Cardiac Care: What a Difference a (30th) Day Makes
In an original and potentially consequential research study, Bryan Maxwell and colleagues from Stanford and Johns Hopkins have unearthed evidence that calls into question whether cardiac patients are getting the full commitment from their providers. Among 600,000 Americans receiving major cardiac surgery in recent years, the authors have found that mortality rates rise sharply just at the time when the patients would no longer be counted by the most common mortality indicator.
For every extra day in the hospital post-surgery—up to the end of the 29th day—the mortality rate rises regularly but very slightly. The rough upward trend shown in the graph below makes sense given that those who need longer inpatient stays are sicker and thus at greater risk. Estimating from data the authors provide, longer stays spell an average risk increase of 2.2% per day. However, day 30 shows a whopping 39% jump that matches the increase occurring over days 15 through 29 combined. This anomaly is statistically significant by almost any standard and makes for a very striking change in the pattern (a mockup is shown below).
Maxwell et al. are careful to note that theirs is a retrospective study and thus cannot prove causation.
We would assume that few patients would be sick enough to require an inpatient stay as long as 29 days or more: sample sizes of patients who stay beyond perhaps 15 days figure to be relatively small (and the authors do not list these numbers). Still, one is hard-pressed to come up with a medical reason why the sudden day-30 change in mortality should occur. This lends support to the authors’ supposition that providers may somehow be contributing to the post-30-day risk increase.