In Reply Dr Hyder raises interesting questions about the interpretation of our findings,1 and we are grateful for the opportunity to clarify. In our Original Investigation, we showed that hospitals that utilize intensive care units (ICUs) more frequently during admissions for diabetic ketoacidosis (DKA), pulmonary embolism (PE), upper gastrointestinal hemorrhage (UGIB), and congestive heart failure (CHF) were more likely to perform invasive procedures and have higher costs without improvements in hospital mortality.1 Dr Hyder suggests that hospital mortality may be affected by discharge bias if hospitals differentially discharge patients with high likelihoods of death to other facilities.2 We agree that using time-specific end points such as 30-day mortality reduces such bias. Unfortunately, only in-hospital mortality was available in the administrative data set used in our study. It is important to note, however, that Reineck et al2 showed that discharge bias generally results in lower in-hospital mortality among smaller hospitals, as they are more likely to discharge high-risk patients to other facilities.2 In our study, there was a greater proportion of smaller hospitals in the higher–ICU utilization group. As such, discharge bias would disproportionately decrease in-hospital mortality in the higher–ICU utilization group. If mortality was actually greater in the higher-utilization group, this would strengthen our conclusion regarding the dangers of ICU overutilization.
Chang DW, Shapiro MF. Lingering Questions Concerning Intensive Care Unit Utilization—Reply. JAMA Intern Med. 2017;177(2):289–290. doi:https://doi.org/10.1001/jamainternmed.2016.8761
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: