FRI0445 IDENTIFICATION OF PSORIATIC ARTHRITIS USING AN ADMINISTRATIVE CLAIMS-BASED ALGORITHM

Abstract

Background Accurately identifying psoriatic arthritis (PsA) in large electronic healthcare database is critical for epidemiological studies. Objectives To develop and validate a claims-based algorithm to identify patients with PsA. Methods We conducted a retrospective chart review of the Partners Healthcare electronic medical record linked to Medicare claims from year 2012 to 2014. 7 claims-based algorithms were developed to identify PsA: 1) ≥2 International Classification of Diseases, Ninth Revision (ICD-9) codes for PsA (696.0) and at least one diagnosis of psoriasis (696.1) by any physician; 2) ≥2 diagnosis of PsA with at least 1 diagnosis by rheumatologist; 3) ≥2 PsA diagnosis with at least 1 diagnosis by rheumatologist and ≤1 diagnosis of rheumatoid arthritis (714.0); 4) ≥2 diagnosis of PsA and at least 1 diagnosis of psoriasis by dermatologist; 5) ≥1 diagnosis of PsA by rheumatologist and ≥1 diagnosis of psoriasis by dermatologist; 6) ≥2 diagnosis of PsA by any physician and ≥1 claims for PsA medication; 7) ≥2 diagnosis of PsA with at least 1 diagnosis by rheumatologist and ≥1 claims for PsA medication. The ICD-9 codes were separated by ≥7 days but Results The 7 algorithms identified 357, 399, 315, 223, 215, 372, and 276 records, respectively. Approximately 45% of the identified records with adequate data were reviewed. The PPV of the algorithms ranged from 75.2% to 88.6% (Table 1). Mean age of identified PsA patients ranged from 72.6 to 73.5 years old. Presence of psoriasis 1 year prior to index date of PsA ranged from 54.2% to 89.2%. Algorithm 6 which captured ≥2 diagnosis of PsA and ≥1 claims for PsA-related medications identified second highest number of patients (n=372) yet still yielded high PPV of 82.4% (95% CI 76.5, 88.3). Conclusion All seven claims-based algorithms had a high PPV of 75-89% in identifying PsA. A claims-based algorithm utilizing two or more diagnosis codes of PsA by any physician with a claim for PsA medication can be a useful and efficient tool to identify the PsA population in large claims databases.