Early ambulation after fibular free flap surgery is associated with reduced length of stay, increased mobility independence, and discharge to home

Fibula free flaps (FFF) are one of the most common bony flaps utilized. This paper describes a quality improvement project aimed at increasing early ambulation.


| INTRODUCTION
2][3][4] Fibula free flap harvest involves resection of the majority of the fibula with or without adjacent soft tissues including muscle and overlying skin paddles.][7][8] While the use of the fibula free flap has become commonplace in the armamentarium of reconstructive head and neck surgeons, questions remain about the optimal postoperative management of the donor site.Historically, patients were allowed limited activity and weight bearing on the donor extremity after a fibula flap harvest and most surgeons opted for some form of splinting or boot placement for the immediate postoperative period.These physical activity limitations are in opposition to recently published literature on the feasibility, safety and success of early mobilization after head and neck microvascular reconstruction. 9,10In their study, Twomey et al. demonstrated a shorter length of stay, as well as a decreased risk for major complications with early mobilization within 24 h after surgery. 10In addition to these studies, extensive work has been done on the implementation and success of enhanced recovery after surgery (ERAS) protocols after complex head and neck reconstructive surgeries, which have revealed a multitude of measurable clinical benefits. 11To date, however, there is a paucity of research specifically examining the success and feasibility of early ambulation and activity after fibula flap harvest in the head and neck patient population.With this study, we aim to investigate the safety and success of immediate postoperative ambulation without use of a postoperative boot versus delayed ambulation in head and neck patients undergoing fibula free flap reconstructive surgeries.

| MATERIALS AND METHODS
A retrospective analysis of patients that underwent fibula free flaps completed by five reconstructive surgeons at a single academic institution was completed.The study included patients from August 2014 to October 2023; during this time a total of 268 patients underwent reconstruction using an osteocutaneous free flap, with 168 (62.3%) of those patients having a fibula free flap.Flaps prior to 2014 were not included, as this was when our institution transitioned to a new electronic medical record, with limited documentation available prior to this.In June of 2018, a quality improvement project was initiated by the Huntsman Cancer Institute, Physical Therapy, and Head and Neck Surgery Divisions, which included a transition to an early ambulation protocol without postoperative boot placement for all fibula free flap patients.Prior to the initiation of this protocol, patients had limited ambulation until the boot was removed at postoperative day 5.This was a long-utilized protocol at our institution, which is similar to previously published ambulation protocols, which encouraged delayed ambulation till postoperative day 5, particularly if there was placement of a split thickness skin graft for closure of the wound. 12While this was a very conservative approach, this was custom at our institution for several years.Regular Activity Measure for Post-Acute Care (AMPAC) scores were recorded for before and after the postoperative boot was discontinued.AMPAC scores are a quantitative way to determine the discharge needs of hospitalized patients, which assesses three domains, including (1) basic mobility, (2) daily activity, and (3) applied cognition. 13In this setting, the AMPAC Mobility (AMPAC-M) Short Form 6-clicks measure was utilized, which predominantly focuses on mobility and assists physical therapists transition patients out of the acute rehabilitation setting. 14The AMPAC scoring system creates a score for each patient ranging from 6 to 24, where a score of 6 indicates a 100% impairment, and a score of 24 indicates 0% impairment. 15After this quality improvement project was initiated, measurement of gait speed (m/s) was also documented as a part of the regular physical therapy documentation to show progress is gait speed over the course of the admission.
An evaluation of length of stay (LOS), gait distance on last day of admission (feet), change in AMPAC-M score, type of bed mobility, transfer and gait assistance, and discharge location was documented for the cohort of patients that had late ambulation (postoperative day 5 = Group 1) with placement of boot versus those that had early ambulation (postoperative day 1-2 = Group 2) with no placement of boot.Patients in both groups were evaluated by physical therapists daily after their fibula free flap and ambulated weight bearing as tolerated after surgery.Gait speed (m/s) was only collected among patients that did not have placement of a boot and started early ambulation after surgery.Other variables collected included the following: age, sex, diagnosis, and type of surgery.Patients <18 years of age were excluded from analysis.Descriptive statistics were formed.A two-sample t-test was utilized to evaluate mean differences between continuous variables.Chi-squared tests were utilized to determine differences in categorical variables.Univariate and multivariate analysis was performed to identify potential confounding factors associated with length of stay and discharge location.For univariate and multivariate analysis, length of stay was dichotomized into ≤7 days or >7 days.Variables that were significant or nearly significant on univariate analysis were then entered into a multivariate analysis.Stata 14.0 was utilized for data analysis.This study was approved by the University of Utah Institutional Review Board.

| RESULTS
A total of 168 patients underwent a fibula free flap reconstruction from August 2014 to October 2023; two (1.2%) patients died during their hospital admission after surgery and were excluded from analysis.A majority of included patients were male (72.3%, n = 120) and the average age was 58.2 years (range, 22-92).The most common diagnosis among study participants was oral cavity squamous cell carcinoma with osseous erosion (59.1%, n = 98), followed by the following diagnoses: osteoradionecrosis (16.9%, n = 28), traumatic gunshot wound (6.0%, n = 10), ameloblastoma (6.0%, n = 10), sarcomas (3.6%, n = 6), salivary gland carcinoma (3%, n = 5), and other (5.4%, n = 9).A majority of patients underwent a segmental mandibulectomy or composite oral cavity resection (79.5%, n = 133), while 19.8% (n = 33) underwent a partial or total maxillectomy, followed by one patient (0.6%) that underwent reconstruction of the anterior spine.LOS was lower among Group 2 compared to Group 1, where the early ambulation group had an average stay of 8.1 (95% confidence interval [CI] 7.7-8.7)days after surgery and the late ambulation group had an average stay of 9.4 (95% CI 7.9-10.9)days after surgery (p = 0.04).In terms of discharge location, there was a higher rate of discharge to a skilled nursing facility in Group 1 (21.3%)compared to Group 2 (11.9%) (p = 0.009).A majority of patients in both groups were discharged home with or without the aid of home health for wound care, but no home physical therapy needs (Group 1 72.3% vs. Group 2 85.5%).Three patients in Group 2 were discharged to an LTAC (2.5%), with no patients going to an LTAC in Group 1. See Table 1 for a summary of patient characteristics based on Group 1 and Group 2.
When evaluating the first AMPAC-M score among Group 1 (delayed ambulation) and Group 2 (early ambulation, no boot), there were statistically lower initial AMPAC-M scores among the delayed ambulation group wearing a boot (Group 1 mean AMPAC-M = 13.9;95% CI 12.7-15.1 vs. Group 2 mean AMPAC-M = 16.2;95% CI 15.7-16.7;p = <0.001).In terms of the last AMPAC-M score prior to discharge, there was no statistically significant differences between the two study groups (Group 1 mean AMPAC-M = 20.8;95% CI 20-21.6 vs. Group 2 mean AMPAC-M = 20.4;95% CI 19.9-20.9;p = 0.42).With regards to the mean change in patient's AMPAC-M scores from the first to last evaluation, there was no statistically significant differences in the mean change of AMPAC-M scores (Group 1 AMPAC-M mean change = 5.6; 95% CI 4.3-7.0 vs. Group 2 AMPAC-M mean change = 4.7; 95% CI 4.2-5.2;p = 0.11), see Figure 1.Among Group 2, the average first gait speed was 0.28 m/s and the average last gait speed was 0.48 (m/s) (mean change of 0.20 m/s).While not statistically significant, there was a higher gait distance on the last day of hospital admission among Group 2 (368 ft) compared to Group 1 (231 ft) (p = 0.13).A higher proportion of patients in Group 2 demonstrated independence during bed mobility, transfers, and gait compared to Group 1 (Bed Mobility 49.1% vs. 39.1%,Transfers 31.9% vs. 17.0%,Gait 18.1% vs. 10.6%)(p < 0.001).Additionally, there was a higher rate of patients in Group 2 not requiring any assistive devices (example cane or front wheeled walker) while walking when compared to Group 1 (Group 1 = 8.5% vs. Group 2 = 33.04;p < 0.001).See Table 2 for summary of physical therapy outcomes.
On univariate analysis, there was no impact of age, sex, benign versus malignant diagnosis, undergoing a maxillectomy versus mandibulectomy, or last AMPAC-M score on length of hospital stay (p > 0.05).However, the first AMPAC-M score was found to be protective against a prolonged length of stay (odds ratio (OR) = 0.87, 95% CI 0.78-0.97),and discharge location to a facility was associated with an increased length of stay (OR = 10.2, 95% CI 2.9-36.0).On multivariate analysis, only discharge location to a facility was still associated with increased length of stay (OR = 12.7, 95% CI 2.8-57.3).In terms of predicators of discharge location, on univariate analysis sex, benign versus malignant diagnosis, undergoing a maxillectomy versus mandibulectomy, and last AMPAC-M score were not associated with discharge location ( p > 0.05).However, again the first AMPAC-M score was found to be protective against discharge to a facility (OR = 0.86, 95% CI 0.75-0.98),while age and length of stay was associated with discharge to a facility (OR = 1.04, 95% CI 1.0-1.07,and OR = 1.3, 95% CI 1.17-1.53,respectively).On multivariate analysis, age and length of stay were still associated with discharge to a facility (OR = 1.05, 95% CI 1.0-1.1 and OR = 1.32, 95% CI 1.14-1.52,respectively).See Table 3 for univariate and multivariate analysis results.

