2006 HSR&D National Meeting Abstract
1001 — Racial Differences in Procedures: Are VAMCs with Large Minority Populations Underusing Technology?
Groeneveld PW (Philadelphia VAMC/CHERP)
Kruse G (Wharton School, University of Pennsylvania)
Chen Z (Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania)
Asch DA (Philadephia VAMC/CHERP)
In the VA healthcare system, blacks receive fewer major medical procedures than whites, but the reasons for this are unclear. Variation in procedure use among VA hospitals may play an important role, particularly if VAMCs with larger black inpatient populations perform fewer procedures than VAMCs with predominantly white inpatient populations.
We examined national VA inpatient, outpatient, and fee-basis procedure utilization data from 1998-2003. Assuming that rapid-growth procedures would be the most likely to vary in use, we selected 10 procedures that have rapidly increased in volume throughout the VA healthcare system recently. Using the cohort of all patients admitted to a VAMC in 1998-2003, we fitted 10 logistic regression models to generate propensity scores for receipt of each procedure based on each patient's relevant diagnoses, comorbidities, and age. Using these scores, we formed 10 cohorts of patients with propensity scores in the top one percentile for procedure likelihood. Finally, a logistic regression model predicting procedure receipt within 90 days was fitted for each cohort. Covariates included race, age, admission to a VAMC with a large black inpatient population, and comorbidities.
Of the ten procedures selected for analysis, bioprosthetic aortic valve replacement (odds ratio [OR] 0.29, 95% confidence interval [CI] 0.18-0.49), endoscopic treatment of gastrointestinal bleeding (OR 0.57, 95% CI 0.43-0.75), colonoscopy (OR 0.62, 95% CI 0.49-0.79), open prostatectomy (OR 0.44, 95% CI 0.37-0.52), total hip replacement (OR 0.51, 95% CI 0.43-0.60), and endoscopic biliary procedures (OR 0.32, 95% CI 0.25-0.41) were performed less frequently at hospitals with larger black inpatient populations. Only percutaneous coronary interventions were performed more frequently (OR 1.23, 95% CI 1.09-1.38) in VAMCs with larger black populations. There were no signficant differences in the use of implantable cardioverter-defibrillators, dual chamber pacemakers, or laparoscopic appendectomies.
VAMCs with larger black inpatient populations performed several major procedures less frequently among hospitalized veterans than predominantly white VAMCs.
Health care equity in the VA critically depends on the performance of VAMCs that care for large numbers of minority patients. These results suggest that VAMCs with larger black inpatient populations may face more barriers in delivering medical procedures to their patients than other VAMCs.