Military Service Members with Major Lower Extremity Fractures Return to Running with a Passive-dynamic Ankle-foot Orthosis: Comparison with a Normative Population

Abstract

Background

Lower extremity fractures represent a high percentage of reported injuries in the United States military and can devastate a service member’s career. A passive dynamic ankle-foot orthosis (PD-AFO) with a specialized rehabilitation program was initially designed to treat military service members after complex battlefield lower extremity injuries, returning a select group of motivated individuals back to running. For high-demand users of the PD-AFO, the spatiotemporal gait parameters, agility, and quality of life is not fully understood with respect to uninjured runners.

Questions/purposes

Do patients who sustained a lower extremity fracture using a PD-AFO with a specialized rehabilitation program differ from uninjured service members acting as controls, as measured by (1) time-distance and biomechanical parameters associated with running, (2) agility testing (using the Comprehensive High-level Activity Mobility Predictor performance test and Four Square Step Test), and (3) the Short Musculoskeletal Function Assessment score.

Methods

We conducted a retrospective data analysis of a longitudinally collected data registry of patients using a PD-AFO from 2015 to 2017 at a single institution. The specific study cohort were patients with a unilateral lower extremity fracture who used the PD-AFO for running. Patients had to be fit with a PD-AFO, have completed rehabilitation, and have undergone a three-dimensional (3-D) running analysis at a self-selected speed at the completion of the program. Of the 90 patients who used the PD-AFO for various reasons, 10 male service members with lower extremity fractures who used a PD-AFO for running (median [range] age 29 years [22 to 41], height 1.8 meters [1.7 to 1.9], weight 91.6 kg [70 to 112]) were compared with 15 uninjured male runners in the military (median age 33 years [21 to 42], height 1.8 meters [1.7 to 1.9], weight 81.6 kg [71.2 to 98.9]). The uninjured runners were active-duty service members who voluntarily participated in a gait analysis at their own self-selected running speeds; to meet eligibility for inclusion as an uninjured control, the members had to be fit for full duty without any medical restrictions, and they had to be able to run 5 miles. The controls were then matched to the study group by age, weight, and height. The primary study outcome variables were the running time-distance parameters and frontal and sagittal plane kinematics of the trunk and pelvis during running. The Four Square Step Test, Comprehensive High-level Activity Mobility Predictor scores, and Short Musculoskeletal Function Assessment scores were analyzed for all groups as secondary outcomes. Nonparametric analyses were performed to determine differences between the two groups at p < 0.05.

Results

For the primary outcome, patients with a PD-AFO exhibited no differences compared with uninjured runners in median (range) running velocity (3.9 meters/second [3.4 to 4.2] versus 4.1 meters/second [3.1 to 4.8], median difference 0.2; p = 0.69), cadence (179 steps/minute [169 to 186] versus 173 steps/minute [159 to 191], median difference 5.8; p = 0.43), stride length (2.6 meters [2.4 to 2.9] versus 2.8 meters [2.3 to 3.3], median difference 0.2; p = 0.23), or sagittal plane parameters such as peak pelvic tilt (24° [15° to 33°] versus 22° [14° to 28°], median difference 1.6°; p = 0.43) and trunk forward flexion (16.2° [7.3° to 23°) versus 15.4° [4.2° to 21°), median difference 0.8°; p > 0.99) with the numbers available. For the secondary outcomes, runners with a PD-AFO performed worse in Comprehensive High-level Activity Mobility Predictor performance testing than uninjured runners did, with their four scores demonstrating a median (range) single-limb stance of 35 seconds (32 to 58) versus 60 seconds (60 to 60) (median difference 25 seconds; p < 0.001), t-test result of 15 seconds (13 to 20) versus 13 seconds (10 to 14) (median difference 2 seconds; p < 0.001), and Illinois Agility Test result of 22 seconds (20 to 25) versus 18 seconds (16 to 20) (median difference 4; p < 0.001). Edgren side step test result of 20 meters (16 to 26) versus 24 meters (16 to 29) (median difference 4 meters; p = 0.11) and the Four Square Step Test of 5.5 seconds (4.1 to 7.2) versus 4.2 seconds (3.1 to 7.3) (median difference 1.3 seconds; p = 0.39) were not different between the groups with an effect size of 0.83 and 0.75, respectively.

Conclusion

The results of our study demonstrate that service members run with discernible differences in high-level mobility and demonstrate inferior self-reported patient functioning while having no differences in speed and biomechanics compared with their noninjured counterparts with the sample size available. This study is an early report on functional gains of highly motivated service members with major lower extremity injuries who use a PD-AFO and formalized therapy program to run.

Level of Evidence

Level III, therapeutic study.

Introduction

Lower extremity fractures represent a high percentage of reported injuries in the United States military and can devastate a military service member’s career. Fractures represent between 12% and 65% of combat injuries, and from 2000 to 2006, there were 157,096 ankle and 92,581 foot injuries in the United States Army alone [20, 21]. Debilitating pain and functional deficits develop in some service members after their injuries, and they are unable to return to their previous level of function, which may preclude a return to active duty.

Using a passive-dynamic ankle-foot orthosis (PD-AFO) with a specialized return-to-run rehabilitation program provides a service member with an opportunity to return to high-level functional activities after severe musculoskeletal lower extremity injury [6, 15, 17]. Previous research has demonstrated that along with a specialized return-to-run rehabilitation program, patients with severe below-the-knee musculoskeletal injuries can return to active duty and high-level functional activities such as running with a PD-AFO [2, 8, 13, 14, 16, 18].

In a previous study, the senior author (KMK) reported on walking biomechanics and spatiotemporal gait parameters in patients who underwent limb preservation and used the PD-AFO to walk; the author found that these patients walk at a slower cadence and have a shorter stance time than patients with amputation who use a prosthesis [8]. However, the stiffness and energy-storing design of the brace are more suited for higher-level function than walking. Another study showed that patients with pilon fractures demonstrated no improvements in gait parameters or pain after application and use of a PD-AFO and concluded that in lower-demand users, the device is unlikely to be cost-effective. Neither of these papers evaluated a group of high-demand brace users who run routinely, for whom a PD-AFO may actually be beneficial [18].

We also wondered how the kinematics and functional outcomes of patients who used the PD-AFO while running, where the PD-AFO was intended for use, differed from a group of uninjured controls. This is particularly interesting to us because part of rehabilitation involves not only strength and endurance training but also understanding what is pathological and different about an injured service member’s gait at self-selected running speeds compared with an uninjured runner. Identifying differences is the next step in improving each service member’s running technique and ability, and this may potentially help them meet their goals and make the most of their brace.

We therefore asked: Do patients who sustained a lower extremity fracture using a PD-AFO with a specialized rehabilitation program differ from uninjured service members acting as controls, as measured by (1) time-distance and biomechanical parameters associated with running, (2) agility testing (using the Comprehensive High-level Activity Mobility Predictor performance test and Four Square Step Test), and (3) the Short Musculoskeletal Function Assessment Score.

Patients and Methods

To answer the question of how high-demand users of a PD-AFO compared with uninjured runners, we performed a retrospective analysis of a prospectively collected data registry of patients who were prescribed a PD-AFO for unilateral lower extremity fractures for high-level functional activities and used it for high-demand activity, specifically running. All patients were treated at a single institution from January 2015 to June 2017 (NMCSD.2014.0026). The study group was compared with a group of uninjured controls, who were active-duty service members, on full duty at the time of their running study without limitations, who passed their biannual military physical fitness test, and who actively ran at least 5 miles a week. These participants were recruited under a separate prospective institutional review board protocol that specified inclusion of healthy military runners (NMCSD.2013.0109). The study group was matched to the uninjured runners by age, gender, height, and weight.

Study Design and Setting

As part of standard clinical care, a multidisciplinary team comprised of an orthotist, physical therapist, physical medicine and rehabilitation physician, and an orthopaedic surgeon determined patient eligibility to receive a PD-AFO at weekly interdisciplinary gait lab meetings. This team specifically identified patients who might benefit from PD-AFO usage in their recovery based on their injury, symptoms, limitations, and goals. Proximal injuries and contralateral injuries were not contraindications for PD-AFO usage. Patients with multiple sclerosis, proximal weakness due to neurological involvement, severe joint laxity or muscle weakness, and/or a diagnosis of complex regional pain syndrome were deemed ineligible to receive the PD-AFO.

Participants

During the study period, the PD-AFO was prescribed for 90 patients, nine of whom were female. Most study participants (n = 77) used the PD-AFO for lower-demand activities such as walking, hiking, and gym use or for sports participation. Only 13 patients actively used their PD-AFO for running on a routine basis and opted to undergo a voluntary three-dimensional (3-D) gait analysis as an additional clinical treatment option to evaluate their running at the end of the return-to-run rehabilitation program. Ten of 13 patients who volunteered for the additional gait analysis were prescribed the PD-AFO to address lower extremity fractures and comprised our current study group.

