Gestational diabetes mellitus (GDM) is an asymptomatic pregnancy-related condition normally identified by initial screening followed by diagnostic testing (Carr, 2001). The condition is a risk factor for the woman and fetus where there is an association between increasing glucose and adverse outcomes (HAPO Study Cooperative Research Group, 2008). In addition, GDM is a risk factor for maternal type 2 diabetes, offspring obesity and cardio-metabolic ill health (Seabra et al, 2015). The evidence to date remains unclear as to the diagnostic procedure of GDM (Farrar et al, 2011) and the impact of hyperglycaemia on adverse pregnancy outcomes (Yogev et al, 2009; Tieu et al, 2010). However, Farrar et al (2015: 795) suggest that ‘diagnostic criteria for gestational diabetes recommended by UK NICE [National Institute for Health and Care Excellence] (2008) might underestimate the prevalence of gestational diabetes' compared with their criteria, or recommendations made by the International Association of Diabetes and Pregnancy Study Group Consensus Panel (2010) for thresholds based on the Hyperglycemia and Adverse Pregnancy Outcomes study (HAPO Study Cooperative Research Group, 2008), which are also recommended by the World Health Organization (WHO, 2013).
The recommendations by NICE (2008; 2015) on diabetes in pregnancy primarily aim to empower women with GDM to have a positive childbirth experience by providing information, advice and support and reducing the risk of adverse outcomes for both the woman and neonate. Recommendations suggest that blood glucose control, assessed by HbA1c levels, is key in ensuring optimum outcomes to avoid the risks of hypoglycaemia by maintaining these levels at 48 mmol/mol (6.5%) (Agarwal et al, 2015; NICE, 2015). However, monitoring HbA1c levels is only recommended for women with pre-existing diabetes and must not be used routinely to assess a woman's blood glucose control in the second and third trimester (NICE, 2015).
Further risk assessment and treatment should include change in diet and exercise undertaken to control GDM and prevent birth complications, such as a macrosomic baby resulting in shoulder dystocia. However, the association between glucose intolerance and increased birth weight is unclear, and some studies suggest that macrosomia is linked to maternal weight gain and ethnicity, and not a result of raised blood glucose levels (Essel and Opai-Tettah, 1995). Others have identified that the probability of GDM for a parous woman increases from 2% to 21% when age increases from 20 to 40 years, and the high rate of GDM among grand-multiparas may be due to the confounding effect of maternal age (Al-Rowaily and Abolfotouh, 2010). Eidelman and Samueloff (2002) suggest that a complex interplay of hyper-metabolic growth factors, which include but are not only due to insulin and glucose levels, may result in adverse neonatal outcomes. Increased monitoring and possible interventions during pregnancy and labour would be additional factors that may have an impact on women's ability to access their planned place of birth, such as a midwife-led birth centre.
There is insufficient evidence to justify universal screening for GDM, according to a systematic review carried out by Scott et al (2002). However, Cosson et al (2006) suggest that universal, rather than selective, screening for GDM may improve fetal outcomes and reduce delay of diagnosis and care. A protocol of screening for GDM was developed from all the evidence available (Tieu et al, 2008), although a Cochrane review has suggested that the evidence is not sufficiently robust to indicate which is the best process to identify women with gestational diabetes (Tieu et al, 2008). A more recent publication from NICE (2015) identified risk factors for GDM and did not recommend the use of fasting plasma glucose for diagnosis. Detecting glycosuria of 2+ and above during routine antenatal urine testing would suggest that further testing is necessary. The use of a meter for self-monitoring of blood glucose and an oral glucose tolerance test (OGTT) is recommended. A 2-hour 75 g OGTT to confirm a diagnosed GDM is used, if plasma glucose level is 7.8 mmol/L or above (NICE, 2015). The purpose of screening is to identify those at risk of GDM, but screening results should not necessarily lead to adverse outcomes if treatment is appropriately managed (Yogev et al, 2009). Yesildager et al (2014) examined perinatal outcomes of borderline diabetic pregnant women and suggested that polygohydramnios remains a significant risk for those women who had an increased 50 g OGTT result, compared with those women whose results remained normal.
A trial of changes in diet and exercise and a referral to a dietitian for women with plasma glucose levels > 7 mmol/L at diagnosis is suggested. Several recommendations by NICE (2015) are documented in advising women how to monitor blood glucose and prevent problematic hypoglycaemia. Urgent testing for ketonaemia is recommended if pregnant women present with hyperglycaemia or are unwell, to exclude diabetic ketoacidosis (NICE, 2015). Alwan et al (2009) suggest treatment should be tailored to improve outcomes, although risk assessment of treatment should indicate any association of this treatment with adverse outcomes.
