A manipulation-free design for single-subject cerebral palsy research

This study describes single-patient research free of treatment manipulation and base-line constraints and a method of analyzing effects and changes occurring over time. The design is presented and illustrated through a case study that compares the relative merits of four treatments administered during the same six-month period to a five-year-old subject with spastic diplegia. Performance measurement was established apart from any of the treatments. The problem of assessing one treatment while controlling for others was absorbed into data analysis by an analysis of variance. Lag sequential analysis was used to investigate the possible lasting effects of treatments. While this manipulation-free design precludes testing one specific hypothesis to discover the absolute effect of a treatment, it is more efficient than traditional base-line designs and it avoids the ethical problem of withholding treatment while establishing base-line data. The manipulation-free design also allows the many nonrandom factors that contribute to subject performance, even during no-treatment periods, to be considered in the analysis.