Research priorities for the future health of multiples and their families: The Global Twins and Multiples Priority Setting Partnership

Over the last 40 years, the incidence of multiple pregnancies has risen dramatically1, due mainly to the increasingly widespread use of assisted reproduction techniques and advanced maternal age at conception. For the purposes of this project, the term ‘multiples’ includes twins and higher-order multiples. The term ‘multiple pregnancy’ refers to twins and higher-order multiples in a single pregnancy. Multiple pregnancies pose unique challenges for families, healthcare professionals and society in general, and they are disproportionately represented in stillbirths2, 3, neonatal deaths4 and cerebral palsy5. Furthermore, it has been shown that parents of multiples are at risk for mental-health problems postpartum and during early parenting6-8. Raising multiple infants of the same age also puts parents at an economic disadvantage9-11. Regrettably, national audits have not yet focused on multiple pregnancies. Moreover, targeted steps for reducing perinatal deaths among multiple pregnancies are lacking from service reviews, commissioning documents and inspection frameworks, including the earlier versions of the UK stillbirth care bundle12. Of note, none of the top 11 priorities for stillbirth research, identified by the UK Stillbirth Priority Setting Partnership, targets multiple pregnancy13. Although several studies have highlighted developmental difficulties for multiples and the mental impact on their parents, research is lacking on modifiable risk factors, protective factors, interventions and the expertise of healthcare professionals14, 15. It is clear that, in order to improve the outcomes of multiple pregnancies, we must focus on the specific needs of multiples, their parents and health professionals who work with them. The Priority Setting Partnership (PSP) for multiples was formed in November 2017 during the Congress of the International Society for Twin Studies in Madrid. The aim of the PSP was to bring together multiples, parents and carers of multiples, clinicians and researchers, from around the world, in order to identify the top 10 unanswered questions in health research for multiples. The long-term intention was to reduce mortality and morbidity in multiples and to improve long-term health outcomes for multiples and their families. Priority-setting initiatives are valuable and potentially powerful, as they guide national and international research-funding policies16. They also give more credibility to research questions, as they take into account the views of multiples, parents, carers, clinicians and researchers. The PSP was led collaboratively by a working group which comprised members of the Twins And Multiple Births Association (TAMBA), UK, Twins Research Australia, Australia and St George's University, UK. Working-group members are listed in Appendix S1. This project followed the James Lind Alliance methodology16. The process is outlined in Figure 1. Priority-setting partnerships do not require ethical approval as they are audits of research questions deemed important through consultation, rather than research projects. However, due diligence was conducted to assess the need for ethics by consulting The Melbourne School of Population and Global Health Human Ethics Advisory Group and St George's University of London Research Governance and Delivery of the Joint Research and Enterprise Office. In addition, the project followed the ethical guidelines of the National Health and Medical Research Council of Australia and the General Data Protection Regulation principles of the European Union. This included requesting consent from participants and advising that their participation was completely voluntary and that personal data would be kept confidential and stored securely for the duration of the project. Demographic data collected described the sample in a non-identifiable format and we advised participants that the information they provided would be used for the purpose of publication. A steering group was formed from representatives of organizations for multiples or parents of multiples, clinicians (such as obstetricians and neonatologists) and researchers with expertise in multiples. This included 32 members from 23 institutions in 13 countries. We aimed to achieve as much diversity as possible in geographic representation and expertise in multiple births. Steering-group members and their institutions are listed in Appendix S1. The steering group developed an initial online survey to identify important unanswered research questions relating to the health of multiples and their families. The survey was designed so that respondents were able to suggest up to three unanswered research questions covering important topics related to multiples' health, such as development, health, disease and emotional and mental wellbeing. The survey also requested basic demographic information including the respondent's gender, age, country of residence and stakeholder group (i.e. multiple, parent, clinician or researcher). The survey design was refined through feedback from a pilot survey involving a small group of parents, clinicians and researchers. The steering group publicized the survey through a PSP webpage, the websites of their own organizations, social media and e-mail. The survey was open to all multiples, families of multiples and those with