It is recognised that the prevalence of gestational diabetes mellitus (GDM) varies according to different populations, but the common trend across the world is rising levels of this condition. Such increases are fast becoming a major health problem [1
] and the trend is thought to be partly explained by the advancing of maternal age and increasing rates of obesity [4
]. In addition, prevalence rates of GDM are higher in ethnic minority populations and, therefore, there is a well-recognised positive association between increasing racial/cultural diversity in Western populations and rates of GDM [5
]. Basildon and Thurrock University Hospital is a University Hospital in the East of England region of the United Kingdom, with an annual delivery rate of approximately 4500 deliveries a year. Routinely collated annual service data at this hospital had demonstrated a 200% increase in the prevalence of diabetes in pregnancy between 2004 and 2013, predominantly due to gestational diabetes mellitus (GDM). Members of the specialist maternity team had increasingly recognised the adverse effect of this increased workload on care quality and in August 2013, we received our first ever written complaint relating to clinic waiting time. A significant sentinel “event” was the increase in the number of diabetes-related stillbirths, which peaked at three deaths over a four-month period (July to October 2013). The root cause analysis investigation reports highlighted a recurring theme of suboptimal clinical decisions in overburdened specialist clinics.
Once concerns were formally raised, a multidisciplinary scoping team was set up to assess the scale of the problem and examine for possible contributing factors, with a view to finding a way forward. This team consisted of a consultant obstetrician, a consultant diabetologist, a specialist diabetes midwife and a specialist diabetes nurse. For a better understanding of the scale of our caseload challenge, the team reviewed the routinely collated maternity diabetes service data for 2013 and the 2013 Public Health England Local Authority Health Profile report for our catchment population. In the UK, Public Health England was an executive agency of the Department of Health and Social Care in England tasked to protect and improve health and wellbeing and reduce health inequalities. This agency was responsible for the production of Local Authority Health Profiles reports that provide an overview of health for each local district in England. The reports contain data on a range of indicators for local populations, highlighting issues that can affect health in each locality. The profiles are intended to help local government and health services make plans to improve the health of their local population and reduce health inequalities.
The team review of our local data came up with some stark findings. The first significant revelation was that the year had marked a major inflection point in our prevalence of diabetes in pregnancy, rising from 67.3 per 1000 deliveries in 2012 to 91.6 per 1000 in 2013.
The second noteworthy finding was that national census data (2001–2011) had demonstrated a three-fold increase in the proportion of ethnic minority residents locally [6
]. In our catchment population of Thurrock, the proportion of ethnic minority residents (Asian, African, Caribbean, Mixed and Other) had risen from 6.1% in 2001 to 18.4% in 2011. This rate of increase was faster than the national average and the trend was predicted to continue. The trend held significance, as the younger child-bearing generation was clearly more ethnically diverse compared to their parents and the data showed that many of the ethnic minority groups were displaying a higher prevalence of obesity (21.5–29.4%) than the national average (19.8%). The variation in obesity prevalence extended across all age groups, including children [7
] (p. 7).
The final concern was an ongoing rise in the overall rate of adult obesity. This trend dated back to 2006 and culminated as worse than the national average in the 2013 health profile reports for both the Thurrock and Basildon districts we serve [8
]. They reported that 28.1% and 26.7% of adults in Thurrock and Basildon, respectively, were obese compared to the national average of 24.2%. This trend remained evident and was highlighted in Thurrock Council’s Joint Strategic Needs Assessment report on “Whole Systems Obesity” in 2017 [7
]. It showed that between 2013 and 2015, the problem of obesity remained alarmingly high in Thurrock, with 69.4% of adults registered as overweight or obese, compared to the England average of 61.3%. Sadly, by 2019, there was still no sign of abatement, with the 2019 health profile reports showing that 72.8% and 69% of adults in Basildon and Thurrock respectively, were overweight or obese compared to a national average of 62% [10
Due to the emerging picture, further increases in the caseload of our diabetes antenatal service were strongly predicted. UK national guidelines [12
] state that all diabetic patients in pregnancy should be routinely reviewed in a specialist diabetes antenatal clinic. However, due to the already overwhelming workload, not to speak of even further anticipated increases due to the local demographic trends, the team realised a crucial need to adapt the service to meet the needs of the increasingly diverse community it served. We, therefore, set ourselves the pertinent challenge of an unfunded remodelling of care pathways to safely deal with increasing caseloads. At a time of ever-widening gaps between hospital income and expenditure, we knew we faced a significant hurdle but also realised that not addressing the issue would potentially be detrimental to long-term patient safety and experience as well as the reputation of our maternity unit. This fuelled a decision to embrace innovation and apply quality improvement (QI) methodology in trying to find a solution.
