Embracing failure in health care delivery and learning from null results
.jpeg)
Amy Finkelstein, scientific director of the Abdul Latif Jameel Poverty Action Lab (J-PAL) North America, recently published a follow-up to a 2020 randomised evaluation of the US healthcare provider, Camden Coalition's, care management programme for high-cost, high-need patients. In an analysis of the new paper in Health Affairs, her J-PAL North America co-scientific director Matthew Notowidigdo notes that while the initial study found that the programme did not reduce hospital readmissions, the new paper found that the programme did achieve important intermediate outcomes for some populations.
Recently, researchers affiliated with J-PAL North America—a research center committed to ensuring that policy is informed by evidence—published a follow-up to a randomized evaluation of the Camden Coalition’s care management program for high-cost, high-need patients.
The initial study, published in 2020, found that the program did not reduce hospital readmissions. Now a few years later, the new paper (on which one of us, Gubb, is a coauthor) was able to dig deeper by pairing Medicaid claims data with a subset of the original study population. This effort found that the program did achieve important intermediate outcomes, getting participants to office visits with their doctors in far greater numbers. But the results also confirm that the program didn’t overcome all barriers to reducing hospital use for this population.
It also demonstrates one of several ways that we can learn from null results when evaluating health care delivery programs.