Association of Surgical Service Line and Frailty with Outcomes after Major Lower Extremity Amputation

Key Takeaways

• Surgical service and patient frailty were associated with outcomes after major lower extremity amputation: orthopedic surgery achieved the highest prosthetic ambulation in both non-frail (60.2%) and frail (72.7%) patients.

Introduction

Major lower extremity amputations (LEA) have become increasingly common, with approximately 2 million people in the United States living with limb loss. ¹ These amputations can occur due to severe infection, ischemia, and/or trauma. ² While the amount of trauma and cancer-related amputations has decreased, amputations related to vascular compromise have increased substantially in recent years. ³ Increased prevalence of diabetes, obesity, smoking, hypertension, and hyperlipidemia, significant risk factors for peripheral vascular disease (PVD), may explain these findings.^3-5 Major LEA can cause profound physical and psychological consequences in a patient’s life. ⁶ Therefore, maintaining ambulatory status with a prosthesis following amputation is imperative for one’s independence, everyday function, and quality of life.

Various surgical services perform major LEA, including vascular (VS), trauma (TS), and orthopedic (OS) surgery. Prior work suggests that surgeon specialty and service line may influence outcomes, with studies in other domains (eg, lung resection and colorectal hepatic metastasectomy) demonstrating differences in mortality and recurrence by subspecialty.^7-10 In the LEA literature, Pei et al reported higher pneumonia, sepsis, and mortality after below-knee amputation when procedures were performed by general rather than vascular surgeons, and Sebekos et al found fewer complications when orthopedic surgeons performed below-knee amputation (BKA) compared to vascular and general surgeons.^11,12

Frailty is increasingly recognized as a key determinant of postoperative outcomes in vascular surgery, including higher rates of wound complications and mortality.^14,15 However, few studies examining major LEA have incorporated frailty, and none to our knowledge have evaluated how frailty and surgical service jointly relate to outcomes across above- (AKA), through- (TKA), and below-knee (BKA) amputations.

This retrospective study evaluated whether the primary surgical service performing major LEA—vascular, trauma, or orthopedic—is associated with complications, return to the operating room (RTOR), re-amputation, mortality, and prosthetic ambulation, and whether these relationships differ by frailty status using the 5-factor modified Frailty Index (mFI-5). We hypothesized that patient frailty would be a strong predictor of adverse outcomes and that service-line variation in outcomes would persist after accounting for baseline differences in case mix.

Methods

We performed a single-center retrospective review of all patients receiving major LEA at a tertiary care center between 2015 and 2022. This study was approved by the Institutional Review Board (IRB) (HP-00085462_4) at the University of Maryland. Because of the retrospective nature of the study, consent of individual patients was waived. Guidelines for Good Clinical Practice were followed in all research methods.

Patient data were obtained from the institutional electronic medical record (EMR) using Current Procedural Terminology (CPT) codes for major LEA (AKA CPT: 27590, TKA CPT: 27592, BKA CPT: 27880). Inclusion criteria were patients ≥18 years of age who underwent a BKA, TKA, or AKA. Patients were excluded if they underwent only minor amputations (transmetatarsal, Syme, or Boyd procedures) or if documentation was insufficient to determine baseline characteristics, operative details, and at least 30-day postoperative outcomes (RTOR, complications, mortality, and ambulation). Data obtained from chart review included patient demographics (age, sex, race, ethnicity, body mass index (BMI), primary insurance coverage, smoking history) and medical comorbidities. Preoperative data included indication for amputation. Indication for amputation was abstracted from the attending surgeon’s documentation and categorized as ischemia, infection, trauma, sarcoma, or other. Patients with mixed etiologies (eg, traumatic injury with subsequent ischemia) were classified according to dominant cause leading to amputation. Perioperative data included primary amputation date, surgical service performing the procedure, amputation level, and operative time.

The primary outcome was prosthetic ambulation at follow-up. Postoperative ambulation was defined as independent community ambulation with a prosthesis, without a cane, walker, or wheelchair, documented by a physician or physical therapist in the medical record. ¹³

Secondary outcomes included the following:

• 30-day RTOR: any return to the operating room within 30 days of the index amputation. Because open guillotine amputations (OGA) often undergo an intentionally planned staged return for definitive revision/closure, we also categorized RTOR as (1) planned staged RTOR for definitive revision/closure after OGA and (2) unplanned RTOR within 30 days, defined as any RTOR excluding planned staged OGA-to-closure returns.

For time-to-event outcomes, time zero was the date of the index amputation; patients were censored at death or end of follow-up.

Frailty status was calculated with the mFI-5. ¹⁴ Patient frailty was scored on a scale of 0 to 5 and was calculated based on the presence of 5 comorbidities: congestive heart failure (CHF) within 30 days before surgery, insulin-dependent or noninsulin-dependent diabetes mellitus, chronic obstructive pulmonary disease or pneumonia, partially dependent or totally dependent functional health status at the time of surgery, and hypertension.

