Diabetes and the human condition

Adapting to a global dilemma

The treatment of diabetes before insulin therapy tested the skills of the clinician and the commitment of the patient. But by the 1920s, insulin became widely available for the diabetic community in North America.1

If diabetes was once considered a simple problem of insulin deficiency it is not seen that way now. Instead it is a window on complex metabolic distortions, signals, substrates and reactions.

A second form of diabetes

An extraordinary twist in the diabetes story has been the epidemic of type 2 diabetes. Rapid changes in the human environment — massive migrations from rural to urban environments with upheavals in diet away from traditional foods — have accompanied soaring prevalence rates in developed and developing economies alike.

Why cannot humans adapt to our new environments? Maybe we can — but it will take ages. Plutella xylostella caterpillars adapt to an obesogenic environment, but in eight generations.2 So we must turn instead to the environmental factors that promote diabetes and hinder its treatment.

What we eat is inextricably linked to the food that is available, affordable and acceptable. In order to make lasting and sustainable changes to the food environment, the food system from paddock to plate must change. This may occur through the accumulation of many small changes.

The role of Big Farmer

Agricultural subsidies for cash crops can be a health hazard. Subsidised corn and soybean oil in the United States lead to cheap ultraprocessed foods. Production of high-fructose corn syrup, used in beverages, is supported by large government subsidies.

But with reduced subsidies corn farmers are now planting orchards and growing vegetables. Cause and effect cannot be disentangled, but changes in agriculture are occurring in parallel with health awareness.3 Farm incomes are recovering in parallel with increased sales of fruits and vegetables.

Improvements must be made not only in food production, but also in food-processing practices. Product reformulation can reduce the levels of salt and sugar and replace trans and saturated fats. Ensuring that a health-promoting replacement is available depends on upstream agricultural production and policies.

Again in the US, where consumer awareness of the hazards of trans fats has grown rapidly, the food production industry has responded. In 2005, the Dow Chemical Company launched canola and sunflower seeds for cooking oils with a longer shelf life and suitable for frying. This shift occurred in parallel with the requirement for all foods to be labelled with trans fat content.4

Food product reformulation followed, and 600 million tons of trans and saturated fats were removed from the US diet between 2005 and 2012.4

So with a blend of informed consumer pressure, legislative support and innovative food industry responsiveness, changes do occur.

In addition to the toxic food environment, many new cities and neighbourhoods are not conducive to walking. Instead, sitting sullen and stressed in traffic jams is normative. Much as with the food environment, promising changes upstream in urban design can be both health-promoting and attractive to developers. More safe bicycle lanes, aesthetic rather than cracked and lumpy footpaths, and accessible public transportation systems increase physical activity levels.5,6

Equity of access to care

Alongside these preventive changes for type 2 diabetes, the environment that limits care for patients with diabetes must change. Each year, thousands of children with type 1 diabetes die in less economically advanced societies for lack of insulin.7,8 Families face the shocking choice between insulin for one child and food for the others.7

We do not know everything about diabetes and the environment but we know enough to advocate for change — at all points in the agri-food chain, in the way we design our cities and in seeking justice for all people who need them to have access to lifesaving diabetes therapies.

Ipilimumab-induced hypophysitis: early Australian experience

To the Editor: We report two men aged in their 60s receiving ipilimumab for metastatic melanoma who presented with headache and constitutional symptoms after the third 3-weekly dose, and were diagnosed with ipilimumab-induced hypophysitis. Ipilimumab is a monoclonal antibody that binds to cytotoxic T lymphocyte-associated antigen 4, resulting in T-cell activation and proliferation. It was the first therapy to yield a survival benefit in metastatic melanoma,1 but at the cost of frequent immune-related adverse events.2

Patient 1 experienced headache, fatigue, postural lightheadedness, anorexia and asthenia. Morning blood test results before steroid administration were consistent with central hypocortisolism (serum cortisol, < 28 nmol/L [reference interval (RI), 70–650 nmol/L]; inappropriately normal adrenocorticotropic hormone [ACTH], 1.2 pmol/L [RI, 0–12.0 pmol/L]), hypothyroidism (low thyroid-stimulating hormone [TSH], 0.2 mIU/L [RI, 0.4–4.0 mIU/L]; low free thyroxine [FT4], 8.9 pmol/L [RI, 9.0–19.0 pmol/L]; normal free triiodothyronine [FT3], 4.7 pmol/L [RI, 2.6–6.0 pmol/L]), and hypogonadism (low testosterone, 2.8 nmol/L [9.5–28.0 nmol/L]; insufficiently raised follicle-stimulating hormone [FSH], 14.0 IU/L [RI, 1.0–12.0 IU/L]; luteinising hormone [LH] within RI, 3.7 IU/L [RI, 0.6–12 IU/L]). Insulin-like growth factor-1, growth hormone and prolactin levels were within RIs. Magnetic resonance imaging (MRI) scans of the brain excluded metastatic disease and demonstrated pituitary enlargement without chiasmal or infundibular involvement (Box). The patient was initially commenced on replacement hydrocortisone.

Patient 2 presented with severe headaches and emesis. Morning blood test results demonstrated central hypothyroidism (low TSH, 0.24 mIU/L [RI, 0.27–4.20 mIU/L]; FT4, 11.4 pmol/L [RI, 12.0–22.0 pmol/L]; and FT3, 2.8 pmol/L [RI, 3.0–7.8 pmol/L]) and hypogonadism (low testosterone [2.1 nmol/L] and inappropriately normal FSH [8.1 IU/L] and LH [3.9 IU/L]). Although morning cortisol and ACTH levels were initially adequate, the patient had been placed on high-dose dexamethasone because of suspicion of brain metastases, which were later excluded by lumbar puncture and brain MRI. The pituitary was not enlarged, but superior convexity was evident.

Both patients received a 3-week tapering course of prednisolone starting at 30 mg daily and weaned to replacement doses. They received thyroxine and, later, testosterone replacement. Ipilimumab was continued to include the standard fourth and final dose. The endocrine abnormalities persisted in both men at 3–6-month follow-up, despite normalisation of MRI findings (Box). For Patient 2, a subsequent insulin-tolerance test demonstrated a flat cortisol response consistent with corticotropic insufficiency due to hypophysitis, although suppression from exogenous steroids may have contributed. Both patients experienced tumour progression over the follow-up period.

Ipilimumab was registered by the Therapeutic Goods Administration in 2011 and funded by the Pharmaceutical Benefits Scheme from August 2013. Ipilimumab-induced thyroiditis and adrenalitis are common differential diagnoses to consider.4 Suprasellar extension is seen frequently,5 requiring urgent ophthalmic evaluation. There are no prospective data on whether high-dose steroids improve pituitary recovery or abrogate the tumour response to ipilimumab. Our cases demonstrated no short-term advantage. Interdisciplinary awareness of this condition is required to ensure that replacement glucocorticoids are started promptly on diagnosis and increased during periods of physiological stress, including surgery or chemotherapy.

Sagittal magnetic resonance imaging (MRI) scans

T1-weighted sagittal MRI scans of Patient 1, showing symmetrical pituitary enlargement up to 9.3 mm in height with superior convexity and loss of the posterior pituitary bright spot at presentation (A) consistent with a diagnosis of autoimmune hypophysitis,3 followed by normalisation 6 months later (B). 

Lower treatment targets for gestational diabetes: is lower really better?

The rationale for management of gestational diabetes mellitus (GDM) is well established — treatment reduces the risk of macrosomia and its attendant complications. This Journal has carried the views of proponents of lower diagnostic and treatment targets for GDM in the context of updated Australian guidelines.1,2 In this article, we focus on the costly and potentially deleterious effects of suggested lower treatment targets. We argue that such targets are based on insufficient interventional data, create potential health and medicolegal risks and pose great problems for implementation, particularly to providers in regional and remote areas such as our own health district, which services an area about the size of Victoria. In our view, the disadvantages of lower treatment targets currently outweigh the limited evidence of benefits.

Evidence of risks and benefits to patients

Three key trials have informed current practice for managing GDM. The Australian Carbohydrate Intolerance Study (ACHOIS), an interventional study of 1000 women enrolled over about 10 years to 2003, used lower diagnostic thresholds as well as lower treatment targets.3 Treatment targets were blood sugar level (BSL) of 5.5 mmol/L (fasting) and 7.0 mmol/L (2 h postprandial) in the intervention group. These treatment targets are now the standard of care in many Australian centres. The comparator control group caregivers were unaware of the diagnosis of “glucose intolerance of pregnancy” (the prevailing terminology at the time). Further investigation and management by the treating clinician was permitted if indicated. The ACHOIS trial showed a significant improvement in the intervention group for the primary composite fetal outcome measure (death, shoulder dystocia, bone fracture and nerve palsy) with increased rates of induction of labour in mothers and an increased rate of admission of babies to the neonatal nursery. Interestingly, the trial did not show a statistically significant reduction in caesarean section rates.

A smaller trial in the United States studied the effect of treatment of mild hyperglycaemia among women recruited from 2002 to 2007.4 Using a 3-hour 100 g oral glucose tolerance test, mild GDM was diagnosed if the fasting glucose level was less than 5.3 mmol/L with two or three timed glucose measurements that exceeded established thresholds of 10.0 mmol/L at 1 h, 8.6 mmol/L at 2 h, and/or 7.8 mmol/L at 3 h. Treatment targets of 5.3 mmol/L (fasting) and 6.7 mmol/L (2 h postprandial) were then applied to the intervention group. The trial did not show a difference in the chosen primary perinatal composite outcome (perinatal death, neonatal hypoglycaemia, hyperbilirubinaemia, elevated cord C-peptide level, and birth trauma). The rate of induction of labour was not different for the intervention group; and caesarean section and shoulder dystocia rates were reduced.4

The large observational Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study showed a continuous relationship between fasting, 1-hour and 2-hour glucose levels obtained on a 75 g glucose tolerance test and the risk of increased birthweight, primary caesarean section, elevated cord C-peptide levels, and neonatal hypoglycaemia.5 No obvious threshold where risk overtly increased was found. GDM diagnostic criteria were subsequently revised: it was suggested that a diagnosis of GDM should be made on the basis of blood sugar levels that correlated with a 1.75-fold increased risk of specific fetal measures (birthweight > 90th percentile, percentage body fat > 90th percentile, and cord C-peptide level > 90th percentile).6 The HAPO study did not specifically address any ongoing glycaemic measures throughout pregnancy, and the subsequently suggested diagnostic criteria have not been uniformly adopted, owing to concerns and debate outside the scope of this article.

The continuous relationship between increased glucose measures and increased risk of complications presents both a scientific and philosophical challenge — where a continuous relationship occurs in medicine there is no firm boundary to the disease entity. The definition of disease is one that the profession decides. We argue that in such areas, interventional data are even more important for establishing cut-off points. Harm to our patients must be captured and weighed against the advantages of further intervention. While there is useful information about normal glycaemic values in pregnancy,7 recent articles, including a meta-analysis, have commented that randomised controlled trials (RCTs) showing benefits of intervention in this area were scarce and data quality poor, particularly with regard to postprandial values.8

Thus, based on the current uncertain interventional data regarding milder hyperglycaemia, it would seem difficult to support treatment thresholds being reduced even further to 5.0 mmol/L fasting, 7.4 mmol/L at 1 h and 6.7 mmol/L 2 hours after meals for Australian women.2 These targets (Box 1) are arguably the most aggressive in the world. Indeed, Nankervis and colleagues state that such targets should, ideally, be examined by RCTs.13 We would greatly prefer such interventional studies be conducted and prove overall benefit before adoption of the revised targets.

It is noteworthy and problematic that a fasting BSL less than 5 mmol/L also conflicts with current Diabetes Australia advice that patients on insulin should have a BSL “above 5 to drive”;14 no driving before breakfast may not be practical for some women.

We are concerned that a treatment target of 5.0 mmol/L fasting will expose women to the risk of hypoglycaemia, particularly given that current international standards for blood sugar monitors allow for a 15% error margin (ie, a BSL of 4.0 mmol/L could be 3.4–4.6 mmol/L). Indeed, in a recent study, a fifth of monitors tested failed the looser standard of a 20% error margin, and half of meters would have failed the new 15% standard without improvements.15 In addition, it is noteworthy that neither the observational HAPO study nor the interventional trial of mild GDM reported maternal hypoglycaemia as an outcome and that overtreatment could potentially lead to an increase in small-for-gestational-age babies among women at risk of placental insufficiency.16

Medicolegal risk and clinical judgement in the real world

Guidelines often allude to notions of clinician judgement as the ultimate authority for patient care.2 Such statements are understandable but can be wishful, particularly in relation to GDM. Many women with GDM will not be seen by specialists, who have greater training in a specific area, better access to diagnostic services (such as ultrasound for determining growth rates) and would be more willing to depart from guidelines. It has been estimated that the new diagnostic criteria will lead to a 35% increase in the number of women diagnosed with GDM.17 Most women will thus be predominantly cared for by diabetes educators, midwives, nurses and general practitioners. This is particularly true in regional and remote areas where specialists are scarce. For reasons of workload and workforce, management of GDM is likely to become even more driven by protocols, adding weight to the importance of guidelines being both workable and correct.

While in Australia it is ultimately courts that determine what constitutes negligent conduct, the opinion of medical experts is persuasive,18 and new lower targets are likely to become the default new standard of care. In practice, many health practitioners will struggle to achieve these targets with many of their patients. Given the large number of women who will be diagnosed with GDM, the range of neonatal medical conditions that can be linked with GDM and the great expectations that accompany pregnancy, a number of adverse outcomes could be litigated in the future. A reduction in the degree of legal protection for Australian health practitioners may be an unfortunate unintended consequence of lower treatment targets that depart from international practice (Box 1).

The Australasian Diabetes in Pregnancy Society (ADIPS) treatment targets are listed as “suggestions” rather than “recommendations”.2 However, it is plausible that health practitioners will refer to the new ADIPS guidelines without appreciating the difference between the two. With the bewildering pace of medical advances in many fields, most health practitioners are dependent on expert guidelines to provide direction in clear terms.

Human resource and economic cost

The human and economic resources required to manage gestational diabetes are considerable. In the Australian setting, such care is resource intensive and often provided by a multidisciplinary approach that may involve GPs, physicians, obstetricians, midwives, nurses, dietitians and diabetic educators. Education and frequent clinical review are the standard of care. The predicted 35% increase in the number of women diagnosed with GDM will lead to a 13% prevalence of the disease17,19 — all women should receive diabetes and dietary education and will require continued review until birth.

Regional centres like our own often serve not just the immediate town but also support remote towns and communities dotted across a vast area. Most women in our centre receive 2 hours of education followed by weekly phone or email contact and regular visits, increasing in frequency until the birth. Women who require treatment with insulin need further intensive education and may have to travel considerable distances to gain this. They usually have more frequent antenatal cardiotocography and ultrasound monitoring, done in a major centre. The baby must be born at a tertiary centre, and labour is usually induced at around 38–39 weeks’ gestation, leading to an earlier delivery date than for women who are “diet controlled”. Neonates are observed in a special care baby unit overnight.

Many of our patients from rural and remote communities must move to live near our hospital in the last month of pregnancy, at considerable cost to themselves and their families. The non-medical cost borne by our hospital system for transport and accommodation alone for the last weeks of pregnancy is at least $6000 per patient.

Cost-effectiveness research has been rather limited. An older Australian study with different diagnostic criteria and treatment targets compared the intervention group to a routine care group who were not made aware of their diagnosis, and thus, the study is of limited applicability to the current clinical context.20 More recent research concluded that the treatment of milder GDM would not be cost-effective if the cost was greater than US$3555 compared to a baseline cost of US$1786 in the different context of the US health system.21

In our region of Australia, a diagnosis of GDM triggers a series of events that shifts care away from peripheral centres to our own tertiary care facility. We doubt this situation is unique, and it creates a number of deleterious social consequences for women in their separation from their full support structures and families. For our Aboriginal and Torres Strait Islander and remote patients, these social and psychological effects should not be taken lightly.

While the effect of new diagnostic criteria has been studied, there is a paucity of data on the effect of new treatment targets. In our centre, the decision to commence pharmacotherapy for GDM is made by endocrinology consultants or advanced trainees. In our prospective audit of 319 patients at our major regional centre over 12 months, we treated women according to our current targets of 5.5 mmol/L (fasting) and 7.0 mmol/L (2 h postprandial), but simultaneously considered what treatment would have been required to treat to targets of 5.0 mmol/L and 6.7 mmol/L. Adopting such a practice would have led to a doubling of patients who needed to start pharmacotherapy by our service (Box 2), with 62% of all women in our clinic requiring pharmacotherapy during pregnancy. Insulin is still the usual first-line treatment in Australia; and studies show that 50% of patients placed on metformin will additionally need insulin to reach targets.22 While it is extremely important that pregnant women are treated optimally, if only 38% of our patients can be managed with diet alone, this will place considerable burden on our already stretched health system. An excessive glucose-centric focus on treating milder GDM may distract from systematically growing contributors to adverse pregnancy outcomes. Obese patients who do not have GDM are contributing to a greater degree to adverse outcomes.23 These factors need to be taken into consideration to avoid misallocation of resources.

Risk of widening the health gap for Aboriginal and Torres Strait Islander Australians

There is some evidence that exposure to hyperglycaemia in utero is linked to an increased risk of developing type 2 diabetes mellitus later in life,24 and, consequently, it has been theorised that management of GDM will reduce this risk. It should be noted that there are no such results from interventional data yet, as these studies take many years to complete. A study of the children of women in the ACHOIS intervention group did not show any reduction in their body mass index at 5 years.25

On the other hand, there is considerable evidence that children of low birthweight also have an increased risk of developing type 2 diabetes. In this group of babies, including in specific studies of North American indigenous populations, the lower the birthweight the greater the chance of developing type 2 diabetes later in life.26,27 Despite high rates of maternal diabetes, babies of Aboriginal and Torres Strait Islander descent have twice the risk of being of low birthweight (< 2.5 kg) compared with the national average (12% v 6%).28 It is unclear whether aiming for a lower birthweight with decreased adipose levels among the Aboriginal and Torres Strait Islander population is healthier than a more normal birthweight. Given the prevalence of low birthweight babies in Aboriginal and Torres Strait Islander populations, many of the benefits of GDM treatment, which are largely mediated through reducing macrosomia, may not occur. The high social and economic costs of GDM treatment at lower thresholds certainly will.

The World Health Organization has noted that the lack of ethnically specific data is a limitation of applying knowledge from the HAPO study, and that adaptation may be required for different ethnic groups.29 While the Aboriginal and Torres Strait Islander population is numerically small, this group deserves strong consideration given their heavy disease burden and disadvantage. The variation in macrosomia that occurs among Aboriginal and Torres Strait Islander babies as a result of GDM may also occur among babies of women from different ethnic backgrounds in our increasingly multicultural country, in which more than a quarter of people were born overseas.30

Conclusion

Although the relationship between maternal blood glucose levels and the risk of macrosomia is a continuous one, there is currently a lack of interventional data to support treatment of GDM to lower targets. To date, studies have not sought to capture the effects of possible maternal hypoglycaemia with lower treatment targets. Some of Australia’s recently revised treatment targets are lower than international practice and impose particular social and economic costs while having limited benefits for Aboriginal and Torres Strait Islander Australians. Neither cost-effectiveness nor safety has been established. In our view, an overall analysis of the advantages and disadvantages of lower treatment targets for GDM in Australia suggests that implementation at this stage is premature.

