Exploring women’s priorities for the potential consequences of a gestational diabetes diagnosis: A pilot community jury

Abstract Background There is no international diagnostic agreement for gestational diabetes mellitus (GDM). In 2014, Australia adopted a new definition and testing procedure. Since then, significantly more women have been diagnosed with GDM but with little difference in health outcomes. We explored the priorities and preferences of women potentially impacted by a GDM diagnosis. Method We recruited 15 women from the Gold Coast, Australia, to participate in a pilot community jury (CJ). Over two days, the women deliberated on the following: (a) which important consequences of a diagnosis of GDM should be considered when defining GDM?; (b) what should Australian health practitioners call the condition known as GDM? Results Eight women attended the pilot CJ, and their recommendations were a consensus. Women were surprised that the level of risk for physical harms was low but emotional harms were high. The final ranking of important consequences (high to low) was as follows: women's negative emotions; management burden of GDM; overmedicalized pregnancy; minimizing infant risks; improving lifestyle; and macrosomia. To describe the four different clinical states of GDM, the women chose three different labels. One was GDM. Conclusions The women from this pilot CJ prioritized the consequences of a diagnosis of GDM differently from clinicians. The current glucose threshold for GDM in Australia is set at a cut‐point for adverse risks including macrosomia and neonatal hyperinsulinaemia. Definitions and guideline panels often fail to ask the affected public about their values and preferences. Community voices impacted by health policies should be embedded in the decision‐making process.


| INTRODUC TI ON
Gestational diabetes mellitus (GDM) is a health condition diagnosed in pregnancy and associated with an increased risk of physical complications for the mother (eg caesarean section, pre-eclampsia) and infant (eg macrosomia, shoulder dystocia). 1 Women diagnosed with GDM have reported emotional and financial consequences such as self-blame and guilt, 2,3 confusion over diet management 4 and the expense of healthy eating. 5 Other women reported the diagnosis as an opportunity to change their behaviours. 6 The definition of GDM has a complex history. Definitions have used different theoretical premises (percentile range like most laboratory tests, the same values as in the non-pregnant or risk based).
Where a risk-based assessment has been used, over time the focus has shifted. For example, in 1964, O'Sullivan and Mahan 7 originally defined it as a way of identifying those women at risk of developing type 2 diabetes, whereas a large international study known as the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study 1 in 2008 focused on foetal outcomes. In addition, the diagnostic test has varied 8 from a 50 g glucose load 1-hour screening test, a 75 g 2-hour test and a 100 g 3-hour test. Recommendations were initially for screening only 'high-risk' pregnant women, but now include all women and diagnostic cut-points have varied over several decades. 9 The varying cut-points arise because the risk of complications of gestational diabetes is on a continuum. The higher the mother's blood glucose level (BGL), the higher the risk of complications to her and her baby. The HAPO study 1 aimed to find a cut-point in the continuum to guide decisions on where the BGL should be drawn to define GDM. Unfortunately, an obvious cut-point was not identified.
Using the available HAPO data, the International Association of the Diabetes and Pregnancy Study Groups (IADSPG) consensus panel recommended a BGL threshold associated with the risk of adverse infant outcomes (such as risk of macrosomia, excess infant adiposity and neonatal hyperinsulinaemia). 8 However, this change was controversial, and there is a lack of international consensus about the appropriate threshold between normal and elevated blood glucose.
Two recent Australian studies 10 and babies. The Cade study 11 also reported a substantial increase in health-care costs: a net cost increase for their hospital alone was over A$560,000 annually. Importantly, many (but not all, eg the Royal Australian College of General Practitioners) medical colleges in Australia adopted the IADSPG new threshold and testing regimen.
Given the absence of consensus and the substantial increase in GDM diagnoses, we believe it is timely to explore the values and preferences of the women who may be directly impacted by the change in GDM diagnosis.
We conducted a citizen/community jury (CJ) to explore women's perspectives. A CJ is a form of deliberative democracy and aims to elicit an informed community perspective on important and often controversial topics. CJ participants are provided with expert presentations and opportunities to question the experts, engage in both facilitated and private deliberation and are asked to form a consensus or majority 'verdict' on the topic question. Community jury participants deliberate on questions requiring a decision that is both informed and ethically sensitive. CJs have been used successfully in research to elicit informed perspectives for several health policy issues, for example screening mammography, 12 screening for prostate cancer, 13 case finding for dementia, 14 quantifying health preferences 15 and more broadly in local governments. 16

| Steering committee
We established a steering committee to plan both the community jury and a consensus panel meeting to discuss the definition of GDM in Australia. The steering committee included a general practitioner, two endocrinologists, a consumer representative and two researchers with CJ and qualitative experience. This committee determined the eligibility criteria of the CJ participants, drafted the CJ questions and considered the expert information necessary for the jurors to make recommendations.

| Participants
We recruited women from the Gold Coast region (Australia) and

| Procedure
The CJ was conducted over two weekend days, 2 and 3 February 2019, at Bond University (see Table 1 for schedule). All sessions, except for the private deliberation on Sunday, were facilitated by RT (a psychologist and experienced facilitator), who moderated discussions and ensured all jurors had equal opportunity to voice opinions. Two researchers, AMS, RS and an observer from Therapeutic Guidelines, LAC, observed CJ processes and jurors' engagement for the sessions except the private deliberation. Only CJ participants were present during private deliberation to maintain confidentiality and prevent bias. Participants provided written consent prior to participating.
Four presentations were conducted on Saturday (via voice-over PowerPoint) and followed with a telephone question and answer session with each presenter (descriptions of topics and experts are provided in Box 1). After the presentations, the CJ participants completed two activities (see below) to assist in answering the two jury questions.
On Sunday, CJ participants reconvened. Following a discussion about reflections from Saturday and overnight, participants were offered the opportunity to further question presenters for additional information and clarification. CJ participants elected a representative to facilitate private deliberation and deliberated on the two CJ questions until consensus or impasse was reached. The CJ recommendations were presented to the Chair of the Steering Committee Paul Glasziou and to the researchers and observers. At the conclusion of the CJ, participants were provided opportunity to debrief on CJ processes and outcomes.

