Preventing all the complications of hyperglycaemia: not a straightforward task

Individualised multifactorial treatment strategies are required to optimise outcomes

Epidemiological studies have found positive relationships between measures of glycaemia and adverse fetal and maternal pregnancy outcomes,1 macrovascular disease2,3 and the microvascular complications of diabetes.2,3 The relationship between glycaemia and microvascular disease is more complicated, as there is an inflection point below which the risk is very low, but beyond which it rises sharply. The inflection point of the relationship between glycaemia values and the prevalence of retinopathy has been used to determine both the glucose tolerance test values and the glycated haemoglobin (HbA_1c) level used to diagnose diabetes (see Figure 1 in reference 3).

Epidemiological studies have also found that the prevalence of hyperglycaemia associated with pregnancy,4,5 macrovascular disease6 and microvascular complications7,8 is modified — usually increased — by other known risk factors. For pregnancy, the independent adverse effects of obesity and maternal weight gain are well documented,4,5 so that the conclusion reached by Abell and colleagues9 in this issue of the MJA is not surprising: “Poor glycaemic control is not wholly responsible for adverse outcomes, reinforcing the importance of other risk factors, such as obesity and weight gain.”

Although the association between hyperglycaemia and the adverse effects of diabetes is clear, an association does not establish cause and effect. However, randomised controlled clinical trials have shown the benefits of better glycaemic control in pregnancy for reducing the incidence of fetal adiposity and macrosomia, of maternal pre-eclampsia, and of microvascular complications in both type 1 and type 2 diabetes. An unexpected finding from long term follow-up studies was that improved glycaemic control early in the course of diabetes had a very long lasting benefit with regard to microvascular and macrovascular complications in both type 1 and type 2 diabetes, even though improved control had been lost at the end of the intervention period.10–12 This long lasting effect has been labelled the “legacy effect”, or “metabolic memory”.

With regard to macrovascular disease, the situation is more complex, particularly in type 2 diabetes. Long term follow-up of the United Kingdom Prospective Diabetes Study7 and of the Diabetes Control and Complications Trial for type 1 diabetes11 found that better early glycaemic control in newly diagnosed patients reduced the long term risk of cardiac infarction or of dying from any cause. On the other hand, more intensive treatment of pre-existing type 2 diabetes in high risk patients in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial was not associated with any reduction in macrovascular disease incidence or mortality in either the long or short term.12 However, a small single-centre study in a different health system showed that a multifactorial intervention in a population at risk of cardiovascular disease, including better glycaemic control (but less than that achieved in the ACCORD trial), resulted in a significant decrease in mortality.12

To add to the complexity, the initial effect of better glycaemic control is to increase the risk of retinopathy. Although this risk declines with time,11 the initial deterioration can threaten vision and may require therapy.

From the above, it can be seen that the benefits of better glycaemic control are not simple, nor are the effects always positive or immediate in onset. Classical cardiovascular risk factors influence the risk of both macrovascular and microvascular complications in patients with diabetes.6,7 Multifactorial interventions that treat the classical macrovascular risk factors of hypertension, hypercholesterolaemia and smoking have been shown to reduce the incidence of both microvascular and macrovascular complications.13

Do these factors influence our treatment of diabetes? What degree of glycaemic control should be targeted? The Australian Diabetes Society recommends that targets for glycaemic control be individualised.14 Factors that will influence the target include the risks to the patient associated with the degree of hyperglycaemia and the potential benefits of tighter glycaemic control, as well as the risks associated with tighter control. In particular, patients with reduced life expectancy (particularly older patients) and those with a high risk of cardiovascular disease probably require less tight glycaemic control. Treatment should, of course, be holistic, including attention to any other relevant risk factors.

The legacy effect6,7,10 should strongly influence the therapeutic targets, particularly in younger, recently diagnosed patients, who can often achieve excellent glycaemic control with little risk of hypoglycaemia. It is crucial that the treatment targets and rationale should be discussed with the patient.

The problem with modern endocrinology

Improved doctor–patient communications and more research are necessary

The problem with endocrinology is that patients and doctors sometimes talk a different language. This is because there are many areas of uncertainty. How aggressively should the management of diabetes or osteoporosis be pursued? Should borderline thyroid function tests or testosterone be treated? What is the best diet for a patient with type 2 diabetes?

A patient with diabetes wants to know if he is going to die young, go blind or end up on dialysis. The doctor tells him that his glycated haemoglobin and LDL cholesterol levels are slightly elevated at 7.2% and 2.1 mmol/L, respectively, and his blood pressure is healthy. A patient with osteoporosis on treatment wants to know if her bones are getting healthier. The doctor says her bone mineral density is increasing. A patient with abnormal thyroid function test results or a slightly low testosterone level wants to know if this indicates illness and needs treatment. The doctor says she does not know but offers hormone replacement because it seems logical, is quick and is often what the patient wants.

Modern research often uses surrogate endpoints because the real outcomes are very hard to measure. Are we fixing the blood tests but not the patient? We are trying to do better. New studies in diabetes look at cardiovascular outcome and sometimes even cardiovascular mortality or overall mortality. Bone research has for years now investigated clinically significant fractures rather than just bone mineral density.

This issue of the MJA tackles these issues in four major areas of endocrinology: diabetes, osteoporosis, thyroid disease and male hypogonadism. It is likely these diseases involve half the population over 50 years of age.

In diabetes, new studies look at cardiovascular outcomes, time to dialysis or transplant, and sometimes overall mortality. New treatments include new drugs as well as new approaches to patient management. To keep up to date, we can read studies that describe real endpoints not surrogates. A cure for either type 1 or type 2 diabetes seems as far away as ever but treatments and outcomes continue to improve. However, this has not stopped the recurrent admission to hospital of patients with diabetes. Better general practice systems for managing patients with chronic disease are required to prevent the merry-go-round of recurrent emergency department visits and hospitalisation of patients with diabetes.1 Behavioural innovation is crucial to improving the health of a million Australians living with type 2 diabetes.2

One of the biggest communication challenges in diabetes management is what to advise the patient about the best diet. This remains unclear — most studies are short term with small numbers. How do we choose among options such as the CSIRO, Mediterranean, low fat, low carb and Paleo diets? In this issue, Andrikopoulos assists by discussing diabetes and the Paleo diet.3 Most recommendations still suggest a low joule, low fat, high complex carbohydrate diet.

In osteoporosis management, the problem is different. How to get the patients to take drugs with proven efficacy but with very rare, but frightening side effects? The currently available bisphosphonates and denosumab reduce fractures by up to 50% but 1 in 100 000 patients will experience dead bone in the mouth (osteonecrosis of the jaw) or an atypical femoral fracture, with the bone suddenly giving way.4 For the doctor, the risk–benefit equation obviously favours treatment. Not so for the patient. We need to improve our communication on this issue.

The management of thyroid disease seems straightforward enough — anti-thyroid drugs for hyperthyroidism and thyroxine replacement therapy for hypothyroidism. But what about patients with either high or low levels of thyroid-stimulating hormone but with normal thyroxine and triiodothyronine levels? Does treatment fix the blood test or the patient? Unfortunately, it is still not clear but the narrative review by Walsh in this issue suggests a practical approach.5 Five per cent of the population have a thyroid nodule. The patient just wants to know whether it is cancer and if they will die of it. Multiple ultrasounds and fine needle biopsies later, the doctor reassures the patient and says “we will watch it”. Frequently, surgical resection reveals benign disease. Better diagnostic pathways are required.

