Detection of adverse drug reactions in a clinical trial using two types of questioning.

Two methods of assessing adverse drug effects-an open-ended question versus a checklist-were compared in a clinical study involving 515 patients being treated with bacampicillin for gonorrhea. Results indicate that adverse reactions are reported more frequently if a checklist is used. However, it was also observed that more serious side effects are usually reported in response to an open-ended question rather than a checklist. Thus it was concluded that the optimal procedure for assessing adverse reactions is to record responses to both an open-ended question and a checklist.