The top 10 research priorities in diabetes and pregnancy according to women, support networks and healthcare professionals (2024)

Novelty statement

• Women report a lack of consistent evidence‐based information to help them manage their diabetes in the period before to after pregnancy.

• The top 10 questions for research in diabetes and pregnancy according to women, their support networks and healthcare professionals were identified.

• Joint top priorities for research were diabetes technology and identifying the best test for diabetes in pregnancy.

• These questions will inform funders of research and researchers towards addressing areas of great need and impact.

2.1. Establishing the PSP

The PSP was overseen by a steering group representing key stakeholders (Supplementary TableS1) and was chaired by a senior JLA advisor to ensure transparency of the process, and fair and equal involvement of all members. The group agreed the scope (Table1) and was responsible for the completeness and appropriateness of the process, ensuring involvement of key stakeholder groups, approval of categorisation, grouping and phrasing of questions and interpretation of data. The protocol was prospectively published online at www.jla.nihr.ac.uk/priority‐setting‐partnerships/diabetes‐and‐pregnancy.

2.2. Initial survey—identifying questions

Women and their support networks (partners, families, friends and carers) with experience of pregnancy or planning pregnancy with any type of diabetes and healthcare professionals were invited via an open survey (26 June–15 November 2019) to suggest up to three questions they felt were important to answer. These could be any questions about the time before, during or after pregnancy with any type of diabetes. The scope was intentionally broad so that the submissions reflected public need.

The survey was available, in English, online and on paper. Targeted efforts to maximise responses, particularly from underrepresented groups, included direct approaches in diabetes and pregnancy clinics, outreach through relevant support groups, professional networks and conferences, diabetes, pregnancy and birth charities’ websites and communication channels and social media platforms. Concerted efforts were made to hear the voices of ethnic minorities working with organisations, support groups and community champions, which aim to address health inequalities. Representation across different ethnic minorities was monitored through broad groupings.

2.3. Categorisation and grouping

The submitted questions were organised using NVivo qualitative data analysis software (QSR International Pty Ltd. Version 12, 2018). Initial data cleaning was manually completed with any issues about the clinical aspects or interpretation of the submitted questions resolved with the steering group. The questions were analysed using content analysis with an initial stage of open coding of the question content, followed by the grouping of codes into categories.¹⁵ To retain the integrity of the initial submissions, some questions were mapped to two or more categories. Independent second checks were conducted with members of the steering group to ensure potential impact of individual bias, and missed or misinterpreted categorisation was minimised. The steering group further consolidated the categories into groups and summarised the initial survey submissions under an indicative question. Indicative questions were formulated to capture the issues raised by the submitted questions within each group, whether originating from single or multiple respondents.

2.4. Evidence checking

A broad level and pragmatic evidence checking strategy was taken (January – May 2020) with the aim of ascertaining whether there was evidence of substantial uncertainty for each indicative question. The search was restricted to the Cochrane Database of Systematic Reviews (www.cochranelibrary.com), systematic reviews published since 2017 using Medline or PubMed and National Institute of Clinical Excellence (NICE) and Scottish Intercollegiate Guidelines Network (SIGN) national diabetes and pregnancy guidelines.^{5, 6} Expanded evidence searches, including evidence highlighted by the steering group, were applied on a question‐by‐question basis after finalisation of the list of indicative questions. Research underway or recently completed but not available as published was not included as evidence. Where part of the question had sufficient evidence, the question phrasing was amended to reflect the remaining uncertainty.

2.5. Interim survey and prioritisation

The interim survey presented the long list of indicative questions in groups by phase. The order of the groups and individual questions within the groups were randomised each time the survey was entered. Participants were invited to pick up to 10 that they felt were most important to answer. Due to the Covid‐19 pandemic and social distancing restrictions, the survey was offered online only. Following a pilot mid‐May, the survey ran for nineweeks (29 May–31 July 2020).

2.6. Interim ranking and shortlisting

Every selection made by an individual respondent had equal weighting and no weighting changes were made if fewer than 10 questions were selected. However, to account for the differences in observed voting patterns and the number of respondents from different groups, ranking was tallied separately for: women and support networks, healthcare professionals, ethnic minorities and diabetes type, namely type 1, type 2 and other, and gestational diabetes. Within each of the groups, the total points for each question were put into rank order. The questions ranked in the top 10 for the two main groups (women/support networks and healthcare professionals), and the top three, and at least eight of the top 10, for each of the other subgroups, were shortlisted. In total, 18 questions were shortlisted for the final workshop, the maximum number considered feasible by the steering group for effective discussion online.

