Understanding the SPARRA Score for Emergency Hospital Admissions in Scotland
Author Information
Author(s): Thoma Ioanna, Rogers Simon, Ireland Jillian, Porteous Rachel, Borland Katie, Vallejos Catalina A., Aslett Louis J. M., Liley James
Primary Institution: Alan Turing Institute, London, United Kingdom
Hypothesis
The performance of the SPARRA score varies across different demographic groups in Scotland.
Conclusion
The SPARRA score is well-calibrated across demographic subgroups, but false positive and negative rates vary significantly.
Supporting Evidence
- The SPARRA score predicts emergency hospital admissions for approximately 80% of the Scottish population.
- False positive and negative rates of the SPARRA score vary widely across demographic groups.
- Older individuals generally have higher false negative rates compared to younger individuals.
- Discrimination of the SPARRA score is stronger in individuals over 65 than in those under 25.
- SPARRA scores are calculated using routinely collected electronic health records.
Takeaway
The SPARRA score helps doctors predict who might need to go to the hospital, but it works differently for different groups of people.
Methodology
The study used retrospective data from electronic health records to evaluate the SPARRA score's performance across various demographic groups.
Potential Biases
The interpretation of ethnicity data is complicated due to non-random missingness.
Limitations
The study relied on historical data, which may introduce observational biases, particularly regarding ethnicity data.
Participant Demographics
The cohort included slightly more than 50% females and over 20% individuals in the most deprived SIMD quintile.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website