OBJECTIVE: Using a Delphi panel method, this study determined and evaluated the clinical treatment patterns of moderate-to-severe gastrointestinal (GI) events associated with NSAIDs in Japan. METHODS: Two resource utilization questionnaires were originally developed in English for a similar survey in the US and Europe (one for GI events requiring hospitalization and another for those not requiring hospitalization). The questionnaires were translated and culturally adapted for Japan and a pilot test was conducted. For the full survey, a Delphi-panel method was used. RESULTS: The pilot test of the translated/culturally-adapted questionnaires indicated that the instruments could feasibly be adapted for use in Japan. Missing data was only 4% in the pilot test. For the Delphi panel, 30 specialists were contacted to participate (10 rheumatologists and 20 gastroenterologists). Physicians were interviewed separately face-to-face, answers were summarized across all participants and the summary results were presented to each participant at a second face-to-face meeting. Participants could then modify their original answers. Average resources and costs required for the treatment and follow-up of GI events will be summarized. CONCLUSION: A lack of observational databases in Japan makes identification of treatment patterns and thus resource utilization estimates quite difficult. Interviewing experts about their treatment patterns using a Delphi panel approach appears to be a feasible option in Japan; individual face-to-face interviews are recommended. Cultural adaptation of questionnaires (originally developed for use outside of Japan) was necessary as the design of questionnaires and the structure of questions in Japan is somewhat different. Additionally, modification of the target specialist physicians was necessary due to the different referral patterns found in Japan.