The effect on health outcomes of COVID-19 pandemic-related disruptions in care is a defining public health question of this generation. In the early phase of the COVID-19 pandemic, admissions to Department of Veterans Affairs Hospitals decreased by 41.9% and outpatient visits, either in person or remote, decreased by 30.4%. While it is possible that the 40.3% fall in admissions for events such as myocardial infarction occurred partially because people were less active, it is likely that many delayed or did not seek care at all. Interestingly, admissions for appendicitis, a disease whose incidence would not likely vary as a result of behavior change, also decreased by 56.7%. Some of this forgone care may have saved patients from unneeded interventions. Careful analyses of these data could provide essential guidance about prioritization of care and social support for this pandemic and for future natural disasters.
The goal of this initial effort is to create a research roadmap to characterize conditions associated with excess mortality during the COVID-19 pandemic in Veterans. The VA experience will be placed in the larger US context by conducting parallel ecologic analyses of VA, Centers for Disease Control and Prevention (CDC), and Centers for Medicare and Medicaid (CMS) data. Changes in healthcare utilization that may be associated with both lowered and excess mortality in VA and Medicare populations will be examined. Available patient-level electronic health record (EHR) data within the VA will then be used to study in depth how much of the excess mortality is likely attributable to COVID-19. We will do this by using a risk index to identify probable undiagnosed cases and examine how these cases are distributed across specific demographic and diagnostic subgroups. Investigators will use the insights gained in this project to determine where further research and public health intervention efforts are most needed.
The main goal of this study is hypothesis generation. We will identify areas of greatest findings and then pursue more in-depth analyses. We will begin by examining broad trends in all-cause mortality overall, within demographic and diagnostic subgroups, and by calendar month, and identify associations between mortality trends and changes in healthcare utilization.
US Mortality data. Data from the 5 years prior to the pandemic, as well as during 2020-21 for US deaths due to all causes and due to specific categories (defined by ICD10 codes) will be obtained from the National Center for Health Statistics mortality surveillance system. Data will be stratified by state, week, and age category.
Medicare Data. Medicare data will be obtained from VIReC (The VA Information Resource Center). These "denominator" files will be used for overall mortality assessment, and the inpatient and outpatient files will provide comparative data for characterizing hospitalizations and procedure-based care (Medicare Part A and B), prescription drug utilization (Part D), and skilled nursing facility use and durable medical equipment files. Medicare claims provide widely generalizable, yet highly granular insights into healthcare utilization across the United States as a whole in the 65 and older population and will provide valuable context and benchmarks for regional and national comparisons to the healthcare of Veterans.
VA Electronic Health Record Data. The Corporate Data Warehouse will provide access to all VA electronic medical record data, including hospital and outpatient diagnoses (recorded using ICD-9 and ICD-10 codes), pharmacy data, laboratory results, clinical notes, and vital signs. Veteran deaths will be ascertained from the VA Vital Status file and the Master Patient Index, which is updated daily. Once available, these VA data will be merged as indicated with Medicare and Community Care data at a patient level to consider outside healthcare utilization.
VA Community Care Data. Over half of VA enrollees are eligible for community care through the MISSION (Maintaining Internal Systems and Strengthening Integrated Outside Networks) Act. In the first year after the start of the pandemic, community care encounters accounted for nearly 40% of all VA provided and paid for care. In comparison, from 2014-18, VA funds obligated for Community Care comprised about 20% of its healthcare budget. The Non-VA Care Program Integrity Tools (PIT) system, a comprehensive set of tools that aggregates many sources of data to check for fraud, waste, and abuse in the VA Community Care program, is the primary data source for VA Community Care. Whenever possible, for all study aims, PIT data will be analyzed in combination with Purchased Care/Fee data for a complete understanding of VA Community Care utilization and costs. During the course of the project outside VA healthcare services including dates, procedures, location, and diagnoses will be utilized.
There are no findings to report at this time.
The work in this area has direct application to everyday care. This experiment is an opportunity to reveal the effects, both positive and negative, of the variations in care that occurred during the early and later stages on the COVID-19 pandemic. The findings have the potential to support broader efforts in US healthcare to further elucidate unmet needs while finding opportunities to safely de-escalate care when appropriate, freeing people from unneeded connections to the healthcare system.
None at this time.
None at this time.
Care Coordination, Organizational Structure
None at this time.