experience in treating or caring for multiples and their families. Participants were able to access the survey between 16 October and 21 December 2018. In addition to the survey, further unanswered research questions were identified by the first author by searching guidelines from around the world relevant to the health of multiples. The search involved 28 bibliographic databases and searchable websites relevant to topics spanning child development. Research questions were collected from guidelines dated between 2011 and 2017, in order to capture the most recent sources, and were limited to those published in English. The search terms used were: ‘twin’, ‘triplet’, ‘multiple’, ‘multiple births’, ‘multiple pregnancy’, ‘research’ and ‘guidelines’. The survey results were downloaded into Microsoft Excel (Microsoft Corp., Redmond, WA, USA) for analysis, and any identifiable information was removed. Exclusion criteria for responses were determined a priori and comprised: questions that were not specific to multiples, those that were too broad (relating to the health of multiples in general, rather than a specific topic area) and incomplete questions. Submitted questions were also excluded when they were translational, i.e. there was existing evidence indicating the question had been answered, but the evidence had not been communicated adequately to stakeholder groups or implemented to improve health outcomes. Following a process agreed a priori, duplicates were removed and questions were categorized into five categories by three working-group members (J.R.L., B.L. and R.B.). Any disagreement was resolved by consensus. The following five categories were used: (1) antenatal care, (2) intrapartum and postpartum care, (3) neonatal and pediatric health, (4) child psychiatry and development, and (5) parental and family health. Questions were then grouped according to similarity and ‘indicative summary questions’ were formulated. Indicative questions clarify the precise question at hand omitting any narrative text, to ensure consistency in language and to make it easier to check the questions against the evidence base to identify whether any of them has already been answered16. When forming indicative questions, the aim was to retain the meaning intended by the respondents but to shape the question into a researchable format. Due to the large number of submitted questions and the extensive resources required to assess questions for which answers are based on qualitative evidence, a decision was made to analyze only indicative questions requiring quantitative evidence to be answered. It was decided that indicative questions requiring qualitative evidence would be analyzed at a later date, when further resources would be available. Indicative questions requiring quantitative evidence to be answered were reviewed and verified by 11 steering-group experts according to their areas of expertise. The indicative questions were checked against the following sources for evidence that may have already answered them: (1) The Cochrane Database of Systematic Reviews, (2) National Institute for Health and Care Excellence guidelines, (3) Scottish Intercollegiate Guidelines Network clinical guidelines, (4) relevant Royal Colleges' guidance, (5) randomized controlled trials and other studies with statistically significant results and (6) ongoing trials and studies. A question was deemed to be answered if a definitive answer was found in any of the above sources of evidence. Verified unanswered indicative questions and research recommendations retrieved from the guideline search formed a list of 89 questions, which were included in a second web-based survey. The aim of the second survey was to shortlist indicative questions for final prioritization at a face-to-face workshop. The survey was promoted through the same platforms used for the initial survey. Multiples, parents of multiples, clinicians and researchers from around the world were asked to select the two questions from each of the five established categories (antenatal care, intrapartum and postpartum care, neonatal and pediatric health, child psychiatry and development, and parental and family health) which they thought were the most important. The survey then asked participants to rank their selected 10 questions in order of importance to them, ‘1’ being the most important and ‘10’ being the least important. To prevent selection bias, the order of the questions was randomized every time the survey was opened by a new participant. As with the initial survey, the second survey requested participants' demographic data. The survey was open for participation between 14 May and 9 June 2019. After the closing date, raw rankings were de-identified and reverse scoring was applied using Stata 15.0 software (StataCorp., College Station, TX, USA) so that non-responses (counted as ‘0’) would not create bias in the total score for each question. Aggregate ranking of the indicative questions was determined separately for (1) multiples and parents of multiples and their representatives and (2) for clinicians and researchers. Aggregate rankings were then combined to ensure that responses of multiples and parents of multiples were weighted equally against clinician and researcher responses. The three indicative questions ranked as most important in each of the five categories were presented at a face-to-face workshop which took place on 16 June 2019 at St George's