1 Comparison of upper limits of treatment targets for GDM, Australian and international guidelines

Guideline

Fasting

1 h after meals

2 h after meals

Fifth International Workshop-Conference on Gestational Diabetes Mellitus9

5.2 mmol/L

7.7 mmol/L

6.6 mmol/L

Canadian Diabetes Association10

5.2 mmol/L

7.7 mmol/L

6.6 mmol/L

UK National Institute for Health and Clinical Excellence11

5.9 mmol/L

7.7 mmol/L

Not specified

US Endocrine Society12

5.3 mmol/L or 5.0 mmol/L*

7.8 mmol/L

6.7 mmol/L

Australasian Diabetes in Pregnancy Society2

5.0 mmol/L

7.4 mmol/L

6.7 mmol/L

GDM = gestational diabetes mellitus. * If this can be achieved without hypoglycaemia.

2 Effect of lower treatment targets on a GDM cohort in a major regional centre*

Cairns Hospital cohort (n = 319, July 2012 – July 2013)	Current practice: 5.5 mmol/L (fasting) and 7.0 mmol/L (2 h postprandial)	Proposed practice: 5.0 mmol/L (fasting) and 6.7 mmol/L (2 h postprandial)

Able to be managed with diet and lifestyle control	183 (57%)	122 (38%)
Required/would require commencement of pharmacotherapy	67 (21%)	128 (40%)
Already on pharmacotherapy when referred	69 (22%)	69 (22%)

GDM = gestational diabetes mellitus. * Ethics approval was granted by the Cairns and Hinterland Health Service District for gathering of the audit data.

Reassessment of the new diagnostic thresholds for gestational diabetes mellitus: an opportunity for improvement

The diagnosis of gestational diabetes mellitus (GDM) has been based on guidelines derived from expert opinion. The original threshold blood glucose levels (BGLs) for diagnosis recommended by the Australian Diabetes in Pregnancy Society (ADIPS) had no scientific validity.1 Subsequently, intervention studies provided validity for these levels.2,3 However, these criteria remained statistically unsupported until the results of the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study were published.

Assessment of the Hyperglycemia and Adverse Pregnancy Outcome study

HAPO analysed data from 23 316 women enrolled in a blinded epidemiological study where neonatal and maternal adverse events were assessed according to prespecified blood glucose ranges after a 75 g oral glucose tolerance test performed at 24–32 weeks of gestation.4 This study showed strong continuous relationships between fasting blood glucose (FBG) levels, 1-hour BGL and 2-hour BGL after a 75 g oral glucose tolerance test and outcomes. HAPO assessed different odds ratios (ORs) for risk of developing adverse neonatal outcomes for each BGL parameter and the percentage of women diagnosed. Subsequently, the International Association of Diabetes and Pregnancy Study Groups (IADPSG) reached consensus that the arbitrary threshold level for risk should be an OR ≥ 1.75 of developing certain adverse neonatal outcomes. These recommended levels were FBG ≥ 5.1 mmol/L, 1-hour BGL ≥ 10 mmol/L and 2-hour BGL ≥ 8.5 mmol/L.5 The diagnosis is established if one or more BGLs equals or exceeds these values. If this threshold is met, a patient presumably will have at least a 75% increased risk of complications. These recommendations have been accepted by many professional bodies, including the American Diabetes Association.6 However, many other organisations have not endorsed them, including the United States National Institutes of Health (NIH).7 They have been recommended for endorsement in Australia by ADIPS.8 Based on these new criteria, there is a significant increase in the number of patients being diagnosed with GDM. In the HAPO study, 16.1% of participants met these criteria.5 The rate of GDM in different countries varies but there is a general concern that the IADPSG criteria will place an increased burden on services.9,10

This increase would not be an issue if these pregnancies were associated with a significant increase in preventable complications. But do these criteria identify pregnancies at risk? This may not be the case. There are at least two problems concerning the new diagnostic thresholds.

The combined adverse neonatal outcomes

The thresholds for diagnosis were based on the average BGLs at which the odds for birthweight > 90th percentile (measure of large for gestational age [LGA]), cord C-peptide levels > 90th percentile, and per cent body fat > 90th percentile reached 1.75 times the estimated odds of these outcomes at mean glucose values. There are several concerns with this approach. Considering the FBG threshold only, the calculated OR for the individual primary and secondary outcome measures of the HAPO study are shown in Box 1.5 The outcome measure of per cent body fat > 90th percentile was neither a primary nor secondary outcome measure but simply an exploratory outcome. There was incomplete ascertainment, with it being assessed in 83% of all babies and in only 73% of babies who had cord C-peptide levels measured.11 The data for this outcome were not published in the original results paper of HAPO.4 Including this outcome would not be considered standard statistical practice.

The primary outcome measure of cord C-peptide levels > 90th percentile was included but it is not a routine test performed in clinical practice. An increase in C-peptide levels at delivery most likely reflects the normal physiological increase in insulin secretion due to increased fetal glucose exposure consequent to maternal hyperglycaemia.4 It may be a marker of insulin resistance with relevance to future adverse metabolic outcomes, but it is not known if it has any long-term clinical significance.11,12 It was the major outcome measure contributing to the lower FBG threshold (Box 1): its OR is 2.02, significantly higher than for adiposity (OR, 1.62) and LGA (OR, 1.68). If increased numbers of women are being diagnosed with GDM because they are at risk of their fetus having high cord C-peptide levels but not LGA, the significance of this abnormality should be established for it to be included in the combined outcome measure.

The combination of these outcome measures appears to have been a post-hoc decision made by IADPSG. These were selected because they had had the greatest difference between the at-risk group and the normal group. Neonatal adiposity > 90th percentile was not a prespecified combined outcome measure.4 This should be considered as an exploratory combined outcome measure at best. Its validity should be proven in other cohorts. It is inappropriate to use it as the basis for establishing new diagnostic thresholds for GDM.

There were four primary outcome measures originally defined for HAPO (Box 1). A more statistically valid approach would be to combine all the four primary outcome measures. This was not done, so the highly relevant clinical primary outcome measures of rates of caesarean section or neonatal hypoglycaemia did not contribute to the diagnostic thresholds. If this had been undertaken, higher BGLs would have been defined to achieve the OR threshold of 1.75, with fewer women identified.

The most recognised clinical problem of gestational diabetes is macrosomia.1–5 The outcome measure assessing this in HAPO was birthweight > 90th percentile. At the threshold FBG level of 5.1 mmol/L, the OR for LGA is 1.68 (95% CI, 1.56–1.80) (Box 1). This implies that the mean risk for LGA at this FBG level is below the IADPSG threshold for diagnosis. A higher FBG level is required to reach the defined OR. Ideally, to ensure with 95% confidence that women diagnosed with GDM are at risk, the lower limit of the 95% confidence interval of the OR for each BGL should be equal to or above 1.75. These values cannot be calculated from the available data.

The effect of normal blood glucose levels

The second problem relates to the effect of normal BGL parameters on risk. When the IADPSG-recommended thresholds are used to diagnose GDM in the blinded cohort of HAPO, 69% of women meet the criteria based on a single elevated BGL parameter only.13 Of these women, 81% were diagnosed on the basis of an elevated FBG or 1-hour BGL.13 The effect on overall risk of having two normal BGLs in these women was not considered. An exploratory analysis assessing this effect suggested that potentially up to 50% of these women may not reach the required risk threshold.14 In response, data from HAPO were reported for women who had only one, two or three elevated BGLs (Box 2).13 The mean OR for LGA of women with an elevated FBG or 1-hour BGL only was 1.76 and the 95% CI crossed the risk threshold, confirming that nearly 50% of these women did not meet the agreed risk threshold. This implies that overall nearly one-third (about 50% of the 69%) of women who had only one elevated BGL did not meet the risk threshold. This also applies to subjects with two elevated postprandial BGLs only (Box 2). Most of the potential increase in clinical demand results from patients who do not meet the IADPSG agreed level of risk for diagnosis.

In contrast, women with an elevated FBG level (a marker of preprandial hyperglycaemia) plus one or more elevated post-challenge BGLs (markers of postprandial hyperglycaemia) have pregnancies at much greater risk of LGA (mean OR, > 3.0; lower limit of the 95% CI, > 2.0; Box 2). These elevated BGLs suggest greater intrauterine exposure to glucose associated with higher rates of LGA compared with women who had one elevated BGL only. These data also suggest that a lower FBG level in combination with a lower 1-hour or 2-hour BGL (or both) may still identify patients at risk.

In absolute terms, for every 100 women in HAPO diagnosed with GDM using the new IADPSG criteria who have only one elevated BGL, nearly 50 do not meet the agreed risk threshold. In these pregnancies, most cases of LGA are not due to GDM.13 Intervention in these pregnancies may not change the rate of caesarean delivery but the rate of induction may be increased. More babies born to women labelled with GDM will be admitted to specialised neonatal nurseries, but with no decrease in the rate of neonatal hypoglycaemia.2,3 It is not known whether any of the secondary benefits of intervention will apply to this cohort given the low rate of LGA attributable to GDM. It is not known whether these women are at increased long-term risk of developing type 2 diabetes (T2DM).

Discussion

This assessment suggests that the new diagnostic criteria are based on a questionable composite end point, resulting in more women being diagnosed with GDM. This increase in numbers is predominantly due to women having a minimally elevated FBG level or 1-hour BGL only. Many of these women do not reach the agreed arbitrary risk threshold for diagnosis. Their long-term risk of T2DM is unknown.

There are many consequences of this. Initially, GDM was based on the risk of developing T2DM.15 It is recommended that women with GDM are assessed regularly for the development of diabetes, resulting in increased long-term costs and inconvenience, especially if they are not actually at risk. Additionally, women may be charged higher life insurance premiums because of a prior diagnosis of GDM.

