Background
One in five pregnant women in the United Kingdom (UK) has multiple, pre-existing long-term physical or mental health conditions (termed ‘multimorbidity’ hereafter) [
1]. Polypharmacy is prevalent in pregnant women with multimorbidity as they may have to manage their health conditions with multiple medication [
2]. Recent studies have demonstrated that maternal multimorbidity is associated with adverse outcomes such as hypertensive disorders of pregnancy, utilisation of acute health services during the perinatal period, preterm birth, severe maternal morbidity and maternal mortality [
3‐
5]. However, this evidence is sparse and the population is under-researched [
3]. The impact of polypharmacy on the pregnancy, the women and her child is also unclear.
Research priorities for multimorbidity in pregnancy include understanding the long-term consequences for mother and child and developing new interventions and models of care [
6]. Both observational and interventional studies are needed to provide information that can help women with multimorbidity make informed decisions with their clinicians, and to develop interventions that will improve outcomes for mother and child. For instance, longitudinal observational studies are crucial to providing evidence on children’s long-term outcomes.
As research in this field gains momentum globally [
3,
7,
8], a core outcome set is needed to avoid heterogeneity of reported outcomes, which can limit the synthesis of research and its usability [
9]. A core outcome set informs what outcomes are important and should be reported as a minimum in all future studies for a particular health condition [
10]. To ensure its relevance, core outcome sets are developed through consensus-setting methods with stakeholders including people living with the health conditions, health and social care professionals and researchers [
10].
There are currently limited core outcome sets available for long-term conditions in pregnancy; examples include core outcome sets for epilepsy [
11], diabetes [
12], heart conditions [
13] and rheumatological conditions in pregnancy [
14]. Core outcome sets for pregnancies in general [
15,
16] and for medication safety in pregnancy [
17] do not have outcomes reflecting challenges specific to women with multimorbidity, such as the impact of pregnancy on their long-term conditions. Conversely, the core outcome set for multimorbidity [
18] does not have pregnancy outcomes. To address this gap, and to guide future studies in this field, a core outcome set specific for pregnant women with multimorbidity is needed. This paper reports the development of a core outcome set for studies of pregnant women with multimorbidity.
Discussion
Main findings
This paper reports the process of developing a core outcome set for studies of pregnant women with multimorbidity. The final core outcome set included 11 outcomes: five maternal outcomes and six child outcomes. Maternal outcomes covered survival, severe manifestation of maternal complications during pregnancy and childbirth, impact on the women’s multiple long-term conditions and mental health and experiential outcomes. Child outcomes covered survival, gestational age and birth weight, separation of baby and mother at birth for health care needs and longer-term neurodevelopmental and quality of life outcomes.
Comparison with the literature
Outcomes that are of importance to all pregnant women are likely to also be important to pregnant women with multimorbidity. Therefore we expected an overlap of the current core outcome set with existing core outcome set for pregnancy, childbirth and maternity care [
15,
16], such as survival of mother and child, gestational age at birth, birth weight and quality and experience of care.
Severe maternal morbidity that arises during childbirth, a composite outcome that is frequently used in recent USA-based studies of maternal multimorbidity, was also included [
3,
4,
40,
42,
44]. However, our study additionally included core outcomes specific to pregnant women with multimorbidity such as
Change in existing long-term conditions (physical and mental).
Strengths and limitations
The core outcome set was developed with a robust multistage approach, balancing the views of all stakeholders including women with multimorbidity, health and social care professionals and researchers. The broad remit of multimorbidity allowed us to work with many national and international patient charities for recruitment. This is reflected in the broad range of study participants, including participants from under-served groups, who provided invaluable perspective on the included outcomes. The multidisciplinary nature of maternal multimorbidity was also reflected in the range of health professionals who participated, including health professionals in women’s health, children’s health and mental health, in both primary care and hospital settings.
Our Patient and Public Involvement Advisory Group was involved at all stages of the study. This is a diverse group of women with lived experience of a broad range of health conditions, disabilities, geographical and ethnic representation in the UK. They advised on the scope of the core outcome set, reviewing and piloting the study materials, recruitment, conduct of the study, interpreting the focus group findings, selection of the initial outcomes and participating in the consensus meetings.
However, a key limitation of this study is the representation of stakeholders. Despite having more women stakeholders in the focus groups, only a third of the Delphi surveys participants were women stakeholders. Although a third of women/partner stakeholders who participated in the Delphi surveys were from non-European countries, all women stakeholders at the consensus meeting were based in the UK. The study findings may not represent the views of participants who do not have digital access or experience care outside of the UK or similar high-income settings.
Despite recruiting for family members, carers and partners of women with multimorbidity, only two partners participated in the focus groups and one partner participated in the Delphi surveys. We were not able to consider the views of children born to mothers with multimorbidity. It may be possible that some of the women participants met these criteria given the hereditary nature of some health conditions, but this information was not captured. The WRISK study highlighted concerns that current pregnancy risk messaging prioritises fetal health over the women’s health outcomes [
70,
71]. Therefore, this study focuses on maternal and child outcomes that are important to women with multimorbidity and information that will help women make informed decisions for their own care during pregnancy and in the postpartum period.