Last, what about communication in the field of men’s health. Low testosterone seems to be a straightforward problem easily solved by testosterone replacement. It is a quick consultation and the patient gets what he wants — patient, doctor and pharmaceutical company are happy. But is it good medicine? Alas no. It takes time to explain that the low testosterone measurements are not a disease and that the likely cause of the symptoms often lies elsewhere. The Endocrine Society of Australia position statement in this issue helps define the problems and suggests a decision making strategy.6

The answers to many of these challenges lie in two approaches — better communication at the clinical level and more research, both basic and clinical.

The Paleo diet and diabetes

Studies are inconclusive about the benefits of the Paleo diet in patients with type 2 diabetes

Type 2 diabetes is characterised by fasting hyperglycaemia as a result of insulin resistance and defects in insulin secretion. Obesity is the major risk factor for the development of the condition and a number of studies — including the Diabetes Prevention Program, the Da Qing IGT and Diabetes Study, and the Finnish Diabetes Prevention Study — have shown that lifestyle modification (diet and exercise) can significantly prevent the progression of glucose intolerance (prediabetes) to diabetes by up to 58%.1–3 In addition, a recent study showed that a very-low-calorie diet for 8 weeks resulted in remission of type 2 diabetes for at least 6 months in 40% of the participants.4 As such, clinical guidelines prescribe lifestyle modification as first-line treatment for type 2 diabetes and indeed throughout the management of the disease process.5 Therefore, it is clear that dietary intervention is a critical component of the glucose-lowering strategy in diabetes.

The Paleolithic or hunter–gatherer diet is currently popular for weight loss, diabetes management and general wellbeing. It recommends avoidance of processed food, refined sugars, legumes, dairy, grains and cereals, and instead it advocates for grass-fed meat, wild fish, fruit, vegetables, nuts and “healthy” saturated fat. In the early 1980s, O’Dea showed that 7 weeks of living as hunter–gatherers and consuming a high-protein, low-fat diet with an energy intake of 5020 kJ per person per day significantly improved or normalised the metabolic abnormalities of Indigenous Australians with type 2 diabetes.6 Thus, in its purest sense, the focus on fresh foods and avoidance of processed foods seems reasonable and consistent with dietary guidelines worldwide. However, what constitutes a Paleolithic diet is often skewed by individual interpretation or bias. This lack of a standard definition further complicates research evidence for or against this dietary approach and is often supported by individual self-reported benefits on health and wellbeing in popular social media channels. Is there scientific evidence that the Paleolithic diet is better for diabetes management than any other diet that advocates reducing energy intake?

Given its popularity, it was somewhat surprising that a PubMed search using the terms “Paleolithic diet and diabetes” resulted in only 23 articles, with many being reviews or commentaries. This is a similar outcome to a recently published systematic review of Paleolithic nutrition and metabolic syndrome.7 Clinical studies in patients with type 2 diabetes have only been performed by two research groups. Lindeberg and colleagues, from Sweden, published a randomised crossover study of the effects of a 3-month Paleolithic diet compared with a diabetes diet (according to current guidelines) in 13 obese (body mass index [BMI] of 30 ± 7 kg/m²) well controlled (glycated haemoglobin [HbA_1c], 48.6 ± 1.5 mmol/mol) patients with type 2 diabetes.8 The data showed that while both diets resulted in a reduction in BMI and HbA_1c, the Paleolithic diet achieved a significantly lower absolute value for these parameters. However, it is important to note that the patients on the Paleolithic diet had a lower BMI and HbA_1c at baseline and at the 3-month crossover, so it is not clear whether the relative reductions were similar with these diets. In addition, although there was no significant difference in oral glucose tolerance, the high-density lipoprotein levels were higher and triglyceride levels and diastolic pressure were lower with the Paleolithic diet. It is interesting that, based on a 4-day diet diary halfway through the intervention, the patients on the Paleolithic diet consumed less total energy. A follow-up study suggested that the Paleolithic diet may well be more satiating in patients with type 2 diabetes.9 In support of these results, Frassetto and colleagues showed, in a 14-day study of patients with type 2 diabetes, that both the Paleolithic diet (including canola oil and honey; n = 14) and standard diet (according to the American Diabetes Association recommendations; n = 10)10 resulted in a small reduction in HbA_1c levels, with no differences in insulin resistance (as assessed with a euglycaemic–hyperinsulinaemic clamp), blood pressure or blood lipids between the diets.11 There was, however, a beneficial effect of the Paleolithic diet only when compared with baseline for fasting plasma glucose, fructosamine, lipid levels and insulin sensitivity. It is important to note that canola oil is generally not considered a component of a Paleolithic diet. Moreover, this study was designed to maintain body weight at the baseline level in both groups of patients, with the result being a small but significant weight loss of 2.1 ± 1.9 kg and 2.4 ± 0.7 kg in the standard and Paleolithic diets respectively. In summary, these small and short-term studies tend to indicate some benefit but do not convincingly show that a Paleolithic diet is effective for weight loss and glycaemic control in type 2 diabetes.

In addition to the above studies of patients with type 2 diabetes, the Paleolithic diet has also been studied in healthy normal-weight individuals.12 Compared with a reference meal (based on the World Health Organization guidelines),13 there was very little effect on plasma glucose and insulin levels during an oral glucose tolerance test, but statistically significant increases were found in plasma glucagon-like peptide-1, glucose-dependent insulinotropic peptide and peptide YY. These hormone changes were associated with a higher satiety score. One of the Paleolithic meals used in this study caused an increase in the glucose excursion associated with a reduction in the insulin excursion during the glucose tolerance test.12 Similarly, in nine overweight healthy individuals, a Paleolithic diet for 10 days resulted in no change in fasting plasma glucose or insulin levels, but it showed reduced plasma lipid levels and blood pressure compared with the baseline usual diet.14 It is interesting that, while insulin levels during an oral glucose tolerance test were lower with the Paleolithic diet compared with baseline, the authors did not report the glycaemic excursions during this test. Moreover, a 2-week study in obese patients (n = 18) with the metabolic syndrome did not show an effect on glucose tolerance, but it resulted in reduced blood pressure and plasma lipid levels associated with a small but significant decrease in weight.15 In patients with ischaemic heart disease plus either glucose intolerance or type 2 diabetes (n = 14), a Paleolithic diet for 12 weeks resulted in reduced glucose and insulin excursions during the glucose tolerance test and was associated with a 26% reduction in energy intake, compared with a Mediterranean-style diet (n = 15).16 Again, in the absence of changes in weight or energy intake, the Paleolithic diet is as effective in improving the above metabolic parameters as a standard diet.

Thus, given that even very short deficits in energy balance can improve metabolic parameters,17 it is difficult to make strong conclusions about the long term benefits of the Paleolithic diet in type 2 diabetes (or any other condition), because of the short duration of the interventions (less than 12 weeks), the lack of a proper control group in some instances, and the small sample size (less than 20 individuals) of the above studies. While it makes sense that the Paleolithic diet promotes avoidance of refined and extra sugars and processed energy dense food, clearly more randomised controlled studies with more patients and for a longer period of time are required to determine whether it has any beneficial effect over other dietary advice.