It is reported16 that the NIH consensus panel has recommended as a priority that a new randomised controlled trial is conducted that evaluates outcomes in women currently classified as “normal” according to current US criteria but meeting IADPSG5 or American Diabetes Association criteria.6 A similar study is suggested for Australia. (William Hague, Head, Obstetric Medicine Group, School of Paediatrics and Reproductive Health, University of Adelaide, personal communication, June 2014.) If these studies enrol women who have only one minimally elevated BGL, there will be fewer adverse events than anticipated with little chance of demonstrating benefits from intervention. Large sums of money and investigator time and effort may be wasted undertaking these trials.

The risk threshold for diagnosis is an arbitrary decision. It has been suggested that the OR threshold could be raised to 2.0 to reduce patient numbers.17 If this threshold is based on the same combined outcome measure, and no adjustment is made for the effect of normal BGLs on risk, the same problems discussed here will arise. This would be unnecessary if an improved diagnostic approach is adopted. Improved identification of at-risk patients could justify adopting a higher OR threshold.

The most clinically recognised complication of GDM is macrosomia, assessed in HAPO as birthweight > 90th percentile (LGA). This analysis suggests that women who have only one elevated BGL parameter require a higher threshold to ensure with 95% confidence that they meet the agreed risk threshold.5 Multiple diagnostic rules based on whether a patient has one, two or three BGLs exceeding different threshold levels would overcome this problem but would be unduly complicated. Alternatively, a risk calculator based on the HAPO data could be an excellent solution. However, the approach recommended by IAPDSG is compellingly simple and a modification of this would seem to have merit.

Recommendation

A statistically justified strategy would be to make some minor modifications to the IADPSG criteria. The currently recommended IADPSG criteria correctly identify most patients with two or more elevated BGLs, including the FBG level. In fact, lower FBG level and post-challenge BGLs in combination could be justified, as the mean OR was significantly elevated above the threshold (mean OR, > 3.0 for all subgroups; lower limit of the 95% CI, > 2.0; Box 2). To establish the diagnosis on the basis of only one elevated BGL, the thresholds must be higher than the currently recommended IADPSG levels to correctly identify women at risk. These thresholds should equate to the lower limit of the 95% CI corresponding to an OR of 1.75 for LGA. This approach would more accurately identify these patients.

1 Primary and secondary outcomes for fasting blood glucose (FBG) testing in the Hyperglycemia and Adverse Pregnancy Outcome study

Outcomes*†

95% CI

Primary outcomes

Cord C-petide levels > 90th percentile

2.02

1.85–2.21

Birthweight > 90th percentile

1.68

1.56–1.80

Neonatal hypoglycaemia

1.24

1.05–1.46

Primary caesarean section

1.18

1.11–1.26

Secondary outcomes

Pre-eclampsia

1.40

1.26–1.56

Shoulder dystocia

1.30

1.07–1.58

Preterm delivery

1.16

1.05–1.28

Hyperbilirubinaemia

1.00

0.92–1.09

Neonatal intensive care

0.99

0.91–1.08

OR = odds ratio. * Fully adjusted calculated ORs (95% CIs) at the FBG threshold; adapted from International Association of Diabetes and Pregnancy Study Groups Consensus Panel (online Appendix).5 † Neonatal adiposity > 90th percentile is not shown as it was not a prespecified primary or secondary outcome measure.11

2 Risk of large-for-gestational-age (LGA) birthweight among women with one, two or three elevated blood glucose levels (BGLs) in the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study*

FBG = fasting blood glucose. * The odds ratio (OR) and 95% confidence intervals for LGA by BGL subgroups in the HAPO study are adapted from Metzger and Dyer.1 3 Also shown is the International Association of the Diabetes and Pregnancy Study Groups OR risk threshold of 1.75 for diagnosis of GDM (dotted line).

The efficacy of bariatric surgery performed in the public sector for obese patients with comorbid conditions

Obesity (body mass index [BMI] ≥ 30 kg/m2) is a growing health problem and is recognised as one of the largest contributors to the chronic burden of disease. Currently, 28% of Australians are obese, placing our nation second for men and fifth for women among OECD (Organisation for Economic Co-operation and Development) countries ranked by prevalence of obesity.1,2 In Australia, an inverse relationship exists between high obesity prevalence and low socioeconomic status; incidence is almost double for areas indexed as the most disadvantaged compared with areas within the highest strata.3 Those living more remotely also have higher obesity rates — 59% versus 32%.2 Among Indigenous Australians, the prevalence of obesity is almost double (34% versus 18%) and that of type 2 diabetes mellitus (T2DM) is triple that of non-Indigenous adults, resulting in sevenfold greater mortality due to diabetes.2,4

Bariatric surgery is an effective treatment of severe obesity (class III [BMI ≥ 40 kg/m2] or class II [BMI ≥ 35 kg/m2] with comorbid conditions) and is purportedly cost-effective, compared with conservative measures.5–7 The evidence supports sustained postoperative weight loss, ameliorating obesity-related comorbid conditions. The Swedish Obesity Study showed a marked reduction in hypertriglyceridaemia, T2DM and hyperuricaemia with surgery after 2 and 10 years.6 The resultant postsurgical weight loss substantially reduces, resolves and even prevents the metabolic complications associated with increasing central adiposity, with 73%–95% T2DM remission rates by 2 years, depending on the type of surgery.8,9

The heavy economic burden of obesity and its comorbid conditions may be alleviated in the long term by surgical management, despite upfront resource costs.10 Severely obese individuals incur twofold higher mean annual health care costs ($2788 v $1472) and use double the number of medications annually (11.4 v 5.3 per person) compared with the general population.7,11 Weight loss surgery can reduce the number of medications required and lower individual health care costs by 26%, a direct saving of $506 per person.7,12 With evidence of reduced mortality and an acceptably low complication profile (estimated mean 30-day mortality < 0.3%), bariatric procedures are supported by national and international health bodies.13–15

The National Health and Medical Research Council (NHMRC) 2013 guidelines recommend bariatric surgery as the most beneficial and cost-effective management for motivated individuals with severe obesity.16 Motivation is an essential factor in considering whether an individual with obesity and comorbid conditions is a suitable candidate for surgery. Surgery is not the final step in the clinical pathway of severe obesity management; postoperative commitment to lifestyle change and regular follow-up are requisite for successful weight loss and continued improvement in health.16 Most individuals are motivated to have surgery for greater control of medical ailments, yet it is unknown whether this applies to those whose procedure is fully funded.17

Since 1992, Medicare has reimbursed the cost of bariatric surgery in the private sector. As most surgery is carried out in private hospitals with large out-of-pocket expenses for those without private health insurance, a significant inequity in obesity management exists.18 Paradoxically, this surgery is least accessible to those who are likely to be in greatest need. Since October 2009, a pilot program has been underway in the Sydney and South Western Sydney Local Health Districts in which bariatric surgery has been publicly funded for a limited number of patients meeting strict inclusion criteria. There are no Australian studies to date that have identified if publicly funded surgical intervention for severe obesity confers the same health benefits seen in private health care. This study aims to assess the efficacy of bariatric surgery for such patients in the Australian public health system.

Methods

Study design

Sixty-eight moderately to severely obese participants with comorbid conditions were deemed eligible for bariatric surgery, and inclusion in our study. Participants were attendees at an ongoing collaborative pilot program run by one of three obesity clinics based within the Sydney and South Western Sydney Local Health Districts. All participants received conservative management for their obesity and related health conditions delivered by a multidisciplinary team trained in obesity management. Participants were seen 6- to 12-weekly for dietary advice, behavioural modification, advice on physical activity and, where appropriate, very low energy diets (VLEDs) and pharmacotherapy. Participants with substantial obesity-related comorbid conditions who had not achieved adequate weight reduction with conservative intervention, and who were interested in surgical management, were assessed by their multidisciplinary team to determine suitability for bariatric surgery. Longitudinal data were collected on those individuals who underwent bariatric surgery for management of resistant obesity and associated comorbid conditions.

The inclusion criteria for bariatric surgery were: age 18–75 years, minimum class II obesity (BMI ≥ 35 kg/m2) with comorbid conditions, completion of at least 1 year of medical intervention during which the participant had demonstrated commitment to lifestyle change, and surgery before 31 May 2013 (enabling a minimum follow-up period of 3 months). Exclusion criteria were: inability to consent, irreversible endocrine causes of obesity and significant comorbid conditions that were expected to result in poor outcomes from surgery, such as unstable cardiovascular disease and uncontrolled psychiatric illness. Participants underwent 2 weeks of VLED before the surgery to reduce abdominal adiposity and liver volume in particular, as liver size can complicate surgical access.19 They were also instructed about the necessary commitment to postoperative care, specifically the importance of nutritional input. Two of us (C J T and D J M) performed either laparoscopic sleeve gastrectomy (LSG) or laparoscopic adjustable gastric banding (LAGB) at a single public hospital. The type of surgery performed was decided by the participant in conjunction with the surgeon, after being briefed on the techniques, risks and expected benefits of each procedure. All participants were screened for nutritional deficiencies before and after surgery.

Ethics approval was obtained from four Human Research Ethics Committees within the Sydney and South Western Sydney Local Health Districts. All participants gave written consent for use of their records for research purposes.

Clincial and biochemical assessments

Surgeries were performed on an as-needed basis from 1 October 2009 to 31 May 2013. Pre- and postoperative results were analysed at seven time points until 1 September 2013. The baseline (preoperative) measurements were recorded at the start of the VLED, at least 2 weeks before surgery (time zero). Postoperative data were recorded from follow-up appointments at 3, 6, 12, 18, 24 and 36 months. The primary end points analysed were weight, BMI and waist circumference, and secondary end points were four common obesity-related comorbid conditions: T2DM, dyslipidaemia, hypertension (HTN), and obstructive sleep apnoea (OSA). Improvement was defined by normalisation of laboratory markers (T2DM and dyslipidaemia), blood pressure sphygmomanometry (HTN) and polysomnography results (OSA). Comorbid conditions were considered “partially resolved” when participants’ measurements fell within normal limits, the number of medications was reduced or the use of the continuous positive airway pressure (CPAP) device was discontinued. Comorbid conditions were deemed “fully resolved” when normal measurements remained after all relevant medications were discontinued.

