Though required to collect data on race, ethnicity and language from patients, many managed care organizations (MCOs) are failing to establish complete demographic data that could help with population health strategies.
Medicare, Medicaid and commercial plans were studied to measure the completeness of data collected to fill Healthcare Effectiveness Data and Information Set (HEDIS) requirements enacted by the Affordable Care Act, with research published in the March 2017 issue of Health Affairs. Researchers looked at data from 2012 to 2015, and found that data on race was the most complete, but other demographic information, including ethnicity, spoken and written language, and other language needs, were incomplete.
“In 2015, one-third of commercial plans, half of Medicaid plans, and nearly three-quarters of Medicare plans reported complete and partially complete data on race. Plans fared worse on ethnicity data reporting, with fewer than half of all plan types reporting complete or partially complete data,” the study’s authors said, adding that patient language data was also very limited from MCOs.
“Fewer than half of commercial, Medicaid, and Medicare plans reported complete or partially complete data on spoken language, and even fewer reported complete or partially complete written language or other language needs data,” the study’s authors said. “Medicare plans reported more complete or partially complete language needs data than did Medicaid plans, while commercial plans reported the least complete or partially complete language needs data.”
The lack of complete demographic data leads to less effective population health strategies, says lead author Judy Ng, research scientist at the National Committee for Quality Assurance, and a visiting research collaborator at the Woodrow Wilson School of Public and International Affairs at Princeton University.
“Without complete demographic data, it is not possible to know whether more vulnerable members are experiencing health gaps or disparities, nor where to target efforts to close such gaps,” Ng says. “Not knowing this information may have important implications for a plan’s healthcare costs.”
Ng says that the study did not assess barriers to data collection, but asserts that patients are more likely to share personal demographic information when they know why it is being collected.
“We know that potential barriers to collection of race, ethnicity or language data include: lack of systemic data collection and use of standardized data categories by health plans, perceptions by plan staff that members might be reluctant to provide such data—a perception that may not reflect reality—and lack of staff training to explain the purpose of collecting the data,” Ng says.
Training staff on how to develop rapport with patients and explain the need to collect personal data is an important step in creating complete demographic data that can then be used to improve care and lower costs, Ng says.
“Directly asking [patients] about race, ethnicity and language is the best approach,” Ng says. “Plans could start by addressing some of the known barriers to data collection, such as providing training to staff on how to ask about this information. There are best practice toolkits for this. They can also use standardized data categories, for example, using HHS standards, to allow for more comparability.”
Researchers note that Medicare plans are required to report HEDIS data, while commercial and Medicaid plans report voluntarily or based on state regulations.
“It is not clear that commercial plans have sufficient incentive to collect this data,” says Ng. “Medicaid reimburses for language service costs for Medicaid patients. Commercial plans are not subject to these same policy requirements or incentives.”