If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Development of a core outcome set for effectiveness studies of breech birth at term (Breech-COS): A systematic review on variations in outcome reporting
Obstetric and Prenatal Medicine Department, Hospital Sachsenhausen, Academic Teaching Hospital of the Goethe-University Frankfurt, Schulstrasse 31, 60594 Frankfurt, Germany
Systematic review of outcomes reported in studies concerning breech births at term.
•
Supported decision-making and evidence synthesis hampered by inconsistent outcome measures.
•
Wide range and heterogeneity in reported outcomes identified.
•
Lack of consensus in definition and measurement observed.
•
Patient-centred care targeted by involvement of PPI group in COS development.
Abstract
Women pregnant with a breech-presenting fetus at term are at an increased risk of adverse outcomes. Although the most common intervention is planned delivery by caesarean section, this is not always possible or desirable. Comparing alternative interventions is difficult due to heterogeneity in reported outcomes and their measurements. Additionally, the evidence, particularly for women in labour with a breech-presenting fetus, is very low quality, with several outcomes viewed as critical and important to decision-making not reported at all. There is a need to develop a core outcome set of minimum outcomes in all studies evaluating the effectiveness of interventions to improve outcomes associated with term breech birth (Breech-COS).
Our objectives were to (1) identify outcomes currently reported in effectiveness studies of breech birth at term using a systematic review of the literature; (2) assess the methodological quality of outcome reporting in the included studies; and (3) engage with members of an established Patient and Public Involvement (PPI) group about the results, to help frame our understanding from the perspective of service users.
We searched three databases (MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials) for all effectiveness studies associated with vaginal breech birth published in English between 2000 and 2020. Two reviewers independently screened and extracted the data. Outcomes were categorised into neonatal, maternal, features of labour, and long-term maternal and hierarchy of outcome classification (primary, secondary, part of composite, or undesignated). Frequency of reporting was calculated for each. An adapted methodological assessment was done for each study investigating whether primary and secondary objectives were clearly stated and defined. The results of the systematic review were then discussed with a PPI group to consider importance and relevance to service users.
A total of 211 outcomes were extracted from 108 included studies, comprising of short (43) and long-term (39) neonatal, short (54) and long-term (39) maternal and features of labour (36) outcomes. The most frequently reported outcome in each category was: APGAR score at 5 min, developmental vulnerability/neurological morbidity, maternal mortality, urinary incontinence, and actual mode of birth respectively. Long-term outcomes were infrequently reported in the included studies, with outcomes for future pregnancies not reported at all, although these were each deemed important by service users. There was a lack of consensus in definition and measurement of outcomes, with only 36% of the included studies having clearly stated primary and secondary objectives.
The observed heterogeneity in reported outcomes, lack of consensus in definition and measurement, as well as desire expressed by service users to have robust risk statistics for outcomes important to them highlights the need to develop a core outcome set for evaluating effectiveness studies of breech birth at term. A Breech-COS will enable useful synthesis of evidence and contribute to supported decision-making for women pregnant with a breech-presenting fetus at term.
]. However, prior planning is not always possible, and some women may prefer to give birth vaginally. Although alternative interventions have been proposed by many providers, heterogeneity in the reported outcomes [
]. Heterogeneity in outcomes measured within the same clinical area, including inconsistency about the broad domains considered, the outcomes themselves, the way these outcomes are labelled and defined, and the methods and timing of measurement, can complicate or prevent useful synthesis and meta-analysis [
Higgins, J. et al., Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Collaboration (Cochrane. Available from www.training.cochrane.org/handbook., 2019).
]. Standardised reporting of outcomes and their measurement enables the direct comparison of effects of different interventions across multiple studies, thereby improving the quality of research and minimising bias [
This literature review was the first step in developing a core outcome set (COS), the first for studies evaluating effectiveness of interventions to improve outcomes in breech births at term, to be titled Breech-COS. A COS is a minimum set of key standardised outcomes, agreed upon by all relevant stakeholders, that should be measured and reported in all trials for the specific clinical field [
Higgins, J. et al., Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Collaboration (Cochrane. Available from www.training.cochrane.org/handbook., 2019).