Diabetes drug lowers risk of heart attack, stroke

A glucose-lowering drug has been shown to safely lower the overall risk of heart attack, stroke or cardiovascular death among type 2 diabetes patients.

Patients who were at risk for cardiovascular disease were found to have a 13% lower risk of cardiovascular death, non-fatal heart attack or non-fatal stroke when they took the drug Liraglutide compared to those who took placebo.

The randomised, double-blind study assigned patients either liraglutide or placebo and followed them for an average of 3.8 years.

Study results found a 22% lower risk of cardiovascular mortality, 15% risk of all-cause mortality and 22% lower risk of new evidence of advanced diabetic kidney disease.

“It is exciting to see such a broad-based benefit for patients who took liraglutide because most prior trials of diabetes medications have not shown such benefits,” lead investigator John B. Buse from the University of North Carolina School of Medicine said.

“Our results should give patients and providers comfort that liraglutide can safely improve outcomes beyond the core treatment of type 2 diabetes.”

The results were presented at the American Diabetes Association’s 76th Scientific Sessions in New Orleans over the weekend and were published in the New England Journal of Medicine.

In another study presented published in NEJM, it was found that empagliflozin was associated with slower progression of kidney disease and lower rates of clinically relevant renal events in patients with type 2 diabetes at high risk for cardiovascular event.

Latest news:

Breaking down the silos of treatment for post-traumatic stress disorder: integrating mind and body

Scalable interventions for PTSD that target mental health and comorbid cardiometabolic health are urgently required

There is increasing awareness of post-traumatic stress disorder (PTSD) among the general community, particularly in relation to the high incidence of the condition and its impact on high-risk populations, such as defence force veterans and emergency service first-responders. PTSD is a highly prevalent and costly condition associated with high rates of comorbid mental disorders, including anxiety and depression, and substance use.

There is growing interest in second-line or adjunctive treatments for PTSD. For example, a recently published randomised controlled trial established the efficacy of mindfulness-based stress reduction (an intervention that teaches individuals to attend to the present moment in a non-judgemental, accepting way) for treating PTSD among veterans.1 Participants were randomly assigned to receive either 8 weeks of mindfulness-based stress reduction therapy, delivered during weekly 2.5 hour group sessions, or an active control condition consisting of group sessions focusing on life problems. The majority of patients were also receiving pharmacotherapy (51/58 in the mindfulness group and 49/58 in the control group). Participants in the adjunctive mindfulness group were significantly more likely to experience clinically meaningful improvements in PTSD symptoms at 2-month follow-up (48.9% v 28.1%). Research among other trauma-affected populations suggests that improvements associated with mindfulness interventions can be maintained for up to 2.5 years following treatment.2

The scalability of the mindfulness intervention (ie, the ability to demonstrate efficacy under controlled and real-world conditions, and the capacity to reach a greater proportion of the eligible population3), combined with low levels of participant dropout and a robust effect size,1 represent a positive step towards establishing mindfulness interventions as an adjunctive treatment for PTSD. This is particularly pertinent where first-line interventions such as trauma-focused cognitive behaviour therapy may not be available.

There is growing evidence that PTSD, along with other mental disorders, is strongly associated with somatic, lifestyle-related comorbidities including obesity, diabetes and cardiovascular disease, with 39% and 49% of patients with PTSD meeting criteria for metabolic syndrome and abdominal obesity respectively.4 In people with psychotic disorders, this overwhelming burden of poor physical health across their lifespan, and its link with premature mortality, has been described as a scandal,5 with numerous calls to arms from clinicians and researchers, and the development of effective prevention-focused interventions. Such initiatives are starting to challenge the view that weight gain and poor cardiometabolic health are inevitable comorbidities for people with mental illness.

The high rate of preventable cardiometabolic disease in PTSD warrants consideration when developing and testing adjunctive or second-line treatments such as mindfulness. In the general population, physical activity is the cornerstone of treatment and prevention of cardiovascular disease, yet people with PTSD are known to be less physically active than the general population, highlighting the need for interventions to address this key modifiable risk factor. A recent systematic review and meta-analysis identified four randomised controlled trials investigating the effect of varying modalities of physical activity interventions for people with PTSD, including structured exercise and yoga.6 Physical activity was found to be more effective than control conditions at reducing PTSD symptoms, while also reducing depressive symptoms. The physical activity interventions ranged from 6 to 12 weeks involving 1–2 supervised sessions per week, including resistance training, yoga-based exercises, aerobic exercise, or a combination of all three modalities. Based on previous research in other mental disorders, the optimal exercise program (frequency, intensity, duration and modality) is contingent on individual factors including previous exercise history, severity of psychiatric symptomatology, somatic comorbidities and motivation. Australian physical activity guidelines7 provide clinicians with a structured framework for increasing habitual levels of physical activity in daily living.

The moderate to high effect sizes reported in trials of physical activity-based interventions in comparison to usual care alone are comparable to those achieved with mindfulness.6 In addition to reducing symptoms of PTSD and depression, structured resistance training and walking have been found to reduce cardiometabolic risk through significant reductions in waist circumference and self-reported sedentary behaviour, which are established independent risk factors for all-cause mortality.8 Further, exercise, particularly resistance-based (strengthening) exercise, can be seen as a mindful activity9 in which the basic principles of mindfulness (attending to the present moment in a non-judgemental and accepting manner) can be utilised. In addition, mindfulness and exercise are both known to be highly acceptable, and may the reduce perceived barriers and stigma that some patients experience when accessing mental health treatment, given that both are considered mainstream and health-benefiting behaviours.

The challenge for future research lies in designing and implementing combined interventions on a large scale, incorporating both best-practice mindfulness and exercise components in addition to usual care. Allied health clinicians with expertise in exercise programming (such as exercise physiologists10 and physiotherapists11) may assist in the design and delivery of best-practice physical activity interventions. Robust economic evaluations also need to accompany such research to establish their cost-effectiveness. Such an approach is likely to contribute to breaking down the silos of PTSD treatment, in order to integrate interventions that address the body and the mind. The physical consequences of PTSD can no longer be ignored and it is time to implement effective multidisciplinary treatments.

Increasing incidence of type 2 diabetes in Indigenous and non-Indigenous children in Western Australia, 1990–2012

An increase in the incidence of childhood type 2 diabetes (T2D) has been reported in several populations worldwide, including Australia, with the highest risk being observed in children of Indigenous descent.1–3 In Western Australia, children throughout the state who are diagnosed with T2D are managed by a single multidisciplinary team at Princess Margaret Hospital, WA’s only tertiary paediatric hospital. In this study, we aimed to determine the incidence and incidence rate trends of childhood T2D in Indigenous and non-Indigenous children in WA.

We undertook a retrospective population-based cohort study of children aged less than 17 years who were diagnosed with T2D in WA between 1990 and 2012, inclusive. Data were obtained from the previously described Western Australian Children’s Diabetes Database.3 T2D was diagnosed according to current guidelines, based on both clinical and laboratory findings.4 Patients identifying themselves as being of Aboriginal and/or Torres Strait Islander descent were considered to be of Indigenous descent.

