Lead/Presenter: Tina Willson, COIN - Salt Lake City
All Authors: Willson TM (Informatics, Decision-Enhancement and Analytic Sciences (IDEAS) Center, VA Salt Lake City Health Care System), Callahan, A (Informatics, Decision-Enhancement and Analytic Sciences (IDEAS) Center, VA Salt Lake City Health Care System; Division of Epidemiology, University of Utah) Stevens, V (VA Office of Clinical Systems Development and Evaluation (CSDE); Informatics, Decision-Enhancement and Analytic Sciences (IDEAS) Center, VA Salt Lake City Health Care System; Division of Epidemiology, University of Utah) Widanagamaachchi, W (Informatics, Decision-Enhancement and Analytic Sciences (IDEAS) Center, VA Salt Lake City Health Care System; Division of Epidemiology, University of Utah; VA Office of Clinical Systems Development and Evaluation (CSDE)) Dalton, C (Informatics, Decision-Enhancement and Analytic Sciences (IDEAS) Center, VA Salt Lake City Health Care System; Division of Epidemiology, University of Utah) Jones, M (VA Office of Clinical Systems Development and Evaluation (CSDE); Informatics, Decision-Enhancement and Analytic Sciences (IDEAS) Center, VA Salt Lake City Health Care System; Division of Epidemiology, University of Utah) Plomondon, M (VA Office of Clinical Systems Development and Evaluation (CSDE)) Box, T (VA Office of Analytics and Performance Integration (API)) Francis, J (VA Office of Analytics and Performance Integration (API))
Objectives:
Most focus on disparity research has been on differences in measured outcomes; however, measurement itself may be biased. We reasoned that relative differences in acuity at clinical presentation would likely be related to relative bias in ascertainment, i.e., if a group presented with higher acuity, then cases were likely relatively under ascertained in that group. To investigate, we examined the presenting acuity for Veterans diagnosed with COVID-19 across different COVID-19 waves and demographic groups.
Methods:
Using data from the VA COVID-19 Shared Data Resource, we identified 547,725 Veterans who tested positive for COVID-19 from March 1, 2020 through May 18, 2022 and presented to the VA for care within 2 days of positive diagnosis. Patient acuity at presentation was extracted from Emergency Department Integration Software (EDIS) data and augmented with clinic visit data. Acuity was ordered from 1 (most severe) to 4+ (least severe). We examined presenting acuity by demographic characteristics including age, sex, race, ethnicity, and by COVID-19 wave. COVID-19 waves were defined as follows: Wave 1 (3/1/2020-9/8/2020), Wave 2 (9/9/2020-6/19/2021), Wave 3/Delta (6/20/2021-11/26/2021), Wave 4/Omicron (11/27/2021-3/21/2022) and Wave 5 (3/21/2021-present). We measured the association of race with acuity at presentation using ordered logistic regression to predict acuity with robust estimation run for each wave, adjusting for age, region, sex, vaccination status, rurality, primary care or other healthcare activity, insurance status, and VA priority group.
Results:
Overall, 4.1% of Veterans diagnosed with COVID-19 presented with the most severe acuity levels. Male Veterans presented with higher severity than females and severity at presentation increased with age. The percentage of Veterans with more severe presenting acuity was highest in Wave 1 and lowest in Wave 4. Black or African American Veterans had more severe acuity at time of presentation across all COVID-19 Waves. After adjustment, Black or African American Veterans had significantly higher levels of severity at presentation during Waves 1-4 although the effect diminished with time (Wave 1: Odds Ratio (OR) = 1.80, 95% Confidence Interval(CI) = 1.69-1.92; Wave 2: OR = 1.64, 95%CI = 1.59-1.69; Wave 3: OR = 1.47, 95%CI = 1.41-1.53, Wave 4: OR = 1.24, 95%CI = 1.20-1.28).
Implications:
Among Veterans diagnosed with COVID-19, presenting acuity differed across demographic groups and pandemic waves. Male Veterans, older Veterans, and Black or African American Veterans had higher levels of severity at presentation. Presenting acuity differed across COVID-19 waves with Black or African American Veterans having higher levels of severity that decreased with each wave. Differences in severity at presentation among demographic groups are likely due to differences in COVID-19 case ascertainment bias.
Impacts:
Barring some unknown cause of differences in disease manifestation due to race, this study suggests that there was initial relative under ascertainment of cases in Black or African American Veterans that decreased over time and thus studies that have identified a high burden of COVID-19 in this population have likely still underestimated its impact. Also, allocation of scarce resources without accounting for this under ascertainment of case estimates may exacerbate inequities.