If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Lower extremity amputation is often performed in patients where both lower extremities are at risk due to peripheral arterial disease or diabetes, yet the proportion of patients who progress to amputation of their contralateral limb is not well defined. We sought to determine the rate of subsequent amputation on both the ipsilateral and contralateral lower extremities following initial amputation.
We conducted a retrospective review of all patients undergoing lower extremity amputation (exclusive of trauma or tumor) at our institution from 1998 to 2010. We used International Classification of Diseases-Ninth Revision codes to identify patients and procedures as well as comorbidities. Outcomes included the proportion of patients at 1 and 5 years undergoing contralateral and ipsilateral major and minor amputation stratified by initial major vs minor amputation. Cox proportional hazards regression analysis was performed to determine predictors of major contralateral amputation.
We identified 1715 patients. Mean age was 67.2 years, 63% were male, 77% were diabetic, and 34% underwent an initial major amputation. After major amputation, 5.7% and 11.5% have a contralateral major amputation at 1 and 5 years, respectively. After minor amputation, 3.2% and 8.4% have a contralateral major amputation at 1 and 5 years while 10.5% and 14.2% have an ipsilateral major amputation at 1 and 5 years, respectively. Cox proportional hazards regression analysis revealed end-stage renal disease (hazard ratio [HR], 3.9; 95% confidence interval [CI], 2.3-6.5), chronic renal insufficiency (HR, 2.2; 95% CI, 1.5-3.3), atherosclerosis without diabetic neuropathy (HR, 2.9; 95% CI, 1.5-5.7), atherosclerosis with diabetic neuropathy (HR, 9.1; 95% CI, 3.7-22.5), and initial major amputation (HR, 1.8; 95% CI, 1.3-2.6) were independently predictive of subsequent contralateral major amputation.
Rates of contralateral limb amputation are high and predicted by renal disease, atherosclerosis, and atherosclerosis with diabetic neuropathy. Physicians and patients should be alert to the high risk of subsequent amputation in the contralateral leg. All patients, but particularly those at increased risk, should undergo close surveillance and counseling to help prevent subsequent amputations in their contralateral lower extremity.
Lower extremity amputation is often performed due to consequences of peripheral arterial disease (PAD), a problem which affects an estimated 5 million adults over the age of 40.
Factors that place patients at higher risk for PAD as well as lower extremity amputation affect not just the limb undergoing the amputation but also the patient systemically. Prior studies have focused on the primary limb undergoing amputation. The fate of the contralateral limb, however, has not been well described in the literature. Published rates of amputation in the contralateral limb vary from 2.2% to 19.8% over periods from 12 months to 10 years.
we feel it is important to further investigate factors that may lead to contralateral amputations.
We performed a retrospective review of all patients undergoing a first time lower extremity amputation at our institution from 1998 to 2010. Patients were identified using the International Classification of Diseases, Ninth Edition (ICD-9) procedure codes for lower extremity amputation (84.10-84.19). Toe or transmetatarsal amputations were considered minor (84.11-84.12) and amputations through the ankle or above were considered major (84.13-84.19). We then performed a thorough chart review of all patients identified with ICD-9 codes to verify amputation levels as well as the laterality of all amputations. Patients were included in our analysis if their primary admission ICD-9 diagnosis code was related to nonhealing wounds with or without PAD or ischemic rest pain. Those whose primary admission ICD-9 diagnosis code was for trauma, tumor, or orthopedic complications were excluded (Supplementary Table I [online only] contains the complete list of ICD-9 diagnosis codes included and excluded in our analysis). Admission ICD-9 diagnosis codes were used to determine the presence of comorbidities (Supplementary Table II [online only] contains a list of all ICD-9 codes used to identify diabetes, chronic renal insufficiency [CRI], end-stage renal disease [ESRD], and hypertension). This method has been shown to be an accurate method of identifying those with diabetes,
but the severity of diabetes and insulin dependence was not specified. The specific set of ICD-9 diagnosis codes used to identify patients with CRI, while not able to provide the stage of disease, have been shown to have a positive predictive value greater than 90% for identifying those with chronic kidney disease.