Aims
The purpose of this study is to examine the rates of adverse outcomes for women who have been diagnosed with GDM and who have controlled their blood glucose levels through dietary change, but who may be excluded from birthing in a midwife-led centre owing to risk of complications. The paper explores the evidence from a clinical audit to identify whether outcomes are the same for women who are able to control GDM through dietary change, as for all other low-risk women giving birth in one local maternity unit. This may have cost implications for the provision of maternity services in this locality.
Methods
A retrospective analysis was undertaken of all women giving birth at one maternity unit in the South East of England between 2014 and 2015, to compare the outcomes of women with GDM who were controlled on diet and those without GDM. The data generated were analysed through descriptive statistics. Data were accessed via maternity records and entered on the audit tool. Confidentiality was maintained as per normal process with maternity notes and data from the tool used in this service audit. Consent for data analysis was not required.
Population
All women who gave birth at the maternity unit were compared to women with diet-controlled GDM. The women were compared for labour and birth outcomes to include women experiencing induced labour, those experiencing spontaneous vaginal delivery (SVD) or assisted vaginal delivery, and those who had emergency lower segment caesarean sections (LSCS) or elective LSCS. Other outcomes that were compared were those for all neonates born in the maternity unit as compared to neonates of women with diet-controlled GDM.
Data collection
Currently, criteria for delivery of low-risk women in the birth centre at one local maternity unit excludes all pregnant women who have been diagnosed with GDM, including those who are able to maintain normal glucose blood levels through diet only (Appendix 1). An audit tool was developed to collect the data on these women from their maternity notes (Appendix 2). The audit tool specifically selected data on women's birth outcomes and neonatal outcomes. This was carried out to identify those women who, as a result of these outcomes, would have been excluded from giving birth in the birth centre as based on the current criteria. A specialist diabetes midwife and a small team of midwives working in the antenatal clinic collated the data at the end of each month in 2014 and the first 4 months of 2015, prior to the birth centre being opened in May 2015. The data were entered on the data collection form and results collated and checked by the diabetes specialist midwife and the university researcher.
Data analysis
Data were analysed in Microsoft Excel and presented as descriptive statistics. Numbers and percentages of women in each group (with diet-controlled GDM or without) and the number of events in each group are presented in Tables 1–4.
| Total number of women giving birth in the maternity unit (n = 4032) | Women with diet-controlled GDM (n = 89) | Amended data for GDM women diagnosed at term or requiring IOL for poor control of GDM at term (n = 67) | |
|---|---|---|---|
| Induced labour | 905 (22.45%) | 32 (35.96%) | 22 (24.72%) |
| Spontaneous vaginal delivery | 2293 (56.87%) | 43 (48.32%) | 32 (47.76%) |
| Assisted vaginal delivery | 901 (22.34%) | 13 (14.61%) | 10 (14.93%) |
| Emergency LSCS | 364 (9.03%) | 15 (16.85%) | 10 (14.93%) |
| Elective LSCS | 474 (11.76%) | 18 (20.22%) (includes those for previous caesarean section) | 15 (22.39%) |
GDM–gestational diabetes mellitus; IOL–induction of labour; LSCS–lower segment caesarean section
| Neonates of all women giving birth in the maternity unit (n = 4032) | Neonates of women with diet-controlled GDM (n = 89) | |
|---|---|---|
| Late fetal deaths (20–23 weeks) | 14 (0.35%) | 1 (1.12%) |
| Stillbirths | 16 (0.40%) | 0 (0.00%) |
| Preterm delivery < 37 weeks | 353 (8.75%) | 21 (23.60%) |
| To NICU at term delivery | 208 (5.16%) | 6 (6.74%) |
| Neonatal deaths (0–6 days) | 12 (0.30%) | 1 (1.12%) |
| Neonates > 4.2 kg | 204 (5.06%) | 3 (3.37%) |
GDM–gestational diabetes mellitus; NICU–neonatal intensive care unit
| Total number of women giving birth in the maternity unit in a 4-month period in 2015 (n = 1362) | Women with diet-controlled GDM (n = 57) | Amended data for GDM women diagnosed at term or requiring IOL for poor control of GDM at term (n = 43) | |
|---|---|---|---|
| Induced labour | 315 (23.13%) | 19 (33.33%) | 14 (24.56) |
| Spontaneous vaginal delivery | 771 (56.61%) | 25 (43.86%) | 24 (55.81%) |
| Assisted vaginal delivery | 282 (20.70%) | 12 (21.05%) | 6 (13.95%) |
| Emergency LSCS | 136 (9.99%) | 7 (12.28%) | 3 (6.98%) |
