home email us! sindicaci;ón

Causation and Credit Challenges in (Health) Management

by Scott MacStravic

With the growing body of reports on the benefits of population health management (PHM) — of employee, health plan member, or government plan beneficiary populations – there is more confusion than consensus on whether and how well it works.  Of course, any attempt to reach a general conclusion about what is a complex variety of different health problems to be managed, and different ways of managing them, is bound to be hopeless, since it is mixing a wide variety of fruits, not just apples and oranges.

But one of the most common problems in reports by suppliers of PHM involves the same set of “delusions” recently written about relative to management and managers in general. [P. Rosenzweig The Halo Effect…and the Eight Other Delusions that Deceive Managers Free Press 2007]  All nine delusions the author discusses relate to the complex challenge of deciding what causes or caused what, and tendencies to leap to simple, but unjustified and dangerous conclusions as a result.

The list of “delusions” includes:

  •    The Halo Effect, a tendency to attribute improvements to “drivers” that had no causal effect at all, or nowhere near as much as given credit for
  •    The Correlation vs. Causality problem, the tendency to believe that if a thing happened after a particular change was made, it happened because of the change
  •    Simple Explanations for Complex Realities, the tendency to look for and find a single cause when there are many, and the causal dynamics involved are complex
  •    Connecting the Dots, the difficulty in isolating reasons for performance differences, particularly when the same information cannot be obtained for good vs. poor performance
  •    Rigorous Research limits, the fact that the most rigorous research and analytical methods won’t yield “truth” if the data input is lousy
  •    Lasting Success delusion, belief that if a firm (or person) has been successful in the past, the future is assured, when regression to the mean is common
  •    Focus on Absolute Performance, when relative performance makes the most difference, as in competitive situations
  •    Wrong End of the Stick, belief that highly-focused firms (or people) are always successful, though they are only often so
  •    Organizational Physics, belief that laws of nature and science apply to organizations (and people), when they rarely do so

These “delusions”, or unfortunate tendencies apply equally to managing health and measuring the “results” achieved thereby.   The most common approach used in evaluating HM results, for example, is before/after comparisons.  While it is certainly possible that an HM intervention in the “after” case is responsible for differences from the “before” case, there may be many other causes as well, unless comparisons are made in a controlled fashion where the only difference between after and before cases was the HM intervention.  This is rarely even possible, the way most such evaluations are made in practice.

Next most common is the side-by-side comparison, usually of those who participated in an HM initiative compared to those who did not. The “delusion” or at least danger here is that there may well have been many differences, in motivation, changeability, etc. between people who chose to participate compared to those who did not.  These individual differences may have more to do with side-by-side differences than the intervention some participated in.

The “Halo Effect” author’s major suggestion was that a healthy skepticism about causes and effects is essential to good management, and evaluation.  Good performance is often achieved through the combination of a shrewd strategy, superb execution of it, and good luck, rather than a simple principle or technique.  Evaluations should be used as much to learn how to do it better next time, including how to manage interventions so as to make accurate and appropriate attribution at least more likely and more credible the next time.

No comments yet »

Your comment

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>