A new study from UCLA has found that a commonly used federal hospital safety metric may be misleading when applied to emergency stroke care, potentially discouraging hospitals from treating the sickest patients.
Published in the Journal of NeuroInterventional Surgery, the research evaluated Patient Safety Indicator 04 (PSI 04), a “failure-to-rescue” metric created by the U.S. Agency for Healthcare Research and Quality to track deaths following treatable surgical complications. Researchers analyzed data from 73,580 stroke thrombectomy procedures performed nationwide between 2016 and 2019, along with in-depth case reviews at UCLA Health.
While PSI 04 is effective for assessing elective surgeries in relatively healthy patients, the study found it to be unsuitable for endovascular thrombectomy—an emergency procedure performed on critically ill stroke patients.
The researchers determined that PSI 04 often flags deaths caused by the severity of the stroke itself rather than by procedural complications. As a result, hospitals delivering high-quality care to the most critically ill patients may appear to perform poorly on safety metrics.
At UCLA’s Comprehensive Stroke Center, expert reviews showed that all PSI 04–flagged deaths were due to complications of severe strokes, not failures in care or technique. None represented preventable safety events.
The authors warn that continued use of inappropriate metrics may unfairly penalize hospitals and discourage lifesaving treatment for high-risk stroke patients. Proposed revisions by CMS to exclude acute stroke admissions from PSI 04 could help correct these shortcomings.