| Elective LSCS | 173 (12.70%) | 13 (22.81%) (includes those for previous caesarean section) | 10 (23.26%) |
GDM–gestational diabetes mellitus; IOL–induction of labour; LSCS–lower segment caesarean section
| Neonates of all women giving birth in the maternity unit (n = 1362) | Neonates of women with diet controlled GDM (n = 57) | |
|---|---|---|
| Late fetal deaths (20–23 weeks) | 3 (0.22%) | 0 (0%) |
| Stillbirths | 3 (0.22%) | 0 (0%) |
| Preterm delivery < 37 weeks | 122 (8.96%) | 4 (7.02%) |
| To NICU at term delivery | 61 (4.48%) | 2 (3.51%) |
| Neonatal deaths (0–6 days) | 4 (0.29%) | 0 (0%) |
| Neonates > 4.2 kg | 73 (5.36%) | 3 (5.26%) |
GDM–gestational diabetes mellitus; NICU–neonatal intensive care unit
Results
The general characteristics of the cohort of women observed in the maternity unit under study are shown in Table 1. Of all 4032 women who gave birth at this maternity unit in 2014, 164 women had GDM. Among these, a total of 53 were assumed to be diagnosed based on their history of GDM in a previous pregnancy. A total of 89/164 women with GDM were controlled with diet alone, of whom 53 women required metformin while 22 required insulin treatment. The national rate of GDM has been estimated at 2–5% (NICE, 2015); 164 women represents 4% of the maternity unit's total in 2014 and, therefore, falls within these estimated statistics. A total of 75 women were diagnosed by targeted modified glucose tolerance test, the remainder being tested as a result of being symptomatic (e.g. repeated glycosuria, polyhydramnios). Of the 89 women with diet-controlled GDM, 32 (35.96%) had their labour induced. The reasons for this were varied, but included women newly diagnosed with GDM at or after 37 weeks, and those whose blood sugar levels were becoming difficult to control at or after 37 weeks.
Figure 1 shows mode of birth of women with diet-controlled GDM at term. ‘Adjusted data’ excludes those women in whom labour was induced specifically because diagnosis was made at term or for poor control of GDM after 37 weeks' gestation. A number of women were identified to have GDM in the late stages of pregnancy, and some who had been controlled by diet found control increasingly difficult to maintain at this time. Once women with GDM reach 37 weeks' gestation, there is a tendency to immediately induce labour rather than commence or increase control measures. It could be suggested that the interventions of induction of labour (IOL), rather than the GDM, may be the root cause of assisted delivery and other complications.
The statistics indicate that women diagnosed with GDM which is well controlled through diet had similar outcomes to the general population of women giving birth at this maternity unit. The higher elective LSCS rate relates to the fact that a number of women in this group had a previous caesarean section, and most of them elected to have a caesarean section again.
Neonatal outcomes for all women, and of those women with diet-controlled GDM, are presented in Table 2.
Data were further examined for a 4-month period in 2015 to assess whether any changes had occurred in the subsequent year. During this period there were 57 women diagnosed with GDM and controlled on diet alone; 17 were on metformin and 6 required insulin therapy. Of the 57 women, 19 were induced, including women newly diagnosed with GDM at or after 37 weeks and those whose blood sugar levels were becoming difficult to control at or after 37 weeks. Again, the data were reviewed to exclude women with GDM where diagnosis was made at term (n = 13) or IOL for poor control of GDM at term (n = 14).
Data from 2015 (Table 3; Table 4) appear to be comparable to data from the previous year for both maternal and neonatal outcomes, apart from a reduction in 2015 for emergency LSCS. In addition, when comparing data for neonatal outcomes, there was a reduction in the number of neonates admitted to the neonatal intensive care unit at term delivery, but a slight increase in the percentage of neonates with birth weight above 4.2 kg. However, the overall percentage for this birth weight is comparable to 2014.
Discussion
It is generally believed that reducing blood glucose levels will improve maternal and neo natal outcomes (Tieu et al, 2008). This is usually managed through dietary control and advice (Tieu et al, 2008), regular self-monitoring of blood glucose levels, oral hypoglycaemic drugs, insulin therapy and exercise regime (Ceysens et al, 2006). In addition, limiting gestational weight gain to 5–9 kg among pregnant women with obesity has been suggested as likely to improve obstetric outcomes (Simmons, 2011). Thangaratinam et al (2012) suggest that lifestyle and, more significantly, dietary interventions are more effective in reducing maternal gestational weight gain and improving birth outcomes.