2. Materials and Methods
The primary objective was to reduce the rate of overbooked specialist diabetes antenatal clinics and the rate of delays in first clinic appointments following a diagnosis of GDM. The second objective was to ensure that any implemented interventions did not result in the worsening of relevant clinical outcomes that were routinely collated as part of the hospital annual service data. These clinical outcomes were therefore regarded as balancing measures, whereby the monitoring of these outcomes would help to ensure that our interventions did not have an unacceptable safety profile. Irrespective of achieving the primary objective, an associated negative trend in any of the balancing measures was judged to be counterproductive. It was therefore agreed from the outset that without positive or static trends in the balancing measures, the project could not be deemed a success.
Following the implementation of our planned interventions, we applied plan, do, study, act (PDSA) cycles [13
] as our QI methodology tool for formative evaluation, and this enabled prospective adjustments during the project time frame.
The definitive final summative evaluation was done in November 2019. The setting for this evaluation was our maternity unit, with an outreach that included community midwifery clinics. The subjects were patients with a working diagnosis of diabetes in pregnancy based on the 2010 International Association of Diabetes and Pregnancy Study Groups (IADPSG)/World Health Organisation (WHO) 2013 [14
] diagnostic criteria between January 2014 and December 2018. The evaluation involved annual reviews of routinely collected data and consisted of 2970 diabetic pregnancy outcomes and corresponding outpatient clinic logs.
The team gleaned baseline information by analysing routinely collated maternity diabetes service data and selected the problematic indicators in the 2013 dataset to constitute a list of chosen measures for monitoring as part of the evaluation process.
The selected process measures were the percentage of months every 6 months when specialist clinic capacity (90 clinic slots per month) was exceeded, and the rate of first clinic attendance within 7 days of GDM diagnosis. The selected clinical outcome measures (balancing measures) were the live birth rate in Type 1 and 2 diabetics, big baby (birth weight > 4 kg) and caesarean section rates in GDM cases.
At the outset, the listed process measures were reviewed every 3–6 months and for the balancing measures, this was done yearly unless earlier reviews were warranted based on safety incident reports. All results were plotted sequentially to aid the visualisation of trends and reported in shared annual service reports, each detailing the range of the plan, do, study, act (PDSA) activities for the year.
For the patient feedback survey, we used a locally designed questionnaire (Supplementary Materials
) that included closed satisfaction statements, and response choices were based on a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). This type of scale was used because it is easily understood by respondents, is useful for gauging opinions and feelings, and offers the option of negatively worded responses. The questionnaire also provided an optional section for open-ended comments. When it came to analysing the questionnaire responses, the main focus was on statements 2–5, which dealt with the four main educational objectives, including psychological adjustment. As this was not a validated questionnaire, a total score was not created for each respondent, and the results for each question were analysed separately.
2.3. Intervention Strategy
This consisted of 3 work streams that ran in parallel throughout the duration of the project.
