A Mixed-Methods Approach to Understanding Diabetes and Peripheral Arterial Disease Related Amputations in a Rural Population

Abstract

Amputation is a devastating but preventable complication of diabetes and/or peripheral arterial disease (PAD). Rural populations have multiple risk factors for amputation; however, there is a significant gap in amputation research in rural communities. This 2-year study funded by the Society for Vascular Surgery Foundation used a mixed-methods approach, combining the strengths of quantitative and qualitative research methods, to understand the risk factors for amputation in patients with diabetes and/or PAD in the highly rural state of West Virginia. A sequential explanatory mixed-methods design was used. Quantitative methods included spatial epidemiology and biostatistical database analyses and informed the recruitment strategy for our qualitative methods. Statewide 2011 to 2016 data from the West Virginia HCUP State Inpatient Database and more detailed longitudinal data from the West Virginia Clinical and Translational Science Institute Integrated Data Repository were used for quantitative analyses. Advanced spatial hierarchical analyses were used to identify geographic hot spots for amputation. Biostatistical analyses using descriptive and multivariable models identified prevalence rates and risk-factors for amputation. Our qualitative approach used semistructured focus groups and 1:1 interviews. We identified a 7/1000 (3/1000 minor, 4/1000 major) prevalence of amputation in patients with PAD and/or diabetes in West Virginia. Our spatial analyses identified zip codes across the state with significantly higher risk of amputation even when controlling for covariates (Figure). Our biostatistical analyses found that insurance type was significantly associated with amputation (self-pay and public vs private), and that patients with PAD alone, or PAD with diabetes had strikingly higher odds of amputation compared to patients with diabetes alone (Table). Our qualitative analysis included interviews with 64 patients and providers and highlighted issues related to (1) lack of education, (2) geographic and cultural barriers, (3) care coordination, (4) patient adherence and (5) depression, diabetes and tobacco use. This mixed-methods study identified that West Virginians are at higher risk for amputation compared to the rest of the country, with a prevalence rate of 7/1000 compared to the national prevalence of 2.4/1000. It also identified high-risk zip codes within the state. Additionally, it identified that patients with public (Medicare/Medicaid) and self-pay insurance, as well as patients with PAD alone or PAD with diabetes, are at significantly higher risk of amputation than patients with private insurance or patients with diabetes alone. Finally, our qualitative analysis provided multiple areas to improve care. This data will be used to work with communities in high-risk areas to build community-level interventions to prevent amputation.Table IMultivariable analysis of odds of major, minor and any amputation in patients with diabetes and/or PAD in West Virginia (From the West Virginia Clinical and Translational Science Institute Integrated Data Repository)VariableMajor (adjusting for minor)Minor (excluding majors)Any amputationMinora11.79 (7.19-19.33)––Rural (defined using rural-urban community area code)1.09 (0.8-1.48)0.82 (0.52-1.29)0.98 (0.77-1.27)Tobacco user0.91 (0.57-1.44)0.9 (0.48-1.71)0.9 (0.62-1.31)Male vs female2.15 (1.61-2.85)2.63 (1.77-3.9)2.43 (1.93-3.07)Medicaid vs private1.79 (1.18-2.71)0.88 (0.51-1.51)1.35 (0.98-1.87)Medicare vs private1.54 (1.04-2.26)0.7 (0.43-1.15)1.12 (0.83-1.52)Self-pay vs private1.83 (1.09-3.07)1.42 (0.76-2.66)1.69 (1.13-2.52)Age0.99 (0.98-1)0.99 (0.97-1.01)0.99 (0.98-1)CAD0.36 (0.27-0.49)0.18 (0.12-0.28)0.26 (0.2-0.33)PAD vs diabetes13.78 (7.75-24.49)8.4 (4.18-16.85)12.29 (7.93-19.07)PAD and diabetes vs diabetes41.07 (23.27-72.46)52.52 (27.43-100.53)51.81 (33.85-79.3)CHF1.11 (0.79-1.54)1.5 (0.94-2.38)1.24 (0.95-1.62)CKD1.83 (1.32-2.54)1.5 (0.95-2.36)1.75 (1.34-2.27)COPD0.99 (0.73-1.34)1.07 (0.7-1.64)1 (0.78-1.29)Hypercholesterolemia0.97 (0.73-1.3)0.76 (0.52-1.13)0.89 (0.71-1.13)Obesity0.92 (0.68-1.25)0.92 (0.61-1.4)0.93 (0.73-1.19)Renal failure1.47 (0.97-2.24)0.85 (0.43-1.7)1.24 (0.87-1.77)C-statistic0.8500.8160.841CAD, Coronary artery disease; CHF, congestive heart failure; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; PAD, 0peripheral artery disease.Values are odds ratio (95% confidence interval).Text in bold indicates statistically significant findings.aFor those with both major and minor we only counted minors occurring before major to see if that led to higher odds of major. Open table in a new tab