Skip to Content

Novel approaches to profiling hospitals on critical illness mortality

PI: Jeremy Kahn, MD, MS

Co-Investigators: Amber Barnato, Derek Angus, Joyce Chang

Funding: NIH R01HL126694 (Project period: 4/1/15 to 3/31/18)

Because of the high clinical burden of critical illness, the ICU is an increasingly important area for quality improvement and comparative effectiveness research. However, current efforts are hindered by the lack of a robust measure of hospital performance for critically ill patients. Current approaches to profiling hospitals on critical care outcomes are based solely on risk-adjusted mortality for ICU patients. However, risk-adjusted ICU mortality measures provide only a narrow window into how well a hospital cares for critically ill patients, since they exclude patients transferred in from other hospitals and ignore severely ill patients who are not admitted to the ICU. Practically, these shortcomings can result in false conclusions about hospital performance. Without more robust measures of critical care performance that address these issues it will be impossible to evaluate the effectiveness of system-wide organizational changes on critical illness outcomes or the impact of large-scale efforts to translate evidence into practice for our highest risk patients.

The overall goal of this project is to develop and validate novel measures of hospital-specific mortality for critically ill patients. We will use a unique state-wide hospital dataset from Pennsylvania containing detailed clinical and administrative data, as well as innovative statistical models that account for care transitions and  have not yet been applied to the field of performance assessment. Our results will provide clinicians, researchers and policy makers with novel measurement techniques for assessing critical care performance on a national scale, addressing a critical barrier to progress in the fields of comparative effectiveness research and T2 translation.