Hospital, London. Participants of the second survey, steering-group members and other parents of multiples and clinician representatives were invited to attend the meeting by the working group. At the workshop, two groups were formed, comprising equal numbers of parents of multiples, clinicians and researchers. Each group was asked to prioritize the 15 indicative questions as a set. To assist decision-making, aggregate rankings of the indicative questions from the second survey were provided for the parents of multiples and clinician/researcher groups. Examples of the original questions from the first survey were also provided to give participants a sense of the alternative perspectives of the indicative questions. After each of the two groups ranked all 15 questions, their rankings were combined into an aggregate ranking for the two groups to consider as a single large group. Subsequently, two groups were formed again, with a different assortment of parents of multiples and clinicians/researchers from that of the first two groups, and the process was repeated once more. A different facilitator chaired each large-group session. The respondents to the initial survey comprised a sample of 1120 multiples, parents of multiples, clinicians and researchers from 31 countries (Table 1), who contributed a total of 2891 research questions. Multiples and parents of multiples comprised 84% of the total sample. Participating clinicians and researchers were from a range of disciplines, with obstetricians (3.1%) and midwives (2.6%) comprising the largest healthcare professional groups. During analysis of the suggested questions, a total of 1744 duplicate questions were identified and 494 questions were considered out of scope as they were either not specific to multiples, too broad or incomplete. A further 37 unanswered research questions were identified from 10 guidelines on health and wellbeing of multiples (Table S1). The working group determined that 455 indicative questions were answerable by qualitative evidence and 235 answerable by quantitative evidence (Figure 1). The steering group reviewed the indicative questions answerable by quantitative evidence and verified 89 of them as being unanswered and, therefore, eligible for inclusion in the second survey (Appendix S2). The second survey was completed by a total of 528 multiples, parents of multiples, clinicians and researchers from 24 countries. Multiples or parents of multiples represented 82% of the sample. Again, a range of professionals participated, with researchers and obstetricians representing the largest professional groups (each 3.8%) (Table 1). The top 15 questions (the three most highly ranked questions from each of the five categories) selected from the list of 89 questions were presented at the face-to-face workshop for final prioritization. Of the 26 participants who attended the workshop to determine the top 10 research questions for the health of multiples and their families, 13 individuals identified themselves as parents of multiples or parent representatives and 13 as clinicians or researchers. Seven countries were represented (Australia, Austria, Belgium, Germany, Nigeria, The Netherlands and UK). The majority of professionals were obstetricians (26.1%) or fetal medicine experts (17.4%), although representatives from neonatology and midwifery were also present (Appendix S1). This article provides an academic summary of the priority-setting process of The Global Twins and Multiples PSP. A lay report is also under preparation for parents of multiples and policy-makers and will be distributed to key stakeholders concerned with the health of multiples. The results of this project represent a range of health research concerns as viewed by multiples, parents of multiples, clinicians and researchers. Many clinician or researcher participants were also multiples themselves or parents of multiples, which brought unique and valuable contributions to the process. The questions identified by the project include clinical topics such as the improvement of multiple pregnancy outcomes through training and medical interventions, maternal complications in multiple pregnancy and the growth trajectories of multiples. There were also some key topics of a psychosocial nature, such as parental postnatal mental health, community support for families of multiples and the development of multiples. The Global Twins and Multiples PSP was unique in that the questions involved addressed the health and wellbeing of multiples specifically. Therefore, the range of topics extended from clinical to psychosocial domains. A few topics overlapped with previous PSPs within the antenatal period. The role of antenatal care interventions in pregnancy outcomes has also been prioritized in the stillbirth and preterm birth PSPs. This highlights the persisting gaps in antenatal interventions for complex pregnancies overall. Moreover, whilst the stillbirth PSP prioritized the question of whether modifiable lifestyle factors contribute to stillbirth risk13, the multiples PSP prioritized the question of what prenatal factors benefit parental and child outcomes. Thus, whilst there has been some overlap between pregnancy-related PSPs, the present PSP supports research specifically for the benefit of multiples and with a more holistic focus on health. This study encompassed a broader scope of research areas evaluated and participants than found typically in studies involving