Descriptive Analysis of Types and Diagnoses Associated with Lower Extremity Amputation: Analysis of the US Veterans Health Administration Database 2019–2023

Abstract

Introduction

Veterans in the US have higher rates of lower extremity amputation (LEA) compared to the general population and these rates have increased between 2008 and 2018. There is limited data which directly evaluate the potential underlying comorbidities associated with LEA in the veterans’ population especially with the most recent data. Such information is critical to help inform clinical management strategies to reduce the risk of amputations among our veterans.

Methods

This was a retrospective observational study of adults in the Veterans Health Administration database who underwent LEA from January 1, 2019 to December 31, 2023. The date of the first LEA procedure was defined as the index date. Index LEA type, patient demographic at baseline, and clinical characteristics (including diagnoses for conditions associated with LEA and other comorbidities) 1 year before and 30 days after the index LEA procedure (except for bacterial infections which the identification period was 30 days before and 30 days after the index LEA procedure) were described.

Results

Of the 27,134 Veterans with LEA, 67.3% were ≥ 65 years of age, 97.0% were male, and 65.3% were non-Hispanic white. The most common type of LEA was transmetatarsal (52.9%), followed by toe (21.9%), above-knee (15.4%), and below-knee (9.8%). The most prevalent diagnoses associated with LEA were diabetes (81.6%), bacterial infections (79.1%), and peripheral artery disease (PAD; 63.3%). Only 15 Veterans (< 0.1%) had a diagnosis for combat-related injuries to lower extremities.

Conclusion

Diabetes and PAD are highly prevalent and among the main conditions associated with LEA among US Veterans. Earlier and more effective preventative and clinical management of these conditions offer an opportunity to significantly reduce the rates of LEA in this population.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12325-024-03005-6.

Key Summary Points

Why carry out this study?

Despite the rising rate of lower extremity amputation (LEA) among Veterans, there exists limited research directly characterizing the possible underlying clinical conditions associated with LEA and their distribution in Veterans using contemporary data.

Our study hypothesizes that chronic diseases—and not traumatic injury sustained in combat—are the more prevalent diagnoses associated with LEA among US Veterans.

What was learned from the study?

The most common types of LEA were transmetatarsal, followed by toe, above-knee, and below-knee among US Veterans between January 2019 and January 2023. Diabetes, bacterial infections, and peripheral artery disease (PAD) were the most prevalent disease conditions associated with LEA.

The characterization of disease conditions associated with LEA highlights the importance of chronic disease management and prevention. More effective management of diabetes, particularly PAD, in Veterans can potentially reduce risk of LEA and lead to improvements in their overall health and well-being.

Introduction

Approximately 150,000 individuals in the US undergo nontraumatic lower extremity amputation (LEA) each year [1, 2], with higher age-specific rates of major amputation among US Veterans than in the general population [3]. LEA is associated with a high risk of morbidity and mortality, and is a serious health burden for the Veteran population.

The Veterans Health Administration (VHA)—the largest integrated health care system in the USA, providing care to more than 9 million US Veterans including more than 96,000 with limb loss [4]. The VHA has implemented programs over the past several years to help prevent and improve outcomes of amputations in Veterans. For instance, the Amputation System of Care program initiated in 2008 aims to enhance the quality and consistency of care and rehabilitation services provided to Veterans with limb loss [4, 5]. The Prevention of Amputation in Veterans Everywhere (PAVE) program provides access to specialist treatment to Veterans for nearly 30 years and is designed to reduce overall amputation rates and reduce the amputation levels performed, such as a toe, foot, or lower leg for at risk for Veterans [6]. Despite these initiatives, LEA rates in US Veterans continue to increase; a recent cohort study found that incidence rates of LEA increased from 12.9 per 10,000 persons to 18.12 per 10,000 persons between 2008 and 2018 [7].

While a recent study examined the factors contributing to the increasing trend [7], there is limited research directly characterizing the possible underlying clinical conditions associated with LEA and their distribution in Veterans. Moreover, there are no current data on the types and associated diagnoses of LEA, although this information is important given the evolving profile of the Veteran population [8, 9].

Understanding the drivers of LEA can inform strategies to reduce the risk and occurrence of amputations among Veterans. In this context, the current study examined LEA types and associated diagnoses among US Veterans in the last 5 years.

Methods

Data Source and Study Design

This retrospective observational study analyzed data in the VHA Corporate Data Warehouse (CDW), which contains the entire contents of the unified national Veteran Affairs electronic medical record, including information on demographics, medical services (e.g., surgery), and hospital admissions (one primary discharge diagnosis and up to 15 secondary diagnoses). The VHA database contains only de-identified patient information and is fully compliant with Health Insurance Portability and Accountability Act. Therefore, institutional review board review was not required.