As stated above, the control group consisted of active-duty service members, who were uninjured, full duty without any restrictions, had passed their biannual physical fitness test, and ran at least 5 miles a week. A total of 29 running controls were enrolled during the study period.

Patients and Controls

Ten male service members with lower extremity fractures who used a PD-AFO for running were compared with 15 uninjured military runners who were matched for age, height, and weight; the former group was called the PD-AFO lower extremity (LE) cohort. With the numbers available, the PD-AFO LE cohort and control group did not differ in terms of median (range) age (29 years [22 to 41] versus 33 years [21 to 42]; p = 0.10), height (1.8 meters [1.7 to 1.9] versus 1.8 meters [1.7 to 1.9]; p = 0.70), or weight (92 kg [70 to 112] versus 82 kg [71 to 98]; p = 0.11). All service members in the study and matched control groups were male due to the sample size and overall small population of females in the military at the time of this study.

The PDA-AFO LE cohort consisted of eight enlisted personnel and one officer (one participant declined to respond). Six of 10 individuals returned to active duty after sustaining their injuries. Injuries included an ankle fracture dislocation, open ankle fracture, talus fracture, Lisfranc fracture, calcaneus fractures, multiple midfoot fractures, pilon fractures, and tibial shaft fractures (Table 1). Four of 10 patients sustained more than one fracture, 8 of 10 patients had unilateral injuries, and 2 of 10 patients had bilateral lower extremity fractures. The mechanisms of injury were an improvised explosive device (two), motor vehicle or motorcycle collision (three), and training accidents or falls (five). The median (range) time from the date of injury to the running study was 5 years (0.7 to 15).

Table 1.

Characteristics of the PD-AFO study group

Agea	PD-AFO Side	Active duty (Y/N) a	Mechanism of injury	Orthopaedic injuries	Total number of relevant surgeries	Time from injury to running study, years
28	Right	Y	Training accident (fall from a height)	Right ankle fracture dislocation (Weber C with syndesmotic disruption)	2	1.3
29	Right	N	Motor vehicle accident	Right open ankle fracture, right femur fracture, right patella fracture, left tibial shaft/fibula fracture	5	8.7
22	Right	Y	Motorcycle accident	Right multiple midfoot fracture (talus, navicular, cuneiform, third and fourth metatarsals, cuboid, calc, hallux)	2	0.7
28	Right	N	Five-story fall	Complex comminuted pelvic fracture, right proximal femur fracture, right distal tibial shaft/fibula fracture, right calcaneus fracture, right distal ulna and scaphoid fractures	6	5.4
31	Right	N	Improvised explosive device	Right comminuted tibial shaft/fibula fracture, right peroneal nerve laceration, right tibialis anterior injury	3	12
41	Left	Y	60-foot fall rappelling from a line	Left pilon fracture, right cuboid fracture	4	15
26	Right	Y	Football	Right Lisfranc fracture	1	1.5
31	Right	Y	Training accident (300-pound tire fell on his heel when he was on his toes)	Right Lisfranc fracture (homolateral dislocation of TMT 1-4)	1	0.7
32	Left	Y	Motorcycle accident	Left talus fracture, left third metatarsal fracture	Unknown	0.8
26	Right	N	Improvised explosive device	Right pilon fracture/distal tibial-fibula fracture	4	5.8

^aDenotes at the time of the running study; TMT = tarsometatarsal.

Treatment Protocols

Patients in the PD-AFO LE cohort must have been fitted with a PD-AFO, finished the return-to-run program, and completed a 3-D running analysis at a self-selected speed at the completion of the program. Specifically, the PD-AFO selected for the study population was an energy-storing, plantar-flexed, supramalleolar, solid ankle-foot orthosis consisting of a foot plate with a posteriorly mounted carbon fiber strut that helps with energy return [6, 17]. The return-to-run clinical pathway involves a multidisciplinary approach to patient care and incorporates orthopaedic surgery, physical therapy, orthotists, physiatrists, mental health, and pain management specialists [1, 2, 13, 14]. This pathway offers a custom, energy-storing PD-AFO in combination with a high-intensity rehabilitation program.

Once a patient was deemed an appropriate candidate for PD-AFO prescription, a single certified orthotist casted and custom fabricated a PD-AFO for each patient. The patient would subsequently work closely with the orthotist to make modifications to ensure a proper fit and function of the brace, then start the return-to-run program. A single licensed physical therapy assistant with a certification in athletic training led the return-to-run program, which involved a 6- to 8-week intensive physical therapy program with three phases of gait retraining, strengthening, and agility using the custom PD-AFO for assistance. This models the established return-to-run clinical pathway that has been demonstrated to be clinically reproducible [11, 17].

