4	N = 27	70.2 ± 17.2	N = 16	65.9 ± 16.4	N = 11	76.5 ± 17.0	.12
Week 8	N = 19	76.9 ± 18.9	N = 11	73.5 ± 22.1	N = 8	81.8 ± 13.3	.48
Week 12	N = 6	83.6 ± 9.7	N = 4	86.9 ± 9.4	N = 2	77.1 ± 8.8	.16
Delta change
Δ BL to week 4	N = 27	23.1 ± 21.9	N = 16	15.6 ± 19.3	N = 11	33.9 ± 21.8	.07
Δ BL to week 8	N = 19	34.6 ± 22.4	N = 11	29.2 ± 22.6	N = 8	42.2 ± 21.0	.15
Δ BL to week 12	N = 6	32.9 ± 17.9	N = 4	27.6 ± 20.5	N = 2	43.7 ± 0	.35

Abbreviations: BL, baseline; MHQ, brief michigan hand questionnaire; SD, standard deviation.

^a Data are presented in means ± SD and the delta change in bMHQ from baseline. A Mann-Whitney U non-parametric test was performed to examine differences between groups.

Figure 4.

Mean Brief Michigan Hand Questionnaire (bMHQ) score at baseline and follow-up visits. Mann-Whitney U non-parametric test was performed to examine difference between groups. Significance was set at P < .05. There were no statistically significant differences between groups.

A linear regression was performed to evaluate the association between age (P = .85), family income (P = .11), employment status (P = .77), and type of work (P = .82) on baseline bMHQ (Table 5). A significant association was found for education status and baseline bMHQ (P = .039, Table 5).

Table 5.

Linear Regression Analysis.^a

	Standardized β	95% CI	P value
Brief MHQ
Age	−0.03	−0.69 to 0.58	.85
Family income	−0.27	−30.71 to 3.11	.11
Employment status	−0.05	−22.25 to 16.70	.77
Education	−0.34	−32.07 to −0.87	.039*
Type of Wwrk	−0.04	−16.75 to 20.94	.82
Grip strength (kg)
Age	−0.28	−0.39 to 0.33	.09
Family income	−0.06	−7.26 to 5.28	.74
Employment status	0.34	−6.16 to 7.46	.85
Education	−0.05	−6.66 to 5.07	.79
Type of work	0.07	5.86 to 8.42	.72
Grip strength % of Canadian Reference Value
Age	−0.26	−0.87 to 0.18	.19
Family income	−0.05	−0.26 to 0.79	.79
Employment status	−0.11	−0.54 to 0.59	.59
Education	0.02	0.11 to 0.91	.91
Type of work	0.002	0.01 to 0.99	.99

^a A linear regression was performed.

Grip Strength

The average baseline grip strength of the injured hand prior to treatment was 17.4 ± 8.6 kg. This correlated with a grip of strength that was 46% of that of the non-injured hand, and 36% of that of the reference grip strength for Canadian men aged 20 to 39 years (48.5 kg). ³⁴ Between the splint group and the elastic bandage group, there was no significant difference in pre-treatment grip strength (15.3 ± 8.9 kg and 19.9 ± 7.5 kg, respectively; P = .15). This represented 45% of the non-injured hand and 32% of the Canadian reference value for the splint group, and 47% of the non-injured hand and 41% of the Canadian reference value for the elastic bandage group (Table 4, Figure 5).

Table 4.

Grip Strength: Percent of Non-injured Hand (Baseline) and Canadian Reference Values.^a