2.7. Final workshop—agreeing the top 10

The final stage involved a 1‐day workshop (2 October 2020) using the established JLA approach, which was adapted to be delivered online.¹⁴ Twenty‐five participants were identified initially through phased targeted approaches to prioritise representation from ethnic minority groups, the devolved nations and Crown dependencies, support networks and specific health professions, for example, psychologists and GPs as underrepresented groups, followed by open invitation. Contacts collated through the surveys, and special interest groups and partner communication channels were used. Participants were screened for possible conflicts of interest and whether they were highly research active in the area. The participants were split into four breakout groups balanced by representation between women, support networks and healthcare professionals, and by experience of diabetes and healthcare specialist.

In breakout groups, the attendees participated in a series of discussion and ranking exercises to jointly rank the shortlist of indicative questions and agree the top 10 most important for future research to answer. The workshop and discussions were facilitated by trained JLA advisors to ensure equal and open participation. Four steering group members joined as observers only. Technical support was made available, and a contact point for emotional support was provided should any participant be upset by the process or discussions. The participants were invited to provide anonymous feedback on the prioritised questions and the workshop generally.

2.8. Ethics

The Medical Sciences Interdepartmental Research Ethics Committee, University of Oxford confirmed that the project did not require ethics committee approval.

TABLE 2

Top 10 priorities for research in diabetes and pregnancy according to women, their support networks and healthcare professionals

TABLE 3

Participant demographics at each stage of the priority setting process

3.1. Initial survey

Four hundred and sixty‐six responses were submitted (64% women and support networks, 32% healthcare professionals and 4% other/not answered) suggesting 1161 questions covering the whole perinatal period (Supplementary TableS3).

Initial questions submitted by women and support networks were mainly in relation to post‐birth effects on themselves and their child, diabetes management during pregnancy and understanding the risks for diabetes in pregnancy. The long‐term effects of diabetes in pregnancy on the child (risks of the child developing diabetes and any wider health effects) being the most frequently asked question (20.1% of women and support networks’ submissions). This group more specifically raised questions about breastfeeding (8.9%) and labour and birth (8.6%) in terms of informed choice, continuity/availability of care and emotional support more generally. Healthcare professionals’ questions were mainly about pre‐pregnancy care, and diagnosis and clinical management of diabetes in pregnancy. How to improve preconception care was most frequently asked (9.7% of healthcare professionals’ submissions), closely followed by the value and methods of diagnosis and management of diabetes late in pregnancy, that is, after 34weeks (8.5%). Modes of delivering care, improving uptake and access to services and motivational interventions were more specifically raised by this group. Common to both groups were questions about individualised and risk‐based care, optimal management of diabetes, prevention of diabetes and safety of medications.

One hundred and forty‐two categories were extracted and broadly organised by the phase of pregnancy: pre‐pregnancy (62 questions, 6.3%), pregnancy (376, 38.2%), labour and birth (87, 8.8%) and post‐birth (373, 37.9%). Technology (20, 2.0%), mental health and well‐being (20, 2.0%) and health services (46, 4.7%) were identified as cross‐cutting categories. A total of 934 questions were within scope, of which 50 mapped to more than one category, and consolidated into 60 indicative questions. Rarely was there a need to specify a type of diabetes within an indicative question, which reflects the significant overlap in priorities regardless of diabetes type. The main distinctions were questions relating to gestational diabetes, due to its transient nature and diagnosis in pregnancy. All 60 indicative questions were considered to have substantial uncertainty following evidence checks. The evidence check summary is provided in Supplementary TableS4.

3.2. Interim survey

Six hundred and fourteen submissions (80% women and support networks and 20% healthcare professionals) were received in the interim survey. In the interim survey rankings, there were notable differences between women and support networks, and healthcare professionals (Figure2a). Four of the top 10 ranked questions for healthcare professionals were below the 45th ranking for women and support networks. Women and support networks ranked the long‐term effects of diabetes in pregnancy on the child's general health (non‐diabetes‐related) highest. Varying standards and advice across hospitals and giving birth at full term were also in the top three.

Voting patterns varied for the main groups between surveys (Figure3) with overall movement towards labour and birth and cross‐cutting categories. For example, despite being rarely asked in the initial survey, the use of technology became the highest ranked for healthcare professionals in the interim survey.

For ethnic minorities, representation was below national population figures for Black and Black British groups in both surveys (1.6% and 0.5% respectively; national 3.0%), and for Asian and Asian British groups (4.1%; national 7.0%) in the interim survey. With low numbers, the votes across 60 questions became too dispersed to discern a strong pattern. However, peaks in voting overlapped the top 10 for women and support networks, and healthcare professionals, with the eight highest voted questions already shortlisted for the final workshop.

Comparing the interim survey rankings by types of diabetes, there were some clear differences (Figure2b). Technology was ranked the top priority for type 1 and type 2/other diabetes groups, but ranked low for the gestational diabetes group. Post‐birth support came in second for type 1, but ranked low for both type 2/other and gestational diabetes groups, whereas testing for diabetes during pregnancy and prevention of developing diabetes ranked high for type 2/other and gestational diabetes groups but low for the type 1 diabetes group.

