How was the “Case-control Study on the Association of Upper Gastrointestinal Bleeding and Nonsteroidal Anti-inflammatory Drugs in Japan” Conducted ? / 薬剤疫学

ABSTRACT

Some case-control studies have been conducted to study the relationship between a drug and adverse events in Japan. However, most such studies adopted hospital controls and rarely adopted community controls. Here, we describe the methodological issues of performing a case-control study using Japanese community controls to illustrate the association between upper gastrointestinal bleeding and nonsteroidal anti-inflammatory drugs.<BR>To ensure efficient and accurate case identification, all candidate patients admitted to the participating hospitals with target conditions (upper gastrointestinal bleeding etc.) were listed by doctors/research coordinators. Summary tables with the candidates information on eligibility and whether or not the patient gave informed consent were reported back to the study office weekly. Using the summary tables, the fraction eligible and the participation rate for cases were estimated as 24% and 76%, respectively.<BR>To select controls, residents matched with each case according to age, gender and district of residence were randomly sampled from the population registries in the municipal office. To achieve a satisfactory participation rate for residents (47%), we used envelopes with prestigious university letterhead and made phone calls in a timely manner rather than waiting for their replies, both of which yielded favorable results.<BR>By telephone interview, a large amount of information was obtained in approximately 30 minutes. To help subjects' recall, a brochure displaying pictures of analgesics and a calendar for recent 4 weeks were sent in advance, and in the interview, questions using 6 kinds of “prompts” that might enhance recall were employed.