Health Insurance Health Reform

Interpreting evidence on value

Value is the latest jargon to hit the health care sector.  One potential way value could manifest itself is through value-based insurance design (VBID).  Under VBID, patients would pay lower copayments for treatments that are highly effective and/or low cost would pay higher payments for treatments that have low effectiveness or high cost to the payer. An editorial by Shrank, Saunders and McClellan (2017) examine the reactions to a study that examined the effect of eliminating cost sharing for certain cardiovascular medications after an acute myocardial infarction.  The results showed a $5000 decreases in cost, but these results were not statistically significant from a control group with standard cost sharing.

How were these results interpreted? Generally, academics concluded that there was no effect–because the results were not statistically significant–whereas private insurers saw a large decrease in cost and some readily adopted the approach.  Why the difference?  The authors explain.

The universities and academic medical centers that conduct evaluations prioritizerigorous, high quality evidence. To be published in the highest-impact journals, researchers aim to conduct studies to minimize the possibility that their results are due to random chance or other confounding factors. This requires a sufficient sample size; well-defined interventions; similar, preferably randomized control groups; and a sufficient period to minimize transitory effects…

In contrast, the business community prioritizes speed and timely information for ongoing adjustments. If a company waited for statistically significant, as well as clinically and financially meaningful, results to be sure an intervention would deliver long-term improvements in outcomes, costs, or both, it would lose precious months or years before it could scale innovations. A health care payer may pilot and expand a new payment model well before there is definitive evidence of its success if the approach is intuitive, is consistent with the mission of the organization, is supported by a compelling business rationale, and has prior evidence trending in the right direction. This can be risky, sometimes leading to wasted resources and effort on “fads.”

In short, academics want to be 95% sure that their conclusions are right.  Businesses just need to be right over half the time; although individual interventions may not work out, as a portfolio, the expected benefits would outweigh the costs on average.


Leave a Reply

Your email address will not be published. Required fields are marked *