A hierarchical model was used and the presence of a dialysis ICD-9 code classifies the patient as having ESRD whether or not a CRI diagnosis code is present. Patients with diabetes were stratified according the presence or absence of diagnosis codes for diabetic neuropathy.
Patients were stratified based on the level of their first amputation, either minor or major. Subsequent amputations were classified as ipsilateral or contralateral as compared with the initial amputation and as major or minor based on the level of amputation. If an initial amputation was performed to control sepsis, with a planned subsequent amputation, the subsequent amputation was not included in our analysis. Our primary outcome was the rate of subsequent major amputations in the contralateral leg. Secondary outcomes included the rate of other subsequent amputations including contralateral minor, ipsilateral minor, and ipsilateral major amputations. Chart review determined the date of the patient's last known follow-up at our institution. The Social Security Death Index was used to determine the date of death.
Preoperative characteristics are reported as proportions of the sample and mean ± standard deviation. Categorical variables were analyzed using χ2 and the Fisher exact test where appropriate. Continuous variables were compared using the Wilcoxon rank sum test. Kaplan-Meier estimates were used to determine rates of subsequent amputations. Cox proportional hazards modeling was used to determine rates of subsequent contralateral major amputations over time as well as to determine rates of contralateral major amputations in those subgroups of patients at increased risk: (1) all patients stratified based on the presence of diabetes; (2) all patients stratified based on the presence of CRI; (3) all patients stratified based on the presence of ESRD; and (4) diabetic patients with and without CRI. Statistical significance was defined as P < .05. All statistical analyses were performed using STATA 12 software (StataCorp LP, College Station, Tex). This study was approved by the Institutional Review Board at Beth Israel Deaconess Medical Center.
We identified 1715 patients who underwent a lower extremity amputation due to PAD at our institution: 575 underwent major amputations and 1140 underwent minor amputations. Of the initial major amputations, below-knee amputations were performed on 431 patients and initial above-knee amputations were performed on 144 patients. Sixty-three percent of all patients were male, and 77% were diabetic. The average age was 67.2 years. Demographics stratified by the level of initial amputation are shown in Table I. Patients who initially underwent major amputations tended to be older and were more likely to be diabetic when compared with those who initially underwent minor amputations. While men made up the majority of those undergoing both minor and major amputations, a greater proportion of women underwent an initial major amputation compared to men (40.8% vs 29.2%; P < .01). The prevalence of hypertension, CRI, ESRD, and nonwhite race was similar between those undergoing initial minor vs initial major amputations.
Table IDemographics and comorbidities of the 1715 patients who underwent an initial lower extremity amputation, stratified by level (minor vs major)
Of the 1715 patients who underwent an initial amputation at our institution during the study period, 559 patients went on to have 729 subsequent amputations. The mean follow-up time was 3 years. Of the subsequent amputations, 415 were on the ipsilateral limb, including 184 major amputations, and 314 were on the contralateral limb, including 134 major amputations. A large proportion of the initial population (1156 patients) did not undergo a subsequent amputation. Characteristics of patients who underwent a contralateral major amputation as well as those who did not undergo any subsequent amputation are shown in Table II. Compared with those who did not undergo a subsequent amputation, patients who underwent a major amputation on the contralateral limb were more likely to be female, of nonwhite race, diabetic, and have renal disease.
Table IIDemographics and comorbidities of patients with no subsequent amputations compared with those with a subsequent contralateral major amputation
Rates of subsequent amputations varied by the level of the initial amputation (Table III). During follow-up, 11.5% of those undergoing an initial major amputation had a contralateral major amputation by 5 years compared with 8.4% of those undergoing an initial minor amputation. Of those undergoing an initial major amputation, 8.4% had an ipsilateral revision amputation or more proximal major amputation by 5 years while 14.2% of those undergoing an initial minor amputation had an ipsilateral major amputation. Table IV lists the rates of subsequent minor amputations. Of those undergoing an initial minor amputation, by 5 years, 15.0% underwent a contralateral minor amputation, while 19.9% had an additional subsequent ipsilateral minor amputation. Nine percent of those undergoing an initial major amputation had a contralateral minor amputation by 5 years. Fig 1 shows freedom from contralateral major amputation based on the level of initial amputation.
