A NBER working paper by Chandra et al. (2021) finds that even small levels of cost sharing have a major impact on pharmaceutical use and mortality.

…we show that an as-if-random increase of 33.6% in out-of-pocket price (11.0 percentage points (p.p.) change in coinsurance, or $10.40 per drug) causes a 22.6% drop in total drug consumption ($61.20), and a 32.7% increase in monthly mortality (0.048p.p.). Second, we trace this mortality effect to cutbacks in life-saving medicines like statins and antihypertensives, for which clinical trials show large mortality benefits. We find no indication that these reductions in demand affect only ‘low-value’ drugs; on the contrary, those at the highest risk of heart attack and stroke, who would benefit the most from statins and antihypertensives, cut back more on these drugs than lower risk patients. Similar patterns exist for other drug–disease pairs, and irrespective of socioeconomic circumstance. Finally…price increases cause 18.0% more patients (2.8 p.p.) to fill no drugs, regardless of how many drugs they had been on previously, or their health risks. This decision mechanically results in larger absolute reductions in utilization for those on many drugs. We conclude that cost-sharing schemes should be evaluated based on their overall impact on welfare, which can be very different from the price elasticity of demand.

This finding is not new, but the magnitude of small cost sharing increases on utilization is very large as is the increase in monthly mortality.

One of the challenges with examining the impact of prices on utilization is that prices are correlated with other factors such as previous medication use and often socioeconomic status (e.g., the Medicare Low-Income Subsidy [LIS] program). Following the work of Aron-Dine et al. (2015) and Kaplan and Zhang (2017), the study uses variation in drug prices as a function of a beneficiary’s birth month. Beneficiary cost sharing (deductible, coinsurance) operates on a calendar basis; but Medicare beneficiaries can enroll in Medicare Part D as soon as they turn 65; thus, in December, 65 year-olds born in January are more likely to be in the “donut hole” benefit phase with 100% out-of-pocket costs than 65-year olds born in December of the same year.

The authors also control for an individual’s predicted spending with no cost sharing, by using a machine learning approach that fits spending projections on Medicare LIS beneficiaries, who face little to no cost sharing. The machine learning algorithm used LIS beneficiary sex, race, zip code, drugs filled, and spending in the first 90 days of enrollment to predict 12 month spending after enrollment. The specific machine learning approaches used were LASSO and gradient boosted trees.

An interesting and important paper. Also covered by Axios.


Source link