3.3. Final workshop and top 10

The diversity and balance in experiences and expertise of workshop attendees were strong (Table3). Participants uncovered unexpected overlaps between questions, for example, technology for diabetes care was considered to include telemedicine which was subsequently ranked lower; labour and birth choices was considered to overlap with safety of giving birth at full term and the need for induction. Groups also highlighted recurring themes that linked with multiple questions. For example, technology was considered to improve understanding of diabetes management and reduce burden, linking with mental health and well‐being. Therefore, with further such links highlighted, both these questions were more highly ranked as the discussions progressed.

Consistently, top ranking questions through the surveys were regarding the long‐term health impact of maternal diabetes on the child. The questions submitted in the initial survey were clearly split into risk of the child developing diabetes or the risk of wider health conditions. Ranking in the top 10 of all groups, both questions were shortlisted for the final workshop. However, neither reached the final top 10 (positions 11 and 12 respectively). Workshop discussions and post‐workshop feedback from participants indicated several reasons for this apparent discrepancy. Firstly, the questions were considered to be addressed within the broader research agenda of child health, and not necessarily within pregnancy or women's health. The other questions all affected or contributed to the child outcomes, so it was felt they would help address these questions too. Some participants considered that the answers to the questions about long‐term child outcomes would add burden to women already concerned by many factors associated with dealing with the responsibility of a pregnancy and diabetes. For some, ‘how can it be prevented’ was important wording, as it balanced the concerns of adding burden of information on risks, with learning about what can be done to prevent these being realised. One of the questions did not include the specific wording about prevention and so was ranked less highly. A further possible reason for these questions receiving less priority was because the votes were split; if presented as a single question, this may have meant a higher final ranking. A final possible reason may have been that the workshop consisted of people with recent experience, that is, within the last fiveyears.

Feedback was received from 10 women and support network representatives, and 10 healthcare professionals from the workshop:

‘It was really good having different perspectives. Moving to a second smaller group was also useful, as it showed how varied priorities can be between 2 groups, despite having a similar mix of people from the different backgrounds’

‘Although the final top 10 was not the same as my personal top 10 that I had prepared, discussing all of the 18 questions, and reflecting on them as a group, meant I was satisfied with it. It was good to hear the views of women with different forms of diabetes and look at the questions from their perspective’.

Further excerpts of the feedback and the JLA’s review of the online format of the workshop are available at www.jla.nihr.ac.uk/development‐of‐online‐priority‐setting‐workshop.htm.

4.1. Main findings

We supported women, their support networks and healthcare professionals to jointly identify the top 10 questions for research in diabetes and pregnancy. The final list includes priorities of relevance to all the stakeholders, all types of diabetes and preconception to long‐term post‐pregnancy. These questions were generated through a robust and inclusive process, which can be trusted and used by funders of research and researchers to inform their research activities towards addressing evidence gaps of great need and impact.

There is consistency across many areas also identified by the modified prioritisation exercises in Canada and the USA focussing on gestational diabetes, such as screening, risk factors, prevention and clinical management of diabetes in the woman.^{10, 11} For example, whether there is a better test for diabetes during pregnancy than the oral glucose tolerance test was ranked joint first for similar reasons of application, interpretation and practicality, with no apparent consensus on the thresholds, methods and timing to test for diabetes in pregnancy, due to small‐sized studies. Our participants raised further questions such as the possible use of testing to predict gestational diabetes, the stratification of testing by risk and testing within special populations, such as those who have undergone gastric bypass surgery, and open up further opportunities for research.

There are also some clear differences in priorities including preconception care, late pregnancy diagnosis, labour and birth experiences and post‐natal care needs, particularly breastfeeding. The role of diabetes technology was ranked joint first position, and while explicably not featuring in the gestational diabetes exercises, technology‐specific questions formed the top three of the JDRF UK type 1 diabetes PSP final top 10. There have been significant developments in diabetes technology in the last few years with ongoing innovations.¹⁶ In the time since the final workshop, the National Institute of Clinical Excellence has completed their review and approved the funding of continuous glucose monitoring for pregnant women with type 1 diabetes, in line with The NHS Long Term Plan.^{17, 18} However, our results endorse their recommendations for further large‐scale research in different monitoring methods and systems (i.e. insulin pumps) in women with diabetes of different types, through all stages preconception to post‐pregnancy, and importantly with wider population diversity. There are also wider facets to consider such as apps, automation, data integration platforms and data sharing in supporting the management of diabetes.

Risks to the child's health was the number one priority identified by women in the Canadian exercise, and ranked highest in the USA exercise, but strikingly fell outside our top 10. However, the possible reasons for this, based on differential prioritisation of two related summary questions, made clear that the impact on the child is still considered highly important for separate study.

