Permutation Trial of Diuretics

Abstract

In order to increase the amount of information per day of observation the permutation trial is introduced in the study of acute effects of diuretic agents. Chlorothiazide and hydroflumethiazide in optimal doses (2 Gm. and 200 mg. daily) and placebo tablets were given, 1 drug on each of the 3 days of the study, in all permutations to 12 patients treated by current methods for right heart failure. The error variance of sodium, potassium, chloride, and water excretion was reduced by taking into account the variation due to permutations and days. Chlorothiazide showed a stronger diuretic action than hydroflumethiazide, whereas the natriuretic effects were not significantly different as judged by analysis of variance. Hydroflumethiazide was less kaliuretic. Magnesium excretion was independent of diuresis. Neither of the diuretics showed any effect on magnesium and calcium excretion. The ratio of sodium and chloride excretion was significantly lower during hydroflumethiazide than during chlorothiazide action. The calculated sodium concentration in the additionally excreted urine during hydroflumethiazide administration was far above serum levels (the mean of 12 observations was 353 mEq. per liter), and significantly higher than during chlorothiazide action. This effect was also demonstrated in 6 control patients during water loading, suggesting that hydroflumethiazide increases free water reabsorption also when antidiuretic hormone is absent. Comparison of the acute effects of diuretic agents is through the permutation trial made possible in the ordinarily equipped hospital without serious interference with current treatment.