What we learned—that providers have a disproportionate impact on how payers are perceived, that seven CMS metrics are key indicators of quality, and that patients don’t necessarily distinguish between health plan and healthcare provider—contain important lessons for MA plans and offer commercial health plans a valuable glimpse into their own transparent, quality-focused future.
Two Maps of the STARS
To test provider impact on Star Rating scores, we compared average overall scores with average physician scores, all on the county level. The similarities between the two maps show graphically the high correlation between the two. The conclusion: Provider scores predict about 90 percent of the variation in Star Rating scores. Click below to view the maps:
1What was the most important thing you learned?
Doctors have much more influence on payer quality scores than you’d expect. They’re supposed to account for about three-quarters of the overall score. In fact they control about 90 percent. When you add in the fact that provider quality scores tend to fall into strong geographical patterns—lots of four-star providers in Minnesota and Massachusetts, almost none in New York, for example—it’s going to be hard for payers to affect ratings on their own.
2Why is that?
It seems that patients don’t really draw a sharp distinction between the insurance company that pays for their care and the doctor who provides it. There’s a very strong correlation between the average provider rating in a given county and the average health plan score. And that means it is extremely important for health plans to include quality as a major factor when they decide which markets they want to prioritize or how they construct their networks.
3Does this research have applications beyond Medicare?
Absolutely. There’s no doubt that U.S. healthcare is moving toward greater transparency and greater focus on value. It’s just a matter of time before all health plans will be scored on quality. The details of how the evaluation is performed will probably be different.