Chronic Kidney Disease: Evaluating Area Level Socioeconomic Characteristics, Individual Characteristics, and Patterns of Care by Census Tract in the Twin Cities Metro Area

Abstract

Chronic kidney disease (CKD) is a major public health problem that affects 15% of the United States adult population. CKD is asymptomatic until advanced stages; therefore, diagnosis and treatment are usually delayed. CKD has been shown to be associated with low socioeconomic status (SES). Disadvantaged socioeconomic populations have a disproportionate burden of CKD and have contributed to the growing CKD epidemic. This dissertation characterizes the interplay between neighborhood characteristics and CKD using the Fairview Health Services electronic health record (EHR) database, a large health system in Minnesota that services the Twin Cities metro area and surrounding communities. For all manuscripts, we used census tracts as our geographical unit of analysis. Data for census tract was obtained from the American Community Survey 5-year data [2008-2012]. We linked each patients’ residence, obtained from their EHR data, to the appropriate census tract and tract characteristics. We defined a low and high SES tract as belonging to the first and fourth quartile, respectively of the distribution of each census tract SES measure in the metropolitan area. In the first manuscript, we evaluated whether adding neighborhood SES to current screening recommendations – patients with hypertension and/or diabetes - improves the sensitivity and/or specificity for detecting CKD. Overall, CKD was prevalent in 13% of our cohort (2008-2019). Additional screening of patients who live in low neighborhood SES improved our sensitivity but decreased our specificity to detect CKD compared to current screening recommendations. In the second manuscript, we examine: 1) if there is an independent association of neighborhood SES and individual insurance type with CKD prevalence and 2) association of neighborhood racial composition, a measure of racial segregation, with CKD prevalence. We observed that patients living in low vs. high SES neighborhood had higher CKD prevalence. Moreover, for patients <65 years being on Medicaid compared to other insurance was associated with higher CKD prevalence. We found no association between neighborhood racial composition (percent blacks in tracts) and CKD prevalence. The third manuscript explores whether neighborhood SES is associated with quality of care received by patients with CKD. Interestingly, we found that quality of care measures (prescribing appropriate medications, testing for proteinuria, or documenting CKD in patients’ charts) are not associated with neighborhood SES. However, overall adherence to CKD guidelines is low (for example, only 27% of patients with CKD had their proteinuria level evaluated). Overall, area SES contributes to CKD burden among patients seen at Fairview Health Services. In the future, health systems may want to consider a multifactorial approach, including neighborhood characteristics and a patient’s individual level SES, to improve detection and management of CKD.