The Value of Sentinel Indicators for Detecting Serious Adverse Events in Hospital Care

Abstract

Objective: The Austrian Inpatient Quality Indicators (A-IQI) are routinely measured quality and patient safety indicators derived from administrative data. A subset of these are sentinel indicators, for which even a single case of death leads to a conspicuous indicator. The purpose of this study was to assess the value of A-IQI sentinel indicators for detecting serious adverse events.

Methods: We retrospectively reviewed all inpatient treatment cases with abnormal sentinel indicators at a large university hospital in Austria from 2013 to 2022 using structured chart reviews with the Institute for Healthcare Improvement Global Trigger Tool. The detected adverse events were classified according to their severity, preventability, and causal relationship with the outcome. Positive predictive values were calculated for the individual sentinel indicators and for all indicators combined.

Results: A total of 189 adverse events in 107 cases (1 to 6 per case; mean: 1.77; SD: 1.4) were identified. 51.9% caused temporary impairment, 3.7% caused permanent harm, 5.3% required life-sustaining interventions, and 36% contributed to the patient´s death. 63.5% of the adverse events detected were assessed as potentially preventable. The positive predictive value of all sentinel indicators combined was 78.5% (95% CI: 70.7%-86.3%) for at least one adverse event and 55.1% (95% CI: 45.7%-64.6%) for an adverse event contributing to the patient's death.

Conclusions: Our preliminary results suggest that sentinel indicators might be useful for detecting serious and preventable adverse events. Further studies with larger case numbers are required to determine the actual value of these indicators for clinical risk management.

Keywords: Global Trigger Tool; adverse events; medical record review; patient safety; patient safety indicators.