To the Editor:

We read with great interest the study by Kaneko et al. [1]. Kaneko et al. first reported the influence of vasopressor in parturients with the hypertensive disorder of pregnancy. The authors correctly performed propensity score matching, which is known to exhibit more empirical power than logistic regression analysis when the number of events is low [2]. In their propensity score matching, parturients who administered phenylephrine or ephedrine were matched with those with ephedrine. To evaluate the effect of choice of vasopressor, those with phenylephrine should be matched with those with ephedrine. Furthermore, although there was no significant difference in the dose of ephedrine after matching, the use of statistical hypothesis test to evaluate balance before and after matching is widely criticized [3]. The simplest accepted method is to calculate the standardized difference. Therefore, we cannot be sure that there was no difference in the dose of ephedrine between the matched groups. Along with these methodological issues, as this was a retrospective study with no strict criteria regarding selection and use of vasopressor, a firm conclusion should be delayed until further prospective randomized trials.