2017 HSR&D/QUERI National Conference
4011 — Identifying and Classifying Health Disparities in VA: Application to Racial Disparities in Minimally Invasive Hysterectomy
Lead/Presenter: Kristen Gray, COIN - Seattle/Denver
All Authors: Gray KE (VA Puget Sound Health Care System, HSR&D; University of Washington School of Public Health, Department of Health Services)
Callegari LS (VA Puget Sound Health Care System, HSR&D; University of Washington, Department of Obstetrics and Gynecology )
Fortney JC (VA Puget Sound Health Care System, HSR&D; University of Washington, Department of Psychiatry & Behavioral Sciences)
Lynch KE (VA Salt Lake City Health Care System; University of Utah, Department of Internal Medicine, Division of Epidemiology )
Zephyrin LC (VA Office of Patient Care Services, Women’s Health Services; VA New York Harbor; NYU Langone School of Medicine, Dept. of Obstetrics and Gynecology)
Uchendu US (VA Office of Health Equity)
Katon JG (VA Puget Sound Health Care System, HSR&D; University of Washington School of Public Health, Department of Health Services)
As patient- and health system (HS)-level factors may contribute to racial/ethnic health disparities, determining their contributions is critical to mitigating disparities. Approaches are needed to characterize HS performance, accounting for both care quality and the magnitude of racial differences. We applied a method of disparity identification and classification to examine racial/ethnic differences in minimally invasive (MI) hysterectomy in VA.
We identified hysterectomies performed at VA for benign indications among women Veterans for 2012-2014. For each HS, we calculated the proportion of hysterectomies performed as MI (vaginal or laparoscopic) as a measure of care quality, as well as the proportion MI by race (black and white), and the difference in proportions between races (white minus black). We plotted racial differences in proportions versus the proportion MI overall and, by using median values across HSs as cutoffs, identified four quadrants. The lower right quadrant reflected higher quality and small racial differences; the upper right reflected higher quality and large racial differences; the lower left reflected lower quality and small racial differences; and the upper left reflected lower quality and large racial differences.
We restricted data to 21 HSs performing > = 10 hysterectomies/year. Quadrants allowed separation of HS-level differences in uptake of MI approaches and patient-level differences by race. This classification system will be used to identify patterns in practice and patient characteristics within and between the four quadrants (no disparity, patient-related disparity, HS-related disparity, both patient and HS-related disparity).
Use of this method facilitated identification of two potential contributors to MI hysterectomy disparities: 1) overall uptake of MI procedures as a proxy for care quality and 2) racial differences in the use of these procedures.
Understanding origins of disparities in MI hysterectomy and developing strategies to address contributors will be important to providing high-quality, equitable care to women Veterans, in line with the VHA Strategic Plan and VA Health Equity Action Plan. Results will help the Office of Health Equity prioritize HSs for delivering interventions that reduce disparities and improve quality. More broadly, this method may be applied to a variety of treatments/procedures to identify health disparities.