Evaluating therapy in cerebral palsy.

The optimal design for most clinical trials of treatment is the randomized controlled group trial. This methodology is not, however, uniformly applicable or appropriate. The evaluation of physiotherapy in the management of cerebral palsy in childhood is a good example of treatment for which modification of this trial design is needed. Some of the problems with eight studies in this field published between 1960 and 1993 are reviewed here and an alternative approach proposed. A representative series of single case studies with individual goal setting and a validated outcome measurement, using randomized treatment, would overcome many of the disadvantages of published studies at a greatly reduced cost. The results would enable more effective group trials to be mounted in due course to evaluate the probable impact of providing treatment to a defined population of subjects.