Because episodes have, by definition, a finite duration, it is now relatively easy to mine data sets to identify which providers can solve specific medical problems more effectively and at a lower cost than other providers. Whether the category being considered is surgeons who perform hip replacements, oncologists who treat prostate cancer, or hospitals that take care of stroke patients, the goal of this approach is to reward providers for delivering a patient’s desired outcome (say, walking, remission, or hospital discharge without readmission) with as few complications as possible and for providing the best experience at the lowest possible cost. Focusing on episodes of care makes it easier to compare performance across providers and encourages productive competition among them. We believe that more than $2 trillion of US annual spending on health care, which currently totals about $2.7 trillion, could be paid through an episode-focused approach.
McKinsey folks think very highly of themselves. I have always resented that.
In this case, Latkovic appears to be comparing the existing compensation system with a system that would not be gamed. That is unrealistic. Instead, ask yourself how doctors would game an outcomes-based compensation system.
1. Over-diagnose. That is, report that the patient has a condition that is worse than it is.
2. Under-treat. Suppose that 10 percent of the variation in outcomes is due to doctor actions, and 90 percent is due to other random variables. Then the profit-maximizing strategy is to pocket the payment for the “episode,” do nothing, and hope for the best. In fact, you should avoid patients where the routine action produces cure with certainty, because the profit margin is likely to be low.
3. Select only patients with high conscientiousness, because they will have much better outcomes than patients with similar ailments and low conscientiousness.
etc.
Most consultants actually do not think all that highly of themselves – quite the opposite (even though on some level they know that they probably graduated 2 stddev above average). They were hired because they were insecure overachievers and they mostly stay that way. In fact, being at a top tier consultancy is an incredibly humbling experience for most – suddenly they are just a small fish in a very much bigger pond compared to what they experienced at university (and that includes top tier ones, while we are at it).
The reason why you believe they think highly of themselves is because consulting (and the business world in general) prizes simplicity in communication (to the extent of consciously omitting details) whereas academics generally do not.
As for the proposal, at the margin, that does seem a like an improvement compared to the current, even more easily gamed system. (possibly the other difference between academics and consultants: the latter are acutely aware that very often, perfect is the biggest enemy of improving the status quo)
I’m with the above commenter…sure, there is no reimbursement system that cannot be gamed, but that doesn’t mean our current system cannot be improved.
On the downside, McKinsey doesn’t have to actually lower costs to be successful; they just have to convince a bunch of Medicaid and ACO/MCO program directors that they have a plausible method for doing so. Sit back and watch those fees roll in!
Easy to underestimate the ‘data mining’ power of reputation and prices in favor of datasets the are poor representations of the real world.
@A consultant is probably right: McKinsey needs to sound confident and assured even when it is not.
But anyway, look at their track record. Not much to be proud of.
http://dealbook.nytimes.com/2013/09/02/in-a-new-book-mckinsey-co-isnt-all-roses/?_r=2