Medicare has embarked on a vast value-based purchasing initiative, which has several components. One is bundled payments, which are being mandated in certain geographic areas for joint replacement procedure, which allows for a controlled study of the effect of the change. Evaluations of the first year of the bundled payment program are published in two studies carried by the Journal of the American Medical Association. (JAMA Studies) (link is to editorial, see also the two studies in the issue) The two articles focus on two different bundled payment types. The first article dealt with the Comprehensive Care for Joint Replacement, covering knee and hip replacements and which began in April 2016 and is to last for five years. For this program CMS randomized some geographic areas to have mandatory use of the bundled payment method. In all 67 MSAs were so designated. This created a natural experiment to track what happened in these MSAs versus ones where bundled payment was not mandated. In the designated MSAs only hospitals that were not participating in some other bundled payment initiative were included in the mandatory one. As with most bundled payments, the joint replacement one includes all services needed for the procedure episode of care, including any post-acute care. The intent is to put one provider in charge of making sure all related services are delivered efficiently. In this program, hospitals could also get a shared savings payment but had no downside risk.
Primary outcomes included discharge to a post-acute facility, days in such facilities, Medicare spending (calculated both before and after savings payments) and some quality measures. The assessment was conducted on episodes commencing through 12/31/2016. Discharge to post-acute facilities was 2.9% lower for episodes at participating hospitals than control ones. Spending on post-acute care was about $300 lower, a significant difference, but total Medicare spending, while lower, was not lower at a significant level. In fact, total spending including savings payments was actually higher. There was no statistically significant difference in quality measure performance. Patient morbidity scores were similar, indicating that there was not patient selection occurring, nor was their greater volume in the participating hospitals, which might have indicated attempts to make up for lower payments with higher volume. So my interpretation of these results is that hospitals with mandatory bundled payments used fewer post-acute care resources, but didn’t save Medicare any money, in fact cost it some due to savings payments. They also did not appear to try to only treat healthier patients (which might be a good thing if it eliminated inappropriate candidates for joint replacement) or try to make up a potential loss in revenue by increasing volume.
The second study looked at a separate bundled payments program, the Bundled Payments for Care Improvement, which started in 2013, and was voluntary. About 320 hospitals at some point have participated for hip and knee replacement. Participating hospitals tended to be larger, urban, non-profit and teaching institutions. The authors here tried to ascertain whether hospitals increased volume or engaged in patient selection as a consequence of participation. Participating hospitals did not appear to have different volume trends compared to participating ones nor did they appear to select healthier patients for the replacement procedures, although there was a statistically significant decline in procedures on patients with prior skilled nursing facility use. The results again suggest that unintended consequences were not occurring. And I again would express the hope that these reforms might actually encourage more appropriate use of these procedures, which have become very widespread. They are beneficial for many people, but of little value to others, particularly the very elderly. From these studies you could perceive that bundled payments have promise, but aren’t yet saving any significant money.