Read on Twitter
Unsurprisingly, I have *thoughts* about these recommendations, as someone who worked at the CMS Innovation Center from 2016-2019. These recommendations are a good start, but missing at least a few important points. https://twitter.com/Health_Affairs/status/1339995134319087617
The recommendations around coordination of model types, and the building of a shared framework are good. CMMI has struggled with the underlying data and technology framework for models for years. Until very recently, every model built a custom data system, for $$$$$$.
Back to technology, data, and interoperability: CMMI should be the absolute forefront of data standards and operationalizing interoperability. CMS interoperability rules should be tested there first. APMs can only be successful with fully electronic health information exchange.
CMMI needs to be at the center of HHS and CMS health reform efforts. The decision to physically place the center on Lord Baltimore Drive, isolated from HHS decision makers, was a bad one, in my opinion. CMMI should be closely linked to the CMS OA and HHS OS.
Contracting needs to be completely overhauled. Federal contracts are always too slow, and too expensive. The same entities are present across every model, but procurement is by-model, and everything is constantly reinvented, and redeveloped based on uninformed preferences.
Challenge Grants--real ones--are barely used (COI alert: I led one, the CMS AI Health Outcomes Challenge).
There are still some great people at CMMI, and in particular, some with great tech and data skills, but attrition has been very significant over the past couple of years. The USDS needs to have a huge presence at CMMI - they should be embedded.
