OUP user menu

Optimal diabetes care—can we afford it?
Evidence-based diabetes care could be highly cost effective

T.A. Chowdhury, P. Bennett-Richards
DOI: http://dx.doi.org/10.1093/qjmed/hct150 983-987 First published online: 3 July 2013


Management of diabetes is expensive and set to get costlier. Managing the condition and it’s devastating complications imposing a huge societal and economic toll on healthcare systems worldwide. While many interventions to reduce complications are available, a number of interventions do not have a strong basis in evidence, and lack cost effectiveness. In a time of economic austerity, and unprecedented pressure to reduce costs of health care in the UK, are there ways improving care, without driving up cost?


Prevalence of type 2 diabetes (T2D) is growing rapidly worldwide,1 and societal costs associated with T2D are immense. In many countries, diabetes is now the commonest cause of end-stage renal failure, blindness in people of working age and non-traumatic lower limb amputation.2 In the UK, 10% (around £11 billion) of healthcare expenditure is on management of diabetes and it’s attendant complications, although the actual cost may be double this figure if indirect costs are included.2 With growing prevalence and increasing costs of treatment, along with unprecedented pressure to reduce costs, are there ways of restraining the financial burden related to diabetes, without deteriorating clinical care?

Ensure all patients with diabetes undergo structured education at diagnosis

The journey of a person with T2D should start at diagnosis with structured education as the bedrock of their subsequent care. There is good evidence that structured education for newly diagnosed patients with T2D will improve glycaemic control, cholesterol and weight, reduce requirement for diabetes medication, increase consumption of fruit and vegetables, knowledge of diabetes, self-empowerment, self-management skills and treatment satisfaction, all at relatively low cost (around £15.00 per patient).3,4 The effect, however, may not be sustained suggesting that frequent refresher education may be required.5

Focus on cardiovascular risk reduction

At time of diagnosis, around 50% of people with T2D have a complication related to diabetes, and cardiovascular disease accounts for 50% of premature mortality seen in people with T2D.6 Around a fifth of all hospital admissions in the UK are due to diabetic complications.7 Multi-factorial intervention aiming for tight control of cardiovascular risk factors has been shown to reduce cardiovascular risk substantially in people with T2D.8 In the Steno-2 study, 12 years of sustained reduction of blood pressure (systolic 130 vs. 145 mmHg), cholesterol (3.5 vs. 5.0 mmol/l) and a modest reduction in glycaemic control (HbA1c reduction of 0.5%) led to around 50% risk reductions in cardiovascular and all cause mortality and 80% reduction in stroke and renal replacement therapy.

Angiotensin converting enzyme inhibitors,9 and statins,10 have a low acquisition cost and proven efficacy in reducing vascular disease in people with T2D, and probably should be prescribed in all people with T2D over the age of 40 years, unless contra-indicated. Evidence around use of aspirin for primary prevention in patients with diabetes is as yet not fully established.11 Smoking cessation must be strongly encouraged in all smokers with T2D.

Improve care of patients in hospital with diabetes

Diabetes is common among hospital in-patients. Scottish data shows that while diabetes prevalence is around 4.3%, people with diabetes account for around a 10th of in-patient costs and a 5th of all in-patient bed days.12 A UK wide estimate from 2010 suggests that around 12.6% of the cost of acute hospital admissions was diabetes related, amounting to a total cost of around £2.3 billion.13 A reduction by just 10% in hospital bed days in people with diabetes could, therefore, potentially save £230 million.

In the UK, the care of in-patients with diabetes appears to be suboptimal, as demonstrated by the findings of the National Diabetes In-patient Audit.7 Observational data of 1500 unselected medical admissions with and without diabetes showed that length of stay, readmission rates and 30-day mortality rates rose with higher blood glucose concentrations.14 While improving care for in-patients with diabetes is likely to be cost effective by reducing length of stay and hospital acquired complications, firm evidence for this is lacking from randomized studies.