| DISCUSSION
Fibula free flaps are the workhorse flap for bony reconstruction of the mandible and midface.While this is a versatile flap that aids in reconstructing complex bone defects, prior literature has raised concern regarding reduced physical activity and quality of life among head and neck surgery patients who have undergone a fibula free flap. 16Notably, these physical struggles may persist after surgery, with a prior study demonstrating reduced peak power and balance ability in a one leg stand a year after surgery. 17Here, we present a retrospective analysis of patients at a single academic tertiary care hospital who have undergone reconstruction using a fibula free flap over the course of nine years, with a dramatic change in our activity restrictions and placement of a postoperative boot midway through the data collection period.Ultimately, this quality improvement project has led to a reduced LOS, increased mobility independence, and increased rates of patients discharging to home.While some authors have reported utilizing a rigid orthotic booth for upwards of 8 weeks after surgery, 18 this study suggests improved physical therapy outcomes without boot placement and aggressive early ambulation.
The use of AMPAC-M scores can aid in making decision about discharge location early in the hospitalization to avoid delays in care. 14To date, no studies have previously evaluated the use of AMPAC-M score in determining discharge location among patients undergoing a fibula free flap.However, AMPAC-M scores have been useful in determining discharge location among patients undergoing total hip and knee arthroplasties. 19Among patients in the early ambulation group, there were significantly higher first recorded AMPAC-M scores, indicating that these patients had a baseline higher level of activity and mobility when compared to the traditional Group 1 that had delayed ambulation and placement of a boot.Ultimately, there were no statistically significant differences between the last AMPAC-M score and the mean differences between the two groups, which suggests that Group 1 patients were able to increase their activity and mobility throughout the hospitalization stay.In the end, Group 2 had a higher rate of discharge to home, which could be a reflection of the impact of their higher first AMPAC-M scores.Univariate analysis also suggests that a higher first AMPAC-M score is also protective against discharge to a facility.These data suggest that among patients undergoing fibula free flaps, a higher first AMPAC-M score is reflective of a patient being discharged to home compared to a skilled nursing facility or long-term acute care hospital.
There was notably a reduced LOS among Group 2 compared to Group 1, which again advocates for early ambulation in patients undergoing fibula free flaps.LOS is a critical piece in assessing success of quality improvement projects, as increased LOS effects overall cost of care and is associated with a rise in complications during the acute postoperative period. 20Not only does decreased LOS reduce postoperative complications, early ambulation does as well.A recent study evaluated outcomes associated with early mobilization after free flap surgery, with mobilization after 48 h being associated with significant major complications, including pneumonia, pulmonary embolism, deep vein thrombosis, delirium tremens, myocardial infarction, hematoma, free flap compromise, and free flap failure. 10It is important to note that there are several factors beyond ambulation that impact LOS after free flap reconstruction.In a retrospective review head and neck patients that underwent free flap reconstruction at a single institution, factors most associated with increased LOS included flap compromise requiring return to the operating room, wound break down or fistula, surgical site infection, and having prior radiation therapy. 21While assessment of postoperative complications related to fibula free flaps is outside the scope of this study, there is a symbiotic relationship between early ambulation and decreased LOS that leads to decreased rates of complications in these medically complex head and neck cancer patients.