Upon completion of the return-to-run rehabilitation program, additional coaching and clinical support in brace utilization was offered for those high-demand participants who used their PD-AFO to run, in the form of a formal 3-D running analysis, if running was clinically relevant for them. The goal of the return-to-run program is to improve the running of the service member, mainly with strength, endurance, and agility training. An additional part of this rehabilitation program involves examining what is pathological about their running technique.

Service members in both the PD-AFO LE cohort and uninjured control groups ran through a 3-D motion capture laboratory consisting of 12 infrared cameras (Motion Analysis Inc), four force plates (Kistler Instrument Corp), and high-speed reference video cameras. The participants were markered with a modified 6-degrees-of-freedom marker set and completed a minimum of five complete gait cycles and three clean force plate strikes per side. Data were processed in Cortex (Motion Analysis Inc) and analyzed in Visual 3D (C-Motion Inc).

Primary and Secondary Study Outcomes

Our primary study goal was to assess time-distance parameters as well as frontal and sagittal plane kinematics of the trunk and pelvis. Utilizing the 3-D motion capture laboratory as described above, velocity, cadence, stride length, trunk lateral flexion, trunk forward flexion, pelvic tilt, and pelvic obliquity were captured and measured at the time of the running study. The running study was performed after completion of the return-to-run program, with the median (range) number of rehabilitation visits with the PD-AFO being 16 (10 to 21) and performed in the gait lab for the control group.

The secondary outcomes of this study assess agility and high-level mobility, as well as self-reported patient functioning on everyday life. Specifically, agility and high-level mobility were assessed using the Four Square Step Test and Comprehensive High-level Activity Mobility Predictor (CHAMP) scores [3, 5, 10]. CHAMP quantified high-level mobility with a combination of four tests: the single-leg stance, which evaluates balance and postural stability; the Edgren side step test, which judges frontal plan agility and body control; the T-test, which measures sagittal and frontal agility, evaluating the ability to change directions rapidly; and the Illinois Agility Test, which considers multidirectional agility in the frontal, sagittal, and transverse planes with an emphasis on acceleration [5]. This agility testing was also performed as part of the final running study for the study group and performed separately in the gait lab for the control group. Self-reported patient functioning was quantified using the Short Musculoskeletal Function Assessment and gathered at the time of the running study for the control group or during participation in the return-to-run program for the study group [22].

Bias

Our study is subject to selection bias and transfer bias. First, not every patient with a unilateral lower extremity fracture in our catchment area received a PD-AFO and specialized rehabilitation protocol. We also understand that not everyone needs a PD-AFO and specialized rehabilitation protocol. The patients who were ultimately prescribed this device had chronic pain and permanent function loss that was unresponsive to standard conservative therapies; they were identified by the multidisciplinary group as potentially benefiting from a PD-AFO and the return-to-run program. None of the studied patients would be able to run without the device. Additionally, 4 of 10 patients in the study cohort are essentially limb salvage patients due to their high mechanism of injury and multiple injuries (2 of 5 bone and soft tissue injuries) or combat blast injury resulting in a functional deficit [8]. With regard to the general population, this is a highly selected group of individuals, but the pool of patients are those who had poor outcomes after routine recovery from their initial injuries and who needed additional intervention. This patient population thus represents a subset of patients who are highly debilitated from their injuries, either because they were limb salvage patients or because they had persistent pain and dysfunction that would not allow them to run at the levels demanded by the military, and yet they were highly motivated. The skew in gender representation also represents a potential risk of selection bias. Although there are females in the registry—about 10%, which is reflective of the proportion of females in the military—none of them utilized the PD-AFO for routine running during the study period, and one of the females went on to amputation. The lack of females who met the specific study criteria may be due to females not getting injured at the severity of their male counterparts (it was in 2017 that the United States Marine Corps graduated their first female infantry officer and had only just begun to transition females into combat roles) or perhaps deciding to pursue another career outside of the military after a severe injury [23] . It could also simply be reflective of the small patient sample size.

Additionally, there may be a component of transfer bias. We did not study all patients who have ever had issues with running because of a unilateral lower extremity fracture in the military. We acknowledge that some may have been medically separated from of the military, some may not have been afforded the opportunity to travel to our tertiary care center for this specialized treatment, and that follow-up across the board for all 90 PD-AFO individuals who were prescribed the brace is unfeasible in the setting of constant movement in the military. To help provide more context and address this bias, we do report on median follow-up after initial injury, to show how far out these patients are running after their injuries. Because of this bias, this paper cannot comment on if and when the PD-AFO should be used, but rather focuses on the best outcome possible for a select group of highly debilitated but highly motivated individuals.

