Quantifying and improving rheumatoid arthritis algorithm performance in biobank settings
Main finding · PPV of RA algorithmsResult — go to the resultIncrease in sensitivity at 90 % PPV 4-11 %Authors' conclusion
Rules-based and eMERGE RA algorithms had worse performance in biobank than administrative settings. Our novel RA algorithm outperformed these standard algorithms.
Study context
- Population
- Participants of the Mayo Clinic Biobank and Mayo Clinic (MC) Tapestry Study; Rheumatoid arthritis (RA) cases identified by ≥2 RA codes or positive anti-CCP plus disease-modifying anti-rheumatic drug (DMARD) prescription as of 7/18/2022, manually verified; all other biobank participants as non-RA controls1,316 RA cases · 82,123 controls
- Exposure
- Rules-based and Electronic Medical Records and Genomics (eMERGE) RA algorithms vs novel LASSO-based RA algorithm
- Outcome
- PPV of RA algorithms
Methodological detailsbiobank setting, manual verification of RA cases by rheumatology physicians, LASSO-based machine learning with five-fold cross validation
- Biobank setting
- Manual verification of RA cases by rheumatology physicians
- LASSO-based machine learning with five-fold cross validation
Main finding
As reported in the abstract
Increase in sensitivity at 90 % PPV
4-11 %
By group
Observational data: associations, not effects.
Further findings
Self-reported RA and family history of RA (MC Biobank)
%
- Self-reported RA, PPV
- 10
- Family history of RA, PPV (worsened performance)
- 3
- Adding self-reported RA only minimally improved algorithm PPV (from 83 % to 85 %)
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Quantifying and improving rheumatoid arthritis algorithm performance in biobank settings. Semin Arthritis Rheum. Published 22 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →