Low-Quality Data Hinders VA’s Efforts to Reduce Care Disparities

January 15, 2020 - While the Department of Veterans Affairs (VA) has taken measures to reduce care disparities, incomplete or inaccurate race and ethnicity data hinder the organization’s ability to effectively close health outcome gaps, according to a report from the Government Accountability Office (GAO).

GAO noted that although the VA has funded several studies to identify health disparities and explore interventions to reduce or eliminate them, health disparities persist among the VA’s patient population.

For example, in 2019, the VA found that African American veterans with cancer and cardiovascular-related illnesses had lower survival rates compared with other minority veterans and white veterans.

Additionally, in 2013, Asian, Native Hawaiian/Pacific Islander, American Indian/Alaska Native and other multiracial veterans reported higher rates of negative experiences at VA medical centers than other minority veterans and white veterans.

The VA has taken steps to address these gaps in care, GAO said. The VA’s Health Services Research & Development (HSR&D) has spent about $12 million to fund research studies related to identifying and reducing disparities in healthcare since 2014. However, data issues have made it difficult to fully eradicate care gaps among minority veterans.

READ MORE: Identifying Care Disparities for Population Health Management

“VA officials and researchers have noted problems with the completeness and accuracy of the data on veterans’ race and ethnicity. These weaknesses, in turn, limit VA’s ability to assess and address health care disparities at the VA medical center level,” GAO wrote.

GAO found that VA officials cannot always ensure that race and ethnicity information labeled in the EHR as “self-reported” is accurate. Self-reported information is the preferred method for obtaining an individual’s race and ethnicity, because it’s more accurate than data collected by observation of a third party. However, intake clerks may sometimes enter information based on observation, which may be unreliable.

“VA research has indicated that observational data is more accurate for Blacks or African Americans and Whites than other racial groups, and that studies focused on other racial groups may be especially vulnerable to misclassification bias. As such, VA lacks reasonable assurance that the identification of race and ethnicity as ‘self-reported’ is accurate,” GAO said.

Data on veterans’ race is also often incomplete, GAO found. The information is collected as two separate categories in the EHR, and ethnicity information – based on veterans’ designation of whether they are Hispanic or non-Hispanic – is often more complete than race data.

VA officials said that one possible reason for this is that the ethnicity question is asked first, and veterans may self-report their ethnicity and then leave the race field empty because they believe they have already provided this information.

Conflicting information can also limit the VA’s ability to address racial and ethnic disparities, GAO noted.

“According to VA researchers we spoke with, because a patient’s race and ethnicity information is uploaded from his or her EHR after each inpatient admission and outpatient appointment, there can be multiple records for each patient’s race and ethnicity data in VA databases,” GAO said.

“These patient records often conflict with one another, and may result from the use of both observational and self-reported data. As such, officials stated that it can be difficult to determine which of the multiple race or ethnicity records are accurate.”

VA researchers have used various approaches to overcome these issues, including using multiple non-VA data sources to supplement VA’s race and ethnicity information, such as Medicare data and data from the Department of Defense’s roster of veterans.

As the number of racial and ethnic minority veterans receiving VA healthcare services continues to grow, it will be essential for the organization to have access to reliable data to reduce disparities.

“Without including performance measures or lines of accountability, VA lacks the means to ensure any action plan will be fully implemented to achieve its goals,” the report concluded.

“Until VA resolves known weaknesses with the completeness and accuracy of its race and ethnicity data, it will be limited in its ability to assess health equity for veterans receiving care at its facilities.”

GAO recommended that the VA evaluate the validity and reliability of data to further reduce disparities.

“To ensure the availability of information on healthcare outcomes by race and ethnicity throughout the VA healthcare system, the Secretary of Veterans Affairs should conduct an assessment to determine the completeness and accuracy of race and ethnicity data captured in VA’s electronic health record, and implement corrective actions as necessary to resolve any identified deficiencies,” GAO concluded.