	All		Splint		Elastic bandage		P
	N	Mean ± SD	N	Mean ± SD	N	Mean ± SD
% of non-injured hand at baseline
Baseline (BL)	28	45.9 ± 18.8	15	45.3 ± 22.8	13	46.6 ± 13.6	.89
Week 4	17	71.5 ± 34.2	12	65.5 ± 37.8	5	85.9 ± 19.1	.19
Week 8	13	89.5 ± 20.6	7	80.4 ± 23.5	6	100.0 ± 10.6	.10
Week 12	2	93.0 ± 19.6	2	93.0 ± 19.6
% of Canadian Reference value for men aged 20-39 (48.5 kg)
Baseline (BL)	36	35.7 ± 17.5	20	31.5 ± 18.2	16	41.1 ± 15.4	.17
Week 4	23	61.4 ± 26.7	16	55.9 ± 24.1	7	73.9 ± 30.0	.20
Week 8	18	74.8 ± 24.5	11	63.0 ± 19.8	7	93.4 ± 19.9	.008b
Week 12	6	69.8 ± 20.9	4	70.1 ± 24.5	2	69.1 ± 19.9	1.00

Abbreviation: SD, standard deviation.

^a Data are presented in % as means ± SD. A Mann-Whitney U non-parametric test was performed to examine differences between groups.

^b Significantly different between groups (P < .05).

Figure 5.

Absolute value for grip strength (kg) of the injured hand at and baseline and follow-up visits. Mann-Whitney U non-parametric test was performed to examine difference between groups. *Significantly different between groups P < .05.

At week 8, there was a significant (P = .008) difference in the percent of the Canadian reference values between the splint (63%) compared to the elastic bandage group (93%). A linear regression was calculated for age, family income, employment status, education, and type of work to predict baseline grip strength with no statistically significant effect (Table 5).

Clinical Outcomes

No patients experienced complications. All patients had a good recovery, and none required hand therapy.

Cost

Elastic bandage

One elastic bandage was provided to each patient. As the cost of a 2″ elastic bandage manufactured by COVIDIEN LLC is 6$ for 12 elastic bandages, the individual patient cost was 0.50 CAN$.

Supportive splinting

We evaluated that each patient required approximately 25 cm of Dynacast Prelude 10 cm width splint manufactured by BSN Medical. As the cost of 4.6 m of splint is 63.95 CAN$, each patient’s splint costed approximately 3.48 CAN$. In addition, one elastic bandage was used to secure the splint, costing 0.50 CAN$. The total cost for the splint group was thus 3.98 CAN$ per patient.

Discussion

This was a prospective national multicenter randomized study that compared both functional and PROs in patients with Boxer’s fractures who were treated with supportive splitting versus elastic bandage with early protected movement. In addition, we performed a characterization of typical patients for which Boxer’s fractures is reported, and we examined the relationship between patient demographics and functional or PROs. Our results demonstrate that Boxer’s fractures tend to occur in young (∼35-year-old), educated men, who holds a permanent (full/part time) position at a sedentary job. From the 37 participants enrolled in the study, no significant difference in each groups’ demographics was observed, indicating a successful randomization.

Regarding the bMHQ, there was no significant difference in the splint or elastic bandage group scores and this was maintained until the last follow-up. At baseline, the average injured-hand grip strength of our cohort was 17.4 kg, which represented 46% of the non-injured hand and 36% of the reference grip strength for Canadian men aged 20 to 39 years. ³⁴ Normalization of injured-hand grip strength to the non-injured hand at baseline was performed to serve as an internal control. Interestingly, at week 8, patients treated with elastic bandage achieved 100% of the non-injured hand grip strength, whereas those treated with supportive splinting reached only 80%. In a similar manner, when normalized to the Canadian reference value, there was a significant difference between splint and elastic bandage groups. At week 8, the splint group achieved 63% and the elastic bandage group achieved 93% of the Canadian reference value. Thus, our results indicate that those treated with elastic bandage with early protected movement, or no treatment per se, experienced an earlier return to normal grip strength and hand function. This indicates that each treatment group had similar functional outcomes subjectively reported by the participants. There was also no effect observed based on associations between income, employment status, and type of work with bMHQ, indicating no confounding of these variables. Similarly, previous studies examining satisfaction and pain ^21,22 observed no differences between participants managed with or without a molded splint. Of note, the study was not powered to detect an association between these variables and bMHQ; future research should confirm the lack of an effect.