Remodelling and implementation of a new care pathway:
The first step was a series of brainstorming meetings of the multidisciplinary team members, and this conceived the idea of the General ownership of Diabetes (GooD) pregnancy network, an innovative initiative that emphasises the contemporary necessity to equip and empower all maternity stakeholders to deliver the basic management requirements of gestational diabetes. Contrary to the conventional limitation of care to specialist diabetes antenatal clinics, it advocates setting up a service hub of midwife-led tele-consultation clinics and the strategic linkage of these to group clinics (Educating Gestational diabetics Group Sessions—EGGS) [16
], specialist clinics and standard antenatal clinics, thereby creating a shared care network. We then took a bottom-up approach by informally engaging with both staff (including community midwives and line managers) and patients about this idea, towards the end of 2013. The majority were positively receptive to the proposal, and these discussions informed the final decision to implement this new care framework in January 2014. After the initial set up, a pilot questionnaire-based feedback report from 17 service users (January–March 2014) showed 100% preferred the option of tele-consultation to physical attendance. An overriding theme of greater patient convenience based on time off work, childcare arrangements, and parking costs became evident. From March 2014 onwards, we proceeded to embed the new pathway and outlined a five-year timeframe, during which we aimed to hopefully demonstrate longer-term benefits and sustainability of the new initiative.
Staff engagement and behavioural change:
Initial consultations with general staff members of the maternity workforce highlighted a need to widen consciousness and clinical knowledge of GDM to enable a positive impact of the new shared care pathway. This proved to be our biggest challenge, especially as management had declined our request to incorporate this domain of care into the mandatory training programme. We coped by grasping “windows of educational opportunity”, such as a rotational attendance of specialist clinics (including EGGS) by non-specialist staff, remote support for non-specialist staff by the specialist team, selecting and training diabetes champions in all clinical areas, holding weekly multidisciplinary team meetings with an open invitation to non-specialist staff members, launching a departmental poster campaign reminding all staff to routinely enquire about glycaemic control when in contact with diabetic women and to escalate when necessary, general circulation of diabetes annual reports, and the development and implementation of home-grown decision support aids. The Diabetes-Related Risk in Pregnancy (DRRiP) score is our simple checklist tool that converts the evaluation of discrete clinical parameters into a single composite score, thus creating a common language between patients, midwives, and doctors. Diabetes and latest delivery date (DIALD) is our decision aid that stipulates the timing of delivery according to the emerging clinical picture.
Collaboration with patients:
Public health data shows that after delivery, women from Basildon and Thurrock face a greater risk of cardiometabolic complications and lower life expectancy. For this reason, we decided to make public engagement a crucial component of this initiative and made a commitment to Patients and Public Involvement and Engagement (PPIE) [20
]. After the pilot survey confirmed a collective endorsement of the new tele-consultation service, all GDM patients were routinely offered participation in feedback surveys, revealing the consequential extent of empowerment and psychological adjustment following the group sessions. Furthermore, patients joined to co-produce a series of 5 newspaper articles since 2015, serving up their stories as case studies promoting “general ownership”. Meaningful public engagement was also evident in 10,882 views of our 2016 social media post and, as a direct consequence, the EGGS group clinic “faculty” has since included a former GDM patient. Her involvement has enabled the group clinics to always end on an encouraging note by showing a short film of her testimony. Despite requiring insulin treatment 16 years earlier, she remained diabetes-free, and both her teenage children were lean and fit, reiterating the tangible linkage of parental lifestyle choices to childhood obesity and glycaemic impairment [21
UK national guidelines [12
] stress the importance of confining the care of diabetes in pregnancy to specialist multidisciplinary antenatal clinics, so the concept of “general ownership” of diabetes in pregnancy is somewhat new. Following implementation using QI methodology, we achieved our primary objective (improved specialist clinic capacity and timely first contact for GDM patients) without any worsening of the balancing clinical outcome measures.