Study Population

The study population comprised Veterans who had undergone at least one LEA between January 1, 2019 and December 31, 2023. The flow diagram of sample selection is shown in Fig. 1. LEA was identified using Current Procedural Terminology, 4th Edition (CPT-4) or International Classification of Diseases 10th Revision, Procedure Coding System (ICD-10-PCS) codes (Supplemental Table S1). The index date was the date of the first LEA procedure during the study period. Additionally, patients had to be at least 18 years old on the index date with continuous VHA enrollment for at least 12 months before, and 30 days after, the index date (observation period) to be included in the analysis.

Fig. 1.

Sample selection. LEA, lower extremity amputation; VA, Veteran Affairs. ^aCodes used to identify LEA can be found in the Supplementary Material. ^bEnd of enrollment was defined as the earliest of the disenrollment date, death, or data cutoff (January 31, 2024)

Study Variables

Patient demographics on the index date were extracted from the CDW dataset, including age, sex, race/ethnicity (e.g., non-Hispanic white, non-Hispanic Black, Hispanic, Other, or unknown), and region of residence (Northeast, West, Midwest, or Southeast).

LEA type for the index LEA procedure was categorized as above-knee, below-knee, transmetatarsal, or toe on the basis of the CPT-4/ICD-10-PCS code. If multiple LEA procedures occurred within a 30-day period after the index procedure, the LEA type was categorized on the basis of the highest level of amputation (with the highest being above-knee and the lowest being toe).

Possible diagnoses associated with LEA were diabetes, bacterial infection, peripheral artery disease (PAD), primary cancer of lower extremities, injuries to lower extremities, and unknown. These conditions were selected on the basis of clinical input and identified in the dataset using an ICD-10 Clinical Modification (ICD-10-CM) code (Supplemental Table S2) in the 12 months before or 30 days after the index date except for bacterial infections, for which the identification period was defined as the 30 days before and 30 days after the LEA procedure to exclude infections that were not directly related to the amputation procedure.

Other comorbidities not directly associated with LEA based on clinical input were also examined using an ICD-10-CM code (Supplemental Table S3). These included non-coronary cardiovascular disease, coronary artery disease, cerebrovascular disease, chronic kidney disease, peripheral neuropathy, and primary cancer (excluding primary cancer of the lower extremities).

Statistical Analysis

Patient demographics were summarized with descriptive statistics, using mean and standard deviation (SD) or median and interquartile range for continuous variables and frequency and proportion for categorical variables. Types of LEA, associated diagnoses, and comorbidities were summarized as frequency and proportion. All statistical analyses were performed using SAS Enterprise Guide 7.1 (SAS Institute, Cary, NC).

Results

Characteristics of the Study Population

Of the 28,499 Veterans with at least one record of LEA during the observation period, 27,134 (95.2%) met the inclusion criteria. The characteristics of the study population are shown in Table 1. The mean (SD) age was 68.6 (9.9) years, and 67.3% of patients were at least 65 years old. Most patients were male (97.0%); the majority were non-Hispanic white (65.3%), and more than one-third (35.7%) of patients were from the Southern USA. Approximately 25.4% of patients had their index date in 2019; fewer than 20% had their index date in each year between 2020 and 2023.

Table 1.

Baseline demographic characteristics of the study population

Characteristic	N = 27,134
Age, years
Mean ± SD	68.6 ± 9.9
Median (IQR)	69.7 (62.8–74.7)
< 65	8872 (32.7)
≥ 65	18,262 (67.3)
Sex
Male	26,323 (97.0)
Female	811 (3.0)
Race/ethnicity, n (%)
White, non-Hispanic	17,724 (65.3)
Black	5855 (21.6)
Hispanic	1579 (5.8)
Othera	642 (2.4)
Unknown	1334 (4.9)
Geographic regionb, n (%)
South	9676 (35.7)
Midwest	7407 (27.3)
West	6132 (22.6)
Northeast	3919 (14.4)
Index year, n (%)
2019	6892 (25.4)
2020	5295 (19.5)
2021	5333 (19.7)
2022	4864 (17.9)
2023	4750 (17.5)

The baseline period was defined as the 12 months preceding the index date. The index date was the date of the first lower extremity amputation procedure

IQR interquartile range, SD standard deviation

^aIncludes Asian, Native Hawaiian or Pacific Islander, and American Indian or Alaskan Native

^bMidwest: IL, IN, IA, KS, MI, MN, MO, NE, ND, OH, SD, WI; Northeast: CT, ME, MA, NH, NJ, NY, PA, RI, VT; South: AL, AR, DE, DC, FL, GA, KY, LA, MD, MS, NC, OK, SC, TN, TX, VA, WV; West: AK, AZ, CA, CO, HI, ID, MT, NV, NM, OR, UT, WA, WY

Descriptive Analysis of LEA in the Study Population

Types of LEA

The most common type of LEA at the index date was transmetatarsal (52.9% of cases), followed by toe amputations (21.9%), above-knee amputations (15.4%), and below-knee amputations (9.8%) (Fig. 2). A total of 2455 patients (9.0%) had more than one type of LEA in the period from the index date to 30 days post index.

