An optimal multi-period crossover design for an application in\n paediatric nephrology

Crossover clinical trials can provide substantial benefits by eliminating\ninter-patient variation from treatment comparisons and by allowing multiple\nobservations of each patient. They are particularly useful when sample sizes\nare necessarily small. These advantages proved particularly valuable in an\nassessment of clot prevention in children undergoing haemodialysis. Only small\nnumbers of children are treated at any given time in any single unit, but each\npatient is obliged to attend two or three times each week, suggesting the use\nof a crossover trial with many periods. Standard crossover trials described in\nthe literature a) typically have fewer than 10 periods and b) are based on a\nmodel of questionable applicability to this study. This paper describes the\nderivation of an optimal crossover trial with 30 periods which was used to\ncompare the treatments using nine patients.\n