Assessment of response to treatment and design of clinical trials
Lusher, J.M.
Progress in Clinical and Biological Research 150: 167-179
1984
ISSN/ISBN: 0361-7742 PMID: 6431425 Document Number: 235151
The assessment of response to treatment and design of clinical trials requires careful consideration of many factors. While there are problems in assessing the response of acute hemarthrosis to treatment, this common, non-life threatening type of bleeding does seem to lend itself reasonably well to clinical trial. Intramuscular hemorrhage involving an extremity (ex: thigh, calf or forearm) also allows objective as well as subjective assessment in a clinical trial setting. Nonetheless one must still determine the intervals for assessment and define what constitutes cessation of bleeding (i.e., how much of an improvement in joint mobility, how much reduction in pain, how much muscle softening or decrease in circumference?). One must also decide whether or not there is sufficient evidence that different joints in the same individual respond differently to justify separate randomization of each joint. If one is to conduct a truly blinded trial, those who will be assessing the response to treatment should not be able to discern which product is being given. The unknown products must appear similar in color, consistency and bottling characteristics. If this is not possible, the unknowns must be given at a center where the dissimilar unknown products are presented to the patient in an opaque syringe. A final but extremely important component in the design of clinical trials is to have adequate attention to appropriate statistical methodology as the trial is being planned. One must decide what difference one wishes to detect, and must determine the sample size needed to demonstrate such a difference if it exists. If one is to conduct a trial from which meaningful, reliable conclusions can be drawn, one must carefully consider statistical methodology, and must be aware of the limitations of the trial.