PRM214 - Identifying Psoriasis and Psoriatic Arthritis Patients in Retrospective Databases When Diagnosis Code Is Not Available: A Validation Study Comparing Medication/Prescriber Visit Based Algorithms To Diagnosis Codes

Abstract

Retrospective database studies rely on the ability to accurately identify patient cohorts of interest within health care databases. Diagnosis code-based algorithms are the primary method of identifying patient cohorts; however, many databases lack reliable diagnosis code information. Our aim was to develop precise algorithms based on medication claims/prescriber visit (MC/PV) to identify psoriasis (PsO) patients or psoriatic patients with arthritic conditions (PsO-AC), a proxy for psoriatic arthritis, in databases lacking diagnosis codes. Algorithms were developed using medications with narrow indication profiles in combination with prescriber specialty to define PsO and PsO-AC. For the study period of July 1,2009 to June 30,2013, algorithms were validated using the PharMetrics Plus™ (PharMetrics) database, which contains both adjudicated medication claims and diagnosis codes. Positive predictive value (PPV), negative predictive value (NPV), sensitivity and specificity of algorithms developed for PsO and PsO-AC were assessed using diagnosis code as the reference standard. In the PharMetrics database, 183,328 patients were identified by diagnosis code or medication claim for validation. The highest PPVs for PsO (85%) and PsO-AC (65%) occurred when a predictive algorithm of ≥2 MC/PVs was compared to the reference standard of ≥1 diagnosis code. The majority of PsO-AC false positives had a diagnosis of PsO and pain or joint symptoms. NPV and specificity were also high (99 – 100%), while sensitivity was low (≤30%). Reducing the number of MC/PVs or increasing diagnosis claims decreased the algorithms’ PPV. We have demonstrated that MC/PV algorithms can be used to identify PsO patients with a high degree of accuracy, while PsO-AC accuracy requires further investigation. Such methods allow researchers to conduct retrospective studies in databases where diagnosis codes are absent.