In the present investigation, we observed no association between participant demographics and baseline grip strength. This, however, was not experienced by those treated with supportive splinting. Intuitively, this is in keeping with basic hand biomechanics and therapy, the more range of motion and use of the hand would conceivably correlate with less muscle atrophy and stiffness and ultimately greater strength. This finding suggests that surgeons may be over treating Boxer’s fractures with splinting. These results may be underestimated, as there were significantly greater attrition rates with the elastic bandage group. Generally, patients stop returning to their surgeon once full function is recovered from hand injuries. ^3,35 Therefore, if there were more elastic bandage group participants lost to follow-up, it may be because their function recovered faster than the splint group. These assumptions need to be further evaluated through research. These data demonstrate that the elastic bandage is a suitable treatment for Boxer’s fracture and may even be superior to supportive splinting.

Previous studies utilizing a pressure bandage for one week followed by mobilization (21), self-adherent wrap covering metacarpal bones II to V and buddy taping (22), and buddy taping alone, ³⁶ shared similar results in that reduction and splinting of these fractures was not necessary. Our study adds to these findings and supports no splinting with an early protected movement of Boxer’s fractures with a degree of angulation less than 70°. This, however, does not include angulation causing a rotational deformity as those patients were excluded from this study and referred for operative management. Furthermore, our data indicate that patients with Boxer’s fractures may, inevitability, be over-treated for their condition. This hypothesis is supported by the number of participants, in both groups, that were lost to follow-up (∼80%), despite various opportunities and requests to meet with their physicians. Similarly, Gamble et al ³⁶ found that patients treated in the ED alone without further face-to-face follow-up had good functional outcomes and patient-reported satisfaction. Given the attrition rates in both our study groups, consideration can be made to minimize follow-up visits in clinical practice unless patients have specific concerns.

All patients, in our study, treated with elastic bandage alone recovered 100% of their grip strength by week 8. These findings are consistent with previous reports of excellent grip strength recovery for boxer fractures managed with buddy taping and soft wrap alone (20,22). Although only short-term functional and PROs were evaluated in this study, longer term follow-up from a prospective cohort study reported that patients managed without splinting remain pain-free with no functional limitations at 3-year follow-up.

Our study is not without limitations. Selection bias was minimized by using parallel group randomization with computer-generated allocation. However, the surgeons and therapists were not blinded to the treatment protocol. This may have contributed to detection bias with over or underestimation in the size of the effect. Despite this, methodological efforts were used to standardize the treatments between the 2 groups with intervention protocols for individuals administering treatment as well as instructions for patients to follow. However, we could not control for adherence of participants to the instructions or account for variability in daily activities and hand dominance. Although there were no significant differences seen in the 2 treatment groups as per the demographic questionnaire, these variables are difficult to control. Lastly, for patients allocated to the elastic bandage group, there was some duration of splinting by the emergency physician prior to assessment and inclusion in the study, albeit a short duration and not significant to allow bony ossification, this may have impacted the results.

Although a sample size calculation was performed, unfortunately, the study did not reach the target recruitment numbers. Attrition was much greater than anticipated (>80% over the entire study period compared to 25%). The attrition rate of the splinted group was 27% per month, whereas it was 29% per month in the elastic bandage intervention group. Our high attrition rate is in line with previous literature ^3,21,27 and continues to be a limitation when investigating treatment for Boxer’s fracture. Multicenter studies with high volumes of Boxer’s fractures are needed to confirm our findings. These studies should consider incentives or alternative assessment techniques (such as virtual follow-up) to decrease attrition rates. Lastly, regardless of the limited number of patients recruited, a significant difference in the grip strength was noted between groups at 8 weeks, a difference which would remain significant with increased study power.

This randomized multicenter study found that patients with a Boxer’s fracture are generally young, educated males, who holds a full- or part-time position. The present investigation highlights that patients treated with early protected movement had improved functional outcomes by 8 weeks compared to those treated with immobilization and splinting as measured by grip strength. Patients did not perceive a difference in hand function based on the bMHQ, a hand PRO. Overall, this group of patients has good subjective and objectives outcomes with minimal treatment. These findings are important in encouraging providers to manage Boxer’s fractures with early protected movement.