In a recent issue of The Lancet, Larissa I van der Windt and colleagues reported the results from APOSTEL 8, a randomised clinical trial comparing atosiban to placebo in 752 participants with threatened preterm birth between 30+0 and 33+6 weeks of gestation.1 Although atosiban prolonged pregnancy beyond 48 h (78% vs 69%; relative risk [RR] 1·13 [95% CI 1·03–1·23]) and led to more completed courses of antenatal corticosteroids (76% vs 68%; RR 1·11 [95% CI 1·02–1·22]), atosiban did not improve neonatal outcomes.

Being a doctor is a treacherous business. I will never forget the medical school Dean, in his welcome address to my cohort of eager, new students, moving swiftly from cordial congratulations to a set of truly hair-raising statistics about our choice of career. With higher rates of divorce, depression, alcoholism, and even suicide, we should be under no illusions, he told us, that patients would not be our only clinical priority. As physicians, we would also have to take care of ourselves.

We thank Lola Loussert and colleagues and Ruben Ramirez Zegarra and colleagues for their thoughtful responses to our study.1 Both Correspondences raise important questions regarding the standard regimen of antenatal corticosteroids, which is often comprised of two doses administered 24 h apart within 48 h.