Validating an HIV Case-Finding Algorithm
Author Information
Author(s): Antoniou Tony, Zagorski Brandon, Loutfy Mona R., Strike Carol, Glazier Richard H., Leslie Dan
Primary Institution: University of Toronto
Hypothesis
We sought to validate a case-finding algorithm for human immunodeficiency virus (HIV) infection using administrative health databases in Ontario, Canada.
Conclusion
Case-finding algorithms generated from administrative data can accurately identify adults living with HIV.
Supporting Evidence
- The specificity of all algorithms exceeded 99%, except for those based on a single physician claim.
- An algorithm with three physician claims over three years had a sensitivity of 96.2% and specificity of 99.6%.
- The application of the algorithm identified 12,179 HIV-infected patients in Ontario.
Takeaway
Researchers created a simple way to find people with HIV using health records, which helps in understanding how many people are living with the virus.
Methodology
The study constructed 48 case-finding algorithms using combinations of physician billing claims, hospital separations, and prescription drug claims, validated against chart data from 2,040 patients.
Potential Biases
The algorithms may not accurately reflect the broader population due to the high prevalence of HIV in the validation cohort compared to the general population.
Limitations
The study's findings may not be applicable to clinics with less familiarity with HIV disease and the validation cohort was not a true population-based sample.
Participant Demographics
The mean age of participants was 47.5 years, with 28.9% being women.
Statistical Information
Confidence Interval
95% CI 95.2%–97.9%
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website