Table IIIPatients with subsequent major amputation, stratified by level of initial amputation
Cox regression analysis revealed several factors that were predictive of a contralateral major amputation. ESRD (hazard ratio [HR], 3.9; 95% confidence interval [CI], 2.3-6.5), CRI (HR, 2.2; 95% CI, 1.5-3.3), atherosclerosis without diabetic neuropathy (HR, 2.9; 95% CI, 1.5-5.7), atherosclerosis with diabetic neuropathy (HR, 9.1; 95% CI, 3.7-22.5), and initial major amputation (HR, 1.6; 95% CI, 1.3-2.6) were independently predictive of subsequent contralateral major amputation (Table V). In isolation, initial below-knee (HR, 1.3; 95% CI, 0.9-1.9) and above-knee (HR, 1.6; 95% CI, 0.9-3.1) amputations were not differentially associated with subsequent contralateral major amputations.
Table VPredictors of contralateral major amputation
For those undergoing an initial major amputation, diabetes was significantly associated with subsequent contralateral major amputations (HR, 2.8; 95% CI, 1.1-7.1; P = .02; Fig 2, A). However, in those undergoing an initial minor amputation, the risk of contralateral major amputation between diabetics and nondiabetics was similar in magnitude and direction but did not reach statistical significance (HR, 1.8; 95% CI, 0.8-3.9; P = .13; Fig 2, B). Patients with CRI had higher rates of contralateral major amputations than those without (HR, 2.7; 95% CI, 1.5-4.8; P < .001 for those undergoing initial major amputations; Fig 3, A; and HR, 2.0; 95% CI, 1.2-3.2; P < .01 for those undergoing initial minor amputations; Fig 3, B). Patients with ESRD had higher rates of contralateral major amputations than those without for those undergoing initial major (HR, 3.2; 95% CI, 1.4-7.2; P < .01) and initial minor (HR, 4.4; 95% CI, 2.5-7.9; P < .01) amputations, respectively (Fig 4, A and B). Among diabetics, those with CRI also had higher rates of contralateral major amputations (HR, 2.5; 95% CI, 1.4-4.7; P < .01 and HR, 1.7; 95% CI, 1.0-2.8; P = .04) for initial major and initial minor amputations, respectively.
Kaplan-Meier estimates of mortality were 17.0%, 29.1%, and 49.0% at 1, 3, and 5 years after initial minor amputations, and 19.2%, 48.7%, and 61.3% at 1, 3, and 5 years after initial major amputations (Fig 5).
Our analysis shows that 8.4% of those undergoing an initial minor amputation and 11.5% of those undergoing an initial major amputation will undergo a contralateral major amputation within 5 years. Approximately 14% of those undergoing an initial minor amputation will need an ipsilateral major amputation within 5 years. ESRD, CRI, atherosclerosis with and without diabetic neuropathy, and an initial major amputation all predict subsequent contralateral major amputations.
Prior studies have evaluated subpopulations of patients, such as studies of outcomes in diabetics only,
; most did not examine the fates of those whose initial amputation was minor. Variation in subsequent amputation rates between those studies and ours may be accounted for by the smaller number of patients in previous works. Studies using data collected prior to the year 2000 show rates higher than ours, with 15% to 53.3% of patients undergoing amputations needing a contralateral major amputation within 2 to 5 years.
Our study further illustrates that those with diabetes and renal disease are a population particularly at risk for limb loss. In our study, the likelihood of progression to contralateral major amputation increases between CRI and ESRD; this suggests that as the severity of kidney disease worsens so does the prognosis for the contralateral limb. Among diabetics, those with renal disease fare poorly.
