Introduction
The Medicaid program constitutes around 28% of state budgets, providing coverage for approximately 77 million people and contributing to 16% of total US healthcare spending. About 70% of Medicaid enrollees are currently in managed care plans. A significant portion of these enrollees, instead of actively choosing a plan, are auto-assigned by the state through a process called “auto-assignment.” Approximately 25 million Medicaid enrollees will be auto assigned to managed care plans by states. This situation presents an opportunity to devise innovative auto-assignment methods leveraging differences in plan quality and networks, aiming to reduce Medicaid spending without negatively impacting enrollee health or satisfaction.
The goal of the project was to identify data gaps that can improve Quality measures that can improve auto-assignment.
Background
A Managed Care Plan had a Quality team and also was involved in a large incentive program for their provider network, but they have seen the membership plateau and also reduce because of the end of public health emergency. Their goal was to maximize the Quality measure score since they suspected that they were providing a solid care management to their members.
Challenge
Identify data gaps in Quality measure calculations and quantify the loss of quality measures scores to enable the health plan to work with their IT teams and vendors to improve the data availability for now and for future years.
At a high level that data required for Quality measures is Membership/Enrollment, claims (medical, Pharmacy, BH etc.) , encounters (claims from capitated entities), Lab results, supplemental data, and Immunization Registries. But the challenge is that there are lot of gaps in the data and that reduced the Quality scores and that directly affects auto assignment.
Solution
The goal was to address the data challenges with Quality measures where the health plan was not doing very well. These were identified to be the chronic condition quality measures like Diabetes HBA1C, Blood pressure control, Preventive measures like Immunizations
CBP – Controlling High Blood Pressure
CCS – Cervical Cancer Screening
CIS – Childhood Immunization Status
HBD – HBA1c Control for Patients with Diabetes
IMA – Immunizations for Adolescents
PPC – Prenatal and Postpartum Care
LSC – Lead Screening in Children
CHL – Chlamydia Screening in Women
Implementation
Review existing HEDIS data sources and assess the completeness of data.
Every piece of data available for Health Plan was analyzed for completeness of data. The analysis included data produced within the health plan like enrollment, eligibility, claims (medical), Lab results and external data that included encounters from capitated entities, supplemental data from providers.
This data was analyzed in the context of Quality measure denominators and numerators from the HEDIS vendor.
Methodology
Results of the Analysis
- This exercise showed some data gaps in the data for Lab data where tests were done, but the lab results were not sent to the health plan.
- The claims sent to HEDIS vendor included only 5 years of history, where some measures needed 7 years of history.
- Claims data also contained mother’s ID in place of new born’s member id
Conclusion:
By identifying and quantifying data gaps across enrollment, claims, lab results, encounters, and supplemental data sources, the health plan gained clear visibility into the factors impacting Quality measure performance. This enabled targeted collaboration with IT teams and vendors to improve data completeness both in the short term and future measurement years. Strengthening Quality measure performance directly supports improved Medicaid auto-assignment outcomes, positioning the health plan for sustained membership stability and growth.