Incidence rates were calculated by age, sex and Indigenous status, per 100 000 person-years at risk, using cases of T2D as the numerator and population data obtained from the Australian Bureau of Statistics as the denominator. Incidence rate trends were analysed using Poisson regression with Stata version 13 (StataCorp).

The study was approved by the WA Health Department Human Research Ethics Committee.

Between 1990 and 2012, 135 eligible cases of T2D were identified, with a mean age at diagnosis of 13.3 years (SD, 2.0 years). Of these cases, 61% (82/135) were in girls, and 56% (76/135) were in children of Indigenous descent. At diagnosis, the mean body mass index Z score was 2.0 (SD, 0.6), with 12% of children being classified as overweight and 61% obese. Their mean glycated haemoglobin (HbA_1c) level at diagnosis was 9.0% (SD, 2.8%) compared with 7.7% (SD, 2.5%) 1 year after diagnosis.

The overall mean incidence of T2D was 1.3 per 100 000 person-years (95% CI, 1.1–1.6 per 100 000 person-years), increasing from 0.2 per 100 000 person-years in 1990 to 3.1 per 100 000 person-years in 2012. The mean incidence in Indigenous children was 12.6 per 100 000 person-years (95% CI, 10.0–15.8 per 100 000 person-years) compared with 0.6 per 100 000 person-years (95% CI, 0.5–0.8 per 100 000 person-years) in non-Indigenous children. Between 1990 and 2012, the incidence increased from 4.5 to 31.1 per 100 000 person-years in Indigenous children, and from 0 to 1.4 per 100 000 person-years in non-Indigenous children (Box). The mean annual rate of increase in incidence over this period was 12.5% per year (95% CI, 8.0–17.0%) in Indigenous children and 10.9% per year (95% CI, 6.1–16.0%) in non-Indigenous children.

This population-based study provides further evidence of an increasing incidence of diagnosed childhood T2D in WA.1 Although a 20-fold higher mean incidence was observed in Indigenous children compared with non-Indigenous children, both groups had similarly high annual rates of increase. As childhood T2D may not present acutely, and population-screening programs are not routine in Australia, the incidence observed in this study is likely an underestimation of the true incidence. Furthermore, as diabetes-related complications occur early in youth with T2D,5 while the disease remains undiagnosed, diabetes-related complications may develop before clinical presentation.

The continued increase in childhood T2D reported in this study highlights the need for early diagnosis and screening for diabetes-related complications in youth at risk of developing the disease.

Box –
Incidence of type 2 diabetes in children aged < 17 years in Western Australia (1990–2012), by Indigenous status

* Per 100 000 person-years at risk.

Diabetic life expectancy 12 years less than average person

Two large studies have revealed that people with type 1 diabetes have a large gap in life expectancy compared to the general population.

The studies, published in Diabetologia (the journal of the European Association for the Study of Diabetes), show there has been little improvement in life expectancy for type one diabetics over the last few decades.

The first study examined 5,981 deaths of type 1 diabetic patients in Australia from 1997 to 2010.

Associate Professor Dianna Magliano and Dr Lili Huo from Baker IDI Heart and Diabetes Institute, Melbourne and colleagues found that deaths for those aged under 60 accounts for 60% of the years of life lost for men and 45% for women.

In the 10-39 year age group, they found that the major contribution to years of life lost was endocrine and metabolic diseases whereas in the over 40 age group, circulatory disease was the main contributor.

Overall, the researchers found that people with type 1 diabetes had an expectant life expectancy of 68.6 years, 12.2 years less than the average population (11.6 years less for men and 12.5 years less for women).

They also found the age when diabetes was diagnosed plays a critical role in determining the overall life expectancy.

“Our study shows a slight improvement in estimated life expectancy with increasing age at diagnosis,” they wrote.

They concluded: “Early onset of diabetes tended to be a predictor of premature mortality. Deaths from circulatory disease and endocrine and metabolic disease contributed most to early mortality in type 1 diabetes. For improvements in life expectancy, greater attention must therefore be paid to both the acute metabolic and chronic cardiovascular complications of type 1 diabetes. A failure to address either one will continue to leave type 1 diabetic patients at risk of premature mortality.”

In the second study, health records from the Swedish National Diabetes Register were linked with death records to examine life expectancy of Swedes with type 1 diabetes.

Dr Dennis Petrie from the University of Melbourne and Professor Björn Eliasson from the University of Gothenburg and colleagues found that although the life expectancy for men at age 20 with type 1 diabetes increased by about 2 years between 2002-06 and 2007 – 11, there was no change for women in the same time period.

They also noted that cardiovascular mortality significantly reduced for both men and women over the period which coincided with a large increase of the proportion of the population with type 1 diabetes who reported being on lipid-lowering medication.

“However, similar relative improvements in the general Swedish population for CVD were also observed, which suggests a similar uptake in lipid-lowering medication in the general population.”

The authors conclude: “There is still some way to go in terms of improvement in care for those with type 1 diabetes in order to close the gap with the general population.”

In a linked comment in Diabetologia, Dr Lars Stene from the Norwegian Institute of Public Health notes that the gap in life expectancy has remained largely unchanged since the turn of the millennium.

However he said that it’s perhaps not surprising that life expectancy hadn’t changed in the years outlined in the studies: “The differences in lifetime exposure to hyperglycaemia and other determinants of survival in the two overlapping groups of people with type 1 diabetes examined for mortality during these recent years may not be very different. We know that glycaemic control has long lasting effects.”

Dr Stene said general populations in Sweden, Australia and other countries have seen a recent reduction in cardiovascular mortality, an integral part of diabetes care.

“It is likely that patients with type 1 diabetes have enjoyed some of the beneficial developments that do not involve glycaemic control alone,” he wrote.

Latest news:

Economic evaluation of Indigenous health worker management of poorly controlled type 2 diabetes in north Queensland

Diabetes and its complications produce significant burdens for the health system in Australia. Between 2000–01 and 2008–09, total annual health expenditure for diabetes increased by 86% to $1507 million (2.3% of total health expenditure in 2008–09), an increase greater than that for all disease during the same period (60%).1

Indigenous Australians experience a disproportionally high burden of diabetes, which is responsible for 12% of the large gap in disability-adjusted life-years between Indigenous and non-Indigenous people.2 Indigenous Australians also have higher rates of hospitalisation for diabetes (3.4–5.0 times higher) and higher mortality rates from diabetes (7.0 times higher) than non-Indigenous Australians.3 They are also more likely to develop type 2 diabetes at an earlier age.4 Persistently high blood glucose levels cause organ damage, resulting in renal, circulatory and ophthalmic disorders. Indigenous Australians experience exceptionally high rates of these complications, including 11.2 times the rate of hospitalisation for renal failure4 and less effective care partnerships with their clinicians.5 It is therefore important to develop clinical programs that better manage diabetes and its complications in Indigenous people.