We also acknowledge there may be a component of assessment bias. We believe in the PD-AFO device combined with the specialized rehabilitation program because we have seen it change lives and keep patients fit and fighting. Although our numbers would suggest that most PD-AFO users utilize their device for activities outside of strictly running, we have a vested interest in helping our patients achieve their goals, with or without the device. Furthermore, none of the authors have any personal or financial interests or investments in the PD-AFO utilized.

Ethical Approval

We obtained ethical approval for this study from the Naval Medical Center San Diego, San Diego, CA, USA (IRB CIP# NMCSD.2014.0026).

Statistical Analysis

Normal distributions of time-distance gait parameters, agility testing scores, and self-reported functioning were not seen among the study and uninjured population, so statistical significance of the median difference from the normative values was determined using nonparametric independent samples testing for nonnormally distributed data. Data were reported in medians, ranges, and p values, with significance set at a p value of 0.05. If p values did not demonstrate a significant difference, effect sizes were also reported. For the current study sample size, a large effect of 0.8 and greater was considered be sufficient to detect a difference with 80% power.

Results

Biomechanical Parameters While Running

With the numbers available, there were no differences in median (range) running velocity (PD-AFO LE cohort 3.9 meters/second [3.4 to 4.2] versus control 4.1 meters/second [3.1 to 4.8], median difference 0.2; p = 0.69), cadence (PD-AFO LE cohort 179 steps/minute [169 to 186] versus control 173 steps/minute [159 to 191], median difference 5.8; p 0.43), stride length (PD-AFO LE cohort 2.6 meters [2.4 to 2.9] versus control 2.8 meters [2.3 to 3.3], median difference 0.2; p = 0.23), stride width (PD-AFO LE cohort 9.8 cm [5.4 to 15] versus control 8.3 cm [3.0 to 16], median difference 1.5; p = 0.43), affected limb step length (PD-AFO LE cohort 1.3 meters [1.2 to 1.4] versus control 1.4 meters [1.1 to 1.6], median difference 0.1; p = 0.23), or sagittal plane parameters such as peak pelvic tilt (PD-AFO LE cohort 24° [15° to 33°] versus control 22° [14° to 28°], median difference 1.6°; p = 0.43), pelvic tilt excursion (PD-AFO LE cohort 8.2° [3.2° to 13°] versus control 7.1° [4.3° to 8.5°], median difference 1.1°; p > 0.99), and trunk forward flexion (PD-AFO LE cohort 16.2° [7.3° to 23°] versus control 15.4° [4.2° to 21°], median difference 0.8°; p > 0.99) between uninjured runners and patients in the PD-AFO LE cohort (Table 2). Only affected limb step length demonstrated a great effect size and no difference between the two groups with a Cohen D effect size of 1.0. However, runners in the PD-AFO LE group demonstrated decreased trunk lateral flexion (PD-AFO LE cohort 3.2° [1.2° to 7.4°] versus control 6.2° [2.0° to 9.9°], median difference 3.0°; p = 0.01) than uninjured runners did (Table 2).

Table 2.

Descriptive statistics for running biomechanics

Parameter	PD-AFO (n = 10)	Control (n = 15)	Median difference	p value	Cohen D effect size
Velocity, m/seconda	3.9 (3.4-4.2)	4.1 (3.1-4.8)	0.2	0.69	0.44
Cadence, steps/minute	179 (169-186)	173 (159-191)	5.8	0.43	0.39
Stride length, m	2.6 (2.4-2.9)	2.8 (2.3-3.3)	0.2	0.23	0.64
Stride width, cm	9.8 (5.4-15)	8.3 (3.0-16)	1.5	0.43	0.24
Affected limb step length, m	1.3 (1.2-1.4)	1.4 (1.1-1.6)	0.1	0.23	1.0
Unaffected limb step length, m	1.3 (1.2-1.5)	1.4 (1.2-1.7)	0.1	0.05	N/A
Trunk lateral flexion stance maximum, °	3.2 (1.2-7.4)	6.2 (2.0-9.9)	3.0	0.01	N/A
Trunk lateral flexion total excursion, °	8.4 (4.4-12)	9.9 (6.3-16)	1.5	0.01	N/A
Maximum trunk forward flexion stance, °	16.2 (7.3-23)	15.4 (4.2-21)	0.8	> 0.99
Peak pelvic tilt, °	24 (15-33)	22 (14-28)	1.6	0.43	0.28
Pelvic tilt excursion, °	8.2 (3.2-13)	7.1 (4.3-8.5)	1.1	> 0.99	0.53
Pelvic obliquity excursion, °	13.4 (9.7-25)	11.7 (7.3-17)	1.7	0.70	0.74

^aData are presented as median (range).