Obesity increases insulin resistance, resulting in a higher incidence of GDM, and has an adverse impact on both maternal and fetal outcomes, leading to a higher perinatal mortality rate (Sathyapalan et al, 2010). Normal physiological and metabolic changes during pregnancy can impair glucose tolerance. Blood glucose levels rise to stimulate more production of insulin in response as the pregnancy progresses. However, in approximately 7% of pregnant women, insufficient maternal pancreatic islet secretion of insulin results in GDM (NICE, 2015). This is usually defined as any degree of impaired glucose tolerance with the onset of pregnancy or first recognised during pregnancy; it is treated through diet modification, use of oral hypoglycaemic agent or the use of insulin, and may persist after pregnancy (McCance, 2011). A systematic review found evidence suggesting that 30–84% of women who experience GDM in their pregnancy are at increased risk of recurrence in subsequent pregnancy and of type 2 diabetes, with lower rates found in non-Hispanic white populations (30–37%) and higher rates (52–69%) found in minority populations (Kim et al, 2007).
Congenital abnormalities are more common in infants of diabetic women than those who do not have the condition. According to some studies (Ray et al, 2001; Platt et al, 2002), a teratogenic process of a multifactorial aetiology in early gestation appears to result in neonatal malformations during the diabetic pregnancy. It has been suggested that future research could develop therapeutic supplementation with antioxidants to prevent abnormalities of the fetus in diabetic pregnancies (Eriksson, 2009). A meta-analysis conducted by Ray et al (2001) concluded that pre-conceptual care of diabetic women could result in a threefold reduction in fetal abnormalities and also lead to a reduction in HbA1c levels in the first trimester. The majority of neonates born to women with GDM do not have complications if the quality of the control of diabetes is optimised (Michael Weindling, 2009; Hawdon, 2011). However, long-term outcomes of offspring of diabetic pregnancies indicate that they have an increased incidence of type 2 diabetes and obesity as adults (Simeoni and Barker, 2009). In addition, stillbirth rates of women with GDM who are not induced after 40 weeks' gestation also increase (Karmon et al, 2009).
The focus of this paper is to assess the risk to women with diet-controlled GDM associated with giving birth in a midwife-led birth centre, and to identify what, if any, adverse outcomes they are likely to experience. The optimum diet to control GDM is unclear, and the evidence is not robust. In 2013, a Cochrane review carried out on the different types of dietary advice for women with GDM was published (Han et al, 2013). The review suggests that dietary counselling is the main strategy in managing GDM, but which dietary therapy is best remains unclear. Based on the available evidence, the authors could not identify which dietary advice was most suitable for women with GDM for improving the health of both women and neonates in the short and long term. No one type of dietary advice was more effective than others in reducing birth weight or adverse outcomes related to neonatal size and weight. However, at the Trust where this study took place, it is accepted practice that all women with GDM are referred for dietary support, and a dietitian is present at their antenatal appointments.
There is evidence that women who give birth in a midwife-led birth centre have a greater likelihood of initiation and continuing duration of breastfeeding, owing to good maternal attachment and support from staff (Finigan and Chadderton, 2014). Some studies have found that, with longer duration of breastfeeding, women with GDM had a lower incidence of developing type 2 diabetes in the long term (Gunderson et al, 2010). Chertok et al (2009) found that neonates of women with GDM who initiated breastfeeding within 30 minutes of birth, were significantly less likely to be diagnosed with borderline hypoglycaemia. The neonates' blood glucose levels were higher compared with formula-fed infants of GDM women who could stabilise blood glucose levels. Other studies indicate that breastfed neonates are less likely to become obese (Harder et al, 2005; Owen et al, 2005) and develop type 2 diabetes (Liu et al, 2010), and that flavours that occur in breast milk as a result of the woman's diet can influence a healthy diet in the child (Mennella et al, 2001).
Milk production commences with lactogenesis I in pregnancy and continues with milk formation through lactogenesis II. It may be suggested that lactogenesis II is delayed in obese women owing to the concentration of progesterone (an inhibitor of lactogenesis II) in increased adipose tissue. In addition, obese women appear to have a decreased prolactin response to neonatal suckling, which decreases the milk supply (Rasmussen and Kjolhede, 2004).
The cost-effectiveness of increased midwifery-led care is evident in the literature, with the recently published recommendations of the National Maternity Review (2016) aiming to normalise pregnancy without increasing risks. The Birthplace study cost-effectiveness findings (Schroeder et al, 2011) show that birth at home was the most cost-effective (£780 per woman) and safe for low-risk women having a second or subsequent baby. For nulliparous women, planned birth in a midwifery-led unit was also safe and less expensive (£968 per woman) compared with planned birth in an obstetric unit (£1142 per woman). Although cost-effectiveness for women with GDM is not included in this analysis, it is discussed in the NICE (2015) guidelines.
