Measuring Mortality in Dutch Hospitals
Author Information
Author(s): Richard Heijink, Xander Koolman, Daniel Pieter, André van der Veen, Brian Jarman, Gert Westert
Primary Institution: National Institute for Public Health and the Environment
Hypothesis
What factors explain the variation in hospital standardized mortality rates (HSMR) in the Netherlands?
Conclusion
HSMR scores vary significantly between hospitals, indicating that they can be used to monitor and compare hospital quality over time.
Supporting Evidence
- The average HSMR decreased yearly by more than eight percent.
- More than two-thirds of the variation in HSMR scores was due to differences between hospitals.
- Academic hospitals had comparatively higher HSMRs, possibly due to high-risk procedures.
Takeaway
This study looked at how many patients die in hospitals and found that some hospitals do much better than others. It helps us understand which hospitals are doing well and which need to improve.
Methodology
The study used binary logistic regression and a two-level hierarchical linear regression model to analyze HSMR data from 89 hospitals over three years.
Potential Biases
Potential biases may arise from differences in hospital policies regarding patient episodes and case-mix adjustments.
Limitations
The dataset was based on hospital episodes rather than individual patients, which may affect the accuracy of HSMR scores.
Participant Demographics
Data included all discharged patients from 89 hospitals in the Netherlands between 2003 and 2005.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website