Agility Testing

Runners in the PD-AFO LE group performed worse in CHAMP performance testing than uninjured runners did in terms of the single-leg stance test (PD-AFO LE cohort 35 seconds [32 to 58] versus control 60 seconds [60 to 60], median difference 25; p < 0.001), the T-test (PD-AFO LE cohort 15 seconds [13 to 20] versus control 13 seconds [10 to 14], median difference 2; p < 0.001), and the Illinois Agility Test (PD-AFO LE cohort 22 seconds [20 to 25] versus control 18 seconds [16 to 20], median difference 4; p < 0.001) (Table 3). There was no difference in CHAMP Edgren side step test with an effect size of 0.83; the PD-AFO LE CHAMP Edgren side step test was 20 meters (16 to 26) compared with the control’s CHAMP Edgren side step test at 24 meters (16 to 29) (median difference 4; p = 0.11). With the numbers available, the Four Square Step Test result was no different between the groups with a medium effect size of 0.75; the PD-AFO Four Square Step Test median (range) was 5.5 seconds (4.1 to 7.2) compared with the control’s Four Square Step Test median of 4.2 seconds (3.1 to 7.3) (median difference 1.3; p = 0.39).

Table 3.

SMFA, CHAMP, and FSST scores

Parameter	PD-AFO (n = 10)	Control (n = 15)	Median difference	p value	Cohen D effect size
SMFA daily activities scorea	20 (2.5-45)	0 (0-0)	20	< 0.001	N/A
SMFA emotional score	45 (11-68)	0 (0-18)	45	< 0.001	N/A
SMFA mobility score	31 (5.6-64)	0 (0-0)	31	< 0.001	N/A
SMFA function index score	25 (4.4-45)	0 (0-3.7)	25	< 0.001	N/A
SMFA bother index score	25 (6.3-42)	0 (0-2.1)	25	< 0.001	N/A
CHAMP SLS, seconds	35 (32-58)	60 (60-60)	25	< 0.001	N/A
CHAMP ESST, m	20 (16-26)	24 (16-29)	4	0.11	0.83
CHAMP T-test, seconds	15 (13-20)	13 (10-14)	2	< 0.001	N/A
CHAMP Illinois Agility Test result, seconds	22 (20-25)	18 (16-20)	4	< 0.001	N/A
FSST, seconds	5.5 (4.1-7.2)	4.2 (3.1-7.3)	1.3	0.39	0.75

^aData are presented as median (range); SMFA = Short Musculoskeletal Function Assessment; CHAMP = Comprehensive High-level Activity Mobility Predictor; SLS = single-leg stance; ESST = Edgren side step test; FSST = Four Square Step Test.

Self-Reported Patient Functioning

Runners in the PD-AFO LE group reported worse median (range) Short Musculoskeletal Function Assessment scores for daily activities (PD-AFO LE cohort 20 [2.5 to 45] versus control 0 [0 to 0], median difference 20; p < 0.001), emotion (PD-AFO LE cohort 45 [11 to 68] versus control 0 [0 to 18], median difference 45; p < 0.001), mobility (PD-AFO LE cohort 31 [5.6 to 64] versus control 0 [0 to 0], median difference 31; p < 0.001), function index (PD-AFO LE cohort 25 [4.4 to 45] versus control 0 [0 to 3.7], median difference 25; p < 0.001), and bother index (PD-AFO LE cohort 25 [5.3 to 42] versus control 0 [0 to 2.1], median difference 25; p < 0.001) than uninjured runners did (Table 3).