‘The midwife's role is to ensure current screening is facilitated and to manage care collaboratively with key practitioners to ensure positive pregnancy and birth outcomes’
The midwife's role
Although the focus of midwifery care is primarily on healthy childbearing women, midwives frequently extend their care to include women with obstetric risk factors and social and medical complications. In the case of GDM, this must include careful monitoring and provision of education and support to determine risk, prevent complications of outcomes and ensure optimum glycaemic control. The midwife's role is to ensure current screening is facilitated and to manage care collaboratively with key practitioners to ensure positive pregnancy and birth outcomes.
Studies have suggested that midwifery-managed continuity of care for women with GDM resulted in women being more satisfied with their overall childbirth experience than those who received shared care (Morrison et al, 2002). In addition, although Turnbull et al's (1996) study is dated, the evidence still supports the finding that midwifery-led care of this group of women resulted in women being less likely to have their labour induced and more likely to have an intact perineum than women who received shared care. Teadora (2010) suggests that despite studies to reduce the risks of GDM, the best evidence is promotion of basic nutritional counselling (Oza-Frank et al, 2014), which has an impact on reduced maternal weight gain and lower rates of complications such as shoulder dystocia, related to macrosomic neonates. It is perceived that there is a need to educate midwives to provide effective counselling about healthy lifestyle choices to women with GDM, to improve pregnancy and birth outcomes (Persson et al, 2011). This is the current practice of diabetes specialist midwives who already support women with GDM at the Trust where this study took place.
‘Effective management of gestational diabetes cannot be achieved without the woman's responsibilities of self-management in partnership and with the support of midwives’
Women's perspective
The effective management of GDM cannot be achieved without the woman's responsibilities of self-management in partnership and with the support of midwives and other health professionals. In studies by Carolan et al (2010; 2012), the authors explored women's knowledge of GDM and their experiences of factors and inhibitors that facilitated self-management. They concluded that women from low socioeconomic backgrounds struggled to manage their GDM and required educational and social support. Other studies suggest that women with diabetes in pregnancy experience higher levels of stress and anxiety in managing their condition in relation to dietary management (Hui et al, 2014) and their perceptions and expectations of GDM may affect their psychological response to pregnancy and their behaviour during and after the pregnancy (Lawrence, 2011).
A meta-synthesis considered the evidence of how women with GDM could prevent the development of type 2 diabetes and suggested that addressing the emotional impact of GDM, providing efficient information about future diabetes risks, and offering healthy lifestyle interventions were key factors in promoting women's control over their future health (Soltani et al, 2008; Parsons et al, 2014). The evidence suggests that women with GDM on a well-controlled diet and supported by experienced health professionals would benefit from choosing the birth centre as their planned place of birth.
Implications for practice
In view of the current situation with ever-increasing health costs, more women giving birth in the birth centre may reduce birth costs. Women with diet-controlled GDM in this Foundation Trust currently give birth in an interventionist, high-risk environment of a labour ward. The findings from this audit are compelling and other maternity units may want to repeat this for their own local Trusts. The authors recommend revisiting the criteria for planned place of birth in a midwife-led birth centre to include women with diet-controlled GDM.
Conclusion
The issues discussed in this paper suggest that it may not be in a woman's best interest to exclude her from the birth centre purely on the basis of a ‘label’ of GDM.
It is important to recognise that 53/89 women in 2014, and 13/57 women in 2015, were labelled as having GDM based on diagnosis in a previous pregnancy. If such a woman has maintained blood glucose levels in the normal range throughout her current pregnancy, it is reasonable to question whether other factors (such as significant weight loss since the previous pregnancy) may mean that she is being labelled and treated at ‘high risk’ unnecessarily.
The mode of delivery and outcomes for the women in this cohort who remained normoglycaemic have been found to be in line with the general outcomes for all women at this hospital in the same time period. The small sample size means further work is needed to confirm this, but the findings indicate that the option of using the birth centre may be considered a reasonable alternative for these women.
Controlling health-care expenditure has implications for the sustainability of maternity services, as is evident in the recommendations of the National Maternity Review (2016). In addition, the increased rate of successful establishment of breastfeeding in women using birth centres could be argued as a reason to recommend this birth environment for women with diet-controlled GDM, owing to the consequent long-term benefits to both woman and baby.