It was proposed that Indigenous health workers (IHWs) who are close to Indigenous communities linguistically and culturally could play an important role in improving the quality of primary health care for Indigenous Australians and contribute to better health outcomes. A trial of a recall system in remote Indigenous communities managed by local IHWs, supported by a diabetes outreach service, reported improved diabetes care and fewer hospitalisations.6,7 A 2006 study of the delivery of diabetes care in remote Indigenous communities found that employing more IHWs was associated with improved diabetes care, but not with better HbA_1c control.8

The Getting Better at Chronic Care Project (GBACC) was a cluster randomised controlled trial (cluster RCT) designed to improve the care of people with poorly controlled diabetes living in 12 rural and remote Indigenous communities in north Queensland. Participants in the six intervention communities received, in addition to standard primary care, intensive chronic condition management for 18 months, delivered by IHWs who had a Certificate III or IV in Aboriginal and/or Torres Strait Islander Primary Health Care. The IHWs received additional training in diabetes management and intensive support from the clinical support team. The Indigenous health worker-supported (IHW-S) model was family-centred and based on community outreach. Control communities received usual care (UC) from a centre-based primary care team (nurses, general practitioners, IHWs etc), but with less intensive IHW support. Service configurations varied somewhat between communities.9

The primary clinical results have been published elsewhere.10 A process evaluation concluded that there was significant implementation failure during the 18-month intervention phase, and six key features were identified as either enabling or hindering implementation. Further, the restructuring of Queensland Health coincided with implementation of the project, generating a number of challenges to the project that had not been anticipated.11

This article reports the economic evaluation of the project. We completed a cost–consequence analysis, in which the costs of implementing the model were compared with differential changes in a range of health outcome measures in the intervention (IHW-S) and control (UC) groups.

Methods

Design

We conducted an economic analysis alongside a cluster RCT. The trial design, participants, sample size, outcomes and ethics approvals have been described elsewhere.9 Study participants were Indigenous people with poorly controlled type 2 diabetes mellitus (HbA_1c levels ≥ 69 mmol/mol) and at least one other chronic condition. The primary clinical goal was a differential (IHW-S v UC) mean reduction in HbA_1c levels of 12.6 mmol/mol during the trial. The intervention was implemented from 1 March 2012 to 5 September 2013.

Measurement of costs

We estimated the per person cost of the intervention on the basis of project costing records. We distinguished between costs for service delivery and support, and for management and evaluation activities related to running the trial. Costs were analysed separately for the central team and the IHWs (Box 1).

Measurement of outcomes

The primary outcome assessed in the clinical trial was the difference in change in HbA_1c levels in the IHW-S and UC groups after 18 months. HbA_1c measurements were extracted from participants’ clinical files. The baseline value was the HbA_1c measurement closest to the participant’s recruitment date; the endpoint was the one closest to the trial endpoint. For the economic evaluation, we also explored the distribution of HbA_1c data, given the limitations on using the mean to describe a distribution. We also estimated HbA_1c outcomes in terms of shift in the numbers of people with moderate, poor and extremely poor diabetes control (as described below), given the relationship between diabetes control and health.

Secondary study outcomes included change in quality of life, disease progression, and rates of hospitalisation. Quality of life was measured with the Assessment of Quality of Life 4D (AQoL-4D) instrument. This has four dimensions (independent living, relationships, mental health, senses), each with three items and four levels. The AQoL-4D was developed in Australia, and the algorithm for estimating the utility score was derived from an Australian population.12 It has not been validated in an Australian Indigenous population.

Disease progression was assessed by allocating a disease stage to each participant, based on clinical markers and hospitalisation data, and using the diabetes severity staging instrument developed by Gibson and colleagues (Box 2).13 Baseline disease stage was based on data for the period 1 July 2010 – 1 March 2012, endpoint disease stage on data for the period 1 March 2012 – 5 September 2013. Once allocated to a stage, there was no possibility of reverting to a less severe disease stage.

Hospitalisation data were derived from the Queensland Hospital Admitted Patients Data Collection, which covers all patient separations (discharges, deaths and transfers) from all public and licensed private hospitals in Queensland.14 Data were obtained for all inpatient episodes for participants discharged between 1 July 2010 and 5 September 2013. This included all inpatient discharges during a pre-intervention period of 20 months and for the 18 months of the intervention. Hospitalisations were categorised into four groups based on International Classification of Diseases, 10th revision (ICD-10) codes (Box 3). Length of stay-adjusted diagnosis-related group hospital costs were also extracted.17

Statistical analyses

The statistical analysis was conducted on an intention-to-treat basis and in accordance with current guidelines for clinical and economic analysis alongside a cluster RCT for assessing differential costs and consequences.18 We adopted methods that take into account within-community clustering and correlation of cost and outcome data. Of the available methods for the economic analysis of cluster RCTs,19 we applied linear multi-level models (MLMs). MLMs acknowledge clustering by including additional random terms that represent the differences between the cluster mean (costs and outcomes) from the overall means in each intervention group. MLMs are efficient and are applicable to RCTs with less than ten clusters in each trial arm.18 Analyses were undertaken with Stata 12.0 (StataCorp).

We used a Markov model to describe disease progression in the IHW-S and UC groups. The probability of staying in the current state or moving between baseline and endpoint to a more severe disease stage was estimated and presented in a transition matrix. This is a simple way of presenting the rate of disease progression and testing for an effect of the intervention.20

Results

One hundred participants were enrolled in the IHW-S group, and 113 in the UC group. Of these, 87 in the IHW-S and 106 in the UC group met the study inclusion criterion (HbA_1c ≥ 69 mmol/mol). At baseline there were no statistically significant differences between the two groups in terms of age, body mass index, smoking or alcohol use (Box 4).

Costs of the intervention

Expenditure for the project is summarised in Box 5. The total cost was $1 991 904, of which $1 006 027 was attributed to intervention delivery. The remaining costs were allocated to research and other non-intervention activities. Total IHW salary cost (including on-costs) was $690 989. Three IHWs were employed full-time and three part-time. After adjusting for IHW involvement in other activities (6–56% of their time), the IHW salary cost attributed to the intervention was $522 421; the attributed cost of the trial manager and clinical support team was $483 606.

One hundred people received the intervention (of whom 87 met the study inclusion criteria), so that the average cost of delivering the intervention was $10 060 per person, or $6707 per person per year. This is the best estimate of the cost of rolling out a model incorporating the same elements as the GBACC.

Effectiveness of the intervention

Results of the incremental effectiveness analyses are reported in Box 6. The mean reduction in HbA_1c levels in the IHW-S group was non-significantly greater than that for the UC group (–10.1 mmol/mol v –5.4 mmol/mol; P = 0.17). This slight difference from our earlier report10 is attributable to our excluding participants who failed to meet the study inclusion criteria from the current analysis. Both groups experienced a minor fall in quality of life (between-group difference, P = 0.62).

There was a statistically significant reduction in the proportion of participants with extremely poorly controlled HbA_1c levels (≥ 102 mmol/mol) in the IHW-S group (from 34 to 19 people, or from 42% to 23%), but a slight increase in the UC group (34 to 36 people, or 35% to 37%; for between-group difference, P = 0.002) (Box 7). If the improvement in the IHW-S group had also been achieved by the UC group, 17 fewer people would have been expected to have had an HbA_1c level ≥ 102 mmol/mol.

