We have read the letter by Mentzelopoulos et al. regarding the possible inaccuracy in a few outcomes of our analysis.1 We sincerely apologize for the possible duplication of a few patients from inclusion of a previous study.2 Although this error is purely unintentional, we are sorry to say that we were unable to find a statement regarding overlapping of the patients in the Methods section of the later study.3 However, a reanalysis of our primary outcome did not change the primary finding of our meta-analysis (fig. 1). Mentzelopoulos et al. correctly reported that I2, which is the heterogeneity value, and CI, are different. We have already used a random effect model in all our analysis even if the I2 value is low due to possible clinical heterogeneity in the included studies. The 95% CI is used to estimate the precision of the odds ratio.4 If a 95% CI for the odds ratio does not include 1 (the null value), then the odd ratios are said to be statistically significantly different. We also accept the possibility of biases from inclusion of the randomized controlled trial by Demory et al.5 However, reanalyzing in-hospital/30-day/intensive care unit mortality excluding Demory et al.’s trial did not change our primary findings (fig. 2). Kindly refer to figure 3 of the study by Young et al.,6 which clearly mentions that the number of survivors at day 30 was 233 in the high-frequency oscillatory ventilation group. After exclusion of the preliminary findings of Mentzelopoulos et al.,2 duration of mechanical ventilation and incidence of barotrauma remain unchanged from our primary findings.
The authors declare no competing interests.