Amputation in Patients with Peripheral Arterial Disease and Diabetes Provides an Opportunity to Improve Hemoglobin A1C and Low-Density Lipoprotein Cholesterol Levels

Amputation is a devastating but preventable complication of diabetes and peripheral arterial disease (PAD). It is a life-altering event that identifies disparities in access to health care and management of disease. In this study we sought to (1) better understand how access to care and/or quality of care represented by hemoglobin A1c (HgA1C) and low-density lipoprotein (LDL) cholesterol levels effect risk of amputation, and (2) identify whether the event of amputation is associated with changes in these levels. The West Virginia Clinical and Translational Science Institute Integrated Data Repository, a longitudinal database of over 2 million patients, containing electronic health record data for all inpatient and outpatient encounters at West Virginia University and affiliated hospitals and clinics comprised the sample for analyses. Adults (≥18 years) with diagnoses of diabetes and/or PAD between 2011 and 2016 were included based on International Classification of Diseases, 9th and 10th editions, codes. Major and minor amputations were identified using Current Procedural Terminology codes. Trauma patients were excluded. Multivariable logistic regression analyses were performed on patients with lab information for both HgA1C and LDL while adjusting for patient factors to examine associations with amputations. In addition, in patients who underwent amputation, we compared average laboratory values before and after amputation using Wilcoxon signed rank tests. During the study period, 50,276 patients were seen for diabetes and/or PAD. The total number of any amputation (major or minor) was 369 (7.3/1000). Overall, we identified 12,930 patients with lab values for HgA1C and LDL. On multivariable analysis (Table I), we found significant associations between odds of amputation and HgA1C, but not for LDL. Specifically, for each 1-unit increase in HgA1C we found a 31% increase in the odds of amputation. In the amputation cohort (n = 369), 103 patients had both before and after laboratory values for HgA1C, and 46 had laboratory values for LDL. When comparing before and after values for patients with amputations we identified significant decreases in median lab values for HgA1C and LDL (Table II). In this study, we found that even small increases in HgA1C can substantially increase the risk of amputation. In addition, we also found that levels of HgA1C and LDL dropped significantly after amputation. These findings highlight the importance of medical optimization and patient education and also suggest that an amputation event may provide an important opportunity for changes in disease management and patient behavior. Surprisingly, only 26% of our patients had laboratory values available, which indicates an opportunity for improvement in disease management across our health care system.Table IMultivariable analysis of risk for amputation based on hemoglobin A1c (HgA1C) and low-density lipoprotein (LDL) levelsLaboratory testPatients average value (no amputation group), median (IQR)Patients average value (amputation group), median (IQR)Wilcoxon testAny amputation, AORa (95% CI)HgA1C6.5 (5.7-7.8)7.3 (6.1-9.6)<.00011.31 (1.15-1.48)LDL93.0 (72.0-117.0)83.4 (63-110.0).00950.98 (0.95-1.01)bAOR, Adjusted odds ratio; IQR, interquartile range.Text in bold indicates statistically significant findings.aAdjusted for rural/urban status, tobacco use, sex, payor type, age, coronary artery disease, peripheral artery disease/diabetes, congestive heart failure, chronic kidney disease, chronic obstructive pulmonary disease, hypercholesterolemia, obesity, and renal failure.bPer 5-unit increase. Open table in a new tab Table IIChanges in hemoglobin A1c (HgA1C) and low-density lipoprotein (LDL) levels after amputationLaboratory testPatients average value before, median (IQR)Patients average value after, median (IQR)Wilcoxon signed rank testHgA1C7.7 (6.4-10.0)7.0 (5.8-8.6)<.0001LDL100.3 (69.5-118.0)84.0 (58.7-113.0).0450IQR, Interquartile range.Text in bold indicates statistically significant findings. Open table in a new tab