Impact of Race-Ethnic and Economic Disparities on Rates of Vascular Dementia in the National Inpatient Sample Database from 2006-2014

Abstract

Background We have previously reported regional differences in the rates of vascular dementia (VD) in the United States (US), with the West having lower rates of VD diagnosis found in the Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS) Database. We aimed to determine if these regional differences were associated with race-ethnic disparities in economic status or prevalence of the stroke risk factors hypertension, diabetes, and hyperlipidemia. Methods A sampling of the NIS database from 2006-2014 was analyzed using ICD-9 codes (290.40, 290.41, and 290.43) to identify patients admitted with primary or secondary VD diagnoses. Multivariable logistic regression was used to identify significant associations between ethnic groups, stroke risk factors and income. Results The NIS sample included 69,273,455 patients with median age of 53 (interquartile range 27 - 72), 55% female and 34% non-Caucasian. The mean VD primary and secondary diagnosis were 0.03% and 0.21% respectively. Black/African Americans had the highest odds of being admitted with a primary VD code [odds ratio (OR) 1.307, confidence interval (CI) 1.248-1.369] or admitted with VD as a secondary diagnosis code [OR 1.998, CI 1.968-2.03]. Black/African Americans admitted with VD as a primary diagnosis were the only group to have increased odds of having all 3 stroke risk factors, hypertension [OR 1.856, CI 1.705-2.019], diabetes [OR 1.166, CI 1.072-1.268], and hyperlipidemia [OR 1.164, CI 1.062-1.276]. Also, Black/African Americans admitted with VD as a secondary diagnosis had increased odds for the most stroke risk factors of the ethnic groups, including diabetes [OR 1.187, CI 1.155-1.22] and hypertension [OR 1.04, CI 1.012-1.068]. Caucasians were the only group less likely to be admitted with a primary VD diagnosis with increasing income [$25,001-30,000, OR 0.782; $30,001-35,000, OR 0.75; $35,000 or greater, OR 0.77; p-value Conclusion We demonstrate how healthcare disparities in race-ethnicity can influence the frequency of VD diagnosis in the US. This highlights the need for investigating the impact of healthcare disparities across the spectrum of cerebrovascular disease.