Fig. 2.

LEA type at index in US Veterans with LEA. LEA lower extremity amputation. LEA procedures were identified using the codes listed in the Supplementary Material. If multiple LEA procedures occurred within the 30-day period after the index LEA procedure, LEA type was categorized on the basis of the highest level of LEA according to the following hierarchy: above knee > below knee > transmetatarsal > toe

Diagnoses Associated with LEA

The most common diagnoses associated with LEA were diabetes (81.6%), bacterial infections (79.1%), and PAD (63.3%) (Fig. 3). Injuries to the lower extremities were only present in 18.4% of the population and only 0.1% (n = 15) had LEA associated with combat-related injury in the lower extremities. Primary cancer of the lower extremities was present in only 0.8% of patients.

Fig. 3.

Diagnoses associated with LEA [1, 2]. LEA lower extremity amputation. ^aAll diagnoses were identified using the codes found in the Supplementary Material. ^bDisease conditions associated with LEA were identified in the 12 months before and 30 days after the index LEA procedure. Patients could have more than one diagnosis associated with LEA. ^cInfection was identified in the 30 days before and 30 days after the index LEA procedure. ^dInjuries to the lower extremities included external injuries to the lower extremities (e.g., ankle and foot, knee and lower leg, hip and thigh) and were not necessarily related to combat trauma. ^ePatients who did not fall into any of the above categories were classified as having an unknown diagnosis related to LEA

Comorbidities

Besides those associated with LEA, the most common diagnoses observed in the study population was non-coronary cardiovascular disease, affecting 76.8% of patients (Supplemental Table S4). Additionally, coronary artery disease was present in 43.6% of patients, chronic kidney disease in 39.4%, and cerebrovascular disease in 19.4%. Other notable comorbidities were peripheral neuropathy (17.5%) and primary malignancy (12.9%).

Discussion

Rates of LEA have been shown to increase in the USA between 2008 and 2018 among the Veteran population [7], with limited contemporary information describing the types of LEA and associated disease conditions. To close this gap and understand the disease conditions associated with LEA, the present study examined the demographic and clinical characteristics of Veterans with LEA and the diagnoses associated with amputations. The results showed that few Veterans who underwent LEA had experienced injury to lower extremities (including during combat); rather, the most prevalent diagnoses associated with LEA were chronic disease, specifically diabetes, PAD, and bacterial infections. Moreover, Veterans had high rates of other medical comorbidities, especially cardiovascular diseases. The burden of LEA remains relatively high among US Veterans; the results of this study suggest that in most cases amputation is associated with modifiable risk factors. Effective management, including preventative measures, earlier detection, and using guideline-directed medical therapy, of these medical conditions offers an opportunity to reduce, or prevent altogether, LEA in our Veteran population.

The demographic profile of the study sample was similar to that of previously studied cohorts (mean age > 65 years and mostly male) [7, 10]. The distribution of LEA type was broadly similar to other reports [7]. Only a small proportion of amputations were above the knee (15.4%) and over half (52.9%) were transmetatarsal, which are less disabling as there is better preservation of mobility and limb function with more distal amputations. The number of patients who had LEA procedures declined over the study period, from 6892 in 2019 to 4750 in 2023. This is in contrast to the increasing incidence of LEA among Veterans reported in the decade prior (by 5.23 per 10,000 persons from 2008 to 2018) [7]. One possible reason for the declining number of patients who had LEA procedures observed in our study is the disruption of healthcare services throughout the VHA during the COVID-19 pandemic (2020–2023) [11]. However, it may also reflect the shrinking size of the Veteran population in the USA [8], or be a result of the PAVE program [6]. Additional studies are needed to determine the impact of the pandemic and prevention programs on rates of and factors contributing to LEA in Veterans.

A previous study reported that on average there were 113 LEAs per year among active-duty military personnel in a 16-year period from 2001 to 2017 [12], compared with 5427 LEAs per year among the Veterans in the most recent 5-year period in our study. In addition, unlike in the active-duty personnel whose amputations were typically related to injuries sustained in combat [12, 13], fewer than 1% of procedures in our study population were associated with a diagnosis of combat-related traumatic injury, with the majority being associated with chronic medical conditions. The results of this study, along with incidence data in the active-duty population, dispel any misconceptions as to combat being the primary cause of amputations in the military community; instead, most LEAs in the military population are performed in Veterans and not active-duty service members.

