Elsevier

The Lancet

Volume 364, Issue 9432, 31 July–6 August 2004, Pages 423-428
The Lancet

Articles
Prevalence, care, and outcomes for patients with diet-controlled diabetes in general practice: cross sectional survey

https://doi.org/10.1016/S0140-6736(04)16765-2Get rights and content

Summary

Background

Tight glycaemic control reduces microvascular complications in patients with type 1 and type 2 diabetes. We aimed to establish the proportion with type 2 diabetes treated through diet only and to determine levels of complications and quality of care received compared with patients on hypoglycaemic medication.

Methods

We undertook a cross-sectional study of 7870 patients with type 2 diabetes from a population of 253 618 patients from 42 general practices in the UK. Our primary outcome was process of care measures and diabetes-related complications.

Findings

31·3% of all patients with type 2 diabetes are being managed with diet only (1% of the total population). More than four-fold variation between practices exists (range 15·6–73·2%). Patients treated with diet only are much less likely to have HbA1c (glycosylated haemoglobin) measurements, blood pressure, cholesterol, smoking, microalbuminuria testing, or screening for foot pulses recorded. 38·4% of patients with type 2 diabetes on medication have a HbA1c above 7·5% compared with 17·3% of those treated with diet only. Compared with those on medication, patients treated by diet only are more likely to have raised blood pressure and less likely to be on anti-hypertensive medication; they are 45% more likely to have raised cholesterol and less likely to be prescribed lipid-lowering medication. Although fewer of those treated by diet (68%) have diabetes-related complications compared with those on medication (80%), the rate is much higher than for the population without diabetes.

Interpretation

Diabetics treated by diet only have significant rates of complications and are less likely than those on medication to be adequately monitored. There is great scope for improved management within general practice.

Introduction

By tradition, a substantial number of people with diabetes mellitus have been managed without medication. They are usually offered dietary advice and, irrespective of whether patients remember or follow the advice, they are referred to being managed on diet only. Not using medication originated in the era when the aim of treatment was to maintain short-term freedom from symptoms of hyperglycaemia.1 For patients with type 2 diabetes, this meant a stepladder from diet to mono-therapy to combined therapy, including the addition of insulin or, more recently, glitazones.2 The use of diet alone was supported by results of early studies that failed to find a convincing association between glycaemic control and the development of complications from diabetes.3 However, the DCCT trial4 published in 1993 provided robust evidence that tight glycaemic control reduces microvascular complications for patients with type 1 diabetes. The UK Prospective Diabetes Study (UKPDS)5, 6, 7 subsequently provided strong evidence that good glycaemic control is associated with a reduction of microvascular complications in patients with type 2 diabetes. The UKPDS also showed the need for good blood pressure control in order to help reduce macrovascular complications8, 9 and that both interventions are also cost effective.10

The rationale for good glycaemic control in all people with diabetes underpins the National Service Framework for Diabetes,11 the new GP contract12 indicators for the management of diabetes, and NICE guidance on the use of hypoglycaemic drugs.2 Since results of studies show that diet alone does not result in adequate glycaemic control,13 it is likely that many such patients need hypoglycaemic medication.

Translating the results from randomised controlled trials into everyday clinical practice takes time. Anecdotal evidence suggests that there is a continuing belief in the existence of “mild diabetes”—a group of people with diabetes at low risk of complications, for whom active therapeutic management is neither indicated nor cost effective. Since other population-based studies of patients with diabetes have focused on patients on medication14 or on those attending secondary care,15 there are virtually no data from primary care regarding the proportion of patients with diabetes managed on diet only. There is very little information about the level of complications experienced and the quality of care received by such patients.

Therefore, we undertook a large population based study to establish the proportion of patients with type 2 diabetes treated by diet only and the interpractice variation in the use of medication, and to determine levels of complications and quality of care they receive compared with patients on hypoglycaemic medication.

We undertook a cross-sectional study of 253 618 patients registered on July 15, 2003, across 42 practices in the former Trent region. Approval was given by the Trent multi-centre research ethics committee. The coded computerised data in the general practice clinical electronic record were extracted centrally in pseudo-anonymised form by a UK computer software company (EMIS). Although no strong patient identifiers (including the patients' postcode) were extracted, every patient had been assigned the Townsend score associated with their electoral ward of residence. The Townsend score is a validated measure of material deprivation derived from census related data of the electoral ward associated with the patients' postcode.16 We validated the resulting database (known as QRESEARCH) by comparing disease prevalence, prescription statistics, population characteristics, referral rates etc, against published data17 and found similar rates per 1000 population. The date of the computer download was July 15, 2003.

We identified patients with diabetes if they had a Read code for diabetes or had more than one prescription for either insulin, sulphonamides, glitazones, biguanides, or for diabetes glucose testing kits.

As in previous studies,18 we classified patients as having type 1 diabetes if they had been diagnosed under the age of 35 years and were receiving insulin. The remaining patients with diabetes were classified as type 2. We then grouped patients with type 2 diabetes into those on medication (defined as one or more scripts of hypoglycaemic agents issued within the last 6 months) and those treated by diet only.

In addition to the patients' year of birth, sex, diagnosis date, and details of hypoglyacemic agents, we identified individuals with hypertension, ischaemic heart disease, stroke, congestive cardiac failure, atrial fibrillation, or peripheral vascular disease. We grouped patients with any evidence of vascular disease if they had any of the conditions listed above since many had more than one vascular co-morbidity.

We used the Read codes library produced by the UK National Health Service Information Authority19 to identify those with complications related to diabetes. This included amputation or leg ulcer; neuropathy (including impotence); evidence of renal impairment (including dialysis, transplant, nephropathy, or creatinine >120 mmol/L); retinopathy; cataract; or glaucoma (including medication for glaucoma). We grouped patients with either retinopathy, cataract, or glaucoma in to those “with diabetes-related eye disorders”, because many patients had more than one condition.

To measure quality of diabetes care, we identified the most recent value for the following: glycosylated haemoglobin (HbA1c); body-mass index; systolic and diastolic blood pressure; serum cholesterol; and serum creatinine. We defined a raised HbA1c as being above 7·4%; obesity as a body-mass index of more than 30 kg/m2; a raised blood pressure as either a systolic blood pressure higher than 145 mm Hg or a diastolic blood pressure above 85 mm Hg; a raised serum cholesterol as greater than 5 mmol/L and a raised creatinine as a value more than 120 mmol/L. We also retrieved data for retinal screening; testing for presence or absence of peripheral pulses; testing for neuropathy; testing for microalbuminuria; smoking status and smoking cessation advice. All these measures are in the new UK General Medical Services contract for general practitioners.12

We examined the use of anti-hypertensive agents for each patient using categories from the British National Formulary (September, 2003). We identified patients receiving medication within the 6 months before the study period. We also identified patients on monotherapy, dual therapy, and triple or quadruple therapy.

We compared the proportion of patients on lipid-lowering agents and the proportion treated with an angiotensin-converting-enzyme inhibitor. We included these treatments since lipid-lowering treatment has been shown to confer vascular protection in patients with diabetes20 and ACE inhibitors are useful in treating congestive cardiac failure, hypertension, and possibly also confer additional renal protection in patients with diabetes.

We used unconditional logistic regression to calculate unadjusted odds ratios with 95% CIs for binary outcomes, comparing patients treated by diet only with patients on medication. In our multivariate analysis, we adjusted for sex, age (<40 years; 40–59; 60–79; ≥80), and fifth of Townsend score (cutoffs were defined using the quintiles for England and Wales). We allowed for clustering by general practice by defining this as a clustered variable and using a robust standard error (STATA version 8.0). We selected a significance level of 0·01 (two tailed).

The funding body had no role in the design of the study, its analysis, in the interpretation of the results, the drafting of the report, or in the decision to submit the paper for publication.

Section snippets

Results

In the study population, there were 8626 patients with diabetes, giving an overall prevalence of 3·4%. Of these, 756 (8·8%) had type 1 diabetes, 5170 (59·9%) had type 2 diabetes treated with medication and 2700 (31·3%) had type 2 diabetes not treated with medication, referred to in this study as “diet only”. The median age at onset was 60 years (IQR 50–68) and 58 years (49–71), respectively. The prevalence of each type of diabetes by sex and quintile of deprivation is shown in table 1. There

Discussion

We found that almost a third of all patients with diabetes are being managed with diet only, that the rate of complications is high, and that routine monitoring in such individuals is much lower than in patients on hypoglycaemic medication. If there were an evidence base showing a balance between side-effects and benefits, or a clear cost-efficiency analysis, this situation might be justified. However, the more than four-fold interpractice variation in the percentage of people with diabetes

References (30)

  • D Nathan

    Some answers, more controversy, from UKPDS

    Lancet

    (1998)
  • H Stovring et al.

    Rising prevalence of diabetes: evidence from a Danish pharmaco-epidemiological database

    Lancet

    (2003)
  • NICE
  • University Group Diabetes Program

    A study of the effects of hypoglycaemic agents on vascular complications in patients with adult-onset diabetes

    Diabetes

    (1970)
  • DCCT Research Group

    The effect of intensive diabetes treatment on the development and progression of long-term complications in insulin-dependent diabetes mellitus

    New Eng J Med

    (1993)
  • United Kingdom Prospective Diabetes Study Group

    United Kingdom prospective diabetes study (UKPDS) 13: relative efficacy of randomly allocated diet, sulphonylurea, insulin, or metformin in patients with newly diagnosed non-insulin dependent diabetes followed for three years

    BMJ

    (1995)
  • United Kingdom Diabetes Study (UKPDS) Group

    Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33)

    Lancet

    (1998)
  • I Stratton et al.

    Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35); prospective observational study

    BMJ

    (2000)
  • ME Cooper et al.

    Optimizing treatment of hypertension in patients with diabetes

    JAMA

    (2000)
  • AI Adler et al.

    Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): prospective observational study

    BMJ

    (2000)
  • A Gray et al.

    Implementing intensive control of blood glucose concentration and blood pressure in type 2 diabetes in England: cost analysis (UKPDS 63)

    BMJ

    (2002)
  • The Department of Health

    The National Service Framework for Diabetes—the delivery strategy

    (2003)
  • The NHS Confederation
  • RC Turner et al.

    Glycemic control with diet, sulfonylurea, metformin, or insulin in patients with type 2 diabetes mellitus: progressive requirement for multiple therapies (UKPDS 49). UK Prospective Diabetes Study (UKPDS) Group

    JAMA

    (1999)
  • C Wills et al.

    Retrospective review of care and outcomes in young adults with type 1 diabetes

    BMJ

    (2003)
  • Cited by (91)

    • The quality of diabetes care: A comparison between patients enrolled and not enrolled on a regional diabetes register

      2011, Primary Care Diabetes
      Citation Excerpt :

      A significantly greater proportion of the non-register type 2 diabetes patients had been prescribed diet only for glycaemic control (41% vs 30%) and mean HbA1c was less (7.2% vs 7.4%, data not presented) than the register type 2 diabetes population. This suggests that their diabetes may have been perceived to be less severe [23], and thus possibly managed less aggressively, as lower proportions had the recommended clinical measures and laboratory tests within recommended intervals (Tables 2 and 4). The notable exceptions were measurement of blood pressure, which was similar in both groups (96%), and HbA1c, where almost all the non-register group had been tested (98% vs 92%).

    View all citing articles on Scopus
    View full text