2004 年 35 巻 6 号 p. 297-303
We investigated deviations reported between April 1, 2002 and March 31, 2003 at Osaka Univ. Hospital. The causes of deviations were identified as investigators/institution (32 cases; 54.2%), subjects (23 cases; 39.0%), and sponsors (4 cases; 6.8%). The deviations were classified as predictable deviations (13 cases; 22.0%), inadvertent deviations (25 cases; 42.4%), deviations identified during source document verification (SDV) (14 cases; 23.7%), and deviations identified after SDV (7 cases; 11.9%). The predictable deviations were generally caused by adverse events or delays in schedules. The inadvertent deviations included insufficient laboratory data and poor medication compliance. The deviations identified during SDV included incomplete understanding of the protocol and poor medication compliance. These deviations occurred during the trials, but the investigators and their trial personnel were not aware of the deviations. During SDV, the monitor communicated such deviations to the investigators. The deviations identified after SDV were caused by delays in schedules. They also occurred during the trial, but the sponsor did not regard them as deviations during SDV.
Different types of deviations occurred within the same trial, suggesting that this kind of clinical trial seemed difficult to conduct in compliance with the protocol. Before the start of a clinical trial, the investigators/institution and IRB need to consider whether it can be conducted in compliance with the protocol. To reduce deviations, the investigators/institution should anticipate various types of deviations and take appropriate action to prevent them before and during the trial. The checklist of the protocol is a useful tool for such prevention.