Importance of internal controls, statistical methods, and side effects in short-term trials of vasodilators: a study of hydralazine kinetics in patients with aortic regurgitation.

We determined that the spontaneous changes in cardiac output (CO) over 12 hr in 21 patients with chronic severe aortic regurgitation averaged +/- 8.9% (p = .03). We then measured changes in CO over time after administering incremental doses of oral hydralazine (50, 100, 150, and 200 mg) every 12 hr and analyzed these changes by several methods. Changes over time of only + 14% were highly significant (p less than .001) when analyzed by t test, but were not significant by repeated-measures analysis of variance (ANOVA). When changes in CO were compared with internal control values (spontaneous changes over 12 hr), only changes of 20% or more were significant (p less than .05). Transient "peak effects" markedly overestimated the maximum effects after all doses. We then compared the incremental doses of hydralazine, given either every 8 or every 12 hr, with respect to (1) the hemodynamic changes induced, and (2) the relative incidence of acute side effects. Maximal increases in CO were similar when hydralazine was given every 8 hr (16 patients) and every 12 hr (21 patients), and ranged from + 14% after 50 mg to + 61% after 200 mg. After the 150 and 200 mg doses, marked sustained increases in CO were present at 8 hr and mild increases in CO were still present at 12 hr. Hydralazine every 8 hr was associated with side effects in 25% to 86% of patients, but when the drug was given every 12 hr it was associated with side effects in only 5% to 19% of patients (p less than .001).(ABSTRACT TRUNCATED AT 250 WORDS)