OBJECTIVES: Access to large-scale public health claims databases is limited due to privacy concerns and lack of a unique identifier in Japan. Since such databases were not originally designed for outcomes research, we assessed the usefulness of one of the commercial databases in identifying treatment patterns in routine clinical practice. METHODS: Treatment patterns for patients with exudative age-related macular degeneration (AMD) using anti-vascular endothelial growth factor (anti-VEGF) was selected as an example to enhance our understanding of real-world clinical practice. A retrospective open-cohort study of patients diagnosed with AMD and treated from January 2010 to December 2012 at community hospitals employing Diagnostic Procedure Combination payment system was conducted by using hospital claims database provided by Medical Data Vision, Co., Ltd. Strengths and weaknesses of using the database were also identified. RESULTS: During the study period, 248 patients were diagnosed with AMD and received ranibizumab, 19 patients received pegaptanib, and two patients received aflibercept. Among the patients who received ranibizumab, the average number of ranibizumab injections per year was 3.21±2.38. The strength of the database was granularity of data including daily medication and medical procedural histories that allow us to obtain key parameters of analyzing treatment patterns. A major weakness was no unique identifier for patients; thus, a patient would be recorded as two different entries in the database if s/he visited two different hospitals. CONCLUSIONS: The results show that the commercial hospital claims database was useful to understand treatment patterns of exudative AMD at non-university hospital settings. Since treatment guidelines are usually written based on published clinical trial evidence, this type of database research provides an understanding of real life clinical practice and its associated patient outcomes, contributing to better adherence and future updates of treatment guidelines.