There were no significant changes in any of the hospitalisation categories. Rates of hospitalisations for all causes (excluding dialysis) and for type 2 diabetes-related diagnoses each increased in both groups; the small differences in favour of the IHW-S group were not statistically significant. Ambulatory care sensitive hospitalisations increased slightly in the UC group but not in the IHW-S group (P = 0.81). The only category in which the change approached statistical significance was for cases in which diabetes was the primary diagnosis: there was a differential net reduction in admission rate of 0.09 per person per year (P = 0.06) (Box 6). The effect size was small, amounting to an estimated eight fewer admissions per year among the 87 IHW participants. This suggests a possible small improvement in morbidity.

The transition between disease stages from baseline to endpoint is depicted in Box 8. The differences between the IHW-S and UC groups were not statistically significant (Markov transition matrix [Appendix], P = 0.73).

Cost-effectiveness of the intervention

Annual hospitalisation costs are reported in Box 6. There was a small reduction for most categories in the IHW-S group, but the difference only approached significance for type 2 diabetes as the primary diagnosis. Additional expenditure of just over $6700 per participant per year achieved no significant improvement in mean HbA_1c levels, rate of disease progression, or quality of life, but realised a statistically sub-significant reduction in hospitalisations for those with type 2 diabetes as the primary diagnosis, yielding an estimated saving of $646 per person per year. The net intervention cost was thus just over $6000 per person per year, or $9000 for the 18-month trial. Taking into account the other significant finding, a reduction in the number of persons with very poorly controlled diabetes, this gives a cost of $42 880 for each person whose HbA_1c level was reduced below the critically high level of ≥ 102 mmol/mol.

Discussion

The study examined the costs and outcomes of the GBACC model as implemented in this trial. It is one of few economic analyses of a new model of primary care for addressing poorly controlled diabetes in Indigenous people, building on a high quality cluster RCT design.

The average annual cost of just over $6700 per person for the intensive IHW-S intervention as an adjunct to regular primary care is high relative to the reported costs of primary care in Indigenous communities in Australia. One study estimated the mean annual primary care costs in 21 mainly remote Indigenous communities in north Queensland (including some of the communities involved in this trial) at $1825 per person in 2004–05, equivalent to about $2700 in 2012–13.21 This estimate included IHWs as well as medical, clinic health, nursing, managerial and clerical staff. The Australian Institute of Health and Welfare similarly reported that total primary care expenditure per Indigenous person was $2648 in 2012–13.22 Despite the higher expected primary care costs of a group with poorly controlled diabetes, $6700 per person per year is a considerable additional cost.

In terms of intervention effect, this economic study explored a range of outcomes, including mean HbA_1c levels and their distribution, disease progression, quality of life, and hospitalisation. There was evidence of only a modest intervention effect, at best, in any of these measures. The only statistically significant improvement was the reduction in the proportion of patients with HbA_1c levels of 102 mmol/mol or more. The difference in the reduction in number of hospitalisations for diabetes (as the primary diagnosis) was close to significant. The study was powered to detect a change in the primary outcome, a mean reduction in HbA_1c levels of 12.6 mmol/mol over 18 months, not to detect changes in secondary outcomes such as hospitalisation or quality of life score.9 The effect of the intervention was assessed from the start of the trial, but there may have been a lag between its start and any impact on hospitalisation. A longer follow-up period than 18 months may have found a greater reduction in the hospitalisation rate.

Nonetheless, given the substantial additional resources that were invested, the outcomes were disappointing, with diabetes still poorly controlled in most patients, as indicated by the continued high levels of HbA_1c, very high rates of disease progression, and increasing rates of hospitalisation.

There are a number of possible reasons for these outcomes. It is possible that intervening in a group of patients with less advanced disease would have been more successful. It was expected that the employed IHWs would devote 100% of their contracted work time to the trial, but other responsibilities within local health services reduced the capacity of some IHWs to support trial clients. While we adjusted for this in the costing of the trial, it will have diluted the intensity of service delivery. Combined with difficulties in recruiting and retaining staff, this meant that two communities received less than 65% of the intended level of intervention (Box 5). Any change in the IHW position will have disrupted the IHW relationship, a core element of the model. Nonetheless, good community commitment was achieved by ongoing community engagement, with the IHW model building on the Apunipima Cape York Health Council (ACYHC) family-centred approach. ACYHC was a partner in the trial, and author MW, who is a public health medical advisor with ACYHC, was a Chief Investigator in this trial.

Investment in the training and upgrading of qualifications of the IHW, as well as in providing clinical support for them is likely to generate value elsewhere in the health system and over the longer term, a likely benefit not captured by our analysis.

Data quality is a common issue in community trials. For example HbA_1c data, which were extracted from participants’ clinical files, included some baseline data gathered well before the trial commenced.

It is also worth reflecting on whether the theory underpinning the trial was correct. In expanding the capacity of IHWs to provide direct and intensive support for Indigenous patients in the community, through both outreach and centre-based care, it was hoped to achieve more effective management of chronic disease because of greater cultural awareness and by improving patient engagement in self-care. While some health gains were identified, the major psychosocial and economic problems that are typical for very disadvantaged populations, and the strong relationship between these factors and chronic disease, mean that it may be necessary to address these factors more directly.23 Most of the IHW-S communities are in the bottom 2% of Queensland communities in terms of socio-economic disadvantage, indicating an extreme level of deprivation, often combined with a range of further serious adverse conditions.24 We did not have data on major life stressors (such as early death of family and friends, involvement with the criminal justice or child protection systems) that affect physical health and, probably, diabetes control; these factors may have affected the intervention and control communities differently.

A separate case study within the GBACC project found that health service providers need to review their systems of care to maximise the value of IHWs as specialist members of the multidisciplinary team.25 IHWs, who participated in regular clinical review sessions, were able to identify examples for improving self-management, which resulted in consistent positive change in HbA_1c levels in patients with the poorest control. Further, IHWs could respond to the problem of patient disengagement.

Conclusions

Our results suggest that the costs of delivering the GBACC model were considerable in absolute terms but achieved only a modest effect. This suggests a need to consider how to improve the effectiveness of the program, reduce its costs, and to increase revenue (eg, through Medicare billings).

The training of IHWs and clinical support workers is generally viewed as positive, but translating it into measurable outcomes for people with poorly controlled type 2 diabetes in highly disadvantaged communities remains a challenge. A more holistic cross-agency approach may be required, one that seeks to directly address the psychosocial, pathophysiological and environmental problems that are common in highly disadvantaged populations. While the need to consider social and economic determinants is understood, there are still major gaps in service delivery. The challenge for the public health community is to devise and implement interventions based on broader understanding of the determinants of health and to test the effectiveness of such interventions.

Box 1 –
Project cost calculations for the central team and the Indigenous health workers (IHWs)

The central team

The central team consisted of the trial manager and the clinical support team responsible for IHW training, which included:
- developing training materials, training delivery;
- enhancing the quality of clinical practice through mentoring, advocacy and reflective practice with IHWs, convening IHW meetings, clinical reference group meetings, team meetings;
- evaluation as an embedded component (data collection, data entry, conference presentations, workshops), and coordination of project activities, including chief investigator and management group meetings.
Costs were extracted from project financial reports for the period 1 January 2011 (commencement of the GBACC project with trial set-up) to 30 September 2013 (trial endpoint). The percentage of time allocated by the manager and the clinical support team to the trial and to the evaluation were determined by the trial manager (BS) after detailed discussion with LS and HN about the type of activities to be classed as intervention and non-intervention (evaluation and trial coordination activities).