Definition of comorbid conditions

T2DM was defined as fasting blood glucose levels ≥ 7 mmol/L, glycated haemoglobin level ≥ 6.5% and/or requirement of at least one oral hypoglycaemic agent. HTN was defined as systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg and/or requirement of at least one antihypertensive agent. Dyslipidaemia was defined as total cholesterol level ≥ 5.0 mmol/L, low-density lipoprotein level ≥ 3.5 mmol/L, triglyceride level ≥ 2.0 mmol/L, or high-density lipoprotein level ≤ 1.01 mmol/L for men and ≤ 1.3 mmol/L for women, and/or requirement of at least one lipid-lowering medication.20,21 OSA was defined according to symptoms (snoring, observed apnoea, daytime somnolence), treated HTN and the apnoea–hypopnoea index record from polysomnography.

Statistical analysis

The results at the seven time points were reported as mean and SD. The differences were tested for significance using paired t tests for continuous variables and the McNemar test for paired categorical variables. The t test was used to evaluate between-surgery differences in mean weight at baseline and subsequent time points, to assess the validity of combining the two surgery types in analyses. Mixed-model regression was used to determine if significant change in weight occurred over time at all seven time points, to account for reductions in sample size with time. All statistical analysis was performed using SAS version 9.3 for Windows (SAS Institute).

Results

Participant characteristics

Of the 68 participants offered surgery, two declined and one moved interstate. Sixty-five participants (41 women and 24 men) with a mean age of 51.5 years (SD, 11.8 years; range, 21–73 years) underwent bariatric surgery between 1 October 2009 and 31 May 2013. The level of comorbid conditions in the group at baseline was high, with a mean of eight conditions per patient and class III (severe) obesity. From baseline to 18 months, only three participants were lost to follow-up at each time point; however, numbers are lower with time as operations were done at different times (ie, only 10 patients had surgery 36 months ago whereas 65 had surgery 3 months ago, 58 had surgery 6 months ago, etc [see Box 1]). Data on all participants who completed follow-up at each respective time point according to surgery date are presented in Box 1 and Box 2.

Intraoperative outcomes

All 65 surgeries were performed laparoscopically and none required conversion to open surgery. The most commonly performed procedure was LSG (57 patients; versus eight patients who had LAGB). Despite it appearing that greater weight change occurred in the LSG group, the difference was not significant at 3 months (P = 0.58), 12 months (P = 0.25) or 24 months (P = 0.17). There were no significant intraoperative complications (one participant had a haematoma that resolved with immediate evacuation) and length of hospital stay was within expected times for all patients (LSG, three nights; LAGB, one night).

Short-term postsurgical outcomes (3–6 months)

There was a reduction in all primary end points in the early postoperative period. By 3 months there was a mean weight reduction of 22.6 kg (SD, 9.5 kg) (Box 3) with a 7.4 kg/m2 mean reduction in BMI and 14.5 cm mean reduction in waist circumference (Box 1). All comorbid conditions showed partial resolution from 3 months (Box 2). Of the numbers of participants who had each comorbid condition at baseline and who had data available, there was full resolution by 6 months in 20/45 with T2DM, 14/43 with HTN, 7/47 with dyslipidaemia and 17/41 with OSA (Box 2). The requirement for all relevant medications also reduced accordingly (Box 4).

Medium-term postsurgical outcomes (12–18 months)

By 1 year, there was a significant reduction in all anthropometric measures. The mean weight loss was 34.2 kg (SD, 20.1 kg), and the mean BMI was reduced by 12 kg/m2 to 36.2 kg/m2 (SD, 7.7 kg/m2). By 18 months, of those who reported the respective comorbid conditions at baseline, almost half of the group had full remission of T2DM and three-quarters had resolution of OSA. Modest resolution of dyslipidaemia was seen with both a significant rise in high-density lipoprotein and reduction in triglycerides occurring transiently at 6–12 months in 7 of 47 patients who had dyslipidaemia at baseline (P < 0.001).

Long-term postsurgical outcomes (24–36 months)

The trend for continued weight loss and partial or full resolution of comorbid conditions appears maximal at 12–24 months. While the smaller group sizes at the 24- and 36-month follow-ups provided less power for statistical analysis, primary and secondary end points remained stable (P < 0.001). The maximal amount of mean weight loss occurred at 24 months (39.9 kg; SD, 31.4 kg). Three-quarters of the cohort had reduced the number of medications taken for HTN and T2DM by 18–24 months, and by 36 months, all had reduced the number of antihypertensive medications (Box 4).

Postoperative complications

Transient nutritional deficiencies were found in three of 65 participants, improving within 3 months with replacement treatment. Of the eight patients who had LAGB, one underwent conversion, due to gastric band erosion, to a Roux-en-Y bypass at 18 months. Of the participants who had LSG, one had a small bowel obstruction secondary to adhesions and required laparoscopic division at 2 years. Three participants required endoscopic stomach dilatation and a further three continued to experience nausea, dyspepsia and/or vomiting at 3 months, but this resolved by 6 months for two and at 12 months for one participant.

Weight change over time

A total of 288 observations for weight from 65 participants were available for linear mixed-model analysis. Significant change in mean weight was observed over time (P < 0.001). Mean weight adjusted for baseline weight decreased from 136.5 kg at baseline to 101.1 kg at 12 months. Stratifying the analysis by sex showed that men had a higher mean baseline weight, 147.5 kg (SD, 36.9 kg) compared with women, 130.1 kg (SD, 23.8 kg), but no difference was observed after 1 year (Box 3).

Discussion

In our study, the significant weight lost by obese participants who underwent bariatric surgery occurred early and was sustained over the first 3 years. The mean maximal weight loss was almost one-third of participants’ preoperative weight and was achieved by 24 months. Despite decreasing sample sizes, our results showed sustained weight loss at 36 months.

There was full or partial resolution of all comorbid conditions tested, except dyslipidaemia, in most participants by 2 years. Different rates of resolution of comorbid conditions occurred, with remission of T2DM being the earliest. By 24 months, there was an associated resolution of T2DM, HTN and OSA, or the need for pharmacotherapy or devices was reduced in most patients, yet dyslipidaemia showed inconclusive results. These findings parallel other bariatric studies.22,23 The Swedish Obesity Study found isolated improvements in hypertriglyceridaemia postoperatively (including after LAGB), although the most consistent improvements in lipid profiles have been seen after Roux-en-Y bypass surgery.6,24

Most LAGB procedures were performed in the first year of the study, until LSG became a preferred option due to its greater weight loss profile with minimal additional complications and the lower need for surgical follow-up. However, throughout the study, participants had autonomy of procedure choice (on surgeon recommendation). While this may suggest an avenue for bias (the participants chose their “intervention”), bariatric surgery is an elective procedure, and it was deemed necessary to replicate the protocol in the private sector. The limitations of our study include its small size, that it is a case series, that it is non-blinded, possible bias due to participants choosing their surgery type, and the use of a clinical database with the possibility of missing or inaccurate values imputed by health professionals. The low rates of adverse events observed in our study are consistent with those reported in the literature.

The direct operative costs of performing these surgeries in the public sector in our study were estimated to be $7000–$9000 (LSG incurred greater up-front theatre costs than LAGB). Perioperative costs, including 2 years of postsurgical visits, may approach $2000 per person, taking the total cost to $9000–$11 000 per person. A 2005 paper reported the annual cost of managing an individual with T2DM as $9095–$15 850.25 Thus, if an obese person with T2DM has bariatric surgery, the operation would pay for itself after about 1 year.

The participants’ baseline characteristics are representative of a typical demographic seen in the public sector, supporting extrapolation of our results to a wider population. The high attendance rate at 2 years shows that the cohort demonstrated adequate motivation to justify surgical intervention. The health potential from bariatric surgery ranges from improved quality of life and amelioration of comorbid conditions to full resolution of complications and reduced mortality for all individuals, paying or not. Strategies to prioritise access are therefore recommended to reduce the apparent inequality that exists. Limited access to surgery discriminates against those who cannot afford the out-of-pocket costs, yet it is likely that this subgroup would benefit most. In conclusion, we hope that our study provides an evidence base for the surgical treatment of obesity in the public health system and, in turn, that consideration will be given to increasing the supply of publicly funded bariatric surgery in Australia.

1 Patient anthropometric data before and after bariatric surgery

Measurement

Baseline (2 weeks before surgery)

Postoperative month 3

Postoperative month 6

Postoperative month 12

Postoperative month 18

Postoperative month 24

Postoperative month 36

N (n)

65 (65)

55 (58)

49 (52)

30 (33)

17 (23)

7 (10)

BMI (SD), kg/m2

48.2 (9.5)

40.8 (8.3)*

38.9 (7.9)*

36.2 (7.7)*

38.2 (12.1)*

35.7 (7.7)*

38.7 (9.4)†

Weight (SD), kg

136.5 (30.3)

113.9 (25.2)*

108.3 (24.8)*

101.1 (22.4)*

99.3 (20.8)*

97.6 (21.9)*

108.6 (25.6)†

Waist circumference (SD), cm

132.2 (16.5)

117.7 (15.2)*

114.9 (14.7)*

109.4 (13.7)*

110.2 (12.3)*

108.4 (15.8)*

114.9 (14.6)*

% weight loss from baseline

17%

21%

26%

27%

29%

21%

BMI = body mass index. N = total number of participants who attended follow-up. n = total number of participants including those lost to follow-up. * Comparison between baseline and follow-up; P < 0.001. † Comparison between baseline and follow-up; P < 0.05.