] all recommend developing a COS to improve synthesis of evidence. Conducting a systematic review of outcomes is an efficient way to identify an inclusive list of outcomes being reported by researchers in the field [
]. Therefore, the COMET Initiative recognises that COS need to include outcomes that are important to patients and carers, in order for care to be person-centred, and recommend incorporating Public and Patient Involvement (PPI) groups in COS development [
We undertook a systematic literature review to identify outcomes, definitions and measurements previously reported in effectiveness studies of breech births at term and involved the study’s PPI group to discuss their importance to service users, as a preliminary exercise to formal participation in the Breech-COS Delphi survey. The results of this review will be included in the international Delphi survey for Breech-COS and will be open to all key stakeholders for consensus development. Breech-COS is part of a larger project that seeks to evaluate an alternative model of care, “The OptiBreech care pathway: evaluating the feasibility and acceptability of team care for women seeking to plan a vaginal breech birth” (NIHR 300582).
Breech-COS has been registered with the COMET Initiative (www.comet-initiative.org; #1749) and has been submitted to the CROWN Initiative.
Objectives
The overall aim of Breech-COS is to identify a core outcome set for effectiveness studies of breech birth at term by an international multi-stakeholder consensus method. Breech-COS is intended as the international standard for randomised and non-randomised effectiveness studies of breech births in term pregnancies. This review is the first stage of the Breech-COS development. The specific objectives are in accordance to the four steps of developing a COS, as outlined by the COMET Initiative, namely: (1) conceptual considerations, (2) finding existing outcome measurement instruments using a systematic review, (3) quality assessment of existing instruments, and (4) recommendations for selection of outcomes included in COS using a consensus process [
The primary objective of this systematic review was:
1.
To identify outcomes currently reported in comparative studies concerning breech birth at term by way of a systematic review of the literature.
The secondary objectives were:
1.
To identify the methodological quality of outcome reporting in existing studies; and
2.
To engage with members of an established Patient and Public Involvement (PPI) group about the results, to help frame our understanding from the perspective of service users
Methods
Search strategy and selection criteria
The methods and criteria for study selection were developed in accordance with the scope of Breech-COS, and in consultation with the research team as well as the Project Steering Committee (PSC), which includes breech-experienced obstetricians, midwives and service users. The databases searched were EMBASE, MEDLINE, and Cochrane Central Register of Controlled Trials (CENTRAL). Full terms of a comprehensive, electronic search strategy are included in the supplementary material (Supplementary Material1).
The eligibility criteria were designed to identify the choice and consistency of outcomes reported in comparative studies concerning vaginal breech birth (VBB). Studies not describing breech birth outcomes, conference proceedings or abstracts without complete trial description were not included. Eligible participants were all childbearing people ≥ 37 weeks pregnant with a foetus in breech presentation. Eligible interventions included planned modes of delivery (planned VBB vs. planned CS), methods of management for VBB (upright vs. supine, expediated vs. conservative, assisted delivery), and models of care delivery (different experience levels, professions, breech teams).
The eligibility criteria were:
•
Settings: both high- and low-income
•
Language: English language,
•
Time: 2000 to 2020
•
Study design: systematic reviews, randomised trials, comparative observational studies
The time period was chosen as the ‘Term Breech Trial’, a single large randomised trial, was published in 2000 [
]. The scoping search revealed no other trials in this time period.
Two of the four review authors (SW, AH, AR, TD) independently screened the title and abstracts returned by the search strategy, followed by full-text screening to determine eligibility. A third reviewer was consulted to moderate and resolve disagreements, in cases of discrepancy. Covidence systematic review software [
] was used to record the screening process. Independent data extraction was then performed for each article using a Microsoft Excel spreadsheet-based extraction form for the following: author and title details, year and journal of publication, study type, setting, key eligibility criteria, study population size, description of intervention, proposed outcomes, primary and secondary effectiveness and safety outcome(s) reported, outcome definition(s), outcome measurement tool(s) and timing(s). Original study authors were contacted in cases of unclear or unavailable data.
Outcomes were categorised into measures associated with each of the five following: short- and long-term neonatal, short- and long-term maternal, and features of labour. Frequency of reporting for each outcome was calculated, and each was designated as a primary, secondary, part of a composite, or reported but not designated; as per the outcome framework by the COMET Initiative [
]. For the features of labour outcome group, an additional designation of group was also assigned, if the measure acted as a subject group in the study. The subsequent Delphi survey for prioritisation of outcomes by stakeholders will also be informed by these categories.
Assessment of methodological quality
The methodological quality of reported outcomes in included studies were assessed using the four questions as below [
] rather than performing a traditional assessment of individual studies.
1.
Is the primary outcome clearly stated?
2.
Is the primary outcome clearly defined so that another researcher would be able to reproduce its measurement (e.g., measurement tools, measurement timing)?
3.
Are secondary outcomes clearly stated?
4.
Are secondary outcomes clearly defined?
A scoring system was devised where each article was scored from 1 to 4 depending on how many of the questions were answered. The frequency of each quality score was then calculated.
PPI group involvement
The summary of findings from the systematic review were brought to the OptiBreech study’s pre-established PPI group in an online meeting led by members of the review team [SW and TD]. The meeting was advertised on the project website (OptiBreech.uk) as well to invite parents with experience of breech birth. The purpose of this stage was to gain insight into (1) how service users with experience of breech birth value the outcomes identified in published literature, including their importance and relevance; and (2) to identify any outcome measures that are deemed important by the PPI group but were not identified by the systematic review. This was to enable the PPI group to have significant influence on the subsequent Delphi process from the start of the survey. The meeting was recorded, and key themes and points of discussion obtained from the meeting transcript. Understanding was checked with participants repeatedly during the meeting. No identifiable information or direct quotes were collected or preserved.
Results
Our search identified 817 records in total, of which 13 were duplicates and 629 were deemed irrelevant based on title and abstract screening. Full text was assessed for the remaining 147 articles and 39 were removed due to various reasons such as full-text not available in English, wrong patient population etc. Finally, data was extracted for 108 studies, as presented in the PRISMA flow chart in Fig. 1. The 108 articles included comprised of reviews with or without meta-analysis, randomised controlled trials, and non-randomised observational studies from various settings: low-, middle-, and high-income countries; at home birth vs. tertiary care centre; and comparing various interventions: mode of delivery (e.g., VBB, external cephalic version, CS), models of service provision (e.g., staff training levels, hospital birth volume), management of delivery (e.g., upright position vs. supine, manoeuvres used, induction), or no intervention. A full record of the characteristics of each article included in this review is provided in the supplementary material (Supplementary material2).
Fig. 1PRISMA flow chart of review screening and inclusion process.
We extracted 211 total outcomes, divided into short-term neonatal (43), long-term neonatal (39), short-term maternal (54), long-term maternal (39), and features of labour (36). Frequency of reporting and outcome hierarchy designation for each of these is provided in detail in Table 1. In the 108 included reviews, the most frequently reported outcome was mode of birth: vaginal birth in 93.5%, elective CS in 76.8%, and emergency CS in 72.2% articles. Some studies compared only vaginal birth vs emergency or elective CS, but not both; others were concerned with the results of vaginal births only. However, most studies reported mode of birth as a subject group, i.e., authors compared various outcomes between vaginal breech birth vs. elective CS vs. emergency CS.
Table 1Full list of outcomes with frequency or reporting and categorisation of outcome.
On average, neonatal outcomes were reported in a more studies as compared to maternal outcomes (11.1% vs. 3.7%), with very few reporting long term or salutogenic outcomes for either. Amongst short-term neonatal outcomes, APGAR score at 5 min was the most commonly reported outcome, reported in 77.7% studies. This was followed by perinatal death (67.5%) and admission to NICU (57.4%). The three most reported long-term neonatal outcomes were developmental vulnerability/special needs/neurological morbidity (18.5%), cerebral palsy (10.1%) and developmental delay/learning disability (10.1%).
The most frequently reported short-term maternal outcome was maternal mortality, measured in 21.2% studies. Postpartum haemorrhage (14.8%) and requiring blood transfusion (14.8%) were the next two most frequent. Long-term maternal outcomes were very rarely reported, with only 6.4% studies having measured urinary, and faecal incontinence each, and 5.5% having reported on breastfeeding complications.
Planned mode of birth, for intention-to-treat analysis, or actual mode of birth was reported in 42.5% and 30.5% studies respectively. Induction of labour was also a frequently reported feature of labour, included in 30.5% studies. The frequency of the top three most reported outcomes in each category is listed in Table 2.
Table 2Top three most commonly reported outcomes in each category.
Out of 108 included articles, 32 (29.6%) scored 4 i.e., answered all four of the methodical assessment questions. However, 16 of these did not analyses any secondary outcomes or objectives. 17 (15.7%) scored 3, 45 (41.6%) scored 2, 6 (5.5%) scored 1, and 8 (7.4%) scored 0 with neither primary nor secondary objectives stated and defined. Table 3 shows detailed frequencies for each criterion. The primary objective was clearly stated in 47 (43.5%) of 108 included studies. 65 (60.1%) articles defined how the primary outcome was measured but did not explicitly classify it as a primary or secondary outcome. Notably, there was a discrepancy observed in definition and measurement for certain outcomes such as measurement period for perinatal death, or outcomes to include in a composite of trauma associated morbidity. For the former, there were 10 different time periods defined, while some left it undefined. Table 4 lists frequency for each of the definitions. 21 (22.2%) of included studies had clearly stated secondary objectives 26/108 (24.1%) studies had defined them. 58/108 (53.7%) did not investigate a secondary objective and did not have any secondary outcomes listed.
