Purpose: Actinic keratoses (AK) diagnosis, billing, and pharmacy codes have not been validated among people living with human immunodeficiency virus (HIV), preventing use in epidemiologic and clinical research. We aimed to calculate the positive predictive value (PPV) of AK diagnosis codes, procedural codes for destruction of pre-malignant lesions, and pharmacy codes for topical 5-fluorouracil.
Methods: Patients diagnosed with HIV within the Infectious Disease clinic at the Atlanta Veterans Affairs Medical Center from 1/1/2002 to 8/5/2017 were eligible. Patients were included if they had any of the following: encounters with a diagnosis for AK (International Classification of Diseases [ICD]-9: 702.0; ICD-10: L57.0), procedural codes for destruction of premalignant lesions (Current Procedural Terminology [CPT]: 17000, 17003, and 17004), and prescriptions for topical 5-fluorouracil. PPV and binomial 95% confidence intervals were calculated.
Results: PPV was 91.9% (89.1-94.7) for 369 encounters with an AK diagnosis. For procedural codes, PPV was 52.6% (48.1-57.2) for 454 encounters with destruction of 1 pre-malignant lesion, 63.7% (58.4-68.9) for 322 encounters with destruction of 2-14 lesions, and 57.7% (38.7-76.7) for 26 encounters with destruction of 15+ lesions. PPV was 72.9% (63.5-82.4) for 85 encounters with a prescription of topical 5-fluorouracil.
Conclusion: AK diagnosis codes are appropriate to use in epidemiologic and health policy research among people living with HIV and may be more reliable than destruction of pre-malignant lesion CPT codes.
Keywords: HIV; actinic keratosis; diagnosis; positive predictive value; procedures; validation; veterans.
© 2022 John Wiley & Sons Ltd. This article has been contributed to by U.S. Government employees and their work is in the public domain in the USA.