Tuesday, July 21, 2009

Community Care of North Carolina, Association vs. Causality and How Insurers vs. Physicians Think

Shiites vs. Sunnis. Lycans vs. vampires. Liberal musicians vs. Republicans. Insurers and health providers. The Disease Management Care Blog wonders about the basis of all the antipathy. Is it how they were raised? Is it the DNA? Perhaps exposure to an environmental toxin in utero? While the DMCB continues to ponder this, it was struck again by how differently insurers and physicians interpret Community Care of North Carolina and what it means for the adoption of the Patient Centered Medical Home.

As readers may recall, CCNC has been credited by physicians with using the Medical Home Model to save the State hundreds of millions of dollars in claims expense while simultaneously paying primary care physicians additional PMPM fees. Even though this was all Medicaid all the time, advocates of the patient centered medical home (PCMH) point to apparent success of CCNC as evidence that this care approach should be implemented by all insurers all the time and the sooner the better. In contrast, commercial insurers have taken a more cautious approach, opting instead to test the medical home with 'pilot' programs. No doubt PCMH devotees are also disappointed by the U.S. House of Representatives’ Tri-Committee proposal to also limit the funding of the PCMH to a pilot.

Which is why the DMCB recommends an insightful article appearing in this month’s Medical Home News (subscription required & a disclosure: the DMCB is a non-compensated member of the Editorial Advisory Board). While a blog summary here won’t do it justice, the very experienced Michael Cousins, PhD walks the no-man’s land between insurers and providers to distill the opposing points of view.

Dr. Cousins points out that the actuaries who authored the CCNC reports used a classic insurance-based paradigm. These were ‘observational’ descriptions of what happened, not explanations about the underlying causes. Actuarial reports can get closer an understanding of the causes if the analysis mathematically neutralizes or adjusts for variables such as patient/physician self selection, cost/service inflation, reimbursement changes or other flaws in the study design. Dr. Cousins read the CCNC reports closely and found they didn't do that and, what's more, that they steered clear of addressing the issue of causality by carefully relying on words such as ‘associated’ and ‘correlated.’ The reports never used the 'C' word and without it, the utility of the CNCC reports, according to Dr. Cousins, is ‘limited.’

The DMCB thinks this is emblematic of one gulf between insurers and docs. The latter think association, while the latter worry about cause. It's 'what' vs. 'why' and 'how.'

‘’What happened?’’ Insurers are in the business of risk transfer. The DMCB wonders if trading risk for money doesn’t necessary require a deep understanding of causality: the observation that characteristic “A” (drivers less than age 18) is associated with “B” (more car accidents) is what is important. The State of North Carolina got an actuarial report that said “A” (paying providers a PMPM fee and pairing it with nurse care management) was associated with “B” (a drop in claims expense). The actuarial 'what' for this version of the Medical Home was amply demonstrated for the State, which continued the CCNC program. Case closed. (There are other equally simple answers when it comes to risk transfer that are more painful, but the DMCB digresses.)

And did it cause savings?’’ In contrast, in order to provide care, physicians need to know the underlying reasons on why and how good and bad things happen to their patients. For example, they’ll ask why adolescent drivers are so vulnerable to risk-taking? The utility of knowing this is what leads providers to constantly search for cures. That is the business of moving from association to causality. And if there is a direct causal link, all stakeholders - insurers, physicians and patients - can be more confident that it will work outside of North Carolina Medicaid.

Unfortunately, understanding causality is a far more difficult exercise and the process is more prone, as has been demonstrated plenty of times, to mistaken assumptions. When circumstances warrant, the best way to address causality is with clinical trials involving a good comparator. Dr. Cousins agrees and points out that concluding there is a causal link between the medical home and lower costs could be a mistake.

The DMCB goes one step further. Since insurers only need ‘what’ in their business decisions, observing a decline in claims expense in their local pilots will be good enough for them. That’s also why private insurers continue to support their local disease management programs: they are also associated with savings. It's an endorsement, but a very shallow one obtained one network at a time.

Doctors, however, continue to look for answers to the questions how and why. They think they’ve found the answers for the Patient Centered Medical Home that will work in all settings.

No wonder the insurers and docs are talking past each other on the topic.

No comments: