Does Digital Nerve Injury Affect Range of Motion Recovery After Zone 2 Flexor Tendon Repair?

Abstract

Background

Evidence surrounding the impact of concomitant digital nerve injury on the outcome of zone 2 flexor tendon repair is sparse and conflicting. The purpose of this study is to assess the impact of digital nerve injury on the range of motion recovery after zone 2 flexor tendon repair. We hypothesized that digital nerve injury is independently associated with decreased motion after zone 2 flexor digitorum profundus (FDP) repair.

Methods

This is a single-institution, multisurgeon retrospective analysis of patients treated with primary zone 2 FDP repair. Patients with or without digital nerve injuries were included. Patients with fracture, extensor tendon injury, dysvascularity, follow-up duration of less than 10 weeks, and younger than 15 years were excluded. The primary outcome measure was Strickland percentage at the last therapy visit. Bivariate analysis was performed using simple linear regression. These results were used to guide backward stepwise multivariable analysis of qualifying exploratory variables.

Results

Forty-one patients with a total of 54 zone 2 FDP injuries qualified. Mean follow-up duration was 24 ± 10 weeks, and mean age was 38 ± 18 years. Thirty-three digits had a concomitant digital nerve injury, 26 digits had multidigit involvement, and 42 digits had combined FDP and flexor digitorum superficialis (FDS) injuries. Both older age and concomitant FDS injury exhibited independent relationships with poorer range of motion outcomes (P < .05). Digital nerve injury, follow-up duration, gender, and multidigit involvement did not influence final digital motion.

Conclusions

In patients undergoing zone 2 FDP repair, concomitant digital nerve injury is not independently associated with poorer postoperative active range of motion.

Keywords

nerve injury nerve diagnosis tendon outcomes research and health outcomes digits anatomy epidemiology

Introduction

Flexor tendon injuries are a debilitating subset of hand injuries with an incidence of 4.8 to 7.0 per 100 000 person-years.^1,2 Although flexor tendon injuries comprise less than 1% of hand injuries, they affect the economy greatly, costing more than $250 million annually to the US health care system.^3,4 Treatment of flexor tendon injuries is a challenging undertaking which requires cooperation between the surgeon, hand therapist, and patient. Flexor tendon rehabilitation is exceedingly difficult in zone 2 injuries due to the tight anatomic relationship between the flexor digitorum superficialis (FDS) and flexor digitorum profundus (FDP) tendons within the fibroosseous sheath.

Given the formidable challenge of regaining motion after zone 2 FDP tendon repair, several studies have examined factors that may affect outcomes. Patient age and injury factors such as fracture, dysvasularity, multiple digit involvement, and associated extensor tendon injuries have all been found to affect range of motion outcomes.^5-10 However, while injury of the obtuse digital nerve(s) is exceedingly common, the evidence surrounding the impact of this concomitant injury is sparse and conflicting.^5,7,10,11 As such, it is the objective of this study to assess the impact of digital nerve injury on active range of motion (AROM) recovery after primary repair of the FDP in zone 2. It was our hypothesis that after controlling for demographic and injury-related covariates in a multivariable analysis, concomitant digital neve injury would be independently predictive of inferior range of motion.

Methods and Materials

Study Design

After receiving approval from an institutional review board, a retrospective chart review was completed for patients who underwent primary zone 2 digital flexor tendon repairs between November 2016 and December 2019. Patients were identified through query of departmental Current Procedural Terminology (CPT) coding databases. Medical records were reviewed for patient demographics, injury details, surgical procedure, and postoperative AROM. Inclusion criteria were patients older than 15 years with acutely repaired zone 2 FDP transections. Patients with concomitant FDS and/or digital nerve injury were included in the study cohort. For patients with multiple qualifying digits, each was considered separately, but status as a multiple digit repair was recorded and analyzed as a separate variable. Exclusion criteria were FDP repairs outside of zone 2, FPL injury, concomitant fracture, dislocation, dysvascularity, or extensor tendon injury in any digit. A minimum of 10-week postoperative follow-up including range of motion information was required for inclusion of any postoperative data point.

Surgery and Rehabilitation

Surgical repairs were performed by 11 hand surgeons at a single institution. Operative technique used for tendon repair varied per surgeon preference. Digital nerve injuries were either repaired directly or bridged with a nerve graft. Similar postoperative therapy regimens were ordered for all patients. On the first day of therapy, a custom dorsal blocking splint was fabricated to protect the tendon repair(s). Patients were instructed in passive digital flexion exercises, followed by active extension to the limits of the dorsal blocking splint. Active digital flexion was introduced based on patient factors such as edema, stiffness, and patient compliance. Patients were gradually weaned out of their dorsal blocking splints around 6 weeks after surgery. Therapy treatment was unchanged whether or not patients had a digital nerve injury.

