Validation of claims-based algorithms for psoriatic arthritis.

Hemin LeeJulia A FordYinzhu JinSoo Kyung ChoAdrian J Santiago OrtizAngela Y TongSeoyoung C Kim

Published in: Pharmacoepidemiology and drug safety (2019)

All seven claims-based algorithms demonstrated a moderately high PPV of 75% to 89% in identifying PsA. The use of ≥2 diagnosis codes plus ≥1 prescription claim for PsA appears to be a valid and efficient tool in identifying PsA patients in the claims data, while broader algorithms based on diagnoses without a prescription claim also have reasonably good PPVs.

Keyphrases

prostate cancer
machine learning
health insurance
deep learning
end stage renal disease
radical prostatectomy
newly diagnosed
ejection fraction
chronic kidney disease
artificial intelligence
patient reported