It’s really good to know we can count on the monthly reports from CMS to keep tabs on Medicaid and CHIP application and enrollment activity. Since CMS released the first data in January, we’ve seen improvements in the number of states reporting, and enrollment data was added to the initial report of application volume and number of eligibility determinations. So hats off to CMS, but the current reporting only scratches the surface of the required performance indicators and leaves a data junkie like me craving for more.

For years, Medicaid and CHIP stakeholders have longed for routine reporting of data needed to assess the overall performance of our public coverage programs. Some states have consistently obliged, but others, particularly those with 30+ year-old mainframe/legacy computer systems found that it was nearly impossible to extract and report key data. CMS has helped remedy this problem with a substantial federal investment in new Medicaid/CHIP eligibility systems. The generous 90% federal match for system development and implementation costs is tied to a number of conditions, including data reporting on key performance metrics.

In mid-September, CMS released guidance on what it called phase I of the performance indicators. The guidance offered few details on the data reporting requirements but it was promising for many reasons as I wrote about in this blog.  However, the specifics were made available on a state-officials only website (the CALT) that cannot be accessed by outside stakeholders. But at some point CMS quietly posted additional resources on Medicaid.gov, which I was delighted to find recently. These resources include not only a detailed list of indicators but also a CMS presentation for states, a data dictionary, and a FAQ.

Looking over the list of performance indicators there is much to anticipate in future reports, including call center statistics, renewals, processing times, and electronic account transfers between Medicaid/CHIP and the marketplace. And we are anxiously awaiting enrollment to be reported separately for children and adults.

But still, I’m eager to talk about what’s needed for phase II. Not only do states need to gear up, but also we hope that CMS will open up the conversation and take comments and feedback from the health policy community before finalizing the next phase of indicators. What additional data would be helpful in maximizing enrollment and retention in Medicaid and CHIP?

• Breakdowns of enrollment, determinations and disenrollment by coverage group. Beyond a breakdown between adults and children in MAGI- and non-MAGI-based Medicaid, state are not required to enrollment, eligibility determinations or renewals broken down by coverage group (i.e. pregnant women or children). Reaching different groups may require different outreach strategies and represent different barriers to coverage so these details are important.
• Retention rates. States are required to report the number of individuals whose coverage is up for renewal, as well as the number of people determined eligible or ineligible at renewal in a given month. But it’s not a simple matter of doing math to calculate the retention rate since determinations may not be occurring in the month of renewal. Surely, our new systems can actually track renewals by disposition and reason for non-renewal. For example, I’d like to see reporting in April for the number of renewals due in March, and of those how many were determined eligible for ongoing coverage and how many were disenrolled (and for what reasons – see next bullet). The key here is to connect the determination to the triggering event, otherwise we can’t get an accurate retention rate.
• Denial and disenrollment reasons. The performance indicators broadly grouped “ineligibility determinations” and require a breakdown between “ineligibility established” and “eligibility cannot be established.” This sounds like caseworker talk. Why can’t we use terms that are more relatable, like reasons for denials of eligibility at new application and disenrollment reasons at renewal? It is critical to distinguish between known reasons for ineligibility (i.e. over income) and when eligibility cannot be established (i.e. did not provide required verification). This level of detail is essential in identifying opportunities to improve how our public coverage programs work. But based on the phase I performance indicators, it appears we will only be able to get topline reporting on a combination of denial and disenrollments based on ineligibility or other reasons. Reporting these data separately for applications, renewals, or when changes are reported at other times is really important.

I’ve written about this previously, but it’s worth pointing out again the great work that Mathematica has done as part of the Maximizing Enrollment project to identify a reasonable number of reason codes for denials and disenrollment. Standardizing these codes would be immensely helpful in identifying where to look for best practices to improve enrollment and retention.

Saying that ACA implementation has been more than challenging for states and stakeholders alike is an understatement. But with systems development work ongoing and the need to plan ahead, it would be helpful to know where we are heading with reporting requirements. Ideally, CMS will establish a robust set of indicators that set minimum standards across states. Short of that, or even ahead of that, health policy advocates will want to encourage their own states to embrace a policy of transparency and routine reporting on a broad set of measures that enable us to assess the performance of our public programs and continually work to make them more efficient and effective.