Measures

Digital range of motion was collected retrospectively from the final visit with the physician or therapist. Total active motion was computed by the sum of flexion at the proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints minus any degree of extensor lag. Strickland percentage was then calculated from total active motion multiplied by 100 and divided by 175, which produces a percentage of normal range of motion.^12,13 Strickland percentages are classified into 4 categories: poor (0%-24% of normal), fair (25%-49% of normal), good (50%-74% of normal), and excellent (75%-100% of normal). Strickland percentage was used as a primary outcome measure in our analysis, but Strickland categories are also presented to provide clinical context.

Statistical Analysis

Patient characteristics were summarized as number (percent) for categorical variables and mean (SD) for continuous variables. Normally distributed continuous data were compared using unpaired Student t test. Categorical variables were compared using χ2 analysis with Yates correction. Simple linear regression analysis was used to examine bivariate relationships between each explanatory variable (age, gender, digital nerve injury, FDS injury, multiple digit involvement, and duration of follow-up) and the outcome variable (Strickland percentage). Variables were initially included and tested in a backward stepwise multivariable linear regression model if their bivariate P values were <.20. Statistical significance was defined as P < .05, and this was used as the criteria for inclusion in the final multivariable model. A post hoc power analysis revealed that 54 participants yielded 96% power detect significant relationships (P < .05) in a 4-predictor model with an observed R² of 28%.^14,15

Results

Our study cohort consisted of 41 patients undergoing 54 qualifying acute zone 2 FDP tendon repairs. The patient characteristics are summarized in Table 1. Mean time to primary repair was 3 days (range, 0-12 days). Thirty-five of the patients were men, 27 were dominant-sided injuries, and 28 were single-digit injuries. Mean age was 38 ± 18 years (range, 15-71 years). Mean follow-up duration was 24 ± 10 weeks (range, 10-52 weeks).

Table 1.

Patient Characteristics.

Patient Variable	Mean/Frequency
Patients	41
Male	35 (75%)
Female	6 (25%)
Age, y (mean +/– standard deviation)	38 ± 18
Single-digit injury	28 (68.3%)
Follow-up duration, wk (mean +/– standard deviation)	21 ± 11

Table 2 summarizes the injury and repair characteristics for each FDP repair. Of the 54 FDP tendons repaired, there were 14 index, 15 middle, 10 ring, and 15 small digits involved. A majority of the FDP tendons (53.7%) were repaired with either a 4-strand locking modified Kessler or 8-strand locking cruciate repair with epitendinous suture, although other locking techniques were also used. Thirty-three digits had a concomitant digital nerve injury in the same digit. Twenty-three digital nerves (69.7%) were repaired primarily and the remaining 10 (30.3%) required nerve grafting. Associated same-digit FDS transection was present in 42 digits (73%), and 35 of these FDS injuries (81%) were surgically repaired. Based on the Strickland categorizing system, excellent AROM was achieved in 14 digits, good AROM in 19 digits, fair AROM in 8 digits, and poor range of motion in 13 digits (Figure 1).

Table 2.

Injury Characteristics by Digit.

Injury Variable	Frequency
Total digits	54
Index	14 (26%)
Middle	15 (28%)
Ring	10 (19%)
Small	15 (28%)
FDS injured	42 (78%)
FDS repaired	35 (83%)
Both slips	30 (14%)
Single slip	5 (86%)
FDS unrepaired	7 (17%)
FDS uninjured	12 (22%)
Digital nerve injured	33 (61%)
Direct repair	23 (70%)
Nerve graft	10 (30%)
Digital nerve uninjured	21 (39%)

Note. FDS = flexor digitorum superficialis.

The clinical context for the nerve injured and non-nerve injured digits is given in Table 3. The nerve injury group had significantly larger rates of total and unrepaired FDS injuries, as well as a higher average age when compared with the no nerve injury group. However, there was no difference in the mean Strickland percentage between the groups (Figure 1). Owing to the differences in demographic and injury characteristics between the groups, the choice was made to use a stepwise multivariable regression to account for any possible confounders. Bivariate analysis (Table 4) revealed that older age and concomitant FDS injury were significantly associated with poor AROM outcomes (P < .05). Concomitant digital nerve injury and follow-up duration did not demonstrate a significant bivariate relationship with range of motion outcomes (P = .163 and .134, respectively), but the P value for these variables fell below the cutoff of P < .2 for entry into the multivariable analysis. Gender and multiple digital involvement were not found to have bivariate relationships with range of motion outcome (P > .2).

