Variations in national health care practices and behaviours and their influence on international research.

Multinational clinical trials are valuable to the understanding of global health problems, but they pose special problems. Our experience with a multinational trial of isoniazid (INH) preventive therapy for tuberculosis revealed marked variation among the seven participating countries in the amount of tuberculosis screening prior to the trial; this variation contributed to the observed differences in the risk of tuberculosis among the countries. The incidence of 'uncooperativeness' and drug side-effects, and the proportion of participants who complied with and completed treatment also varied significantly from country to country. These differences in completion and compliance served to differentially alter the expected risk of tuberculosis among the three regimens being studied. For all factors investigated, variation from country to country was greater than variation from dispensary to dispensary within a country. This suggests that cultural and other national characteristics are more potent determinants of health care practices and behaviours than patient and health care practitioner characteristics.