Unpacking Unwinding Data: What’s with All the Different Numbers?

As of June 1st, most states have started terminating Medicaid coverage for children, low-income families, and others for non-eligibility, procedural reasons.  We are trying to get our hands on as much data as we can to get a sense of the outcomes of Medicaid renewals and what the unwinding process looks like across states. Transparency and access to unwinding data, as we discussed last month, are critical to monitoring how the unwinding is going and provide important insights when things are not going well to advocate for process changes at the state level.

We are encouraged to have data from around 30 states, with most being shared publicly by the state itself, on the outcomes of their Medicaid redeterminations so far. While we applaud state efforts in transparency, the lack of standardized data reporting and, in some cases, multiple iterations of data for a single month have left us scratching our heads.

States are required to submit monthly Unwinding Data Reports to CMS by the 8th of every month showing the outcome of Medicaid renewals due in the previous month. In addition to these reports, some states have been sharing their unwinding data in a variety of different formats, including press releases and data dashboards. The variety and lack of standardization of these many data points have made it hard to compare how states are doing.

Even more puzzling, though, has been states publishing data in multiple formats with numbers that don’t match up. We have seen several instances of this so far, as we explain below. It doesn’t always mean that the numbers are wrong (although this is certainly a possibility). So what could cause these data discrepancies?

  1. States may be reporting on different populations. In some states, the data being shared outside of the required CMS reports is focused on their “Covid protected” or “maintained” population – i.e., those individuals states have identified as potentially ineligible or having retained Medicaid coverage without a renewal in the past twelve months due to the continuous coverage provision. For example, Pennsylvania’s unwinding data dashboard focuses on their “maintained” population of about 1.3 million, rather than their entire Medicaid population (of about 3.7 million). Their CMS Unwinding Report, on the other hand, reflects renewals conducted for any enrollee in their Medicaid program within a month, not just that the “maintained” group.
  2. Data reports may include metrics that do not give a full picture of the renewal outcomes. The only standardized data reporting for states is the required CMS reports, which CMS reviews and ensures states are completing correctly and plans to release publicly in the summer. Other data that states choose to share are at their discretion, which can lead to inconsistencies in either the metrics reported or numbers themselves. For example, Arkansas’s May CMS report  indicates 47,725 individuals were terminated. However, the state’s May press release on the unwinding, which goes into further detail on their terminations, indicates 68,838 total closures. Currently, CMS requires states to report on renewals in cohorts, meaning that the state only reports on the renewals due for the month of the report. (Only renewals due in May are reported on in the May data to CMS). The state is taking a different approach in its press release and is instead reporting data on all processed renewals for May, not just those due in that month.

    The press release also only lists the reasons for these closures for 59,446 individuals, compared to the approximately 68,838 terminations for the month. So, we can only assume the state is not reporting some closure reasons beyond the five detailed in the state’s release.

  3. Many states have high numbers of pending renewals. In the report to CMS, states are required to report the number of incomplete, or pending, renewals that were due for the month. Just a few months into the unwinding and we’ve already seen troubling numbers of pending cases. High numbers of pending renewals may indicate that state Medicaid agencies are already experiencing backlogs. Due to the nature of the CMS reports – only reflecting outcomes for those renewals for the month being reported on – we aren’t able to see the outcomes of these pending cases in future CMS reports, unless states share their own updated data to reflect renewal outcomes as these pending cases are processed. Arizona, for example, updated its April renewal outcomes data when updating its data dashboard in May showing that the majority of the state’s pending April cases had been processed, and the outcomes numbers were been adjusted accordingly. As a result, Arizona’s dashboard data for April no longer matches the data in its April CMS report.
  4. States are using different definitions to describe renewal outcomes. Kansas has described procedural terminations as “Individuals in a 90-Day Window,” which reflects that individuals whose coverage is terminated for administrative reasons have 90 days to return their renewal, and if found eligible, have their coverage reinstated and possibly backdated to the day coverage was terminated. Virginia’s data dashboard offers a “redetermination completed” number; although this may sound like the total number of renewals that have been conducted in the state, the state’s definition indicates that “redetermination completed” actually refers to those individuals who have completed their renewal and been determined eligible. The differing definitions for reporting metrics makes comparing data difficult for states where the CMS monthly report is not available.

Accurate, timely data is key to the ability of leaders and advocates to identify problems early and take or advocate for corrective action as needed. It has already proven to be essential in monitoring the unwinding so far. However, the ability to use the data in a meaningful way is limited when there isn’t a clear and complete picture of what is going on within individual states or across the country.

[Editor’s Note: This is the 37th blog in the Unwinding Wednesday series. For more information, visit our PHE Unwinding resource page where you’ll find other blogs in this series, reports, webinars and the 50-state tracker.]