| Presentations
Four experts provided presentations to the CJ participants.
Presenters, their topics, a short biography and their presentation URLs are provided in Box 1. Participants were provided with biographies of the presenters and copies of PowerPoint presentations. No reimbursement (financial or otherwise) was provided to the experts.

| Activity 1: Ranking the consequences of GDM
The Steering Committee identified physical and emotional consequences often reported in GDM as important for the CJ participants. To quantify the physical consequences (eg pre-term birth, caesarean section, macrosomia, shoulder dystocia), we used risk ratios identified by McIntyre et al 21 to construct icon arrays for CJ participants. We conducted a systematic review of qualitative data 22 to identify reports of emotional consequences. Twelve identified consequences were described on A4 size laminated paper and were presented and explained to the CJ participants. A further four blank consequence pages were also provided. A smaller version of each consequence and its description is provided in Supplementary File S1.
On Day 1, CJ participants individually ranked the consequences on a worksheet. The group then discussed their rankings and their reasons which were documented on a whiteboard. On day 2, the participants discussed their rankings again and deliberated in private.  The descriptions were as follows:

TA B L E 1 Schedule of events
A Katie: developed diabetes as a result of pregnancy (ie this woman now met the criteria for non-pregnant diabetes); B Jenny: has higher than usual blood sugar levels as a result of the pregnancy and is at increased risk of complications; C Emily: has higher than usual blood sugar levels as a result of the pregnancy and is at normal risk of complications; and D Sofia: had diabetes before pregnancy and still has diabetes in pregnancy.
The risk of complications was deliberately varied in scenarios B and C to examine its impact on the label. See Supplementary File S2 for clinical descriptions provided to the CJ participants.
Participants were given scenarios and a list of the potential labels with two blank options describing the four different clinical states.
The descriptors and the labels were developed in consultation with the Steering Committee, GPs and research colleagues.
On day 1, CJ participants were asked to identify their preferred label for each of the scenarios A-D followed by a discussion about reasons. CJ participants were asked to individually post each label to a laminated vignette which had been arranged on a wall. All labels had to be used, either with a vignette description or in a personal 'discarded' pile. Participants could also develop alternative labels.
On day 2, CJ participants deliberated in private.  The women did not identify any additional consequences to include than those synthesized from the available evidence.

Potential reasons from reflections and deliberations
At the start of day 1, the women had individually ranked the potential consequences of GDM on day 1, most women ( The women also articulated the distinction between the instinctive and emotional and more deliberative thinking processes.

P4: Well our first reaction's always an emotional one,
isn't it; we don't automatically look for numbers, we automatically react emotionally.

(day 2 page 133)
The women recognized that because their views were based on evidence presented by the experts on day 1, they would likely differ from those unfamiliar with the evidence.

| Question 2
What should Australian health practitioners call the health condition currently known as GDM?

CJ recommendation
The women's final recommendations and their reasoning for what to label the four different clinical states of GDM is described in Box 3 from most preferred to least preferred for each descriptor. The women did not identify any other potential labels for GDM than those suggested and provided by the Steering Committee and clinical and research colleagues. (day 2 page 120)

Potential reasons from reflections and deliberations
The two scenarios with higher than usual blood sugar but different risks of complications (scenarios B and C) had the most discussion and the most concern about unnecessary harms. For example, when discussing the scenario depicting raised blood sugar with normal risk (scenario C), the women deliberated:

| D ISCUSS I ON
While the participants in our pilot community jury initially agreed that the most important consequence of a GDM diagnosis (rated as highest priority) was the 'opportunity to minimize the risks to the unborn baby', after reviewing the level of risks and upon reflection, the women changed their opinions and countered the 'knee jerk' (day 2 pg 133) reaction they believe they had on day 1. people. 23 However, we had fewer participants than we anticipated, and the findings should be interpreted cautiously. We cannot claim that women's views from this CJ represent broader community views. Because we wanted to recruit the population most affected by the question of GDM definition, we recruited mothers, who are challenging to recruit as they experience many competing demands on their time. When contacted, many women could not attend the required 2 days, and some had sick children during the CJ weekend.
Recruitment limitations from this study serve as key learnings, and future CJs on GDM will be adapted to suit the demands of young families. For example, for young mothers, it may be more prudent to conduct CJs on weekdays when childcare may be more accessible or conduct the CJ over multiple days/weeks (1 day a week for 2 weeks). Despite recruitment challenges, the women who did attend produced consensus recommendations and transcripts reflecting cohesive group discussion and no dissention. However, we upheld robust CJ methods. We recruited from the population we considered most affected by the question, and the jury verdict was unanimous.

ACK N OWLED G EM ENTS
The authors would like to thank Professor David McIntyre for his contribution to the steering committee, expert presentation to the CJ participants and his availability and willingness to answer their questions. We also thank the commitment of the women who gave up the weekend with their families, worked tirelessly to provide us with recommendations to our questions and humbled us with their knowledge, insight and wisdom. Funding for this project was provided by

AUTH O R CO NTR I B UTI O N S
PG and RT conceived the project design. All authors provided project assistance in preparation and conduct. AMS and RT extracted, analysed and interpreted the data. All authors contributed to the drafting of the manuscript and approve the final version.

S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the Supporting Information section.
How to cite this article: Thomas