Table 3.

Demographic and Injury Characteristics in the Nerve Injured and Nerve Uninjured Groups.

Explanatory Variables	Nerve injured (n)	Nerve uninjured (n)	P value
Digits	33	21
Gender			.617
Male	27	16
Female	6	5
Strickland percentage (mean)	50.3	61.7	.131
Age, y	44.3	33.6	.025
Single-digit injury	13	13	.106
Follow-up duration, wk (mean)	22	20.5	.652
FDS injured	30	12	.004
FDS repaired	24	11	.001
FDS unrepaired	6	1
FDS uninjured	3	9

Note. FDS = flexor digitorum superficialis.

Table 4.

Bivariate Analysis for Association With Strickland Percentage.

Variable	Mean Strickland %	R ²	Unstandardized β	SE β	P value
Gender
Male	53	0.12	−9.03	10.53	.395
Female	62.03	0.12	−9.03	10.53	.395
Digital nerve injury
Yes	50.32	0.199	−11.6	8.17	.169
No	61.92	0.199	−11.6	8.17	.169
Multidigit injury
Yes	56.5	0.068	−3.89	8.01	.629
No	52.62	0.068	−3.89	8.01	.629
Flexor digitorum superficialis injury
Yes	67.73	0.124	−20.61	7.8	.010
No	46.98	0.124	−20.61	7.8	.010
Age		0.159	−0.621	0.201	.003
Follow-up duration		0.044	−0.622	0.408	.134

Because multiple explanatory variables were found to have significant (age, FDS injury) or near-significant (digital nerve injury, follow-up duration), bivariate relationships with Strickland percentage, multivariable analysis was indicated to assess the independent predictive value of each. Stepwise multivariable regression demonstrated that FDS injury and age each have significant independent relationships with the Strickland percentage (P = .034 and .005). The β coefficients for patient age and FDS injury were −0.56 and −18.05, respectively, and the R² for the model was 25.1% (Table 5). Follow-up duration and digital nerve injury did not have significant independent associations with outcome.

Table 5.

Stepwise Multivariable Linear Regression Analysis of Strickland Percentage: Final Model.

R ²	Variable	Unstandardized β	SE β	P value
0.251	Flexor digitorum superficialis injury	−18.05	7.35	.018
0.251	Age	−0.56	0.19	.006

Discussion

The central aim of this study was to determine whether concomitant digital nerve injury has an impact on AROM after zone 2 flexor tendon repair in the same digit. At present, the evidence surrounding this question is sparse and conflicting. In a randomized controlled trial comparing active and passive motion therapy protocols after flexor tendon injury, Trumble et al⁷ found a significant decrement in digital AROM for patients with concomitant digital nerve injuries. Specifically, digital nerve injury predicted 16° less motion (at the DIP and PIP joints combined) when compared with digits without nerve injuries. In addition, patients with digital nerve injury had larger flexion contractures, decreased satisfaction, and worse function on the Disabilities of the Arm Shoulder and Hand questionnaire. However, because this article was not focused on digital nerve injury per se, multivariable analysis was not performed, and it remains possible that factors such as age and FDS injury could confound this relationship.

Although children were excluded from this study, it is interesting to note that Elhassan and his group at the Mayo Clinc came to a similar conclusion as Trumble et al⁷ when studying pediatric zone 1 and 2 flexor tendon repairs.⁵ In their comparison of 23 isolated tendon and 18 combined nerve/tendon injuries in patients aged 2 to 16 years, they detected a statistically significant 13% deficit in total AROM for patients with combined injury.⁵ However, multivariable analysis was again not included, and it remains possible that other factors could confound this relationship.

In contrast to the American literature, a duo of Swedish studies found no impact of digital nerve injury on AROM recovery after flexor tendon repair. Edsfeldt et al¹⁰ investigated a host of prognostic factors in 311 flexor tendon repairs within zones 1 and 2. In their analysis, there was no significant bivariate relationship between digital nerve injury and AROM recovery. However, the inclusion of multiple zones and lack of multivariable analysis in this study are significant limitations. Rosberg et al¹⁶ examined zone 2 flexor tendon injuries for the purpose of analyzing the impact of patient and treatment variables on health care cost. As part of a secondary analysis of AROM outcomes, they found that digital nerve injury had no significant impact on motion in a multivariable model. However, the inclusion criteria for this study are unclear, and it is difficult to determine whether patients with confounding injuries such as fracture, dysvascularity, or extensor tendon injury were included. Furthermore, because of technical and theoretical advancements affecting the processes of digital nerve repair, flexor tendon repair, and flexor tendon rehabilitation, it is unclear whether the outcomes of this study, which includes patients treated from 1989 to 1998, are still applicable today.

Recognizing the limitations of the existing literature, this study attempts to isolate the impact of digital nerve injury on AROM after flexor tendon repair using methodological and statistical means. This begins by using narrow inclusion/exclusion criteria which eliminated children (who seem to fare disproportionately well) and patients with complex injuries involving bone, extensor tendons, and/or vascular structures. In addition, the effects of gender, age, FDS injury, follow-up duration, and multidigital injury were controlled using sequential bivariate and multivariable analyses.

Preliminary bivariate analysis clarified that sex and multidigital injury had little or no impact on AROM. Conversely, significant bivariate relationships were detected for age and concomitant FDS involvement. Follow-up duration and digital nerve injury had nonsignificant bivariate relationships with AROM. However, given their marginal bivariate P values for these variables and the possibility of interaction with the other covariates, entry into the multivariable modeling process was deemed appropriate. Backward stepwise multivariable regression beginning with these 4 variables clarified that in our series, only age and FDS injury have significant independent predictive value for AROM, whereas follow-up duration and digital nerve injury do not.

Based on this analysis, the hypothesis of this study is refuted. This is in agreement with findings of Edsfeldt et al¹⁰ and Rosberg et al,¹⁶ who found no effect of digital nerve injury, and inconsistent with the findings of Trumble et al⁷, who detected a significant negative impact of digital nerve injury. It is surprising that our findings align with the 2 Swedish studies, and diverge from the study performed at another major US trauma center and teaching institution. This suggests that social and cultural variables are unlikely to explain the different conclusions of these prior studies. Rather, it is likely that the slightly differing inclusion criteria, therapy protocols, and outcome assessment methods used explain the conflicting evidence.

Secondary findings of this study include the substantial influence of age and FDS injury on AROM outcome. Both these findings are consistent with the existing literature.^8,9 For each year of increased age, we detected a decrement of 0.56 Strickland percentage points achieved at the final follow-up. The reason for this negative impact of advanced age is incompletely understood, although alterations in the inflammatory response to injury and superimposed degenerative joint disease may play a role.^9,17 Concomitant FDS injury was associated with a decrement of 18% of normal motion in this series. Because most FDS injuries in this series were repaired, most of which with both slips of the FDS, we are unable to assess the impact of FDS repair versus single slip repair or benign neglect. This topic remains controversial, and there is evidence to support either practice.⁸,^18-23

Limitations of this study include its retrospective cohort design and modest population size. While randomization is not possible when examining the impact of an injury variable (digital nerve transection), this design leaves several possible confounders uncontrolled. Among these are the surgical method of tendon repair and rehabilitation protocol. While the prescribed protocol was similar for all patients, individual implementation may have varied, especially when orders were sent to therapists outside the institution. The presence or absence of a nerve injury did not change the postoperative therapy protocol. In addition, we have studied only a single measure of outcome (AROM), and deeper understanding could be gained from additional outcome measures. Furthermore, this study included patients with multiple digital injuries, but analyzed each digit separately. While it seems unlikely to have a substantial effect, this approach does allow the possible confounder of a nerve injury on 1 digit affecting the outcome of an adjacent digit without a nerve injury. Finally, given the evidence that outcomes measured as early as 8 weeks postoperatively correlate well with the late follow-up, we chose a relatively permissive minimum follow-up duration of 10 weeks.^8,9 While we found no relationship between follow-up duration and outcome in our multivariable model, it remains possible that further improvement in AROM could occur after this time point, and that stricter inclusion criteria could alter our conclusions.

Evidence surrounding the impact of concomitant digital nerve injury on AROM recovery after zone 2 flexor tendon repair and rehabilitation is sparse and conflicting. In this narrowly focused study, multivariable analysis revealed no independent association between digital nerve injury and postoperative AROM. This information can be used to reassure patients that digital nerve injury is unlikely to affect their ability to regain motion and adds to our understanding of how and why we continue to see variability in attained outcomes. In addition, our results suggest that patients with and without digital nerve injury can reasonably be pooled in future studies examining the impact of various factors (ie, therapy protocols or surgical techniques) on AROM recovery after flexor tendon repair. Further study is warranted to clarify whether digital nerve injury may affect outcomes other than AROM, such as patient satisfaction, function, and pain.

