What Do You Mean, “Better Health Outcomes”?

It suddenly occurred to me, at some point during my ongoing rambling about better health outcomes, that I didn’t really know what constitutes “outcomes” in health in the first place. Fearful of falling into that trap of missing the trees from seeing too much of the forest (which Swatties seem especially susceptible to), I decided to a little digging.

It appears that many health care providers aren’t really sure themselves. “In practice, quality usually means adherence to evidence-based guidelines, and quality measurement focuses overwhelmingly on care processes,” writes Michael Porter, a professor at Harvard Business School and god of business and economics (at least in terms of number of citations). It seems that he’s recently shifted from a life time of working on competition and company strategy to cracking the health reform conundrum plaguing our nation, and his ideas on improving value of care are outlined in this New England Journal of Medicine article.

Here is his framework for the determinants of health outcomes, distilled into a pretty diagram:

Initial conditions refer to patient characteristics such as severity of condition or demographic factors that may influence the outcome of care (which risk-adjustment programs are trying to account for). Providers treat patients of various initial conditions with appropriate processes of care, such as angioplasty for clogged arteries or antibiotics for an infection. But providers are operating within a definitive structure, which may include the organizational unit within the hospital or the mechanism and amount of care integration. These treatments, together with the degree to which patients comply with treatments, determine a number of health indicators (such as HbA1c levels for diabetes, or blood pressure for hypertension) and ultimately health outcomes.

But how exactly should we define “outcomes”? In a supplemental paper, Porter outlines a three-tiered definition of health outcomes:

1. Health status achieved. This includes not only survival rates but also other measures such as restoration of function.
2. Process of recovery. How long does it take to complete treatment, and how much discomfort does the process cause?
3. Sustainability of health. Even after treatment is completed, how long does wellness last before illness recurs, or before treatment-induced complications arise?

Any measurement of outcomes, Porter argues, would be incomplete without considering all of these dimensions. Once we can define outcomes, we can craft incentive mechanisms such as Pay-for-Performance to reward (or punish) providers based on their performance in achieving these outcomes.

However, over on KevinMD.com, Dr. Vipan Nikore raises a warning flag: should providers really be held accountable for outcomes they may not have complete control over? He makes an analogy to the mixed fixed salary plus bonus structures employed by other industries:

“The percentage of [a] bonus is dependent on the opportunity for the employee to influence the outcome,” writes Nikore. “For instance, some organizations such as those in the fast food industry probably do not want to have too high a company bonus for its employees doing straightforward tasks since they have a relatively small influence on the bottom line, however, a senior VP at a small startup probably may have a large bonus tied to the performance of the organization. The bonus must be commensurate to the potential to influence.”

The logical conclusion, it would seem, would be to incentivize each player in health care for what she or he does have control over: providers for following best practices in processes of care, and health care administrators for ensuring the proper structure in which providers can function.