The attrition rate in the follow-up survey was high (44%). Previous studies have reported attrition rates ranging from 21 to 48% [
29,
60]. The survey burden presented by the long list of outcomes is likely to have contributed to the difficulty in recruitment and retention. To avoid further imbalance of stakeholder representation, we terminated the Delphi survey after the second round. This meant participants did not have the opportunity to reflect on the scores for the four new outcomes added from the first Delphi survey.
Research implications
A core outcome set lists the minimum standard list of outcomes that should be measured (‘what to measure’). Once this is defined through consensus-setting methods, a separate piece of work is needed to reach consensus on how the core outcomes should be defined and measured (‘how to measure’) [
10], following the guidance of the Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN) initiative [
72]. This includes a literature search to identify existing measurement instruments for each of the core outcomes, quality assessment of the instruments and a consensus process to agree on one instrument per core outcome. In this study, key points were raised in the consensus meetings on defining the core outcomes, these should be taken into consideration when developing a consensus on how to measure the 11 core outcomes. The next step is to disseminate the core outcome set for use in future observational and interventional studies in line with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) statement and the CROWN initiative [
9,
73]. As this core outcome set is also applicable to observational studies using routine health records, it can be considered by those designing data collection tools within the healthcare services. This can provide consistency in data collection across healthcare providers, allowing for clinical audit and secondary analysis.
To reduce the survey burden, some outcomes were combined into broader categories when designing the Delphi surveys. For instance, vaginal, caesarean and instrumental births were combined as Types of birth. Preterm births, small and large for gestational age, are captured by Gestational age and Birth weight. Some outcomes were considered so important they were kept as standalone outcomes alongside broader outcomes, such as Cerebral palsy, General motor, cognitive and social ability alongside Neurodevelopmental conditions (child); Post-traumatic stress disorder, Suicide, Self-harm alongside Perinatal mental health. Although we have grouped all types of Congenital anomaly and Neurodevelopmental conditions into one outcome respectively, that does not mean they should be researched as one entity. Depending on the research question and granularity of the data source, further subclassification of the types of congenital anomaly and neurodevelopmental conditions may be required.
Some of the outcomes were process measures. The second consensus meeting offered the opportunity to consider whether these process measures or the associated longer-term impact are more important. For instance, the quality of Care for long-term conditions and Perinatal mental health support would ultimately determine the status of the women’s long-term conditions or mental health outcomes; Requiring intubation/ventilation (neonate) and Neonatal resuscitation matter if the baby required admission to the neonatal unit is separated from the mother or develops longer-term complications. Consequently, many of these process measures were not included in the final core outcome set.
In the consensus meeting, some women stakeholders were concerned about the introduction of ableism through child outcomes such as
Neurodevelopmental conditions. Ableism is a value system that discriminates against people with disabilities [
74]. Disabled people have differing views, some may find research aimed at preventing impairment offensive whilst others are supportive [
74]. The term neurodevelopmental conditions itself has been widely debated. Within the spectrum of neurocognitive function, there are neurodivergent individuals whose neurocognitive differences fall outside societal norms but are not considered impairment, whilst a diagnosis of neurodevelopmental conditions is for those with significant functional impairment [
75]. It is, therefore, imperative to keep an open conversation with disabled people and maintain sensitivity and awareness about this [
76]. It is also important to involve people with neurodevelopmental conditions in research about the condition itself [
77].
The inclusion of perinatal mental health outcomes is important as it is one of the commonest complications of pregnancy, with suicide being the leading cause of maternal death, especially in high-income countries [
78‐
80]. Severe mental health conditions were proposed as an umbrella outcome for perinatal mental health outcome and were discussed at length. Health professionals wanted to focus on mental health conditions that are severe. However, women participants were concerned that this would not capture birth-related post-traumatic stress disorder. There is no international consensus on the definition of severe mental illness/health conditions [
81]. Conventionally, two approaches are being used: narrow (three-dimensional) and broad (two-dimensional) operationalised definitions of severe mental health conditions [
81,
82]. The three dimensions consider the following: (i) a diagnosis of non-organic psychosis, (ii) duration and (iii) disability [
81,
82]. The first approach includes a narrower list of health conditions (e.g. bipolar affective disorder, schizophrenia, psychosis) and is widely used in health services and research [
83,
84]. The second approach uses the latter two dimensions and would include any mental health conditions resulting in serious functional impairment [
85]; it was advocated by health professional participants.
As discussed by Zumstein et al., although international consensus for severe mental health conditions can facilitate large-scale epidemiological studies, definitions that are context-specific may be more useful [
82]. For example, in the context of perinatal mental health, health professional participants raised the difficulty with the duration criteria, which may exclude acute perinatal mental health conditions which are nevertheless severe. Ultimately, two of the included core outcomes will capture perinatal mental health outcomes:
Change in existing long-term conditions will capture improvement, worsening, or relapse of existing mental health conditions;
Development of new mental health conditions will capture new onset antenatal and postnatal mental health conditions, such as birth-related post-traumatic stress disorder, self-harm and suicide attempts, postnatal depression and puerperal psychosis.
Finally, just because an outcome is not included in the core outcome set does not mean it is not important. Additional study-specific outcomes can still be measured depending on the research question. This can be guided by the preliminary list of 52 outcomes prioritised by stakeholders through the Delphi surveys and first consensus meeting. For instance, studies of medication safety in pregnant women with multimorbidity may want to include
Congenital anomaly (child) [
17]. As more studies are conducted for pregnant women with multimorbidity, an update of this core outcome set may be indicated in the future [
10].
Acknowledgements
We would like to thank the following individuals, organisations and many others for helping with the recruitment of the Delphi surveys: 4M Mentor Mothers, African and Caribbean Support Northern Ireland, Alopecia UK, Ammalife, Association of South Asian Midwives, Attention Deficit Hyperactivity Disorder UK, Autism Connected, Balachandran Kumarendran, Birthrights, Black Female Doctors UK, Black Mothers Matter, Bliss, Breast Cancer Now, Bristol, North Somerset and South Gloucestershire Maternity Voices Partnership, British Adult Congenital Cardiac Nurse Association, British Association of Perinatal Medicine, British Human Immunodeficiency Virus Association, British Intrapartum Care Society, British Maternal and Fetal Medicine Society, British Thyroid Foundation, Cardiff Lupus Support Group, Cardiomyopathy UK, Chelsea and Westminster Maternity Voices Partnership, Community of Cultures Sheffield Maternity Cooperation, Core Outcome Measures in Effectiveness Trials Initiative, Crohn’s and Colitis Canada, Crohn’s and Colitis UK, Dads Matter, Diabetes UK, Disability Maternity Care (Australia), Elly Charity, E-MOTIVE Trial, Epilepsy Foundation of America, Epilepsy Society, Fair Treatment for the Women of Wales, Fibromyalgia Action UK, General Practitioners Championing Perinatal Care, Global Kidney Foundation, Graham Mcllroy, Haemophilia Foundation Australia, Hereditary Spastic Paraplegia Support Group, Institute of Health Visiting, International League Against Epilepsy (Africa), Irish Neonatal Health Alliance, Juvenile Diabetes Research Foundation, Katie’s Team, Kidney Patient Involvement Network, Kidney Wales, LGBT Mummies, MacDonald Obstetric Medicine Society, Malaysian Obstetric Medicine, Maternity and Midwifery Forum, MIDIRS Midwifery Digest, Midlands Maternal Medicine Network, Milena Forte, MQ Mental Health Research, Multiple Sclerosis Australia, Mums Like Us, Mum’s Pride, Mumsnet, Muslim Women’s Network UK, National Childbirth Trust, National Human Immunodeficiency Virus Nurses Association, National Kidney Federation, National Rheumatoid Arthritis UK, Newport Yemeni Community Association, Niina Kolehmainen, Obsessive Compulsive Disorder Action, Obstetric Anaesthetists’ Association, Organisation for Sickle Cell Anaemia Relief and Thalassaemia Support Birmingham, Parathyroid UK, Parent Voices in Wales, Parents 1st, Positive East, Positive Life Northern Ireland, Postural Tachycardia Syndrome UK, Psoriasis Association, Raham Project, Royal College of Midwives, Royal Surrey County Hospital Maternity Voices Partnership, Scottish Perinatal Network, Scottish Research Nurse, Midwife & Coordinators’ Network, Section for Women’s Mental Health Institute of Psychiatry, Psychology and Neuroscience (King’s College London), Sjogern’s India, Society of Obstetric Medicine of Australia and New Zealand, Society of Obstetric Medicine (India), Somerville Heart Foundation, Sophia Forum, South African Nephrology Society, South Asian Health Foundation, South London Applied Research Collaboration Maternal and Perinatal Mental Health Research Patient and Public Involvement, Stockport Foundation Trust, Taraki, The Black Wellbeing Collective, The International Marcé Society for Perinatal Mental Health, The Pituitary Foundation, Thyroid Patients Canada, Tommy’s, Turner Syndrome Support Society UK, UK Audit and Research Collaborative in Obstetrics and Gynaecology, UK Preconception Early-and Mid-Career Researchers Network, UK Teratology Information Service, University of Bristol Centre for Academic Primary Care and Patient and Public Involvement Panel, Vasculitis Ireland Awareness, Verity Polycystic Ovarian Syndrome UK, Wales Perinatal Mental Health Network. We would also like to thank Clare Evans for her input in reviewing this manuscript.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.