Continuous variables were summarized as mean ± standard deviation (SD) or median [interquartile range, IQR] as appropriate; age is reported as mean ± SD and LOS as median [IQR]. Categorical variables are presented as counts (percentages). Group comparisons used χ² tests or Kruskal-Wallis tests (two-sided α = 0.05). Adjusted analyses included logistic regression for ambulation, surgical site infection (SSI), 30-day readmission, and 1-year re-amputation, and negative binomial regression for LOS. Covariates were prespecified: age, sex, race, BMI ≥30, mFI-5, indication, amputation level, OGA, end-stage renal disease (ESRD), PVD, and year of surgery; service line was the exposure. Multiple amputations per patient were accounted for using cluster-robust standard errors. Analyses were performed in Jamovi v2.3 (Sydney, Australia) and Python (statsmodels 0.14).

Results

Discussion

In this single-center retrospective study of 689 major LEAs, we evaluated whether surgical service line and frailty were associated with complications, RTOR, mortality, re-amputation, and prosthetic ambulation. We observed service-line variation in functional outcomes: OS had higher prosthetic ambulation. TS had the highest overall 30-day RTOR; importantly, after separating planned staged OGA-to-closure returns from unplanned RTOR, differences in overall RTOR were largely driven by planned staged returns, while unplanned RTOR did not significantly differ by service. Unadjusted mortality differed by service (lowest in OS), which likely reflects differences in baseline case mix and comorbidity burden.

VS performed 44% of major LEAs, TS 37.3%, and OS 18.7%. Ischemia was the most common indication, and amputations for ischemia or infection were frequently performed by TS, likely reflecting our trauma center’s case mix and referral patterns and, at other institutions, the availability of dedicated vascular services or reliance on general surgeons. Patients had substantial comorbid burden—most commonly hypertension (67.5%), diabetes mellitus (52.3%), and PVD (45.4%)—and overall lower baseline functional status.

OS patients had higher prosthetic ambulation, particularly among non-frail patients, and a numerically lower composite complication rate than TS and VS, although this composite difference was not statistically significant. SSI differed by service, highest after TS and lowest after VS, paralleling prior work by Sebekos et al showing fewer overall complications, including wound and cardiovascular complications and prolonged LOS, when orthopedic surgeons performed BKA. ¹² In our broader cohort including AKA, TKA, and BKA across 3 services, we likewise observed higher ambulation after OS and higher 30-day RTOR after TS. The OS cohort was younger and less comorbid, which likely contributed to these findings. Even after adjustment for measured comorbidities, frailty, and amputation level, unmeasured factors such as pre-amputation functional status, social support, and rehabilitation intensity likely influence service-line differences and limit causal inference.

Adjusted analyses confirmed that higher ambulation after OS was not explained by age, case mix, comorbidities, or frailty. In multivariable models adjusting for demographic, clinical, and operative factors, VS and TS had roughly one-third the odds of prosthetic ambulation compared with OS. This supports a role for service-specific factors—such as patient selection, operative technique, perioperative pathways, and early rehabilitation—in shaping functional recovery after LEA. Older age and more proximal level remained strong predictors of lower ambulation, emphasizing careful limb-preservation and level selection. The absence of an independent effect of continuous mFI-5 in adjusted models likely reflects overlap between frailty, comorbidity, and service-level practice patterns rather than a lack of prognostic value.

Frailty, assessed with the mFI-5, was a strong determinant of postoperative outcomes: patients with mFI-5 ≥ 3 had higher complications, RTOR, and mortality than non-frail patients. These findings parallel those of Karam et al, who reported increasing wound infection and mortality with higher frailty indices in vascular surgery. ¹⁵ The prevalence of comorbidities in our cohort underscores the importance of frailty assessment and risk stratification preoperatively, as well as tailored perioperative planning and goals-of-care discussions.

These results parallel work linking surgical specialization to outcomes in other domains, reinforcing that service-specific training and pathways influence results.^9,10,16-18 Standardizing LEA care—particularly OGA-to-closure protocols and early prosthetics/rehabilitation—and embedding routine mFI-5 screening may reduce variability and improve function, while acknowledging that pre-amputation function and rehabilitation practices likely also contribute to service-line differences.

This study has limitations, including its retrospective single-center design, potential misclassification of indications, limited data on preoperative function, time to amputation, and service-specific rehabilitation pathways, and the potential for residual confounding, which may limit generalizability of findings to other settings. We did not capture time from admission to amputation or the number of procedures before the index amputation, which may differ by service and influence outcomes. Ambulation was abstracted from clinician documentation rather than standardized testing, and there was no institutional LEA rehabilitation pathway; prosthetics referral and physical therapy were at the discretion of the treating service. Nonetheless, surgical service line and frailty appear to be important correlates of postoperative outcomes after major LEA. Service-line differences in prosthetic ambulation, coupled with similar mortality and modest differences in 1-year re-amputation, suggest potentially modifiable processes of care—such as timing of amputation, perioperative optimization, and early prosthetics/rehabilitation—that could be standardized. Future multicenter studies should clarify which elements of LEA care pathways most strongly influence functional recovery in diverse practice settings and how frailty-informed, standardized protocols can be implemented to reduce variation and improve patient-centered outcomes.