The occurrence of potential patient safety events among trauma patients: are they random?

Abstract

Objective: The Patient Safety Indicators (PSIs) from the Agency for Healthcare Research and Quality are validated measures of quality of care. The pattern of PSIs among adult trauma patients is unknown.

Hypothesis: The occurrence of PSI events should be random and have no identifiable pattern across age, gender, and racial groups in trauma, because trauma services are designed to be an equal-access system.

Design: : Retrospective analysis of a nationally representative dataset.

Setting: Nationwide Inpatient Sample (representative 20% sample from 37 states) for 5 years (2000 through 2004).

Patients: Patients aged > or =18 admitted primarily for trauma.

Outcomes: Occurrence of at least one of the applicable PSIs on multiple logistic regression analysis, with confirmation by sensitivity analysis.

Results: A total of 1.35 million trauma patients were identified, with 19,338 patients (1.43%) experiencing at least one of the applicable PSIs. On multivariate analysis, controlling for injury severity and disease comorbidity, the adjusted odds ratios (ORs) for occurrence of at least 1 applicable PSI were noted to increase for patients who are 1) above age 35, 2) male gender (OR 1.25, 95% CI 1.19-1.31), and 3) black (OR 1.20 vs. whites, 95% CI 1.10-1.30) but not for any other racial groups. These results did not change significantly on sensitivity analysis.

Conclusions: Patients who are above age 35, male gender, and black are associated with increased likelihood of experiencing a patient safety event in trauma care. When all else is equal, black patients are approximately 20% more likely than any other racial groups to experience a patient safety event, even after controlling for injury severity and disease comorbidity. These findings can help institutions prioritize chart review-based investigations to determine potential targets of systems improvement.