The Journal of General Internal Medicine has just published a short research letter## that I wrote with Sih-Ting Cai, Jean Abraham and Coleman Drake. We were curious about what insurer characteristics were associated with better or worse care quality and care effectiveness ratings for ACA plans.
We used Center for Medicare and Medicaid Services (CMS) Quality Rating System (QRS) data to get four quality of care metrics:
- Are kids tested appropriately for pharyngitis?
- Are kids treated right for upper respiratory infections?
- Are antibiotics given to adults for acute bronchitis?
- Is imaging used appropriately for lower back pain.
We thought that there were a few notable charactersistics of insurers that could influence these ratings in a consistent manner. We used Robert Wood John Foundation Plan Participation Tracker and National Association of Insurance Commissioner data to identify each plan as having the following four binary attributes:
- Member of Blue Cross/Blue Shield Association
- Medicaid Managed Care dominant entity
- Offers a PPO
I had fairly strong priors. I had thought that non-profits were more likely to have good quality ratings. I thought PPO plans would have weak control on their networks. I thought Blues would be notably better than Medicaid Managed Care entities. This was based on the assumption that there is both a premium-quality trade-off and that Medicaid Managed Care dominant entities are making a premium subsidy benchmark cream skimming and risk adjustment arbitrage play.
I was wrong.
Non-profits were a bit better on two metrics. PPOs were a bit worse on two metrics. Everyone was bad on back pain. None of this was shocking.
However, Blues and Medicaid were about the same.
That surprised me. I love doing research and having to update my priors.
We have more research in this general area. Some will be coming out soon(ish) and other chunks are still under review but this has been a fascinating line of poking at the data and changing my mind as to what the markets are doing.
## J Gen Intern Med DOI: 10.1007/s11606-020-06248-5