Footnotes

Ethical Approval

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Because this was a purely retrospective minimal risk study, our institutional review board waived the necessity for informed consent. All patient data presented are completely deidentified

Statement of Human and Animal Rights

This article does not contain any studies with human or animal subjects.

Statement of Informed Consent

Informed consent was obtained when necessary.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Grace Keane

Mitchell A. Pet

References

Clayton

RAE

Court-Brown

. The epidemiology of musculoskeletal tendinous and ligamentous injuries. Injury. 2008;39(12):1338-1344.

Manninen

Karjalainen

Määttä

, et al. Epidemiology of flexor tendon injuries of the hand in a Northern Finnish population. Scand J Surg. 2017;106(3):278-282.

Mehrzad

Mookerjee

Schmidt

, et al. The economic impact of flexor tendon lacerations of the hand in the United States. Ann Plast Surg. 2019;83(4):419-423.

Hill

Riaz

Mozzam

, et al. A regional audit of hand and wrist injuries. A study of 4873 injuries. J Hand Surg Br. 1998;23(2):196-200.

Elhassan

Moran

Bravo

, et al. Factors that influence the outcome of zone I and zone II flexor tendon repairs in children. J Hand Surg Am. 2006;31(10):1661-1666.

Kasashima

Kato

Minami

Factors influencing prognosis after direct repair of the flexor pollicis longus tendon: multivariate regression model analysis. Hand Surg. 2002;7(2):171-176.

Trumble

Vedder

Seiler

III , et al. Zone-II flexor tendon repair: a randomized prospective trial of active place-and-hold therapy compared with passive motion therapy. J Bone Joint Surg Am. 2010;92(6):1381-1389.

Rigo

Røkkum

Predictors of outcome after primary flexor tendon repair in zone 1, 2 and 3. J Hand Surg Eur Vol. 2016;41(8):793-801.

Silfverskiöld

May

Oden

. Factors affecting results after flexor tendon repair in zone II: a multivariate prospective analysis. J Hand Surg Am. 1993;18(4):654-662.

10.

Edsfeldt

Eklund

Wiig

Prognostic factors for digital range of motion after intrasynovial flexor tendon injury and repair: long-term follow-up on 273 patients treated with active extension-passive flexion with rubber bands. J Hand Ther. 2019;32(3):328-333.

11.

Daluiski

, et al. The epidemiology of reoperation after flexor tendon repair. J Hand Surg Am. 2012;37(5):919-924.

12.

Strickland

Glogovac

SV.

Digital function following flexor tendon repair in Zone II: a comparison of immobilization and controlled passive motion techniques. J Hand Surg Am. 1980;5(6):537-543.

13.

Fujihara

Ota

Watanabe

Utility of early active motion for flexor tendon repair with concomitant injuries: a multivariate analysis. Injury. 2018;49(12):2248-2251.

14.

Soper D. Post-hoc statistical power calculator for multiple regression [Software]. 2001. https://www.danielsoper.com/statcalc/calculator.aspx?id=9

15.

Cohen

Statistical Power Analysis for the Behavioral Sciences. New York, NY: Academic Press; 2013.

16.

Rosberg

Carlsson

Höjgård

, et al. What determines the costs of repair and rehabilitation of flexor tendon injuries in zone II? A multiple regression analysis of data from southern Sweden. J Hand Surg Br. 2003;28(2):106-112.

17.

Ackerman

Bah

Jonason

, et al. Aging does not alter tendon mechanical properties during homeostasis, but does impair flexor tendon healing. J Orthop Res. 2017;35(12):2716-2724.

18.

Lister

Kleinert

Kutz

, et al. Primary flexor tendon repair followed by immediate controlled mobilization. J Hand Surg Am. 1977;2(6):441-451.

19.

Moriya

Yoshizu

Maki

, et al. Clinical outcomes of early active mobilization following flexor tendon repair using the six-strand technique: short-and long-term evaluations. J Hand Surg Eur Vol. 2015;40(3):250-258.

20.

Nielsen

Jensen

. Primary flexor tendon repair in “no man’s land.” J Hand Surg. 1984;9(3):279-281.

21.

Brunelli

Vigasio

Brunelli

Slip-knot flexor tendon suture in zone II allowing immediate mobilisation. Hand. 1983;15(3):352-358.

22.

Ikuta

Tsuge

Postoperative results of looped nylon suture used in injuries of the digital flexor tendons. J Hand Surg Br. 1985;10(1):67-72.

23.

Tang

JB.

Flexor tendon repair in zone 2C. J Hand Surg Br. 1994;19(1):72-75.