Important overarching issues were noted around continuity of care and support (particularly post‐birth), consistency in care standards and advice, joint decision making (particularly in labour and birth) and the burden for women of being diagnosed with and managing diabetes. The consequences on women's well‐being and mental health were at the core of much of the discussion at the final workshop. The initial survey submissions also raised further unanswered questions about the impact of a ‘medicalised’ pregnancy, withdrawal of intensive clinical involvement postpartum and support needs of women if adverse outcomes in their child linked to diabetes are realised. There is ongoing lack of evidence on the support pregnant women with diabetes may need as previously echoed by the gestational diabetes prioritisation exercises. However, the Diabetes UK Too Often Missing report highlights pregnancy as a time of particular high risk for emotional and psychosocial impact of diabetes requiring increased awareness and support.¹⁹

Table S1

Table S2

Table S3

Table S4

1. Feig DS, Hwee J, Shah BR, Booth GL, Bierman AS, Lipscombe LL. Trends in incidence of diabetes in pregnancy and serious perinatal outcomes: a large, population‐based study in Ontario, Canada, 1996–2010. Diabetes Care. 2014;37(6):1590‐1596. [PubMed] [Google Scholar]

2. Davies SC. Annual Report of the Chief Medical Officer, 2014, The Health of the 51%: Women. London: Department of Health; 2015. [Google Scholar]

3. Lowe WL Jr, Scholtens DM, Lowe LP, et al; HAPO Follow‐up Study Cooperative Research Group. Association of gestational diabetes with maternal disorders of glucose metabolism and childhood adiposity. JAMA. 2018;320(10):1005‐1016. [PMC free article] [PubMed] [Google Scholar]

4. Scholtens DM, Kuang A, Lowe LP, et al; HAPO Follow‐up Study Cooperative Research Group; HAPO Follow‐Up Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcome follow‐up study (HAPO FUS): maternal glycemia and childhood glucose metabolism. Diabetes Care. 2019;42(3):381‐392. [PMC free article] [PubMed] [Google Scholar]

5. National Institute for Health and Care Excellence. Diabetes in pregnancy: management from preconception to the postnatal period (NG3). London: NICE; 2015. [cited 2020 Nov 27]. (Clinical guideline NG3). Available at: https://www.nice.org.uk/guidance/ng3. [PubMed] [Google Scholar]

6. Scottish Intercollegiate Guidelines Network. Management of diabetes: A national clinical guideline (116). Edinburgh: SIGN; 2017. [cited 2020 Nov 27]. Clinical guideline 116. Available at: https://www.sign.ac.uk/assets/sign116.pdf. [Google Scholar]

7. Crowe S, Fenton M, Hall M, Cowan K, Chalmers I. Patient’s clinicians’ and the research communities’ priorities for treatment research: there is an important mismatch [published correction appears. Res Involv Engagem. 2015;1:2. [PMC free article] [PubMed] [Google Scholar]

8. Gadsby R, Snow R, Daly AC, et al. Setting research priorities for Type 1 diabetes. Diabet Med. 2012;29(10):1321‐1326. [PubMed] [Google Scholar]

9. Finer S, Robb P, Cowan K, Daly A, Shah K, Farmer A. Setting the top 10 research priorities to improve the health of people with Type 2 diabetes: a Diabetes UK‐James Lind Alliance Priority Setting Partnership. Diabet Med. 2018;35:862‐870. [PMC free article] [PubMed] [Google Scholar]

10. Rees SE, Chadha R, Donovan LE, et al. Engaging patients and clinicians in establishing research priorities for gestational diabetes mellitus. Can J Diabetes. 2017;41(2):156‐163. [PubMed] [Google Scholar]

11. Bennett WL, Robinson KA, Saldanha IJ, Wilson LM, Nicholson WK. High priority research needs for gestational diabetes mellitus. J Women's Health. 2012;925‐932. [PMC free article] [PubMed] [Google Scholar]

12. Boddy K, Cowan K, Gibson A, Britten N. Does funded research reflect the priorities of people living with type 1 diabetes? A secondary analysis of research questions. BMJ Open. 2017;7:e016540. [PMC free article] [PubMed] [Google Scholar]

15. Hsieh H‐F, Shannon SE. Three approaches to qualitative content analysis. Qual Health Res. 2005;15(9):1277‐1288. [PubMed] [Google Scholar]

16. Beck RW, Bergenstal RM, Laffel LM, Pickup JC. Advances in technology for management of type 1 diabetes. The Lancet. 2019;394(10205):1265‐1273. [PubMed] [Google Scholar]

21. van der Zande ISE, van der Graaf R, Oudijk MA, van Dleden JJM. Vulnerability of pregnant women in clinical research. J Med Ethics. 2017;43:657‐663. [PubMed] [Google Scholar]