2 Proportions of patients* who continued to have conditions found at baseline

Comorbid condition

Baseline (2 weeks before surgery)

Postoperative month 3

Postoperative month 6

Postoperative month 12

Postoperative month 18

Postoperative month 24

Postoperative month 36

T2DM

36/52†

25/45†

21/42†

12/23†

8/17‡

3/6

HTN

45/51‡

29/43†

18/40†

10/27†

4/12‡

2/6‡

Dyslipidaemia

45/58‡

40/47

33/42

13/24‡

8/15‡

4/6

OSA

37/41

24/41‡

15/41†

6/24†

4/17‡

2/5

OSA requiring CPAP

18/27

14/23‡

8/22†

4/13‡

2/8‡

1/3

CPAP = continuous positive airway pressure. HTN = hypertension. OSA = obstructive sleep apnoea. T2DM = type 2 diabetes mellitus. * Numerators are number of participants who had that condition at that time point. Denominators are total number of participants who had that condition at baseline and had data at that time point. † Comparison between baseline and follow-up; P < 0.001. ‡ Comparison between baseline and follow-up; P < 0.05.

3 Patients’ weight change over time after bariatric surgery*

* Weight change at all seven time points is significant (P < 0.001). Despite differences in weight lost for men and women, there were no significant between-sex differences.

4 Number of patients with a reduction in medications used at each time point after bariatric surgery

	Antihypertensive medications		Glucose-lowering medications		Lipid-lowering medications
Postoperative month	Patients at each time point	Patients with reduced use of medication	Patients at each time point	Patients with reduced use of medication	Patients at each time point	Patients with reduced use of medication

3	65	31	60	28	64	6
6	54	34	54	29	53	10
12	48	37	47	25	44	12
18	29	22	23	17	24	12
24	16	12	16	15	13	3
36	6	6	6	4	4	1

Working knowledge of diabetes

THIS PUBLICATION is now in its fifth edition. It was first published in 1996, with the next two editions being published over the next 13 years, but the last three editions have been published since 2009, presumably reflecting the dynamic changes in diabetes recently.

As its title suggests, the book sets out to succinctly provide facts about diabetes. In this regard it is very successful. The authors do not provide detailed and in-depth reviews of diabetes, its complications and its management recommendations, which would clearly be inappropriate for this type of book. It is designed instead to educate readers who have a low level of knowledge of diabetes. The book gives the reader a good basic understanding about all aspects of diabetes and its management. It is written in a very easy style which is a pleasure to read and is appropriate for its target audience.

The book is well presented with 12 chapters covering all aspects of diabetes from epidemiology to complications and management. Each chapter has useful summary tables and key references, and excellent illustrations, figures and pictures. The section on retinopathy stands out in this regard. The chapters cover each topic adequately, providing succinct summaries of issues rather than heavy, complex, extensively referenced discussions.

The book would be ideal for general practitioners, medical students, allied health professionals and nursing staff who are commencing work in diabetes and are keen to acquire a sound level of knowledge. A health professional who has been working in diabetes for several years would find the book a little superficial, but it would potentially be ideal for some patients and their relatives to understand diabetes and its management. Indeed, there could be a major role for this book in patient education, although the language used is probably more suitable for persons with a basic medical knowledge. All in all, this is a very useful book on diabetes for all health professionals involved in the management of persons with diabetes.

Newborn bloodspot screening: setting the Australian national policy agenda

The recent article by Maxwell and O’Leary1 is timely in outlining the obstacles to introducing newborn screening tests in Australia, and the need for a nationally consistent approach, where the benefits of screening are proven. These obstacles exist despite clear policy developed by the professional newborn screening community.2

The absence of newborn screening for congenital adrenal hyperplasia (CAH) is the clearest example of the impact from the absence of any national mechanism, where initiatives to introduce such testing have bounced between state and federal bodies for many years, despite clear evidence of benefit.3 It is likely that a number of Australian children have died as a result of missed diagnoses while these initiatives have floundered.4 In addition, the incidence of CAH in Aboriginal children is about 2.5 times that in non-Aboriginal children, suggesting an even greater need for national screening.4

CAH newborn screening has benefits additional to reduced mortality. There is a great difference for families in taking onboard the complexities of managing a child with CAH who is well, having been diagnosed through newborn screening, rather than a critically unwell neonate unnecessarily in adrenal crisis in the intensive care unit. Those of us who manage children with CAH in New Zealand, where screening exists, have observed this crucial difference.

Working groups and governmental announcements supporting screening for CAH are welcome; however, action is required now to introduce newborn screening for CAH Australia-wide. All that is required is political will, without which more Australian children will die unnecessarily.

Black bones: minocycline-induced bone pigmentation

An 82-year-old man with bilateral knee osteoarthritis underwent consecutive total knee arthroplasty 5 months apart. During both procedures, he was noted to have black subchondral bone with otherwise normal architecture and normal-coloured cancellous bone. At the time of surgery, bone specimens sent for pathology testing were histologically normal. The patient had been treated with minocycline for rosacea for 7 months before the first procedure. Minocycline is an uncommon cause of skeletal pigmentation and is not known to affect bone quality.1 Discolouration may also be owing to ochronosis, metal deposits, sequestrum and metastatic disease.2

Who’s responsible for the care of women during and after a pregnancy affected by gestational diabetes?

Gestational diabetes mellitus (GDM) is the strongest single population predictor of type 2 diabetes,1 and current Australian prevalence is 10%–13%, depending on the criteria used.2 Poor health outcomes extend to children of mothers who had GDM, due to increased risk of obesity and abnormal glucose metabolism during childhood, adolescence and adulthood.3

Antenatal lifestyle intervention is shown to improve short- and long-term maternal and infant health outcomes.3 In addition, it can effectively prevent type 2 diabetes among women who have had GDM.1 However, although some centres of excellence exist, in many cases, antenatal care is not delivered systematically.4

After their babies are born, women who have had GDM can be described as falling into a health care “chasm”.5 When these women leave hospital, their obstetricians and endocrinologists feel that their work is done. Lack of coordination between the hospital and primary care sectors can mean that no one assumes responsibility for the care of these women.

The opportunity to prevent or delay type 2 diabetes in this high-risk population through primary care was noted more than a decade ago.6 However, defined care pathways and coordination remain elusive; implementation of evidence has not occurred. In many cases, general practitioners may not be aware that the woman has had GDM, and may not have a clear pathway directing responsibility for follow-up care.

There is an urgent need to implement a widespread and coordinated approach to prevent progression to type 2 diabetes in this population. Rectifying this situation requires cooperation and collaboration between all care providers.

Antenatal care: navigating the new gestational diabetes landscape

The health care sector operates under guidelines with conflicting content and differing levels of comprehensiveness and professional endorsement (Box). The Australasian Diabetes in Pregnancy Society (ADIPS) recently released revised consensus guidelines for testing and diagnosing GDM in Australia and New Zealand.7

Women with GDM are managed in hospitals because they are identified as having pregnancies at higher risk of adverse outcomes. The ADIPS guidelines recommend an oral glucose tolerance test (OGTT) for all women (unless already diagnosed with GDM in early pregnancy) at 24–28 weeks’ gestation.7 These guidelines were informed by several studies, including the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study, which indicated a strong continuous association of maternal glucose levels with increased diabetic fetopathy.14

A change to testing protocols will be introduced in July 2014 and diagnostic criteria on 1 January 2015 (Aidan McElduff, Clinical Associate Professor of Medicine, University of Sydney and ADIPS President; personal communication). Concerns exist about their potential workload implications and evidence base.

Health service and pathology database analyses have resulted in equivocal projections of the potential workload increases; it is most likely that many will see a doubling of cases.2,15 Workload projections can be difficult, as true prevalence is not known, but it has been suggested that the increasing rate reflects the prevalence of abnormal glucose metabolism in the general population.16

In considering potential workload costs and changes, we need to consider the results from two well executed randomised controlled trials, which demonstrated that treatment of GDM can prevent adverse outcomes.17,18 For most women (80%–90%), GDM could be managed through dietary counselling delivered by a dietitian. In some centres, this proportion will be lower, depending on population characteristics. Medical nutrition therapy is a cornerstone intervention for women with GDM,19 and its appropriate delivery results in reduced insulin requirements and improved blood glucose control.19 However, systematic, evidence-based dietetic care of women with GDM does not occur in many centres in Australia.4 Australian health services require clinician leadership and commitment to partnership and change in (re)allocation of resources to support a multidisciplinary team in providing evidence-based care for improved maternal and infant outcomes.

Some clinicians raise concerns about diagnostic criteria changes based on observational study outcomes, but the previous diagnostic criteria were the product of an ad-hoc working party and lacked the strong evidence base that underpins the current criteria.20

Postnatal follow-up: who’s taking responsibility?

Australian guidelines recommend that all women who had GDM should undertake a 75 g OGTT between 6 and 12 weeks after delivery.7 International guidelines also highlight the importance of lifestyle modification, breastfeeding, birth control and risk counselling to improve health outcomes for these women and their children.12,13

The extent to which these recommendations are integrated into postnatal GP visits is not known, but some studies suggest diabetes testing is suboptimal.21 Self-report surveys of women with prior GDM indicated that about half of participants returned for OGTTs, but only a quarter in the appropriate period.21,22 The potential use of glycated haemoglobin testing instead of the OGTT appeals to many, but the approach may not change until it is approved on the Medicare Benefits Schedule.

Appropriate strategies to engage women in screening are paramount, as the motivation to manage a GDM diagnosis transforms to apathy once GDM resolves.23 Barriers to ongoing screening include a lack of awareness of the need for screening, difficulty attending screening with an infant, dislike of the OGTT process, being a mobile population, and inconsistent advice from health care providers about testing, lifestyle modification and risk.21–2 3 Findings from the United Kingdom suggest that health care professionals need to balance between reassurance of likely resolution of GDM and adequate information about potential progression to type 2 diabetes.2 3 Perception of risk is an important motivator; a lack of perceived risk of developing type 2 diabetes is common and can be related to timing, content and tone of messages.2 3,2 4

Prevention of diabetes in primary care

Which guidelines?

