Comment: Tocolysis in placenta previa.

above which the study drug must demonstrate improvement. For the specific study question of the effectiveness of propranolol (or another beta-blocker) in the management of frequently occurring negative behaviors that are not adequately controlled by psychotropic agents, the following criteria should be met: I. Patients who are severely disabled and require long-term hospitalization are the most suitable candidates for the studies. Sixto 12month studies would be feasible in this subpopulation. 2. A lengthening of the individual blocks of the study design, perhaps in the range of 6-12 weeks, would allow adequate time for the observation of the slowly changing outcome measure. 3. Carryover effects can be handled in a manner similar to that of Greendyke et al. in their A-B design case study, which is very much in the model we are advocating.' In essence, one builds in upward and downward titration periods for introduction and removal (washout) of the study drug. In summary, aside from the difficulty of maintaining staff interest in the quantitative measurement of negative behaviors, there seem to be only advantages for applying rigorous standards to the evaluation of drug effects in individual patients.