Indigenous health workers

IHW salaries (including wage on-costs) in the six intervention communities were identified from project records. The proportions of their time allocated to intervention and to non-intervention activities were determined from detailed time logs kept by the IHWs. The IHW cost was calculated from their total wage costs and the percentage of time allocated to the project by each IHW.

Box 2 –
Diabetes vascular severity staging employed in this study, based on reference13

Type 2 diabetes with no evidence of microvascular or macrovascular risk factors.
Type 2 diabetes with screen-detected microvascular comorbidities and/or risk factors for macrovascular disease.
Type 2 diabetes with moderate microvascular or macrovascular complications.
Microvascular or macrovascular complications of late stage type 2 diabetes.

Box 3 –
Categorisation of admissions to hospital in this study

All hospitalisations.
Hospitalisations with principal or other diagnoses related to type 2 diabetes (ICD-10 E11 code in the principal or other diagnoses).
Ambulatory care sensitive (ACS) hospitalisations related to chronic disease (used by the Australian Institute of Health and Welfare to estimate ACS hospitalisations for Aboriginal and Torres Strait Islander people).15
The top three ACS condition categories (type 2 diabetes as principal diagnosis, cardiovascular diseases, and infections).16

Box 4 –
Baseline characteristics of the study participants

	Usual care	Indigenous health worker-supported	P

Number of participants	106	87
Mean HbA_1c level (SD), mmol/mol	95 (19)	99 (17)	0.12
Mean age (SD), years	47.6 (8.7)	47.5 (10.6)	0.958*
Sex (female)	70 (66%)	53 (61%)	0.533^†
Daily smoker	38 (36%)	34 (39%)	0.654^†
Current drinker	39 (37%)	36 (41%)	0.511^†
Mean body mass index (SD)	32.6 (6.2)(n = 43)	31.2 (6.3)(n = 44)	0.522*
Obese^‡	28 (65%)	23 (52%)	0.280^†

SD = standard deviation. * Results of t test for equal means, adjusted for within-group clustering. † Results of χ² test for equal proportions, adjusted for within-group clustering. ‡ Body mass index ≥ 30.

Box 5 –
Total cost estimates for the Getting Better at Chronic Care (GBACC) project

	Total trial expenditure	Time and cost allocated to GBACC intervention
	Total trial expenditure	Time*	Expenditure

Central team
Clinical support team	$626 091	57%	$357 353
Management	$234 624	10%	$23 462
Operation	$440 200	23%	$102 791
Sub-total	$1 300 915	37%	$483 606
Indigenous health workers
Community A	$151 551	78%	$118 210
Community B	$151 551	64%	$96 993
Community C	$75 775	44%	$33 341
Community D	$78 028	89%	$69 445
Community E	$156 056	84%	$131 087
Community F	$78 028	94%	$73 346
Sub-total	$690 989	76%	$522 421
Total expenditure	$1 991 904	51%	$1 006 027

Source: Project financial reports. * The allocation of project team time to research and service delivery was determined by the program manager. Allocation of Indigenous health worker time to GBACC was based on time records.

Box 6 –
Summary of the incremental effectiveness analyses (change between baseline and trial end)

	Usual care (n = 106)			Indigenous health worker-supported (n = 87)			Difference of differences^† (95% CI)	P
	Baseline	Endpoint	Change*	Baseline	Endpoint	Change*	Difference of differences^† (95% CI)	P

HbA_1c level (SD), mmol/mol	94.7 (19.0)	89.3 (24.1)	–5.4 (n = 97)	99.0 (17.4)	88.8 (25.7)	–10.1 (n = 81)	–4.7 (–11.6 to 2.1)	0.174
AQoL-4D, mean utility score (SD)	0.80 (0.18)	0.79 (0.21)	–0.01	0.75 (0.18)	0.72 (0.28)	–0.03	–0.02 (–0.08 to 0.05)	0.623
Rate of hospitalisation (per person per year; total number of admissions in parentheses)
All causes, excluding dialysis^‡	1.02 (172)	1.24 (176)	0.22	0.98 (135)	1.07 (124)	0.09	–0.13 (–0.68 to 0.41)	0.633
Type 2 diabetes, any diagnosis^§	0.53 (88)	0.92 (128)	0.39	0.47 (64)	0.78 (88)	0.31	–0.08 (–0.20 to 0.03)	0.150
Ambulatory care sensitive
All^¶	0.33 (58)	0.44 (60)	0.11	0.31 (45)	0.30 (36)	–0.01	–0.11 (–1.04 to 0.81)	0.811
Type 2 diabetes as principal diagnosis**	0.15 (26)	0.18 (23)	0.03	0.17 (23)	0.11 (13)	–0.06	–0.09 (–0.18 to 0.00)	0.063
Cardiovascular disease^††	0.01 (1)	0.08 (12)	0.07	0.02 (3)	0.04 (5)	0.02	–0.05 (–0.13 to 0.02)	0.149
Infections^‡‡	0.13 (21)	0.14 (20)	0.02	0.10 (14)	0.09 (11)	–0.01	–0.03 (–0.10 to 0.04)	0.362
Mean hospitalisation cost (per person per year)
All causes	$5438	$7421	$1982	$8010	$9866	$1856	–126 (–5024 to 4771)	0.960
Type 2 diabetes, any diagnosis^§	$4248	$6582	$2335	$4921	$8595	$3674	1340 (–2724 to 5404)	0.518
Ambulatory care sensitive
All^¶	$1665	$2132	$467	$2967	$2677	–$290	–757 (–2130 to 616)	0.280
Type 2 diabetes as principal diagnosis**	$907	$1245	$338	$1553	$1245	–$308	–646 (–1348 to 56)	0.071
Cardiovascular disease^††	$23	$163	$140	$239	$383	$144	4 (–749 to 757)	0.992
Infections^‡‡	$623	$609	–$14	$1040	$451	–$589	–574 (–1490 to 342)	0.219

AQoL-4D = Assessment of Quality of Life 4D score. * Only participants for whom baseline HbA_1c levels were measured after 1 January 2009 and endpoint levels after 1 March 2012 were included. † Estimates for incremental difference in outcomes between usual care and IHW groups using linear multi-level models adjusted for within-community clustering. ‡ Two people in the IHW group had dialysis after the intervention commenced (starting July 2012 and March 2013); their dialysis records were excluded. § International Classification of Diseases, revision 10 (ICD-10) code in principal or any other diagnoses starting with E11. ¶ All potentially preventable hospitalisations (ICD code in principal diagnosis: D501, D508, D509, E101–E108, E110–E118, E130–E138, E140–E148, E40–E43, E550, E643, E86, G40, G41, H66, H67, I10, I119, I110, I20, I240, I248, I249, I50, J02, J03, J06, J20, J312, J41–J44, J45, J46, J47, J81, K02–K06, K08, K098, K099, K12, K13, K250–K252, K254, K255, K256, K260–K262, K264–K266, K270–K272, K274–K276, K35–K37, K522, K528, K529, L03, L04, L08, L88, L980, L983, N10–N12, N136, N390, N70, N73, N74, O15, R02 or R56).16 ** ICD-10 code in principal diagnosis starts with E11. †† ICD-10 code in principal diagnosis: I10, I110, I119, I20, I240, I248, I249, J81 or I50. ‡‡ ICD-10 code in principal diagnosis: H66, H67, J02, J03, J06, J312, L03, L04, L08, L980, L88, L983, N10–N12, N136, N390, N70, N73, N74, or R02.