Prior work has shown that more men than women undergo lower extremity amputation,
Our results are similar, as the majority of the amputations performed at our institution were in men. However, women were more likely to have a major vs minor amputation compared with men. We also found that women are more likely than men to progress to a subsequent contralateral major amputation. There are many possible reasons as to why this occurs. Women have been found to be more likely to present with occlusions as well as multilevel disease as compared with men, even when adjusted for age, smoking, and diabetes.
In a study of gender differences among all lower extremity revascularization or amputation procedures, women were shown to have lower rates of hospitalization for revascularization procedures but higher rates of mortality after amputation, and women tended to present at later stages of disease.
Hormonal changes as women age may produce a different vascular environment from their male counterparts, which may lead to some of these changes. Elucidating the reasons for their elevated risk in this study and others necessitates further study, potentially on both a basic science and clinical level.
Our survival rate is comparable to previously reported estimates, suggesting that this patient cohort is comparable to others in the literature.
This corresponds with our finding that minor amputees have better survival initially than those undergoing major amputation.
As with all single-institution studies, our data and conclusions may not extend to all patients. However, chart review allowed us to definitively state whether a subsequent amputation was ipsilateral or contralateral; this is not the case in studies of large databases such as Medicare or the Nationwide Inpatient Sample where laterality has to be inferred. The large sample included in our study, over 1700 patients, makes it more likely that our results are applicable to other populations. Our institution is located in a major metropolitan area; we cannot account for amputations that patients may have had elsewhere. Additionally, our institution is well known for aggressive revascularization for limb salvage.
Additionally, the study is biased toward patients whose operations occurred earlier in our study period due to a longer follow-up time; our data may underestimate the effects of changes in practice over the 12-year period during which it was collected. We also do not have presenting clinical symptoms and data on prior or simultaneous revascularization, which may affect initial treatment decisions and risk for subsequent amputations. We do not have ankle brachial index data for our patients, which prevents us from determining the severity of PAD, and we did not account for the level of disease when examining rates of amputation. This prevents us from stratifying patients based on the extent or severity of their PAD, but we feel that none of these limitations significantly detracts from our overall conclusions.
In this study, we have shown that the rate of a contralateral major amputation is 8.4% and 11.5% within 5 years for all patients undergoing initial minor and major amputations, respectively. Specifically, those with renal disease, atherosclerosis, and diabetic neuropathy are at even greater risk. All patients, but especially these high-risk groups, should have close surveillance of their contralateral limb.
Conception and design: JG, RB, RH, MS
Analysis and interpretation: JG, RB, RH, SD, AH, MW, EC, MS
Data collection: JG, RB, RH
Writing the article: JG, RB, RH, MS
Critical revision of the article: JG, RB, RH, SD, AH, MW, EC, MS
Final approval of the article: JG, RB, RH, SD, AH, MW, EC, MS
Statistical analysis: JG, RB, RH, SD
Obtained funding: Not applicable
Overall responsibility: MS
JDG and RPB contributed equally to this work.
Supplementary Table I (online only)ICD-9 diagnosis codes used to identify patients with amputations due to non-healing wounds with or without peripheral arterial disease or ischemic rest pain
Complications of vascular procedures
996.1, 996.62, 996.74, 997.6
Gangrene and sepsis
Peripheral vascular disease
171-3.x, 198.x, 215.x
Fractures and dislocations
858.3-8, 928.x, 906.4
Pelvic pyogenic arthritis
ICD-9, International Classification of Diseases, Ninth Edition.
Supported by the National Institutes of Health T32 Harvard-Longwood Research Training in Vascular Surgery grant HL007734 and the Sarnoff Cardiovascular Research Foundation Fellowship for Medical Students.
Author conflict of interest: Dr Schermerhorn is a consultant for Endologix, Boston Scientific, and Medtronic.
The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest.