PReCePT: Equity in antenatal MgSO4 treatment
Senior Research Associate, Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK and National Institute for Health and Care Research (NIHR) Applied Research Collaboration (ARC) West, University Hospitals Bristol and Weston NHS Foundation Trust, Bristol, UK
Consultant Neonatologist and Senior Lecturer in Population Medicine, University Hospital of Wales, Cardiff, UK, and Cardiff University, Cardiff, UK
Research Fellow in Public Health Evaluation, Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK and NIHR School for Public Health Research, Newcastle University, Newcastle upon Tyne, NE2 4HH, UK
Professor in Epidemiology and Public Health, Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK and National Institute for Health and Care Research (NIHR) Applied Research Collaboration (ARC) West, University Hospitals Bristol and Weston NHS Foundation Trust, Bristol, UK
Consultant Neonatologist and Professor in Neonatal Medicine, Translational Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK and National Institute for Health and Care Research (NIHR) Applied Research Collaboration (ARC) West, University Hospitals Bristol and Weston NHS Foundation Trust, Bristol, UK
*Corresponding author; email: hannah.edwards@bristol.ac.ukAbstract
This protocol describes the rationale, plan, and methodology for a study investigating sociodemographic inequities in use of antenatal magnesium sulphate (MgSO4) in England, for neuroprotection of the preterm infant. It is a secondary data analysis study using routinely collected healthcare data from the National Neonatal Research Database. The focus of our analysis is to identify whether any inequities in treatment changed from before to after implementation of PReCePT, a national Quality Improvement programme to increase antenatal MgSO4 use in England. This work is motivated by, and in the context of, wider evidence of persistent inequities in maternal and perinatal care in England, and evidence that QI programmes, even those that successfully improve overall levels of care, do not necessarily address – any may even worsen – inequities in care.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
Funding Statement
The research is part of a study sponsored by the University of Bristol and is organised and funded by The AHSN Network, The National Institute for Health Research Applied Research Collaboration (NIHR ARC West), and The Heath Foundation. The views expressed in this article are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. This study has been reviewed by University of Bristol Health Sciences Faculty Research Ethics Committee (FREC).
Introduction
This study is part of the PReCePT Devolved Nations study, and readers are referred to the original protocol for details of the wider study.
Preterm birth is the leading cause of brain injury and cerebral palsy (CP).(1) Around 1.5% of births are very preterm at less than 30 weeks gestational age (GA).(2) Progress in perinatal care has led to greater survival rates in infants born preterm.(3) However, while 90% of very preterm infants survive beyond the postpartum period, approximately 35% subsequently develop neurodisabilities, including CP, blindness, deafness and cognitive impairment.(4)
Antenatal magnesium sulphate (MgSO4) given to women at risk of preterm birth reduces the risk of CP in their child by 30% (relative risk (RR) 0.68; 95% Confidence Interval (CI) 0.54 to 0.87).(5) It has been estimated that around 200 cases of CP in England could be avoided by consistent administration of MgSO4 during labour each year.(6) At under 33 weeks gestation, the number needed to treat (NNT) to prevent one case of CP is 42 (95% CI 26 to 187), and this reduces to a NNT of 37 in births under 30 weeks gestation.(6) The impact of CP is significant and lifelong for individuals, families and health services.(7) Administration of MgSO4 can be highly cost-effective with an estimated £1M lifetime societal savings per case of CP avoided.(8, 9)
Since 2015 the National Institute for Health and Care Excellence (NICE) has recommended administration of MgSO4 in very preterm deliveries (under 30 weeks GA) as a core part of maternity care.(10) Failing to comply with this guideline is considered sub-optimal care. However, until recently, uptake of MgSO4 in the UK has been low. For infants below 30 weeks GA, uptake was 9%, 24% and 38% for 2012, 2013, and 2014 in the UK respectively compared to 46%, 51% and 56% for the international network.(11) By 2017 UK uptake was at 64%. High regional variation (range 49% to 78%) indicated inequalities in perinatal care.(12) This slow uptake, as was also seen in the case of antenatal steroids, which took decades to become embedded in clinical practice, is associated with high health and economic costs. Investment in initiatives to accelerate uptake is likely to benefit patients and families.