Box 7 –
Distribution of HbA_1c level categories at baseline and endpoint*

* Only participants for whom baseline HbA_1c levels were measured after 1 January 2009 and endpoint levels after 1 March 2012 were included.

Box 8 –
Distribution of disease stages at baseline and endpoint

Warning over diabetic ketoacidosis after man’s death

GPs are being told to consider diabetic ketoacidosis after a man died less than 24 hours after presenting to his doctor.

According to the Victorian Coroner’s report, a 29-year-old male visited his GP complaining of increased urination, thirst and difficulty sleeping.

He had a fever of 38.6 with high blood pressure and the GP prescribed cephalexin antibiotics for a urinary tract infection.

A urine sample was taken and sent to pathology as well as requests for blood tests including full blood exam, LFT, TSH, urea and electrolytes and fasting blood glucose.

However the patient was found dead in his home at 9pm the following day. A taxi was booked at 5:30am that morning however he didn’t respond to the taxi’s arrival at between 7:40am and 8am.

The pathology report dated the day after testing found ketones and glucose in the urine sample.

The coroner found that the patient died from diabetic ketoacidosis. Post mortem toxicological analysis showed a glucose concentration of 45mmol/L, combined with raised acetone levels in the blood and vitreous. The coroner said this indicated that the man had died of undiagnosed diabetes.

The patient had reported considerable weight gain over the previous year, resulting in him weighing 122.9kg when he presented to his doctor. The GP concluded that the patient likely had diabetes and ranked the likelihood as Type 2 ahead of Type 1 diabetes. However he didn’t check the patient’s blood glucose at the time, instead referring him for fasting blood tests the next day.

It is noted in the coroner’s report that the GP was regretful that he didn’t perform a fingerprick test at the time of consultation and that it would have been the appropriate action.

The report stated that the doctor was under the impression that fasting blood tests was the most accurate way to diagnose diabetes. “(The doctor) did not consider that (the patient) was in any immediate danger from diabetes,” the report said.

CPD active learning module: Addressing the challenges in the fast moving field of endocrinology

The coroner recommended the following:

The Royal Australian College of General Practitioners provides a clinical update to GPs to highlight the importance of recognising hyperglycaemia and ketosis in adult diabetic patients, as an uncommon but potentially serious complication of type 2 diabetes, or indication of newly recognised adult-onset type 1 diabetes.
The Royal Australian College of General Practitioners advise GPs that although uncommon in adults and clinically subtle in its earliest states, evolving diabetic ketoacidosis may produce a dangerous metabolic decompensation and require escalation of care to a hospital setting for further assessment and management.

RACGP President Dr Frank R Jones told doctorportal: “The RACGP will certainly review and respond to the coroner’s recommendations regarding diabetic ketoacidosis (DKA) as published in the findings into the death of (the patient) and will communicate a clinical update to members.

“The RACGP does have some existing guidance on DKA in its resource General practice management of type 2 diabetes.”

Latest news:

Vertebroplasty is not a do-not-do treatment

Vertebroplasty has been controversial but remains clinically useful and new evidence awaits publication

Duckett and colleagues have classified vertebroplasty as a do-not-do treatment.1 They referenced two randomised controlled trials (RCTs)2,3 as definitive proof of this. However, the authors failed to heed our clinical opinion published in the MJA that these two trials were “not relevant to the patient group that we treat with vertebroplasty”.4 We have the largest clinical vertebroplasty experience in Australia, yet our published advice was apparently ignored. In the article by Duckett and colleagues, Box 1 illustrated the selection process that the authors used to determine do-not-do procedures. The process supposedly excluded evidence which was “contested” or “which was not supported by consulted clinical experts”. Accordingly, vertebroplasty should have been deleted from the list.

The authors used the United Kingdom National Institute for Health and Care Excellence (NICE) for clinical guidance. Current NICE guidance5 states that “vertebroplasty and kyphoplasty can be considered appropriate interventions for people with recent, unhealed osteoporotic vertebral compression fractures in whom the pain is severe and ongoing despite optimal pain management”.

From 1208 potential treatments, the authors excluded 1200, leaving five apparently incontrovertible do-not-do treatments. The fact that at least one of the five is wrongly included (by the authors’ own criteria) demonstrates the failure of the proposed model and the danger of adopting this kind of formula to influence clinical practice in hospitals.

The evidence for and against vertebroplasty is inconclusive. There is disparity in measured outcomes between blinded RCTs2,3 of vertebroplasty for fractures up to 12 months old and a larger, open-label RCT6 of fractures less than 6 weeks in duration. The blinded trials found no significant benefit of vertebroplasty over placebo, whereas the open-label RCT found significant benefit of vertebroplasty over conservative care. This disparity is well described in the NICE guidance.5

For the past 10 years, my vertebroplasty practice has been confined to treating fractures less than 6 weeks old.7 It is clear to me that the published blinded trials tested a different approach and are not relevant to the patient group that my practice treats with vertebroplasty for two principal reasons: the fractures were mostly non-acute; and the volume of cement used in these trials (2.6 cm³ on average in both trials) would have been insufficient to stabilise an acutely collapsing vertebral fracture.

Attempting to answer the acute fracture conundrum, the authors of the blinded RCTs published a meta-analysis of 52 patients from both trials with fractures less than 6 weeks duration.8 Only outcomes at 2 weeks and 1 month were presented and the evidence is hardly definitive.

The onus was placed on vertebroplasty practitioners to provide high-quality blinded data in this group of patients. For this purpose, my co-investigators and I embarked on the Vertebroplasty for Acute Painful Osteoporotic fractURes (VAPOUR) trial.9 Five years ago, we reconfigured the protocol from the INvestigational Vertebroplasty Efficacy and Safety Trial (INVEST),2,10 the larger of the two blinded vertebroplasty trials. We excluded crossover, which was permitted at 1 month in INVEST. We changed the selection criteria to include only fractures less than 6 weeks duration (average fracture age in the VAPOUR trial was 2.6 weeks compared with 18 weeks in INVEST) with pain scores greater than 7/10 and with either magnetic resonance imaging or single-photon emission computed tomography evidence of acute fracture. In-patients, already hospitalised with acute fractures, comprised 59% of the VAPOUR trial enrolment but were excluded in INVEST. The procedural technique was different in the VAPOUR trial, where we attempted maximum fill of the vertebral body to stabilise the fracture and prevent ongoing collapse. The average cement volume of 7.5 cm³ in the VAPOUR trial was three times that in INVEST. The method of blinding and data collection was similar for the two trials.

Our trial team included four Sydney centres with established vertebroplasty programs. The VAPOUR trial completed enrolment of 120 patients in December 2014 and is the largest RCT and the only acute fracture RCT of vertebroplasty in Australia. Statistical assessments of outcomes are nearing completion and the results of the trial will soon be published.