Reevaluation of clinical trials of antimicrobials from viewpoints of protocol-defined schedule

A new guideline for good clinical practice (GCP) of Japan was introduced in 1988. Thereafter, progress in clinical trials has tended to be slower. Since strict adherence to the guideline is a 'must' in clinical trials, a study was undertaken as to whether protocol-defined schedule according to GCP was strictly adhered to in clinical trials. A total of 748 cases which were enlisted in clinical trials of twenty-five antimicrobials from 1985 through 1997 in our fifteen affiliated hospitals were subjected to reevaluation as to whether the protocol-defined schedule was strictly adhered to during the trials. Also protocols of 1,010 cases and 1,693 cases which were enrolled in two different parenteral cephem trials were reexamined from a similar point of view. The ratio of non-adherence to protocol-defined schedule (hereafter non-adherence ratio) was 7.6% and 9.9%, respectively, in the above trials of two different parenteral cephems developed in late 1980's in a nationwide survey of Japan. In the above two trials, the non-adherence ratio exceeded 0.5% in cases with diseases which were out of indication of trial antimicrobials (out-of-indication cases), in cases which received a partner drug that was not registered in the protocol, in cases with severe underlying diseases, in cases in which the dosage and usage of trial agents were not in line with the instructions of the protocol, and in cases without signs and symptoms characteristics of infection for which the trial antimicrobials were applied. As regards phase of clinical trials, the non-adherence ratio was high in the dose-finding and double-blind studies, and low in the open clinical trial. On the other hand, the non-adherence ratio was 16.3% in our affiliated hospitals. It was high in out-of-indication cases, in cases of which not all the necessary laboratory data were available (insufficient- lab-data case), in cases in which the dosage and usage of trial agents were not in line with the instructions of the protocol. As regards the phase of clinical trials in our own cases, the non-adherence ratio was high in the open clinical trial and in the double-blind study, but low in the dose- finding study. The non-adherence ratio exceeded 20% in three of fifteen institutions and was below 10% in three other institutions. It was found that out-of-indication cases and insufficient-lab-data cases were enrolled in institutions where the non-adherence ratio was high. The non-adherence ratio was 18.2% from 1985 through 1987, 17.3% from 1988 through 1990, 16.3% from 1991 through 1993, and 12.5% from 1994 through 1997, respectively. The out- of-indication cases and insufficient-lab-data cases have been decreasing in number. However the number of cases of which the age range was above (or below) the limit stated in the protocol remained almost unchanged. Finally, three representative cases were discussed. In conclusion, it is necessary that only the cases, in which the clinical effectiveness and safety can be evaluated, be enrolled into the clinical trial.