Development and Validation of Claims-Based Algorithms for Non-Melanoma Skin Cancer in Patients With Psoriasis Using...

Accurate identification of non-melanoma skin cancer (NMSC) using ICD-10 claims-based algorithms is largely uncharacterized. This study aimed to develop and validate claim-based algorithms for NMSC.

Adult patients with psoriasis were identified in Optum Market Clarity between Oct 2015 and Dec 2020. Eligible patients had at least 1 year of claim enrollment prior to cohort entry, linked claims-EHRs, overlapping follow-up periods, and no prior malignancy. Data were randomly split into the development and validation datasets. Broad NMSC encompassed narrow NMSC (basal cell carcinoma, squamous cell carcinoma, sebaceous cell carcinoma, and malignant neoplasms of the sweat glands), Merkel cell carcinoma, carcinoma in situ, Kaposi sarcoma of skin, and cutaneous T-cell or B-cell lymphoma. The five developed claims-based algorithms consisted of 1+ diagnosis code (1) at any position; (2) at the primary position; (3) by a dermatologist; (4) at any position plus treatment procedure code; (5) at primary position plus treatment procedure code. NMSC cases and non-cases were adjudicated using EHR records. Algorithms performance measures were estimated.

Of 25 647 patients (mean age 51.4 ± 15.2 years, female 53.6%), 321 broad and 281 narrow NMSC patients were identified in the development dataset (n = 12 824). Algorithm sensitivity was 89.9% (algorithm 1) and 81.9% (algorithm 2) but under 65% for algorithms 3-5 for broad NMSC. PPV ranged from 98.2% to 99.1%. Results were consistent in the validation dataset.

This study developed and validated high-performing ICD10-based algorithms for NMSC to facilitate future dermatology research using health data.