Three Australian guidelines exist for the follow-up of women who are at risk of type 2 diabetes (Box).7–9 Their core messages are similar, but they vary in several areas, diluting GP awareness and implementation. Beyond the timing of testing regimens, recommendations regarding lifestyle interventions to prevent type 2 diabetes progression are absent from the ADIPS guidelines, but the Diabetes Australia/Royal Australian College of General Practitioners (RACGP) Diabetes management in general practice 2014–20158 and Guidelines for preventive activities in general practice (the “red book”; also distributed to GPs in Australia)2 5 outline diabetes management and dietary advice for diagnosed cases in general practice and for diabetes prevention.

Many similarities exist between the diet for GDM and diabetes prevention (ie, focus on low glycaemic index, low saturated fat, high fibre content). However, during a pregnancy complicated by GDM, there is a major focus on tightly controlled blood glucose levels, although appropriate diet quality for pregnancy requirements and gestational weight gain is also paramount. By contrast, diabetes prevention diets have a greater focus on weight reduction. Currently, there is no effort to explain to women who have had GDM the difference in approach.

A missed opportunity?

Although GPs view follow-up care as their role within the broader context of general health screening and promotion, this is often opportunistic.2 6 Advice from GPs is a powerful motivator for women to adopt lifestyle modification.27 However, GPs report not being well versed in guidelines for GDM follow-up care, potentially reflecting the lack of clarity in the literature and their varying knowledge and confidence in provision of lifestyle advice and interventions.28 GPs generally give appropriate exercise advice, but can be less clear about dietary or weight loss goals.26

These practices are reinforced by systems and process barriers of prioritisation of issues during a consultation, a lack of integration of recall tools and intervention resources in daily workflow, and uncertainty about responsibility for screening, as well as poor communication between secondary and primary care sector and fragmentation of pre- and postnatal care services.28

Right information, right people, right time

Clinical trials have demonstrated that lifestyle modifications with weight loss and moderate exercise can reduce the incidence of type 2 diabetes by up to 58% for people at high risk, with an impact still evident 8 years from the intervention onset and 4 years after the active intervention ceased.29 Real-world implementation in the Australian health care system has achieved 40% reduction in the risk of progression to diabetes.30

Agreement between and willingness to work in partnership with key stakeholders — such as ADIPS, Diabetes Australia, the RACGP and the Royal Australian and New Zealand College of Obstetricians — is required for a collective approach to delivering diabetes prevention to this high-risk population.

However, despite convincing evidence about effective programs in Australia, postnatal support after a pregnancy with GDM is lacking and is without coordination. Interventions using technologies such as telephone, SMS and the internet have been trialled for diabetes care and may be useful in prevention. These must be underpinned by behaviour change theories and address barriers to making changes regarding future risk.23 Women have been identified as being receptive to messages several months after birth, which may align with “transition times” (eg, introduction of solids).23 Further efforts are urgently needed to develop lifestyle strategies that meet the specific needs of this group of women.

Diabetes Australia’s National Gestational Diabetes Register (NGDR), part of the National Diabetes Services Scheme, was launched in 2011 as a free service to women with a Medicare card to help those who have had GDM to manage their health and prevent progression to type 2 diabetes. One function of the NGDR is to send regular reminder letters to women and their GPs regarding diabetes checks (at registration, 12 weeks after birth, and annually thereafter). These reminder letters also include general information for the women and their families to help them continue a healthy lifestyle.

Although the NGDR outlines what testing to undertake, its potential to allow implementation and dissemination of a comprehensive, consolidated set of guidelines is perhaps underused. It could facilitate effective connection of women with a history of GDM with specific, effective, evidence-based lifestyle advice as well as clinical guidance for their GPs.

A call to action: the need for a collaborative approach

A clear pathway, developed between all stakeholders, with delineated roles and responsibilities to ensure that best-practice care is delivered along the continuum of antenatal, postnatal, interconception and longer-term care is required. Delivery of coordinated, effective programs is essential for this group of women. Without such clarity, and in the absence of a systems approach to care, we are failing to seize an opportunity to reduce the incidence of type 2 diabetes and promote the wellbeing of these women and their children. After a diagnosis of GDM, women view their GP as the most appropriate source of follow-up care,24 so it is imperative that GPs are given the right guidelines and education to advise these women about preventing or delaying progression to type 2 diabetes.

A comparison of current gestational diabetes mellitus diagnosis (GDM), treatment and follow-up guidelines

	Details of guidelines
Guideline/society (country)	ADIPS (Australia and New Zealand)7	Diabetes Australia and RACGP (Australia)8	Therapeutic guidelines: endocrinology (Australia)9	ACOG (US)1 0	ADA (US)1 1	NICE (UK)1 2	CDA (Canada)1 3
Antenatal testing protocol	Universal OGTT at 24–28 weeks; earlier if clinically indicated	Universal screening at 26–28 weeks. Two-step approach recommended (GCT then OGTT).	Universal GCT or OGTT at 26 weeks. Early screening if high risk	—	Universal OGTT at 24–28 weeks in women not previously diagnosed with overt diabetes	At 24–28 weeks if the woman has any risk factors or earlier if GDM in a previous pregnancy	Universal screening 24–28 weeks. Two-step approach preferred (GCT then OGTT)
Timing of first postpartum follow-up visit	6–12 weeks	6–12 weeks	6–12 weeks	6–12 weeks	6–12 weeks	6 weeks	6 weeks – 6 months
Which test(s) for postpartum screening	75 g OGTT	75 g OGTT	75 g OGTT	FPG or 75 g OGTT	75 g OGTT; not HbA1c	FPG	75 g OGTT
Who with?	—	GP	—	—	—	—	—
Frequency of follow-up and recommended test	Dependent on future pregnancy plans and perceived risk of type 2 diabetes, yearly OGTT if planning pregnancy. 1–2 yearly FPG (low risk); OGTT/HbA1c (higher risk)	3-yearly; with FPG	If postnatal test normal: annual fasting or random blood glucose or OGTT every 2 years and before subsequent planned pregnancies	3-yearly, as above	Minimum 3-yearly; with OGTT. If IFG or IGT, yearly	Yearly; no blood test specified	At least 3-yearly and before each pregnancy; not specified
Other postnatal advice included	No recommendations	Increase physical activity, weight loss/healthy diet. Refer to dietitian and/or physical activity program. Preconception advice.	Risk counselling for future type 2 diabetes. Lifestyle advice: diet/physical activity. Subsequent pregnancy: early screening 12–16 weeks repeated at 26 weeks.	Weight loss and physical activity counselling as needed	Women with a history of gestational diabetes found to have prediabetes should receive lifestyle interventions or metformin to prevent diabetes.	Lifestyle advice: weight control, diet and exercise	Lifestyle advice to prevent diabetes and cardiovascular disease should begin in pregnancy and continue postpartum. Encourage breastfeeding for at least 3 months postpartum. Provide risk and preconception counselling.

ACOG = American College of Obstetricians and Gynecologists. ADA = American Diabetes Association. ADIPS = Australasian Diabetes in Pregnancy Society. CDA = Canadian Diabetic Association. FPG = fasting plasma glucose. GCT = glucose challenge test. HbA1c = glycated haemoglobin. IFG = impaired fasting glucose. IGT = impaired glucose tolerance. NICE = National Institute for Health and Clinical Excellence. OGTT = oral glucose tolerance test. RACGP = Royal Australian College of General Practitioners. UK = United Kingdom. US = United States.

Primary care of women after gestational diabetes mellitus: mapping the evidence-practice gap

Gestational diabetes mellitus (GDM) is the strongest population predictor of type 2 diabetes mellitus,1 with the cumulative incidence of type 2 diabetes ranging from 2.6% to 70% from 6 weeks to 28 years postpartum.2 Women who have had GDM are also at greater risk of a recurrence of GDM, cardiovascular disease and metabolic syndrome.3

General practitioners have a key role in providing postpartum and long-term preventive health care.4–6 While appropriate care and preventive health approaches in the weeks and months after childbirth provide an opportunity to improve health outcomes for mothers and infants, there are few comprehensive, evidence-based guidelines available.7 Women who have had GDM, and their infants, are even more likely to benefit from proactive care during this period,8 and there are several guidelines that cater to this group. For example, Australasian Diabetes in Pregnancy Society (ADIPS) guidelines (current at the time of study)9 recommended an oral glucose tolerance test (OGTT) within 6–8 weeks (now 6–12 weeks10) of birth for women who had GDM. International guidelines also highlight the importance of lifestyle modification, breastfeeding, contraception and risk counselling to improve health outcomes for these women and their infants.11,12

There are several guidelines available to GPs when providing care after a GDM-affected pregnancy. Beyond the timing of testing regimens, recommendations regarding lifestyle interventions to prevent type 2 diabetes progression are absent in ADIPS, but the Diabetes Australia/Royal Australian College of General Practitioners (RACGP) Diabetes management in general practice,1 3 and the RACGP Guidelines for preventive activities in general practice (the “red book”)1 4 outline diabetes management and dietary advice for diagnosed cases in general practice and for diabetes prevention.

What informs, and the extent to which preventive health practices are consistently integrated into, postpartum GP visits in Queensland, are unknown.

We aimed to evaluate GPs’ awareness, perceived knowledge, and use of GDM guidelines, and to determine the extent to which care within the first 12 months postpartum of a woman with a history of GDM is delivered according to guidelines.

Methods

We surveyed GPs who participated in a shared care arrangement with a south-east Queensland maternity hospital and undertook a retrospective chart audit of their patient records for women who were provided with maternity shared care between July 2011 and June 2012. Data collection occurred throughout 2013.

