Background Gout outcome studies have used administrative and claims databases to assess It is unknown whether administrative-derived data are accurate and specific for gout-related utilization.
Objectives The goal of the study was to assess the accuracy of Veterans Affairs (VA) administrative and clinical databases claims for gout-related health care utilization.
Methods This retrospective study utilized the VA administrative and clinical databases claims in the fiscal year 2006. The data programmer identified a cohort consisting of randomly mixed sample of two types of visits: those visits with gout as primary or secondary diagnosis and those with diseases other than gout as primary and secondary diagnosis. A well-trained senior epidemiologist (JS) blinded to the database information related to the visit performed review of electronic health record (EHR). The gold standard was medical record documentation of gout or gout-related terms (gouty arthritis, tophaceous gout, acute gout, chronic gout, podagra) in the chief complaint, history of present illness or assessment and plan for the visit. This indicated that gout was the main or one of the main reasons for the visit. We assessed the accuracy of database-derived gout-related utilization by calculating sensitivity, specificity, positive and negative predictive values (PPV and NPV).
Results Of the 108 potential visits, 85 visits to a health care provider (in 85 patients: 84 men, 1 women with mean age of 63 years) in one of the three settings (outpatient, inpatient or urgent care/emergency room) and retrievable data from medical records, constituted the analyzed dataset. According to the gold standard of chart documentation, 21 visits were related to gout and 64 were not. Administrative claims for visits related to gout were accurate with excellent PPV of 86%, sensitivity of 86%, specificity of 95% and NPV of 95%. There were three visits coded as gout-related visit in databases, that did not have medical record documentation as related to gout: one visit each to discuss blood pressure medication, regular follow-up for multiple medical problems and for increased blood sugar. Three visits coded as not related to gout in administrative databases were related to gout based on medical record documentation: one patient with continuing acute gout, one with a new diagnosis with documentation of urate crystals in knee joint fluid and one with chronic gout stable on allopurinol.
Conclusions VA databases can be used to identify gout-related visits with acceptable accuracy. This finding should allow for studies of health services outcomes to identify gout-related utilization using VA databases.
Disclosure of Interest J. Singh Grant/Research support from: takeda, savient, Consultant for: takeda, savient, URL, Novartis, Ardea, URL