Antiemetic studies: a methodological discussion.

Methodological issues encountered in reviewing the antiemetic literature are discussed. Variables affecting antiemetic assessment include those relating to the patient, the emetic stimulus, and the differences in action of the antiemetic drugs. Study design questions including the desirability of randomization, double blinding, use of controls, and stratifications for factors such as previous exposure to chemotherapy and strength of the emetic stimulus are discussed. From a statistical viewpoint a parallel subject design with adequate sample size is recommended. The problems of standardizing the evaluation of nausea and vomiting are considered and simple ordinal scales grading severity are suggested. The patient's overall assessment of tolerance of therapy is of primary importance.