Eligible GPs were identified from the hospital’s database and invited by mail to participate in the study. A week after the mail-out, practices were telephoned to explain the study requirements. Hard to reach or undecided practices were contacted by a GP member of the research team (some, numerous times).

Consenting practices were sent a survey, medical chart audit form, instructions, and a reply-paid envelope. Each practice was contacted after 1 week to confirm receipt and reiterate the instructions. Two follow-up reminders were made to non-responding practices by telephone at 2-week intervals. A gift voucher valued at $100 was offered as an incentive for participation.

The GPs were asked to complete a one-page self-administered survey regarding postpartum management approach of women with a history of GDM. Before distribution, the survey was pilot-tested with two academic GPs independent of the study. Practice managers and/or nurses completed a one-page audit for each identified patient medical chart. The audit form was developed and revised based on a review of related literature, pilot testing at one practice, and review by two GPs independent of the study. The time frame of the audit covered a review of all GP consultations in the 12 months after the birth of the baby. The audit took about 10 minutes to complete. Patient names were not included on audit forms.

Ethics approval was granted by the Mater Health Services Human Research Ethics Committee.

Outcome measures and data analysis

Outcomes in the GP survey included: awareness and usefulness of specific practice guidelines in addition to their nominated guidelines, information on use and effectiveness of postpartum reminder systems for patient follow-up, and information on recommended timing and type of test for postpartum diabetes testing.

Audit outcome measures were checking and recording of preventive health indicators including: blood testing for diabetes, weight, body mass index, blood pressure, breastfeeding, and mental health status. Other outcomes, such as provision of advice on contraception, diet, exercise and relevant referral to specialist or allied health services were also assessed.

Survey and audit responses were entered into SPSS, version 22.0 (SPSS Inc) and checked twice for accuracy. Descriptive statistics were calculated and reported as frequencies or medians and ranges.

Results

General practitioner survey

We identified 38 GPs from 35 practices who shared care for a woman with GDM (n = 43) within the study period. Of these, 18 consented and completed the questionnaire (47% response rate). No other demographic information was collected from participants.

Box 1 shows GPs’ ratings of their familiarity with, and their perceived usefulness of, various guidelines. All GPs were familiar with the hospital’s GP maternity shared care guideline and rated it somewhat useful, useful, or very useful.

GPs had excellent knowledge of which diabetes test to order and timing of testing postpartum, with 100% stating that they order the OGTT and recommend testing within 6 to 8 weeks.

Fifteen of the 18 GPs used reminder systems to monitor postpartum women with prior GDM, with all but one GP indicating that it worked well. Although three GPs did not use a reminder system, all used record systems that had the capacity to set up reminders and recall. One GP indicated that they did not think the reminder system worked well, and stated it was because they had to remember to click “reminders” in the electronic medical record system as it did not come up automatically and thought it was easy to miss. Other barriers to the reminder system working well included patient non-compliance and the patient’s choice as to whether to attend their follow-up appointment.

Chart audit

Eighteen GPs completed one chart audit and one GP completed two. The total number of completed audits was 19 (19/43 audits). No pregnancies were recorded during the 12-month postpartum period.

The median number of times that a woman consulted her GP during the year after her pregnancy with GDM was five (range, 1–14). All women visited their GP at least once in the 12 weeks after the birth. All women were offered type 2 diabetes screening by their GP (18/19) or the hospital (1/19). The most frequently ordered test was an OGTT (15/19). Other tests ordered included glycated haemoglobin (HbA1c) (1/19), fasting blood glucose level including full blood count (1/19) and electrolyte and liver function tests (2/19).

More than half (10/19) of the women had their OGTT ordered between 6 and 12 weeks (9/19 ordered between 6 and 8 weeks). The test was ordered earlier than 6 weeks for about one-third of the women (6/19), and after 12 weeks for two women. Of the women who had the OGTT performed, more than half had their OGTT between 6 and 12 weeks (10/19) (8/19 between 6 and 8 weeks). One woman had her test before 6 weeks, and three after 12 weeks. Five women did not have a test result recorded in their chart.

The chart audit indicated that each of the additional elements of care were recorded at least once in the 12-month postpartum period (Box 2). Body mass index, weight, diet, exercise and breastfeeding status were generally checked in the first 3 months, but not subsequently, while mental health status was checked within the first 3 months, and often had a second follow-up recorded. Blood pressure was checked regularly over the 12 months and contraception had more follow-up than other elements of care. Only one woman was referred to a dietitian.

Open-ended responses indicated each consultation generally focused on presenting symptoms or requests for tests or vaccinations.

Discussion

Our study demonstrated that GPs who participated in a shared care program with a major maternity hospital have an excellent awareness of the timing and practices around the OGTT for women who have had GDM. GPs in our study were informed by a range of guidelines, and placed a great emphasis on guidance from the maternity hospital with which they collaborate. Knowledge, opinions, and practices regarding other postpartum preventive health indicators also reflect behaviour previously documented among maternity patients and the wider population,1 7,1 8 with blood pressure measurement, and discussions about contraception and infant feeding/breastfeeding occurring in most consultations. Mental health assessments and discussions occur less often, and measurements and discussion around lifestyle indicators occur much less frequently.

In previous research, GPs reported not being well versed in guidelines for GDM follow-up care, potentially reflecting the lack of clarity in the literature and their varying knowledge and confidence in provision of lifestyle advice and interventions.19

Further, although GPs viewed follow-up care as their role and within the broader context of general health screening and promotion, it was often opportunistic.19 We found that women generally presented for another issue, rather than a post-GDM check-up, resulting in a conflict in priorities. Consequently, the discussion about preventive health measures, particularly within the context of limited consulting time and the current remuneration system, is often overlooked. These results are partially reinforced by previous findings that GPs generally give appropriate exercise advice, but can be less clear about dietary or weight-loss goals,19 despite the fact that advice from GPs is a powerful motivator for women to adopt lifestyle modification.20

The need for a systematic approach to delivery of care to women after a pregnancy affected by GDM is recognised.21,22 Another article in this supplement calls for a clear pathway and one recognised and endorsed guideline to ensure best-practice care is provided to this high-risk group by all health care professionals.21 A review of the topic noted that systems-based approaches are associated with a larger potential impact to improving testing rates, and appear easily generalisable.2 2 Essential elements of a system to increase a woman’s engagement with her GP in the postpartum period should be aligned with other key milestones in the postpartum period2 3 and involve proactively contacting patients.2 2 Diabetes Australia’s National Gestational Diabetes Register, part of the National Diabetes Services Scheme, already functions to remind women and their doctors of relevant testing and lifestyle modifications, and could be adapted to provide further clinical guidance to GPs.21

A targeted approach to translate guidelines into practice is required to complement the systemic approach to care, as awareness and dissemination of guidelines alone does not change practice.2 4 The assessment of influencing factors and implementation and evaluation design must be theory-driven.2 5 The implementation should address the knowledge gaps in guideline identification and content, and other barriers, such as time constraints and recall of new protocols. Recognition of other locally relevant barriers and enablers that facilitate implementation of clinical guidelines should also be identified and explored.

Despite our survey and chart audit returning a 47% response rate, slightly below the return rate noted for primary care surveys,2 6 it reflects the challenge of research in the context of the heavy workload of many practices. The chart audit demonstrated quite clearly the concerted efforts of many GPs to provide best-practice care to women in the postpartum period. Another limitation of this study is capping the analysis of the provision of preventive health care at 12 months. A longer follow-up period could have provided a stronger insight into the delivery of care to this patient cohort.

Our research provides new knowledge around care provided to women by GPs in the 12 months after birth. We have demonstrated that GPs from our surveyed cohort knew guidelines around the timing and type of test for women who have experienced GDM in their pregnancy, and our chart audit has demonstrated that this knowledge is translated into practice. Less ideal were the practices and beliefs around the provision of other preventive health behaviours. This problem exists due to the absence of a systems approach to care, resulting in a lost opportunity to work systematically to reduce the incidence of type 2 diabetes and promote the wellbeing of these women, who are at high risk of chronic disease, and their infants.

1 General practitioners’ ratings of their awareness and usefulness of various guidelines (n = 18)*

Guidelines or approaches used

Usefulness of the guideline/approach

Not familiar

Not useful

Somewhat useful

Useful

Very useful

Diabetes management in general practice: guidelines for type 2 diabetes 2012/13 (Diabetes Australia)1 3

2/16

5/16

4/16

5/16

Therapeutic guidelines: endocrinology, version 4 (2009)1 5

14/16

1/16

ADIPS consensus guidelines for the testing and diagnosis of gestational diabetes mellitus in Australia (Australasian Diabetes in Pregnancy Society, 2013)10

8/17

4/17

5/17

Gestational diabetes mellitus – management guidelines (ADIPS; 1998, 2003)9

9/16

3/16

2/16

Hospital discharge summary

1/18

7/18

4/18

6/18

Hospital general practice maternity shared care guideline (August 2012)

3/18

5/18

10/18

Other (please specify) — Diabetes in pregnancy: women’s experiences and medical guidelines1 6

1/1

* Missing responses were excluded from totals.

2 Preventive health care indicators recorded as discussed by a general practitioner with a woman who had gestational diabetes mellitus within 12 months of birth (n = 19)

	Body mass index recorded	Weight recorded	Blood pressure recorded	Mental health assessed	Breastfeeding status recorded	Contraception discussed	Diet discussed	Exercise discussed

Number of women with health care indicator recorded	4	6	14	7	18	15	9	8
Median time of first discussion, weeks postpartum (range)	2.5 (1–6)	2.5 (1–20)	2.0 (1–37)	6.0 (1–31)	2.0 (1–13)	7.0 (1–33)	7.0 (2–27)	6.5 (2–27)
Number of women with health care indicator recorded more than once in the 12-month period	2	3	8	3	10	4	3	3