The PReCePT (Preventing Cerebral Palsy in Pre-Term labour) Quality Improvement (QI) toolkit and implementation guide was developed to improve maternity staff awareness and thereby increase use of MgSO4. Following a positive pilot study in 2018,(13) this was scaled-up and rolled-out across England as the National PReCePT Programme (NPP). The aim was to increase MgSO4 use in England to 85% by 2020. Units received the PReCePT QI toolkit, comprising of an implementation guide and additional resources (pre-term labour proforma, staff training presentations, parent information leaflet, posters for the unit, learning log)(14) Each unit had a lead midwife as ‘PReCePT champion’, and the NPP provided up to 90 hours funded backfill for this role. Coaching and support was available through the 15 Academic Health Science Networks (AHSNs), via a regional clinical lead (obstetrician and/or neonatologist).
We evaluated the effectiveness and cost-effectiveness of the NPP in increasing MgSO4 uptake in England. The first evaluation covered the first 12 months of the programme and found average MgSO4 uptake increased by 6.3 percentage points (95% CI 2.6 to 10.0 percentage points) to 83.1% post-implementation (adjusted for unit size, maternal, baby, and maternity unit factors, time trends, and AHSN region). The health gains and cost-savings associated with the NPP generated a net monetary benefit of £866 per preterm baby. The probability of the NPP being cost-effective was greater than 95%.(15) The second evaluation (the Devolved Nations extension study) aims to evaluate the longer-term effectiveness and cost-effectiveness of the programme over the first four years following roll-out, and additionally compare MgSO4 use in England with Scotland and Wales. The full study is detailed elsewhere(16) (protocol and SAP also available).
One workstream within the extension study is to evaluate individual-level risk factors for mothers receiving MgSO4, to elucidate any inequities in treatment, and changes in equity over time. The protocol outlined here details this workstream specifically. There is evidence from previous work that there are inequities in maternal and perinatal care. For example, the MBRRACE-UK Perinatal Confidential Enquiries reports have consistently found racial differences in care of women who experienced stillbirth or neonatal death, and in maternal death rates (most recent 2024).(17) The Royal College of Obstetrics and Gynaecology National Maternity and Perinatal Audit Inequalities report 2021 found differences by race and level of socio-economic deprivation in risk of caesarean birth, blood loss after childbirth, baby being small for gestational age (SGA), Apgar score, admission to a neonatal unit, and breastfeeding.(18) A UK national cohort study in 2021 found socio-economic and ethic disparities in stillbirths, preterm births, and births with fetal growth restriction.(19) Another study from England in 2020 found socio-economic, ethic, and maternal age disparities in maternal mortality, with ethnic disparities worsening and socio-economic disparities improving over time.(20) As well as these socio-demographic variations, the latest annual report from the National Neonatal Audit Programme(21) indicates geographical (regional) disparities across England in many outcome metrics including neonatal mortality, brain injury, bronchopulmonary dysplasia (BPD), necrotising enterocolitis (NEC), antenatal steroid delivery, MgSO4 delivery, deferred cord clamping, normal temperature, and breastmilk feeding.
In this study we aim to explore disparities in maternal receipt of MgSO4 specifically.
Research Aim
Identify any disparities in pregnant people (hereafter referred to as mothers) receiving antenatal MgSO4 in England, and any changes in disparities over time.
Objectives
Primary
- Identify sociodemographic disparities in receipt of MgSO4 by maternal age, ethnicity, and area deprivation.
- Identify geographic disparities in receipt of MgSO4 by North versus South of England.
- Identify any changes in disparities from before to after the launch of the National PReCePT Programme (NPP).
Hypotheses
- Sociodemographic disparities in receipt of MgSO4 will have reduced post-NPP compared to pre-NPP.
- Geographic disparities in receipt of antenatal MgSO4 will have reduced post-NPP compared to pre-NPP.
Study design
This is a cohort study using routinely collected, longitudinal observational patient data.
Methods: participants, intervention, and outcomes
Study setting
Maternity and neonatal units in England.
Study duration
We anticipate the data time frame to include babies born between January 2014 and October 2024.
Eligibility criteria
Inclusion criteria
- Mothers with a baby born prematurely between 24+0 and 29+6 weeks gestation
- The pregnancy was booked in an English hospital and baby born in England
- Singletons and first infant of multiple births
Exclusion criteria
- Mothers with babies born under 24 weeks, or born at 30 weeks or more. This is because the NICE guidance on antenatal MgSO4 recommends treatment for babies under 30 weeks gestation. (They recommend ‘consideration’ of treatment for babies 30-34 weeks gestation; this older group is not included in this study.)
- Pregnancy not booked in England or baby not born in England. This is because study approvals cover England only.
- Mother or baby missing unique identifier code in the data.
Intervention
The intervention being evaluated is the National PReCePT Programme, in terms of its potential effect on inequities in receipt of antenatal MgSO4.
Outcomes
The primary outcome is receipt of antenatal MgSO4 (yes versus no) at the individual (mother) level.
Sample size
Data from all mothers meeting the inclusion criteria above are included in this study. We anticipate 155 English maternity/neonatal units to be included. Data from the original study suggests that we can anticipate around 3000 per year age 24+0 to 29+6 weeks gestation (and around 6700 per year age 30+0 to 33+6 weeks gestation).
Recruitment
Recruitment is not applicable as this study uses only routinely collected healthcare data.
Methods: data collection, management, and analysis
Data collection methods
Routine pseudonymised patient-level data from the UK National Neonatal Research Database (NNRD) is used. Consent to use data for each unit is obtained centrally by the Neonatal Data Analysis Unit (NDAU), which manages the NNRD.
Data management
Storage of all data will comply with the General Data Protection Regulation (2018) and University of Bristol’s data protection policies. Storage will be on secure University computer systems within a Safe Haven folder. The Chief Investigator acts as data custodian.
Analysis
All quantitative data will be analysed using STATA version 18 (Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC).
Variables
Descriptive analysis
The distribution of demographic, geographic, and clinical factors will be described for pre and post-NPP. Continuous variables will be summarised using means and standard deviations (SD) (or medians and interquartile ranges (IQR) if the distribution is highly skewed), and categorical data will be summarised as numbers and percentages. The number and percentage of mothers receiving MgSO4 by year will be presented. For consistency with nationally reported data (National Neonatal Audit Programme (NNAP) reporting), MgSO4 uptake is defined as the number of mothers recorded as having received MgSO4 at a Unit divided by the total number of eligible mothers at that Unit, excluding missing values from the denominator. This will be expressed as a percentage.
Main analysis
Logistic regression models will be developed for each characteristic, generating an odds ratio (OR) representing odds of receiving MgSO4 if a mother was exposed, divided by odds of receiving MgSO4 if a mother was not exposed. ORs below 1.0 indicate that the characteristic is associated with lower odds of receiving MgSO4. ORs above 1.0 indicate that the characteristic is associated with higher odds of receiving MgSO4. ORs will be presented with 95% confidence intervals (95% CI) and p-values. If the 95% CI excludes 1.0, that will be interpreted as evidence for an association between the characteristic and receipt of MgSO4.
Unadjusted and adjusted multivariable models will be developed for each characteristic and the outcome of receiving MgSO4. To evaluate changes in equity from pre- to post-NPP, both unadjusted and adjusted models will include a binary variable for pre-versus post-NPP, and interaction between this and the characteristic of interest. Where there is evidence of statistically significant interaction, the coefficients for pre- and post-NPP will be presented separately.
The variables included in the adjusted models have been decided a priori based on the minimally adjusted set from Directed Acyclic Graphs (DAGs)(22), developed by the research team, using the DAGitty software programme (https://www.dagitty.net/)(23). For each exposure of interest, we will show two different adjusted models, one showing the total effects of that exposure, and one showing the direct effects of that exposure. Separate confounder sets have been considered, and developed, for each exposure of interest and to estimate both total and direct effects. These adjustment sets for each exposure of interest, together with their DAGs, are included as an appendix in this protocol. Full coding of each model within the DAGitty programme is available from the research team.
