A significant and growing performance gap exists between dual eligible and non-dual eligible members that cannot be attributed to a health plan's quality of service, a new study by Inovalon has found.
A significant and growing performance gap exists between dual eligible and non-dual eligible members that cannot be attributed to a health plan's quality of service, a new study has found.
"An Investigation of Medicare Advantage Dual Eligible Member-Level Performance on the Centers for Medicare and Medicaid Services (CMS) Five-Star Quality Measures" from Inovalon, provider of data-driven healthcare solutions, is the largest-ever analysis of dual eligible member data. Results show that the presence of low education, low income, and other social and demographic factors associated with poverty not only affects a person’s health and health-related outcomes, it also affects how a person scores on measures of health plan performance, independent of plan characteristics. Dual eligible beneficiaries are indidvuals who qualify for coverage from both Medicare and Medicaid.
The study's objective was to identify and quantify the clinical, sociodemographic and community risk factors most associated with the worst-observed outcomes for a majority of Medicare measures used to assess the performance of Medicare Advantage (MA) health plans in delivering quality of care.
Teigland“This is an important, ground-breaking analysis of data not previously achievable in the industry,” said Christie Teigland, Ph.D., director of statistical research at Inovalon and principal investigator of the study. "Previous studies have shown that plans specializing in care of low-income Medicare beneficiaries generally scored lower on the Centers for Medicare and Medicaid Five-Star Quality Rating System than plans serving a healthier, more-educated population with higher income. However, studies to date have not determined whether these findings are a result of poor plan performance or enrollee characteristics such as sociodemographic status-related factors, or a combination of both.”
Results showed that a significant association exists between dual eligible status and lower performance on specific Part C and D measure Star ratings. They validate the integral role that income, race/ethnicity, and gender play on the HEDIS and CMS Part D measures used in the Five-Star rating system, according to Inovalon.
As evidenced by this analysis, the gap has widened in reported Star ratings for 2012 and 2013 compared to previous findings. When scored by either the Charlson Comorbidity Index or CMS MA risk score, dual eligible members were found to be consistently more complex to manage.
Additionally, examination of 80 CMS MA contracts indicated that dual eligible members performed worse on nine of the 10 Star measures that were investigated.
NEXT: Results suggest need for refinement of Star quality measures
Further, the study found that multivariate analyses controlling for demographics, socioeconomic characteristics, and severity of illness confirm dual members consistently underperform in eight of the 10 measures investigated. According to Inovalon, this is an important finding demonstrating a significant performance gap exists between dual eligible and non-dual eligible members even after adjusting for other important socioeconomic and clinical risk factors. It suggests that the Five-Star rating system, in its current state, may penalize MA plans serving a high proportion of dual eligible beneficiaries. Lower Star ratings result in lower incentive payments and may lead to reduced services to dual eligibles, suggesting a need for further research into the benchmarking and refinement of Star quality measures to assure fair comparisons of performance across MA plans serving different populations, says Inovalon.
The findings demonstrated that dual eligible members enrolled in MA plans have significantly lower-quality scores, but the less positive outcomes for those members does not appear to be due to the quality of care provided by a health plan, according to Teigland. "Instead, the increased prevalence of poor outcomes is due to a variety of clinical, sociodemographic and community resource factors, such as living in a high poverty area or an area with a shortage of primary care physicians,” she says.
NEXT: Research can be used to help guide future measure development efforts
The study used member-level MA data extracted from Inovalon's Medical Outcomes Research for Effectiveness and Economics Registry (MORE² Registry). The MORE² Registry provides visibility into the medical utilization of more than 98 million unique and de-identified individuals nationwide covering more than 3.1 billion member-months of data from 2002 through September of 2013.
Within this study, from the 11.8 million MA enrollees present within the MORE² Registry, Inovalon identified 1,335,709 enrollees in 2011 (16.6% dual eligible) and 1,605,644 enrollees in 2012 (16.2% dual eligible) from 80 individual CMS contracts who met the study inclusion criteria. Rates for nine Star measures were calculated independently for the dual and non-dual eligible members and then within each of those groups, stratifying by various demographic, clinical, and socioeconomic characteristics. In addition, a tenth measure-plan - all-cause readmission rate (PCR) - was calculated using the National Committee for Quality Assurance (NCQA) risk adjustment model for MA members age 65 and older, which controls for chronic conditions and factors impacting likelihood of readmission.
Bringewatt“There is long-standing research that shows a person’s income, level of education, and other social demographic factors affect a person’s overall health and use of healthcare services when they are sick,” says Rich Bringewatt, president, National Health Policy Group, Special Needs Plan (SNP) Alliance. "As a result, individuals who are in low socioeconomic circumstances have factors that negatively affect their care and their outcomes compared to individuals in higher socioeconomic circumstances.” The SNP Alliance is a national group of specialty health care plans and programs exclusively focused on improving SNP and Medicare-Medicaid Plan policy and performance.
“This study provides insight for plans, providers and other managed care executives about what some of these factors are and opens up options for how to more efficiently and effectively improve health outcomes for these disadvantaged persons," says Bringewatt. "It is critical that public reporting of plan performance take these factors into account. It also imperative that agencies that provide financial incentives or penalties associated with a plan’s performance rating account for these factors as well. Otherwise, plans that specialize in the care of people who are poor may be unintentionally penalized by quality measurement programs.”
This research can be used to help guide future measure development efforts to improve the validity, reliability and usefulness of performance measures, according to Teigland, and “also help guide the development of methods that help plans address the adverse effects of low income, low education, and other social and community factors associated with poverty.”