Of episiotomies, implementation science, and repeating trials

Abstract

Guidelines from the UK, France, USA, and Canada all concur to recommend that the use of episiotomy should be restricted to selected cases. The first trials comparing routine versus selective policies were published in the early 1980s, the first meta-analysis was published in 1989, and the latest Cochrane Systematic Review, in 2017, included 12 randomised controlled trials with 6177 women. Knowledge transfer has been difficult in some countries, requiring implementation activities such as a randomised trial in Latin America (Althabe et al. N Eng J Med 2008;358:1929–40) or a strong national guideline in France (Gachon et al. Gynecol Obstet Fertil Senol 2019;47:627–36). Sangkomkamhang et al. (BJOG 2019; https://doi.org/ 10.1111/1471-0528.15982) report on a new trial in Thailand, replicating the initial trials and challenging their own practice of routine episiotomy, with 87% episiotomy in the routine arm and 30% in the restricted arm, with, as in previous trials, no benefit found for routine episiotomy. This new publication opens two lines of debate: (i) is there an ideal rate of episiotomy and should it be adapted to the patient, in particular Asian women; and (ii) is it legitimate to perform a repeat trial when there is already robust evidence from previous trials. The discussion turns essentially around episiotomy and the risk of obstetric anal sphincter injuries (OASIS). At present some countries have rates of episiotomy lower than 10%, with conflicting observational data regarding the effect on OASIS. Another topic is the increased risk of OASIS in Asian women delivering in Western settings, described since 2003 at least (Goldberg et al. Am J Obstet Gynecol 2003;188:1063–7). Various authors have reported and hypothesised on this, but the evidence is weak and conflicting, both for anatomical differences and for the potential benefit of a liberal episiotomy approach for these women. These two questions and more regarding episiotomy warrant new trials: restrictive versus very restrictive episiotomy and individualised versus universal policy for episiotomy. The work of Sangkomkamhang is important and thought-provoking, also because it goes against the quasi-universally accepted dogma that, where evidence exists, implementation science rather than randomised trials is warranted. This is mainly for two ethical reasons: one is that one arm will receive what is accepted to be less good management, and the other is that resources are limited and should be used to fill in gaps rather than replicate studies. A certain degree of justification might apply here because of the reports of increased risk in Asian women, although the latest Cochrane review includes one trial in Malaysia (Jiang et al. Cochrane Database Syst Rev 2017;CD000081). It can be hypothesised that for Sangkomkamhang and colleagues, mounting a randomised trial and professional participation was an effective way of introducing change. If this is the case, it may constitute an epidemiological milestone.