Models will be multi-level, with maternity unit as a higher level variable; to account for clustering by maternity unit. Here the individual-level data is the first level (fixed effects), nested in maternity unit as the second level (random effects). To account for the underlying time trend, time (in months) will also be included in the models as a first level variable.
Sensitivity analysis
As a sensitivity analysis will re-run the models excluding mothers with a record of pregnancy hypertension. This is because in theory (although uncommonly in practice) MgSO4 can be given to a mother with hypertension as a prophylactic against seizures. As this is in theory a distinct group from mothers given antenatal MgSO4 for neuroprotection of the infant, we will check that main results are robust if we exclude them.
Data Availability
NNRD data dictionaries are publicly available at https://digital.nhs.uk/data-and-information/information-standards/governance/latest-activity/standards-and-collections/dapb1595-neonatal-data-set/. NNRD data is accessible via formal application.
Glossary of Abbreviations
- AHSN
- Academic Health Science Network
- ARC
- Applied Research Collaboration
- CI
- Confidence Interval
- CP
- Cerebral Palsy
- FREC
- Faculty Research Ethics Committee
- GA
- Gestational Age
- GCP
- Good Clinical Practice
- HRA
- Health Research Authority
- MgSO4
- Magnesium Sulphate
- NDAU
- Neonatal Data Analysis Unit
- NHS
- R&D National Health Service Research & Development
- NICE
- National Institute for Health and Care Excellence
- NIHR
- National Institute for Health and Care Research
- NNAP
- National Neonatal Audit programme
- NNRD
- National Neonatal Research Database
- NPP
- National PReCePT Programme
- ODN
- Operational Delivery Network
- OR
- Odds Ratio
- PreCePT
- Prevention of Cerebral Palsy in pre-term labour
- RCT
- Randomised Control Trial
- SAP
- Statistical Analysis Plan
- SOP
- Standard Operating Procedure
- VON
- Vermont and Oxford network
- WEAHSN
- West of England Academic Health Science Network
Ethics, Regulatory issues, and dissemination
Research Ethics
No informed consent is sought from individual women delivering in maternity units for their participation in this study, as health care provided is according to current practice, following the clinical practice NICE Preterm labour and birth guideline (NG25) [7], at the discretion of the obstetrician/midwife. Only routinely collected pseudonymised data will be used, which are not subject to NHS ethical review. The wider PReCePT Devolved Nations study, of which this is a part, has received research permissions and approvals from the HRA, and Sponsor faculty approval, as detailed on the covering page.
Indemnity
For the wider PReCePT Devolved Nations study, the University of Bristol has arranged Public Liability insurance to cover the legal liability of the University as Research Sponsor in the eventuality of harm to a research participant arising from management of the research by the University. This does not in any way affect an NHS Trust’s responsibility for any clinical negligence on the part of its staff (including the Trust’s responsibility for University of Bristol employees acting in connection with their NHS honorary appointments).
Study Sponsorship
The research is sponsored by the University of Bristol and is organised and funded by The AHSN Network and The National Institute for Health Research Applied Research Collaboration (NIHR ARC West). This extension is funded by The Heath Foundation and ARC West. This study has been reviewed by University of Bristol Health Sciences Faculty Research Ethics Committee (FREC).
Patient and Public Involvement and Engagement (PPIE)
The original PReCePT1 intervention was developed using a co-design and co-production approach, developing partnerships with BLISS, a support organisation for mothers experiencing pre-term births, and two mothers who had experienced pre-term births. These two mothers are part of the steering group for the wider PReCePT programme evaluation, enabling the project to benefit from the knowledge and experiences they have developed over the life of the PReCePT implementation.
Dissemination
All outputs from this research will be written and presented according to the NIHR ARC West publication policy. We plan to publish and share the results of study in relevant professional network publications, including peer reviewed journals. We will present the work by disseminating it to a range of networks for both clinical and academic audiences. We aim to achieve regional and national impact through presentations at national and international scientific conferences and workshops.
Appendix Group
Appendix 1
Adjustment sets for models
Appendix 2
Directed Acyclic Graphs for models
Appendix 3
DAG models key