Using a National Representative Sample to Evaluate the Integrity of the 30-Day Surgical Mortality Metric

Abstract

The 30-day surgical mortality metric is endorsed by the National Quality Forum for value-based purchasing purposes. However, its integrity has been questioned, as there is documented evidence of hospital manipulation of this measure, by way of inappropriate palliative care designation and changes in patient selection. To determine if there is evidence of potential manipulation, we retrospectively analyzed 1,725,291 surgical admissions from 158 United States hospitals participating in the National Inpatient Sample from 2010 to 2011. As a way of evaluating unnecessary life-prolonging measures, we determined that a significant increase in mortality rate after post-operative day 30 (day 31–35) would indicate manipulation. We compared the post-operative mortality rates for each hospital between Post-Operative Day 26–30 and Post-Operative Day 31–35 using Wilcoxon signed-rank tests. After application of the Bonferroni correction, the results showed that none of the hospitals had a statistically significant increase in mortality after post-operative day 30. This analysis fails to impugn the integrity of this measure, as we did not identify any evidence of potential manipulation of the 30-day surgical mortality metric.