Our study showed that the most common diagnoses associated with LEA were bacterial infections and chronic diseases such as diabetes (81.6%) and PAD (63.3%). Diabetes and PAD are known risk factors for LEA [7]; together, they are estimated to lead to more than half of the LEAs performed in the USA each year [14]. The increased risk of amputation from these conditions may be due to peripheral neuropathy, infection, and ulcers (diabetes) and impairment of blood flow to the extremities (PAD) [14]. Diabetes and other cardiometabolic conditions are more prevalent among Veterans than in the general population [15]. Additionally, non-coronary cardiovascular disease was observed in 76.8% of the study sample; coronary artery disease was present in 43.6%, and chronic kidney disease in 39.4%. Other notable comorbidities were cerebrovascular disease (19.4%) and peripheral neuropathy (17.5%). The high rates of chronic diseases in this study are consistent with the poor overall health of US Veterans [16] and underscores the need for better risk-factor modification strategies for this high-risk population.

The observation that chronic diseases—and not traumatic injury sustained in combat—are the more prevalent diagnoses associated with LEA demonstrates that LEA in Veterans is potentially preventable with appropriate management [2, 17, 18]. While an annual foot examination is required by the PAVE program in all patients with diabetes [19], which has been linked to lower rates of LEA in Veterans with diabetes [20], a recent analysis demonstrated more than 30% of Veterans did not receive vascular assessment in the year prior to LEA [21]. Enrollment in the PAVE program also increased adherence to guideline-recommended therapies such as antiplatelet and lipid-lowering agents in patients with a diagnosis of PAD [22]. Additionally, a dose-dependent relationship was observed between the use of statins and the risk of LEA and mortality in Veterans with PAD [23]. Recent lower extremity PAD guidelines recommend ankle-brachial index screening for those at risk for PAD and low dose anticoagulation is recommended in patients with asymptomatic PAD to prevent amputation [24]. Thus, more effective management of diabetes and, particularly, PAD, where recent guidelines increase focus on detection and prevention, in Veterans can lead to improvements in their overall health and can potentially reduce their risk of LEA.

Strengths and Limitations

This is one of few studies to date that has investigated potential reasons for LEA among US Veterans. Besides the large sample size, a strength of this study was that it used current data (from the most recent 5 years) from the nationwide VHA database. The study had certain limitations. First, LEA and diagnoses of medical conditions were identified from the claims data using diagnosis codes, and there may have been misclassification (e.g., of LEA types) based solely on diagnosis codes and limitation of using diagnosis codes that are not readily used (e.g., combat injury-related codes). Similarly, for patients with multiple LEA procedures, the LEA type was identified on the basis of the highest level of amputation in the 30 days post index, which could have led to misclassification if a higher level of LEA occurred outside of the 30-day window. Second, as the observation period was limited to 12 months before and 30 days after the index LEA procedure, the study did not include Veterans with less than 12 months of continuous enrollment in the VHA or Veterans who may have died during index LEA procedure. The trends observed may not be reflective of entire Veteran population. However, as most Veterans remain in the VHA system for a long time, the proportion who did not meet this inclusion criterion was relatively small (3%). This study focused on amputations of lower extremities across all severity levels and age groups. Future studies that focus on a major amputation subgroup or stratified by age groups may be warranted to shed more insight on the heterogeneity of the risk profile for this population.

Conclusions

Diabetes and PAD are the main chronic diagnoses associated with LEA in US Veterans, with only a minor proportion of LEA procedures related to traumatic injuries sustained in combat. These findings highlight the need for earlier preventative and clinical management strategies to prevent or significantly reduce LEA among the US Veteran population.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

All authors (Brajesh Lal, Chi Gao, Fan Mu, Grace Chen, Qi Hua, Jared Calish, and Marie Parker) contributed to the study conception and design, drafting the manuscript, and critically reviewing the manuscript for intellectual content. Fan Mu, Chi Gao, Grace Chen, and Qi Hua contributed to data collection and analysis. All authors read and approved the final manuscript.

Funding

This study was funded by Janssen Scientific Affairs, LLC, a Johnson & Johnson company, which also funded the journal’s Rapid Service Fee and Open Access Fee.

Data Availability

The data that support the findings of this study were used under license and are not publicly available.

Declarations

Conflict of Interest

Brajesh Lal has nothing to disclose. Chi Gao, Fan Mu, Grace Chen, and Qi Hua are employees of Analysis Group, Inc., a consulting company that has provided paid consulting services to Janssen Scientific Affairs, LLC, a Johnson & Johnson company, which funded the development and conduct of this study and manuscript. Jared Calish and Marie Parker are employees of Janssen Scientific Affairs, LLC and stockholders of Johnson & Johnson.

Ethical Approval

The VHA database contains only de-identified patient information and is fully compliant with Health Insurance Portability and Accountability Act. Therefore, institutional review board review was not required.