Improvements in care might include regular reviews of glucose control for in-patients, a diabetes care plan for all in-patients, regular foot checks in hospital and targeted referral of appropriate patients to in-patient diabetes specialists. Randomized controlled trials have shown that short-term tight glycaemic control pre-operatively for cardiac surgery patients can lead to reduced post-operative complications, such as sternal wound infections.15

Consider carefully who might benefit from more intensive glucose management

There is growing concern about the value of very tight glucose control on the basis of some randomized trial evidence. The UK Prospective Diabetes study (UKPDS) demonstrated that improved glucose control for a period of around 10 years from diagnosis of T2D could provide significant protection from microvascular disease, but had little impact on cardiovascular complications.16 Subsequent 10-year follow up of the UKPDS cohort showed significant relative risk reductions in all cause mortality and myocardial infarction, suggesting a ‘legacy effect’, of tight glucose control early in the disease.17 More recent data from randomized controlled trials of patients with longer duration of diabetes (around 10 years from diagnosis) suggest that there is little benefit in very tight glucose control later in the disease, and possible evidence of harm.18–20

Taken together, current evidence suggests that early tight glucose control from diagnosis, may provide some benefit in reduction of complications, but around 10 years from diagnosis, attempts to tighten glycaemic control may lead to adverse outcomes. While we must avoid therapeutic nilihism when treating glucose in people with long standing diabetes, the UK National Institute for Health and Clinical Excellence (NICE) guidelines suggest individualized targets for glycaemic control.21

A number of factors may be considered when deciding who might benefit from tight glucose control, such as age, duration of diabetes, motivation, capacity for self-care and social support. In particular, tight glucose control among elderly people with diabetes lacks a firm evidence base. Mortality among people with diabetes diagnosed over the age of 70 is not increased above that of the baseline population,22 and most clinical trials of glycaemic control have not included large numbers of patients over the age of 75 years. In the UK, pay for performance schemes such as the quality outcomes framework incentivize tight glucose control in all age groups, irrespective of age or co-morbidity. This can lead to un-necessary prescribing of expensive therapy, which may also increase risk of adverse effects, particularly that of hypoglycaemia, which is undesirable particularly in the elderly. A method of excluding these patients from such targets, or perhaps incentivizing individualization of glucose targets might be preferable, and cost saving.

Ensure all patients get regular screening for complications

Screening for complications of diabetes is cost effective as early intervention can lead to reduced morbidity, such as blindness or renal failure. Nine care processes are recommended by NICE for all people with diabetes (body mass index, blood pressure, foot check, retinal screening, glycated haemoglobin, cholesterol, serum creatinine, urine microalbuminuria and smoking status). The UK National Diabetes Audit of all primary care trusts showed that on average 56.4% (range 15.9–71.2%) of people with T2D achieved all nine checks in 1 year.23 Careful call and recall processes and chasing of non-attendance may help detect patients with complications early, and lead to treatment to prevent progression of complications.

Use human insulin rather than analogue

Insulin therapy may be deemed necessary in some people with T2D with poor glucose control, particularly if they have symptomatic hyperglycaemia. In recent years, use of analogue insulins has been favoured as first line insulin therapy in many units in the UK, with an incremental cost of £625 million per year, although no significant improvement in glucose control has been demonstrated with these agents.24 A modest effect on nocturnal hypoglycaemia may be seen, but this does not impact significantly on cost effectiveness. NICE guidelines recommend the use of human insulin first line for people with T2D, and is based on that fact that hypoglycaemia is uncommon in patients with T2D treated with insulin.25 Use of human insulin first line in all patients with T2D in unlikely to have major disbenefits, and could save a significant proportion of the incremental costs seen with these drugs.

Reduce use of self-monitoring of blood glucose

Self-monitoring of blood glucose (SMBG) is clearly helpful in circumstances where knowledge of glucose levels is essential to safely manage T2D. Thus, patients on insulin therapy, people who drive for a living and people with possible symptomatic hypoglycaemia on sulphonylureas may all benefit from SMBG. There is, however, no evidence that routine SMBG improves glucose control, or morbidity from complications. The cost of SMBG in the UK is in excess of £160 million, and systematic reviews suggest that there is little benefit in people not treated with insulin.26 Indeed some studies suggest there may be an adverse impact on quality of life.27 There is little justification for their routine use due to their high cost, and only targeted use should be encouraged. In addition, certain glucose monitors have much less expensive consumables compared with others. Reducing self-monitoring of glucose to more appropriate individuals has the potential to save a significant proportion of the costs of self-monitoring in the UK.

Take care with high cost new drugs for glucose control

In recent years, a number of new classes of anti-diabetic agent have been developed, increasing the therapeutic armamentarium for glucose control in people with T2D. Not all new drugs have had a positive impact, however. The debacle of the withdrawal of rosiglitazone has proved chastening for many specialists. It is now a decade since pioglitazone was launched, an only recently have the possible adverse effects of bladder cancer and post-menopausal fractures been highlighted.28

Costs of prescribing glycaemic therapy have risen very significantly in recent years. Between 1997 and 2007, prescribing costs for T2D rose from £39.00 per patient per year to £740.00 per patient per year. During the same time period, however, mean HbA1c has dropped by only 0.1%.29

Newer drug classes such as dipeptidylpeptidase-IV inhibitors, glucagon-like peptide-1 (GLP-1) analogues and most recently sodium glucose transporter-2 inhibitors have a significant benefit in terms of promotion of weight loss. Recently, some concern around the pancreatic effects of the former two classes has been expressed (pancreatitis, pancreatic dysplasia and possibly pancreatic cancer).30 It has been suggested by some that licensing of new anti-diabetic medications should only occur once they have proven cardiovascular outcome benefits.31

Perhaps a pragmatic view would be to limit the use of such drugs to NICE guidelines,25 to ensure that they are used as a trial of treatment for a limited time, and stopped if targets of weight loss and glucose control are not achieved within a certain time interval. NICE guidelines suggest continued use of GLP-1 analogues only if HbA1c falls by 1% plus a 2.5% weight loss is achieved. Adherence to this guideline could save significant expenditure on futile but expensive therapy.

Use of GLP-1 analogues with insulin therapy may have some benefits in reduced weight gain and reduction in insulin dose, but their use cannot yet be routinely advocated in the absence of more robust randomized trial data. In 2011, prescribing of gliptins and GLP-1 analogues cost the UK £87 million, and costs are rising year on year, despite lack of evidence for their long-term efficacy and safety.

Focus on prevention of diabetes

T2D is preventable. There is strong randomized trial evidence that intervention either by pharmacological or lifestyle methods in patients with pre-diabetes, can reduce risk of incident diabetes.32 The seeds of diabetes are, however, sown in childhood and nurtured by environmental factors which require concerted action to break. It is clear, therefore, that prevention of diabetes should start earlier than the phase of pre-diabetes. NICE have published guidelines on prevention of diabetes, extolling local and national action to tackle obesity and physical inactivity, such as provision of culturally appropriate community based weight management programmes for people who are overweight or obese.33

The guidelines also promote national governmental action to prevent diabetes suggesting that national and local government work with food manufacturers to improve composition of foods, develop clear nutritional labelling information and reduce costs of healthier foods. The UK governments’ response of ‘nudge’, and information programmes such as ‘Change 4 life’, lack scientific rigour and may at best offer a minor short-term benefit. There is also some concern about their sponsorship by commercial companies producing sugar sweetened beverages and unhealthy cereals. Relying on food companies to lead efforts to improve health is unlikely to be effective,34 and while more concerted public health intervention has been considered in New York (banning supersized sugar sweetened beverages) and tried in Denmark (fat tax), both have been successfully lobbied against by the food manufacturers.

The health benefits of reducing calories and improving physical activity particularly in children will not just be seen in diabetes prevention but also in prevention of cardiovascular disease and cancer. While the costs may be significant, the health and financial benefits will accrue over the long term. In the short term, a pragmatic and cost effective intervention might be greater use of metformin in people with pre-diabetes, which reduces risk of incident diabetes by around 30%.32


Diabetes is costly, and set to get costlier. Many health systems are struggling to restrain spiralling costs of managing the expensive complications of the condition, although a number of expensive, but unproven interventions are commonplace. Spiralling costs may be inhibited by prudent, evidence-based prescribing of drugs and other interventions, and in the longer term by effective public health measures to prevent diabetes.

Key points

  • Structured education, smoking cessation, blood pressure and cholesterol lowering and rigorous attention to screening for complications should be prioritized.

  • Tight glucose control may not be suitable for everyone—targeted personalized therapy may be more effective. Where insulin is required in T2D, human insulin should be prescribed.

  • Judicious use of new drugs for hyperglycaemia should be undertaken until cardiovascular outcomes studies prove these drugs to be beneficial.

  • Improving care for in-patients with diabetes is likely to reduce complications and hospital bed usage.

  • More resources should be focussed on preventing diabetes.

Conflict of interest: None declared.


View Abstract