CMS’ long-standing Hospital Readmission Reduction Program incorrectly penalizes hospitals or overlooks hospitals that should receive a penalty, according to a new study.
The findings, published Wednesday in JAMA Cardiology, are the latest to uncover flaws in the Hospital Readmission Reduction Program, which has been heavily criticized by hospitals and quality researchers since its inception in 2012 under the Affordable Care Act.
The new study found CMS incorrectly penalizes hospitals due to margins of error associated with the 30-day risk-adjusted readmission measure the program relies on. The study evaluated three of the six conditions part of the program: acute myocardial infarction, pneumonia and heart failure. Using CMS data from 2014 to 2017, the study found 20.9% of hospitals should have been penalized for their readmission rates for acute myocardial infarction but weren’t, while 13.5% should have received a penalty for their readmission rate for heart failure and 13.2% for their readmission rate for pneumonia. At the same time, 10.1% of hospitals received a penalty for their performance on readmissions for acute myocardial infarction but shouldn’t have, while 10.9% were incorrectly penalized for heart failure and 12.3% for pneumonia.
In the penalty program, the readmissions rate for hospitals is an estimate because CMS is using a measure with a margin of error. Other outcomes measures CMS uses for value-based payment programs also have a margin of error such as 30-day mortality rates.
Changyu Shen, lead author of the study and a senior biostatistician at the Smith Center for Outcomes Research in Cardiology at the Beth Israel Deaconess Medical Center, said he isn’t aware of published research until this study that demonstrates how the margin of error actually impacts penalties in the readmissions program.
Acute myocardial infarction had the highest percentage of wrong penalties because it had the smallest amount of discharges in the CMS dataset and therefore a high margin of error. “When you have a higher margin of error, of course you are more likely to make mistakes,” Shen said.
The findings show the magnitude of incorrectly penalizing hospitals due to the margin of error “is not small,” Shen said.
Researchers should now look into how hospitals and patient outcomes are affected by the incorrect penalties, Shen said. “The next step would be to understand the consequence, does it matter?” he said.
Shen and his colleagues offer some potential solutions such as extending the data collection period CMS uses to decrease the margin of error. The authors suggest lengthening the data collection period from three years to 12 years for acute myocardial infarction and six years for pneumonia and heart failure, although Shen said he acknowledges that time period isn’t realistic.