Table 3Frequency of methodological assessment criteria.
The PPI group meeting was attended by service users, practicing midwives, health educators, and breech birth activists. Overall, the participants emphasised that neonatal risk outcomes should be provided during consultation, substantiated by robust statistics, particularly for commonly occurring negative outcomes. Women prioritised neonatal mortality as measured from the onset of labour, as they felt this would reflect a difference in risk compared to cephalic (head-first) birth, rather than with caesarean section. They emphasised the need for accurate statistics in taking informed decisions, particularly being able to compare risk of outcomes in vaginally born breech babies versus those born vaginally in a head-first position. There was an issue raised of health workers using hyper-clinical or very emotive language to describe possible negative outcomes. For example, several PPI members reported ‘head entrapment’ described as an outcome, and although the image was very emotive, they were unsure what the actual impact on their baby or themselves would have been had this occurred in their births.
The time of clamping the baby’s umbilical cord was another common concern, and mothers agreed that the baby being taken away immediately was distressing, though no studies reported timing of umbilical cord clamping nor separation as outcomes. This was also true for admission to NICU, with mothers reporting that they needed to know why the baby had to go to the NICU, for how long, and if they themselves could join. There was agreement that features of labour outcomes such as rates and risks associated with induction, manoeuvres used etc. need to be more prominently reported in general. They also agreed on the importance of measuring long-term maternal outcomes particularly mental health effects for women and their partners, and rates and outcomes of future pregnancies. Service users stressed the need for improved psychological support available to women, to discuss the aftereffects of a possibly traumatic birth experience and feeling isolated during pregnancy. These concerns may also be important outcomes to measure.
Discussion
In designing a robust clinical trial, outcomes must be chosen carefully, reflecting clinical importance for service providers and meaningful endpoints for service users [
]. This systematic literature review has identified a comprehensive list of outcomes reported in current comparative studies of breech birth at term and assessed their variation and frequency; the first step of developing a Breech-COS.
Summary of main findings and implications
The results show that there is substantial inconsistency in the outcomes currently being reported. Mode of birth (vaginal breech birth, planned, and emergency CS) was the most commonly reported outcome. However, mode of birth is measured in almost every study as a subject group, i.e., to classify risks associated with each of the three modes of birth for breech presentation at term; this provided little information on the birth outcome based on intention-to-treat. Neonatal outcomes were more commonly reported than maternal ones: APGAR score, perinatal mortality, and admission to NICU; maternal mortality, PPH, and PPH requiring transfusion were reported most frequently in each of the categories respectively.
Our results partially correspond with outcomes identified as ‘critical’ or ‘important’ in the recent NICE evidence review for breech presenting in labour, informing the guideline for Intrapartum care for women with existing medical conditions or obstetric complications and their babies [
NICE Intrapartum care for women with existing medical conditions or obstetric complications and their babies. Evidence review for breech presenting in labour. NICE guideline [NG121] 2019 NICE
]. Concerning maternal outcomes, the evidence review identifies maternal mortality and major morbidities as critical, defined as obstetric anal sphincter injury (OASI), PPH and systemic infection. In our review, these were also the most common reported maternal outcomes. However, none of them was reported in over 25% of studies, and the heterogeneity with which they were measured would preclude meaningful comparisons. For example, studies reporting genital tract trauma seldom specifically reported rates of OASI, and the threshold for reporting PPH as an outcome differed considerably between studies. If a consensus agreement is reached that these, and similar major morbidities described for babies, are indeed critical to enable meaningful comparisons, agreement will also need to be reached on their definition and measurement.
The results demonstrated a lack of consensus in the measurement period for perinatal mortality, also identified as ‘critical’ by NICE. Some studies measured deaths in the entire perinatal period (antepartum and intrapartum stillbirths and neonatal deaths) [
Is vaginal breech delivery associated with higher risk for perinatal death and cerebral palsy compared with vaginal cephalic birth? Registry-based cohort study in Norway.
]. This heterogeneity in reporting increases the likelihood of measurement inconsistency, inability to compare evidence, and can lead to research waste.