Discussion

Service members who sustained major lower extremity fractures can have life-altering sequelae that jeopardize their ability to return to an active-duty military career; however, substantial advances in rehabilitation medicine and orthotic technology resulting from conflicts in Afghanistan and Iraq have established a practical pathway for military service members to return to a high level of physical functioning that is required for active duty [12–14, 16, 19]. Our study sought to compare two groups: (1) patients with unilateral lower extremity fractures, who were initially debilitated by their injuries because of chronic pain and dysfunction, who then returned to running after wearing a PD-AFO and completing a specialized rehabilitation program, and (2) uninjured, full-duty, active service members who performed at self-selected running speeds. Overall, we found that this group of highly motivated patients who strongly desired to run at their highest level ran with slightly more lateral truncal stiffness and poorer agility, though in-line running was overall achievable. This suggests that in select groups of patients, running at a high-demand military level is possible with the PD-AFO and specialized rehabilitation pathway, though agility is limited. Finding and comparing injured patients with civilian occupations and recreational activities with comparable physical demands of the military members may identify a role of this treatment modality outside of the active duty population. Based on the performance of this select population and the comparison to the normative population, the data can help decide how and where this treatment modality can be practically incorporated into civilian practice.

Limitations

The primary limitations in the study stem from our small sample size with inherent concerns for selection and transfer bias. Although 90 patients were prescribed a PD-AFO, only 13 continued to run in the rehabilitation program at a median follow-up of 5 years after their initial injuries, and only 10 participated in a formal 3-D gait analysis. With regard to the remaining 77 patients, many only use the PD-AFO to walk and no longer run, some went on to amputation, some were medically separated from the military, and many went back to their duty stations (patients are often sent temporarily to our institution for 2 months to complete this program and leave abruptly and continue therapy if needed with a standard physical therapist at their home duty station). We acknowledge this transfer bias in our methods section. We still believe this cohort of 10 patients is still worth studying because this subset of patients is truly what drives the heart of military medicine. To take debilitated Marines or sailors and return them to a high level of activity, even full, active duty, is one of the primary purposes of military medicine. It motivates us as providers, and it motivates the units that support and cherish these individuals.

With regard to the small sample size, we attempted to mitigate this using nonparametric independent samples testing for our nonnormally distributed data, reported our statistics in terms of medians, ranges with median difference, and we also reported effect sizes for parameters that did not demonstrate statistically significant differences with the numbers available. An effect size of greater than 0.8 was considered a large effect size for our sample size, with an 80% power to actually detect a difference. Although some of our effect sizes were large, most were medium or small, and thus we do not make any firm conclusions on similarities between the study and control groups with smaller effect sizes. Future studies would involve a larger sample size of both the study group and the control group.

Another limitation is a concern about the lack of applicability to the civilian population. The military inherently has a different trauma population than observed at a comparable civilian center [7]. This study specifically identifies motivated individuals who continue to run years after their injuries and have extensive resources to pursue their goals. The civilian population at a trauma center is likely older with multiple medical comorbidities. Although this poses a question about applicability, ongoing research, which was presented by Hsu et al in the 2019 Limb Lengthening and Reconstruction Society annual meeting, is attempting to address this with the variable-cadence ankle-foot orthosis, which comes with a modular kit that civilian orthotists can construct and customize; it has a scaled-down cost that insurance carriers may be willing to cover, and it can help highly motivated, older individuals with well-controlled medical comorbidities go back to running if that is a firm goal. Additionally, this cohort of patients is not purely limb salvage. The injuries in our cohort had a mix of high-energy and low-energy mechanisms and not all were sustained in combat. Therefore, the results of this study may be more pertinent to the civilian population. Future work should continue to focus on the translation of this work to the civilian sector.

Other limitations include cost, the need for a multidisciplinary team, fluctuations in body weight and calf size, and the possibility of lifetime maintenance. Although the price of a custom PD-AFO and a multidisciplinary team dedicated to identifying those most in need of extra support may cost a lot of money for a standard hospital, the Department of Defense deems this necessary for our current military forces. This again ties in with applicability into the civilian sector. As stated above, a variable cadence AFO, which is modular and can be modified by a civilian orthotist, may be a reasonable application of this work. Most tertiary referral centers have multiple providers to include a physical therapist, orthopaedic surgeon, and physiatrist, and can thus form the beginnings of a multidisciplinary team for patients who perform physically demanding activities as part of their job. This intensive program is reproducible; other facilities and studies have successfully replicated the Intrepid Dynamic Exoskeletal Orthosis and return-to-run physical therapy program, with outcomes that remain consistent across different centers [6, 11, 17]. Finally, with regard to the possibility of lifetime maintenance, some patients may outgrow the need for the brace as they get stronger and continue to run [17] or grow older and no longer have the need or desire to run, in contrast to amputees who often require lifelong prosthetics even to ambulate. These limitations are important to keep in mind when interpreting our results, as it is important to remember we are reporting on a group of highly selected individuals who are extremely motivated and demand much of themselves and those around them; however, this type of person is not unique to the military.

