Benefit-harm assessment in childbirth care, with breech presentation as an illustration

Introduction: Childbirth care is aiming at safe births and positive birth experience. Obstetric decision-making is often complicated. Breech presentation is an example of such complexity. Judgements about individual cases and quality of care in general can be affected by choice of outcome measures. The aim of this project was to study use of obstetric outcome measures, and to explore if changes in the choice can lead to re-evaluation of results of some interventions for breech presentation and help to improve obstetric research, audit, and counselling.

Methods: I. Guidelines for external cephalic version (ECV) from all Swedish labour wards were analysed, and effectiveness of the procedure was compared between hospitals with liberal and restrictive approach. II. All births 2017-2020 recorded in the Swedish Pregnancy Register were stratified by Robson classification. Five outcomes were assessed: caesarean section, operative vaginal delivery, postpartum haemorrhage, obstetric anal sphincter injury, and Apgar score <7 at 5 minutes. III. Systematic methodological review of Cochrane Systematic Reviews focused on management of labour at or near term. We collected short-term perinatal outcomes and classified them regarding their importance to patients. IV. Consensus study performed with all relevant stakeholders. A list of outcomes, based on the results of Study III, was entered into a two-round Delphi survey, followed by a consensus meeting. V. Register-based nationwide cohort study including a cohort of women with singleton breech first births in Sweden 2000-2019, followed by second births. Women with vaginal and caesarean breech first births were compared regarding maternal and infant outcomes of their first and second births.

Results: I. Large differences between hospitals were found in several aspects of ECV. In more liberal hospitals, success rate was significantly higher (54.0% vs 50.5%, p=0.015), and the proportion of remaining breech births significantly lower (2.81% vs 3.01%, p=0.009) compared to restrictive hospitals. II. We identified Robson groups with the highest rate of each outcome. The largest contribution to four of the five outcomes was made by nulliparous women with spontaneous or induced labour and, notably, multiparous women with spontaneous labour. III. Our search resulted in 806 Cochrane Systematic Reviews, of which we included 141. We identified 348 unique outcome measures, of which only <5% were frequently used. Patient-important outcomes were rare. IV. Response rates were 82.1% and 92.7% for the first and second Delphi rounds. The final core outcome set created by study participants included 17 outcomes, whereof two were suggested by patient representatives. V. Infant risks were higher for vaginal compared with caesarean first breech birth (aOR 7.06, 95% CI 2.91–17.1). However, outcomes for the second child were better in mothers with previous vaginal compared to previous caesarean breech birth (aOR 0.26, 95% CI 0.08–0.84). There was no significant difference between groups when the two siblings were assessed together (aOR 1.19, 95% CI 0.68–2.09). Maternal outcomes were similar between the groups.

Conclusions: I. A liberal approach to ECV is more effective than restrictive in reducing breech births and is preferable to ungrounded restrictivity. II. Parallel interpretation of a balanced set of important outcomes of births stratified by Robson groups has potential to improve childbirth care. III. There is too large variation in the choice of outcome measures for labour management and discrepancy between those preferred by reviewers and reported by trialists. Harmonization is needed to facilitate research synthesis. IV. An agreement was reached on a Swedish Perinatal Core Outcome Set to be recommended in future studies on management of term birth, regardless of the specific population or condition studied. V. Vaginal compared with caesarean breech birth of the first child was associated with higher risks for this child, but lower risks for its younger sibling, giving equal outcome burden in total for the two children together. Benefit-harm assessment in childbirth care includes many interconnected aspects. A careful choice of outcome measures facilitates overview and ability to shift focus from the immediate results to ultimate goals. Several management options of breech presentation are reasonable, enabling individualised decisions.

List of scientific papers

I. Savchenko J, Lindqvist PG, Wendel SB. External cephalic version for breech presentation: The guideline landscape and a quest for an optimal approach. Eur J Obstet Gynecol Reprod Biol. 2020 Dec;255:197-202.
https://doi.org/10.1016/j.ejogrb.2020.10.019

II. Savchenko J, Ladfors L, Hjertberg L, Hildebrand E, Brismar Wendel S. A step towards better audit: The Robson Ten Group classification system for outcomes other than cesarean section. Acta Obstet Gynecol Scand. 2022 Jul;101(7):827-835.
https://doi.org/10.1111/aogs.14350

III. Savchenko J, Lindqvist PG, Brismar Wendel S. Comparing apples and oranges? Variation in choice and reporting of short-term perinatal outcomes of term labor: A systematic review of Cochrane reviews. Eur J Obstet Gynecol Reprod Biol. 2022 Sep;276:1-8.
https://doi.org/10.1016/j.ejogrb.2022.06.017

IV. Savchenko J, Asp M, Blomberg M, Elvander C, Hagman A, Pegelow Halvorsen C, Lindqvist P, Nelander M, Skiöld B, Brismar Wendel S. Key outcomes in childbirth: Development of a perinatal core outcome set for management of labor and delivery at or near term. Acta Obstet Gynecol Scand. 2023 Jun;102(6):728-734.
https://doi.org/10.1111/aogs.14560

V. Savchenko J, Pegelow Halvorsen C, Lindqvist P, Brismar Wendel S. Happy family after breech – outcomes of the two first consecutive births whereof the first one in breech presentation: a Swedish nationwide register-based cohort study. [Submitted]