METHODOLOGICAL ISSUES IN THE DESIGN AND IMPLEMENTATION OF CONTROLLED CLINICAL-TRIALS IN SEVERE SEPSIS AND SEPTIC SHOCK - DISCUSSION

The conduct of clinical investigations requires intense planning to en sure appropriate patient selection, realistic end points, and a fair e valuation of the potential benefit and risk of the intervention. Criti cal reviews of the methodology used in the conduct of clinical trials are typically encountered after the completion of the study and are no t constructive in effecting change and improving study design. The cli nical protocol should clearly state the hypothesis to be tested and su pply a rationale to support the interventions under investigation. The trial should incorporate prospectively randomized patients from multi ple centers in a placebo-controlled, double-blind evaluation with clea rly defined end points that have a realistic chance of being obtained. It is also important to clearly define the population to be studied, and it is desirable to evaluate each patient's severity of illness. Wi thin reasonable limits, a similar overall approach to patient manageme nt should be assured at all investigative sites involved in the conduc t of the trial. The concept of prior submission, in which clinical tri als undergo intense scrutiny by the interested and knowledgeable medic al community before initiation, may provide a constructive basis for s tudy improvement and help to effect needed changes in the experimental design. This process may then impact favorably on the eventual conduc t of the study, translating into refinements in the analysis of the re sults; it is likely to be cost-effective for the sponsor of the study and will ensure that there are few, if any, controversies concerning t he methodologies used in the acutal performance of the trial. In today 's economic and scientific climate, this process appears to be particu larly attractive.