Although not identified by NICE, APGAR score at 5 min is likely to be of clinical importance to service providers, as it is reflective of long-term cognitive outcomes for the baby such as reduced IQ [
The third commonly reported neonatal outcome, admission to NICU, is likely to have economic implications, increasing hospital and health service costs [
]. The importance of including this outcome is consistent with the findings of the PPI group meeting, where premature cord-clamping and admission to NICU were identified as a source of concern and distress for new mothers. The findings of this review are also concomitant with recent outcome sets developed in evaluating maternity care, where a 3-round Delphi study with key global stakeholders identified maternal death, mode of birth including use of instruments/manoeuvres, neonatal death, NICU admission, and PPH as the most important outcomes [
The results also showed that long-term maternal outcomes were very rarely reported (less than7%), particularly positive health focused, or salutogenic, outcomes. Where they were reported, most were measured by healthcare workers rather than patient-centred and self-reported. Participants in the PPI group expressed that their birth experience had long-term emotional effects, both positive and negative, and they would like more acknowledgement of these effects in research design. Long-term health outcomes and outcomes for subsequent pregnancies were also deemed important to service users, although these were less likely to be reported in current studies.
Less than 30% of the included studies clearly defined and stated the primary and secondary (where relevant) objectives. Furthermore, there was an inconsistency in outcome definition and nomenclature. PPI group members stressed the importance of availability of robust risk outcome statistics and inclusion of outcomes related to features of labour such as manoeuvres used and long-term effects on women, particularly mental health related effects.
Strengths and limitations
Breech birth at term is an under-researched topic within maternal and child health, with no clinical trial published since 2000. Therefore, evidence synthesis depends heavily on the results of non-randomised observational cohort studies. To the best of our knowledge, there exists no published COS to investigate outcomes of breech birth at term [
]. This systematic review, and Breech-COS once developed, will bridge this gap in the literature, enable improved synthesis of evidence, and alleviate some of the methodological issues discussed above in comparing effectiveness studies on breech birth at term. This study design benefits from its inclusion and involvement of PPI members, ensuring that research findings are relevant and meaningful to service users.
A strength of this review includes the comprehensive search criterion that yielded a wide-ranging list of outcomes. However, not all of these might be of clinical importance, and many might not be feasible to collect in a low-resource setting. Effectively combining, prioritising, and comparing these results in the future Delphi study, can help prevent unnecessary confusion and wasted resources [
]. The adapted methodological assessment design for this review allowed appraisal of outcomes but does not speak of limitations and inherent bias present in each of the included studies. Therefore, there is a risk of reporting bias present. In particular and most relevant to our work: outcome reporting bias which can be defined as selective reporting of some outcomes but not others, due to the influence of the results or statistical insignificance, which may compromise the validity of results [
Higgins, J. et al., Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Collaboration (Cochrane. Available from www.training.cochrane.org/handbook., 2019).
This systematic review provides a comprehensive assessment of outcomes currently reported in effectiveness studies of breech birth at term. Data was extracted from 108 studies which yielded 211 outcomes in total. Heterogeneity in reported outcomes and lack of consensus in definition and specifics on measurement indicates a need to develop a COS in this field, in order to inform future effectiveness studies and the development of a trial protocol currently undergoing feasibility testing.
Author’s contribution
SW is the Chief Investigator; she conceived the study, led the proposal and protocol development. SW and TD wrote this manuscript. SW, TD, AH, and AR contributed to screening, data extraction and analysis. Professor Jane Sandall, Professor Andrew Shennan, Dr Catey Bunce and Phoebe Roberts (service user) contributed to study design and development of the proposal. All authors read and approved the final manuscript.
Funding
This study is funded by a National Institute of Health Research (NIHR300582) Advanced Fellowship and supported by the National Institute for Health Research (NIHR) Applied Research Collaboration South London (NIHR ARC South London) at King’s College Hospital NHS Foundation Trust. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. The COS Delphi was registered with the King’s College London Research Ethics Office (MRA-18/19–13879). Study design and development, as well as data collection, management, analysis, interpretation, and writing of the report was conducted independent of the study sponsor and funder.
Declaration of Competing Interest
Shawn Walker, the lead author, is a co-Director of the Breech Birth Network, a community interest not-for-profit company.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Higgins, J. et al., Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Collaboration (Cochrane. Available from www.training.cochrane.org/handbook., 2019).
NICE Intrapartum care for women with existing medical conditions or obstetric complications and their babies. Evidence review for breech presenting in labour. NICE guideline [NG121] 2019 NICE
Is vaginal breech delivery associated with higher risk for perinatal death and cerebral palsy compared with vaginal cephalic birth? Registry-based cohort study in Norway.
00300-6&id=gr1.jpg){kind=link}