Related Experiment Videos
Using severity-adjusted stroke mortality rates to judge hospitals
L I Iezzoni1, M Shwartz, A S Ash
1Department of Medicine, Harvard Medical School, Boston, MA, USA.
Summary
Hospital performance evaluation using death rates requires adjusting for patient severity. Different severity measures can lead to varied conclusions about which hospitals have higher or lower mortality rates.
Area of Science:
- Healthcare quality assessment
- Clinical informatics
- Public health
Background:
- Mortality rates are a key metric for assessing hospital performance.
- Accurate comparison necessitates controlling for patient severity differences.
- Numerous patient severity tools are available in the US market.
Purpose of the Study:
- To determine if different patient severity measures identify distinct hospitals as having higher or lower than expected death rates.
- To assess the impact of severity measurement selection on hospital performance evaluations.
Main Methods:
- Applied 11 distinct, widely-used severity measures to a single database.
- Utilized data from 9407 medically-treated stroke patients across 94 hospitals.
- Analyzed in-hospital deaths (916, or 9.7%) to calculate mortality rates.
Main Results:
- Unadjusted hospital mortality rates varied widely, from 0% to 24.4%.
- 27 hospitals showed significantly different observed vs. expected mortality rates based on at least one severity measure.
- Agreement between severity measures in identifying top/bottom performing hospitals was fair to good.
Conclusions:
- Evaluating hospital performance using severity-adjusted stroke patient mortality is highly sensitive to the chosen severity measure.
- The selection of a specific patient severity tool can influence the perceived performance of hospitals.
- Standardization or careful consideration of severity measurement is crucial for reliable hospital quality assessment.