questionnaires. This is likely to provide a more balanced representation of research topics concerning the health of multiples. The high engagement of multiples is a marker of the project's success as a participatory PSP. Combined with the use of existing evidence, we can be confident that this degree of stakeholder engagement produced a credible outcome. Despite its strengths, the limitations of the PSP approach should also be discussed. Given that the participant group was not equally weighted between multiples/parents of multiples and clinicians/researchers, potential bias cannot be excluded. However, when prioritizing questions, we attempted to mitigate this bias by using aggregate ranking to weight evenly contributions from multiples and parents of multiples, and clinicians and researchers. In addition, due to the international nature of the project, not all steering-group members were able to attend the final workshop, which may have contributed to bias at the latter stage of the project. Due to the participatory and collaborative nature of PSPs, the list of prioritized questions is broad. Each question was formed to capture the meaning intended by the participants and remain accessible not only to clinicians and researchers, but also to the community. Thus, researchers will be required to refine and format the questions prior to conducting research. We sought global representation of parents of multiples, multiples and clinicians in this PSP; however, the two surveys were online and in English. Therefore, this process excluded participants who did not understand English or did not have internet access, and may have excluded those living in remote areas. One way of addressing this issue would be to work with community organizations to hold locally based focus groups to prioritize the indicative questions. PSPs generally require a high standard of evidence for a research question to be considered ‘answered’ (for example, evidence from systematic reviews or randomized controlled trials). Therefore, there may be some evidence available for some of the prioritized questions, but this evidence was not deemed sufficient by the experts verifying these questions during the priority-setting process. Moreover, some of the suggested questions were of more relevance to translation, reflecting the need for researchers to communicate existing evidence more effectively to stakeholder groups, rather than indicating that researchers need to conduct more research in these areas. The results of this PSP clearly point to the top priorities for further quantitative research in clinical care and psychosocial health for multiples and their families. This includes care before, during and beyond pregnancy for the short- and long-term wellbeing of multiples and their parents. Although only 10 questions were prioritized in the final list, we acknowledge that many of the other verified 89 questions (Appendix S2) were closely ranked and are also of high importance for improving the health of multiples. We believe that these priorities, derived from a collaborative, participatory and multidisciplinary approach, provide credible direction to those driving forward the research for the health of multiples, their families and communities around the world. We thank the steering-group members for dedicating their expertise and time towards this project. We would also like to acknowledge Joel Marsden and Helen Regan (parent participants) for their contribution to the process undertaken at the final workshop. We also thank Lucas Ferreira at Twins Research Australia for contributing towards the establishment of the PSP and Jess Tyler at Twins Research Australia for her generous technical assistance with the survey. The PSP was funded by TAMBA UK and Twins Research Australia. Twins Research Australia is supported by the National Health and Medical Research Council through a Centre of Research Excellence Grant (ID: 1079102). H. Bartley, London, UK A. Baschat, Johns Hopkins Center for Fetal Therapy, Department of Gynecology & Obstetrics, Johns Hopkins University School of Medicine, Baltimore, MD, Centre Fetal St George's University of London, London, UK Research Australia Twins Research Australia, Centre for and Melbourne School of Population and Global The University of Australia Multiple Births University, UK Institute for University College London, and Department of Care Health University of UK The Netherlands Twins and Multiple Births London, UK Multiple International Council of Multiple Multiples of Institute of Multiple Births University, UK H. It UK, London, UK University Medical Center University, A. of Obstetrics, Gynecology and and The University of Health The Fetal Hospital, TX, London, UK A. Fetal St George's University of London, London, UK University of London, UK Twins Research Australia, The University of Australia University Multiples International Council of Multiple Fetal St George's University of London, London, UK The Netherlands International Council of Multiple Twins and Multiple Births London, UK St George's University of London, London, UK Medical University of A. Multiple Australia London, UK London, UK Department of Medicine, The Royal Hospital, and The University of Australia H. University of The Netherlands of and and and University of UK and Hospital, UK University of Australia University Medical The Netherlands The is not for the or of any information by the Any than should be to the author for the