The recent recalculation by the Centers for Medicare and Medicaid Services (CMS) of 2024 Star measures for Medicare Advantage (MA) shows just how important it is for MA plans to closely scrutinize CMS’s calculations and push back when they see issues. CMS offers that chance during the Star Plan Preview period. Unfortunately, many plans do not invest enough into planning for that process throughout the year. More to the point, many plans are constrained in terms of accessing the right data, having the right analysts, and coordinating throughout the organization.
Plan Preview’s short window and the challenges outlined put immense strain on even the most sophisticated Stars operations. In that vein, Lilac Software has created this best practice approach to planning for and engaging with CMS during plan preview to ensure the most accurate and robust Star scores possible.
Background
Plan Preview is divided into two periods – Plan Preview 1 (early to mid August) and Plan Preview 2 (early to mid September)
In Plan Preview 1 (for 2024 Star, it was August 9 to 16, 2023), CMS provides data and reports that show its estimates for the percentage attainment/other values for each measure. Plans should then compare the calculations for each measure to the internal data it has – submitted via HEDIS filings, via Part C and D reporting, external CMS vendor data, or internal analysis of certain measures. Plans have the opportunity in this preview round to alert CMS to discrepancies/errors/anomalies in data and measure calculations between CMS’ calculations and their calculations. Plans report back to CMS on any errors or discrepancies.
In Plan Preview 2 (for 2024 Star, it was September 8 to 15, 2023), CMS provides data and the actual Star calculations for the coming year, including final percentage attainment on measures/measure values, the cut points, each measure’s Star score, and the aggregate rollups (Part C, Part D, and overall score). In this round, plans need to continue to examine percentage attainment on scores/measure values, examine cut points, examine the measure score, and assess the rollups for Parts C, D and the overall score. This round offers plans a final window to make CMS aware of any remaining data issues as well as issues related to the calculations of measures and overall scores.
Overall Best Practice On Data Anomalies/Errors/Corrections
CMS is especially strict on Star measure corrections when data anomalies, errors, or corrections are not reported to CMS in a timely fashion. Plans need to practice data oversight strategies continuously throughout the year to demonstrate a track record of reporting data issues as soon as they are known. Early notification of data discrepancies greatly improves the chances that CMS will accept the challenge. As an example, plans should not expect Star recalculations if an issue is reported for the first time in Plan Preview 2. Plans must report data errors throughout the year and be ready to document data discrepancies/errors/anomalies in Plan Preview 1. Lilac recommends that plans adopt the following due diligence strategies.
- Establish an ongoing dialogue with CMS and key CMS vendors based on ongoing data review and raising issues with data while the data is fresh. It is important that plans are scrutinizing data and the calculations of underlying data throughout the year to quickly identify any inaccuracies. Documentation of conversations and reports are very important. The following areas are examples of where this approach should be applied:
- Part D PDE processing and medication adherence calculations
- Disenrollment data
- IRE timeliness and overturns from IRE published site. Compare to internal data and Part C report (2023 year sent in February 2024)
- MTM/CMR calculations
- Call Center Performance
- Medicare Plan Finder accuracy
- Plan Preview 1 should be focused on data anomalies/errors/corrections, with ties to previous reports of data issues throughout the year if not corrected or still suspect.
- Overall challenges during both Plan Previews should be focused on questionable accuracy and validity based on arguable data elements. Challenging self-reported data or measure calculation is very difficult during the overall process unless an issue was reported immediately to CMS.
Lilac’s Support of Plan Preview
In order to facilitate a plan’s reviews during plan preview, Lilac Software consumes data for in order to calculate measure scores as well as the Part C, Part D, and overall ratings.
For Plan Preview 1, Lilac uses the data submitted by the plan to detail measure attainment against what is reported by CMS. From here, Lilac and the plan undertake any analysis needed to investigate discrepancies and report any errors/discrepancies/anomalies to CMS. In addition, the CAI component is reported in Plan Preview 1. Lilac examines the CMS reported CAI calculation against plan data/previous CAI plan reports/global published CMS reports.
For Plan Preview 2, Lilac uses the data submitted by the plan to calculate measure scores as well as aggregate summary and overall scores (including applying all adjustments such as disaster, CAI, Reward Factor, etc.). This includes its projections of cut points. For 2025 Star, Lilac is doing the following:
- Analyzing the 2022, 2023, and 2024 published “Tukey applied” simulation data
- Analyzing the reported 2022, 2023, and recalculated 2024 cut points
- Simulating a trend for 2025 cut points with the Tukey outlier formula included and applying 5% guardrails for 0-100 measures/other guardrails as applicable for not 0-100 measures to arrive at expected 2025 Star cut points.
Post release of Plan Preview 2, Lilac compares all measure and aggregate CMS scores against Lilac anticipated. We also compare calculated cut points against the guardrails prescribed in regulation. From here, Lilac undertakes any additional analysis needed to investigate discrepancies and have supporting documentation to submit to CMS.