There was also an increased level of independence among patients in Group 2, including mobility, transfers, and gait, in addition to lower rates of assistive devices being used.There is limited data on the use of assistive devices, including canes and front wheeled walkers, after fibula free flaps.One study cited 4% of patients utilized an assistive device postoperatively, however, the time period these were used after surgery was no clearly delineated. 22While gait speed was not a part of our regular physical therapy evaluation in Group 1, there was an increase in gait speed over time in Group 2. There is also limited data on changes in gait speed after fibula free flaps, but prior data shows a decreased gait speed at 1 month after surgery, which then improves to baseline gait speed after 3 months from surgery. 23Unfortunately, our data did not collect preoperative gait speed, but improvements seen over the course of one's admission is encouraging.
While this study is the first to describe changes in physical therapy outcomes, including AMPAC-M scores and discharge location, among patients who did not have a boot placed postoperatively after fibula free flap surgery, there are several limitations to or study.This study data was collected retrospectively and was based on chart review completed using the electronic medical records, thus, charting errors can affect data analysis.Furthermore, we started collecting new data after implementation of our quality improvement project in 2018, including gait speed.While this allow us to further understand outcomes related to early ambulation without boot placement, we are unable to evaluate how this change compares to patients in the late ambulation group.We plan to start evaluating preoperative gait speed in the future to aid in assessing changes during the perioperative period.Patients were also evaluated by various physical therapists throughout their stay, which could create inconsistencies in physical therapy outcomes and data reporting for AMPAC-M scores.Furthermore, age is likely a confounding factor that is associated with discharge to a facility as seen in the multivariate analysis.Given that there was a higher mean age in Group 1, this could be related to a higher rate of discharge to a facility in this group.This study had a relatively low number of participants, thus, the study could be underpowered to detect the true impact of these confounding variables.Despite these limitations, these data do support the integration of early ambulation and demonstrate that boot placement after fibula free flap is not necessary, and may in fact hinder patients from reaching their physical therapy potential during their hospital stay.

| CONCLUSION
Given that the fibula free flap is commonly utilized in head and neck reconstruction of the midface and mandible, understanding how these patients can be optimized after surgery is of critical importance.In this study we demonstrate that after a quality improvement project which implemented early ambulation (1-2 days after surgery) and elimination of the boot placement after surgery, there was a reduced LOS, increased mobility independence, and higher rates of discharge to home.These findings advocate for collaboration with physical therapy teams in order to enable patients to gain increased levels of mobility and activity after fibula free flap reconstruction.

F I G U R E 1
Figure demonstrating the mean AMPAC-M scores at the first and last day of patient's hospitalization, in addition to the mean change in scores.Group 1 had a delayed ambulation with placement of a boot, while Group 2 had no boot placed and were encouraged to ambulate starting postoperative day 1.
T A B L E 1Note: Bold values are statistically significant.Abbreviations: CI, 95% confidence interval.;SD, standard deviation.
T A B L E 2 Physical therapy outcomes.Univariate and multivariate analysis results for length of stay and discharge location.
a Note: Bold values are statistically significant.a One patient who was in a wheelchair and required maximal assistance at baseline.T A B L E 3