Biomechanical Parameters While Running

Patients sustaining a lower extremity fracture who use a PD-AFO with a specialized rehabilitation program differed only slightly in running biomechanics from uninjured service members acting as controls. The PD-AFO LE cohort demonstrated more lateral trunk stiffness, which may represent the unilateral stiffness of the brace and a decreased ability to compensate in the coronal plane, with no significant difference in velocity, cadence, and stride with the numbers available and medium effect sizes. It is interesting to note that there was no difference in affected limb step length between the two groups, with a large effect size, suggesting that these PD-AFO runners likely have a similar step length as uninjured, full-duty runners. This is an important professional point because many service members must perform some running to remain on active duty in the US military. The senior author’s (KMK) prior study compared spatiotemoporal gait parameters at self-selected walking speeds such as velocity, cadence, and stride length in a case series of patients with pilon fractures, before and after PD-AFO fitting and completion of the specialized rehabilitation program [18]. The authors of that paper did not find any differences, and they suggest that the PD-AFO may not provide much benefit at self-selected walking speeds. In contrast, our current study looks at PD-AFO use in higher demand activities such as running, which none of the patients were able to do before using the PD-AFO and return-to-run program. Thus, running is where the PD-AFO and return-to-run program truly make a difference.

Agility Testing

With regard to the CHAMP performance measures, runners in the PD-AFO LE group performed worse than controls, indicating they had less agility than the control group. However, they performed similar to the 75th percentile of service members who underwent amputation for traumatic lower limb loss, as reported elsewhere [5], indicating that they have good function and adequate performance. Again, CHAMP quantifies high-level mobility with the single-leg stance to assess balance and postural stability, the Edgren side step test to assess frontal plan agility and body control, the T-test to evaluate the ability to change directions rapidly, and the Illinois Agility Test to assess multidirectional agility in the frontal, sagittal, and transverse planes with an emphasis on acceleration [5]. The PD-AFO runners had worse balance and ability to change directions rapidly, but they could side step as well as controls could with no difference in CHAMP Edgren side step test with a large effect size. Thus, complex agility is likely not comparable, but single-plane agility may be more achievable. Additionally, a prior study looking at agility gains before and after the PD-AFO in patients with unilateral lower extremity fractures found that overall total CHAMP scores improved [9]. So although these patients may run with less agility compared with uninjured, full-duty runners, they still likely improve in agility compared with their own postinjury baseline.

Self-Reported Patient Functioning

Patients using a PD-AFO with a specialized rehabilitation program reported worse perceived function than uninjured service members. They experienced more issues with daily activity, mobility, function, and even emotions. This may reflect the complex, multifactorial nature of recovery in the trauma patient population, the patient’s own perception of their recovery, or how they match up to their peers, which is a daily experience in the life of an active-duty service member. However, even though the quality-of-life scores for the PD-AFO LE group were worse than those of the control participants, they were better than those in the limb salvage cohort from the Military Extremity Trauma Amputation/Limb Salvage (METALs) study [4]. The current study includes only two blast injuries and four total patients who meet criteria for limb salvage patients, while the METALs study included all combat-injured patients, thus accounting for the improved quality-of-life scores in the current patient group. This also demonstrates a transition in military medicine toward using the PD-AFO and return-to-run rehabilitation program for more nonbattle-related injuries, such as isolated lower extremity trauma.

Conclusion

The results of our study demonstrate that patients with lower extremity fractures who have completed the return-to-run pathway with a PD-AFO can run with slightly more trunk stiffness in the coronal plane compared with uninjured, full-duty service members. Although our study was underpowered to detect differences in velocity, cadence, stride length, and stride width between the study and control groups, we found no significant differences with overall medium effect sizes. We suspect that with a larger sample size, differences for in-line running would remain minimal. Furthermore, every patient who was prescribed this device and rehabilitation program had chronic pain and permanent function loss, which would render them incapable of doing the activities they wished, and running falls into that category. None of the patients in the study group could run before using this device and the rehabilitation program, and they now can.

Additionally, the PD-AFO runners ran with less agility and had poorer self-perceived functioning compared with their uninjured counterparts. This is important to manage expectations. In-line running may be an achievable goal, but the device and rehabilitation program may not be able to meet all expectations of what patients perceive as full running to include quick cutting maneuvers and multidirectional changes in acceleration. This device also does not improve daily activities of life. It simply can help highly motivated individuals with unilateral lower extremity fracture causing chronic pain and dysfunction return to in-line running at a level that can be compatible with full duty.