The strengths of this project are the reporting of patient satisfaction data, the application of PDSA cycles, and an overall time frame that provides evidence of sustainability. The main limitation is the lack of data on other neonatal morbidity outcomes, such as neonatal hypoglycaemia and admissions to neonatal intensive care units. We also recognise the limitation of not applying any statistical tests to explore the significance of the positive trend in the clinical outcomes. We deliberately omitted this step because our project was a simple before-and-after evaluation, and we were highly cognisant of the inherent bias of such evaluations in relation to clinical outcomes in the absence of an unexposed control group. For this reason, the clinical outcomes remained relegated to the status of balancing measures. That said, the noticeable positive trends were an interesting observation that now creates a fresh hypothesis that these trends may be a direct consequence of the interventions. We are of the opinion that this hypothesis should now be explored in larger prospective multi-centre-controlled trials.
Staff engagement and education proved to be our biggest challenge during this operational journey. Sadly, at the outset, the management team declined our request to incorporate this domain of care into the mandatory training programme. In hindsight, we now feel we should have worked harder at trying to convince them otherwise, as it became clear that instituting mandatory training for non-specialist clinicians before commencement would have enabled a smoother transition.
One of the criticisms of the IADPSG/WHO 2013 diagnostic criteria for GDM is the association with increased caseloads. However, the proven connection of these criteria to long-term risks of maternal and childhood glycaemic impairment [31
] makes it a crucial component of contemporary public health initiatives for the ethnically diverse population we serve. It is especially noteworthy that during the corresponding time period (2014–2018), there was a 44% surge in the caseload of patients with diabetes in pregnancy, and our official district health profile reports showed concurrent increases in rates of overweight adults and obese children, peaking to become the highest rates in the East of England region and 10th highest of 325 national districts. This poor ranking on the health profile scale adds far-reaching value to our project as a long-term prevention initiative. Therefore, the GooD Pregnancy Network initiative offers a pragmatic balance of immediate minimal additional resource costs against the potentially markedly reduced costs of improved perinatal outcomes and long-term health. This organisational journey demonstrates the cost-effective realignment of the clinical capacity to increasing demand by minimising the duplication of care and improving patient empowerment. Furthermore, aside from upholding stipulated values of inter-professional education, the group clinic provides an additional dimension by enabling patients, partners, and professionals to learn with, from, and about each other [33
Although our maternity unit is based in the UK, we have applied the WHO 2013 diagnostic criteria for gestational diabetes since 2013 and serve a highly diverse multi-ethnic pregnant population. We, therefore, believe the applicability of this initiative extends well beyond the borders of the UK, and so we strongly encourage all clinicians to consider which components of the range of interventions we used would best suit the populations they serve. We also recommend linkage to publicly available information resources about the health and demographics of the local population, as we did, before proceeding with an intervention strategy.
On reviewing the contemporary literature, we found only a few published studies of antenatal group clinics for gestational diabetes [34
]. However, a common thread was that they all reported improvements in patient satisfaction, knowledge, empowerment, and other outcomes. A similar tendency to improvement was also evident when examining the literature on telemedicine usage for gestational diabetes [26
]. Our experience, therefore, adds to this growing body of evidence but offers an extra dimension by demonstrating the potential benefits of applying a composite of different elements, including active patient collaboration, to generate the kind of fundamental cultural shift that improved operational performance and has kept our service afloat during the COVID-19 pandemic. To the best of our knowledge, this report on the sustainability of a composite approach for gestational diabetes is the first of its kind in the medical literature.
In conclusion, our described inclusive network structure is an alternative care framework that offers a “right care, right place, right time” model for delivering the stipulated elements of care. It features a modified role for the specialist team that concentrates on the highest risk cases, whilst attentively overseeing the care of all others by their usual carers. Essentially, system leadership has successfully eliminated elitism by demystifying definitive care, embedding a culture of wider consciousness and connectivity that enables consistency during holistic care.
The final message is simple. “The prevalence of diabetes is set to skyrocket. Rising to this challenge requires basic involvement of all health professionals in prioritising early prevention strategies that target the parents of unborn children, embodying larger societal needs”.