Retrospective database studies of PAH using US payer claims data have limitations due to lack of specific ICD-9 codes for PAH and ability to identify patient severity.   Previous studies used an algorithm which includes patients with non-specific PH codes, along with a claim for an advanced PAH drug therapy. This study attempts to validate the algorithm and identify patient disease severity through linkage to data abstracted from medical charts. OBJECTIVES: To evaluate validity of a retrospective review of payer database along with chart abstraction for confirmation of PAH diagnosis and identification of World Health Organization Functional Class (FC). METHODS:  Medicare patients who received an (1) endothelin-receptor antagonist, phosphodiesterase type 5 inhibitor,  or prostacyclin AND (2) had a diagnosis of pulmonary hypertension, other chronic pulmonary hypertension or chronic pulmonary heart disease OR (3) medical claim indicating right heart catheterization (RHC) were identified from pharmacy and medical claims data.  A random sub-sample of 110 patients was chosen and the providers contacted to provide medical charts. Charts were reviewed to abstract data indicating PAH diagnosis, FC, and/or symptoms, diagnostic tests, and treatments to enable classification. RESULTS:  Of 110 charts requested, 41 were received and abstracted. Twenty-one charts (51%) came from a specialist.  All 41 charts documented a confirmed diagnosis of PAH. Of those, 18 (44%) explicitly identified PAH class.  Physical symptoms were reported, with dyspnea (66%) being most frequent, while walk test results, documentation of RHC and pulmonary diagnostic tests were reported in less than 20% of cases. CONCLUSIONS: The identification algorithm successfully identified diagnosed, confirmed cases of PAH. Refinements to provider selection algorithm could result in an increase in provider response rate, charts with documented FC and overall chart data quality.