How diabetes care is set to change under new contract
As a greater burden for diabetes care falls on GPs, Dr Tonia Myers and Dr Rina Davison discuss the way forward
hat are the implications for primary care of the diabetes national service framework?
The NSF shifts care of type 2 diabetics almost totally towards primary care, except for patients with bad complications. Secondary care will be limited to educating and training health professionals in general practice and treating selected groups, such as pregnant women, young children and acute emergencies. Local networks of people with diabetes will be set up to determine services and priorities.
GP practices will shoulder responsibility for primary education of newly diagnosed type 2 diabetics, planning the annual review and treating diabetes according to NICE guidelines.
The aim is for a network of primary care nurses to educate patients early on about lifestyle and diet, blood glucose monitoring and complications so they take responsibility for their own care. The NSF recommends each patient has a personalised plan.
By 2006 all practices must have a diabetic register with treatment regimes and targets, data such as the percentages of diabetics who have had HbA1cs, lipids, blood pressure, urinary albumin, creatinine ratios and smoking status measured in the last year.
Practices must record what has been done about abnormal results and write follow-up plans (systematic treatment regimens). PCTs will be responsible for setting up a digital retinal screening programme including call and recall systems using the practice's diabetic register, so all patients with diabetes will be screened by 2007. Many practices already provide these services, but the aim of the NSF is to standardise care across districts.
GP-led 'interface' clinics are being encouraged to form a seamless service between primary and secondary care.
New contract targets for risk factors in diabetics
The new GP contract has targets linked to the NSF and NICE documents on diabetes. Much is about data collection, but points are also awarded for tight control of glycaemia/blood lipids and blood pressure. In my experience, getting blood pressure down to 145/85mmHg or less in diabetics needs multiple agents and often causes postural hypotension. What is your approach to improve compliance and minimise risks?
I go for once-daily drugs with fewest side-effects. My first-line in diabetes without microalbuminuria or proteinuria is an ACE inhibitor, then a thiazide, calcium channel blocker, and vasodilating ?-blocker (carvedilol).
For diabetics with nephropathy, my first-line is ACE inhibitors or angiotensin receptor type 2 blocker (A2 blocker) then a thiazide, calcium channel blocker etc.
Fixed-dose combinations such as lisinopril plus hydrochlorothiazide, or irbesartan with chlorothiazide, or ACE inhibitors with calcium channel blockers are available and should improve compliance.
The new contract awards points for the percentage of diabetic patients whose total cholesterol is less than 5.0mmol/l. Many diabetics also have a high triglyceride level. How do you suggest we approach drug therapy?
NICE recommends a full lipid profile at diagnosis. If abnormal, exclude secondary causes by checking liver, thyroid and renal function. Repeat after three months of diet, which also gives glycaemic control an opportunity to improve. If it is still abnormal, you need to assess cardiovascular risk using tables (NICE guidelines for type 2 diabetes, table 4, page 9).
For secondary prevention, all patients must be on a statin at a dose that reduces LDL cholesterol to <3.0, even="">3.0,><2.5 in="" patients="" who="" have="" had="" coronary="" artery="" interventions.="" if="" hdl="" cholesterol="" is="">2.5><1.0 and/or="" triglycerides="">2.3, add a fibrate. If on dual therapy patients must be warned to stop immediately and without reference to a doctor initially if there are side-effects, especially generalised muscle pains.
For primary prevention using cardiovascular risk tables, treat with a statin if 10-year risk is >15 per cent and total cholesterol >5.0mmol/l. There is currently no strong evidence for using fibrates in primary prevention, only if triglycerides >5 because of the risk of pancreatitis. There are no specific risk tables for south Asian patients and risk tables underestimate the risk for diabetic patients so some may need a fibrate as well as a statin.
Should virtually all diabetics be on aspirin?
For primary prevention, the guidelines state if patients have a cumulative cardiovascular risk per year of more than 15 per cent they should be on aspirin. Of course, anyone who has ischaemic heart disease should be taking aspirin if they can tolerate it.
I prescribe aspirin to any diabetic patient with any of the following problems: hypertension, raised lipids, nephropathy and retinopathy if there are no contraindications.
What do you do when there is still sub-optimal control and you cannot get HbA1c down to the elusive 7.5 per cent or less?
Metformin, sulphonylureas and glitazones are equally effective in lowering HbA1C. If patients are taking two agents and there is still not good control, check they are sticking to their diet. It may be helpful to look at their sugar profile as an HbA1c only gives an average over three months.
Is the patient starting the day with a sugar of 11mmol/l pre-breakfast? If so, whatever you give them in the day cannot change the fact that they started badly. This would be an indication for once-daily insulin before bed. Post-prandial peaks is the main indication for the post-prandial regulators repaglinide and nateglinide.
A suitable patient would be an elderly person on no medication or perhaps a low dose of metformin, whose sugar goes up to 10 or 11mmol/l after a meal. There is a definite advantage to getting the sugar down to 7 or 8mmol/l post prandially, for symptomatic relief. Repaglinide and nateglinide are insulotrophic drugs, which must be given three times daily, with each meal, which can create compliance problems. They are less effective than a sulphonylurea in lowering HbA1C and cannot be given with a sulphonylurea.
Where does acarbose fit in?
Acarbose gives dire gastrointestinal side-effects. Patients often do not comply with 50mg three times daily. Metformin is preferable. It also causes gastrointestinal side-effects, but is more effective in reducing HbA1c.
What factors should GPs consider before contemplating starting a patient on insulin?
Absolute indications for insulin are poorly controlled diabetes, with weight loss and symptoms such as polyuria or polydipsia. Others are acute diabetic amyotrophy or bad neuropathy.
Some long-standing type 2 diabetics may eventually become insulin dependent rather than resistant because they lose their ? cell reserve.
Poorly controlled patients on maximal therapy with a high blood sugar before breakfast are suitable for a once-daily pre-bedtime dose of long-acting insulin.
This provides a basal degree of overnight insulin with an aim to get fasting glucose in the morning of 5-7mmol/l. A rough guide is to start with eight units at night.
Initiating insulin should usually be done in secondary care as there is a risk of hypoglycaemia and the back-up of diabetes nurses is still needed. GPs may still be able to do annual reviews in these patients, but it is usually the hospital doctor, GP specialist or community diabetes nurse who manipulates doses. Remember with type 2 diabetics, the problem is insulin resistance rather than insulin deficiency. These patients should remain on metformin because it improves insulin sensitivity and in theory reduces the need for insulin. Insulin is added in either once or twice a day.
Sulphonylureas should be reduced or stopped in these patients because they can cause hypoglycaemia, as can insulin.
What advice do you give patients on varying their insulin when sugars become high or during illness?
The rules apply to both type 1 or 2 diabetics. If sugars are consistently high the dose of short-acting insulin can be increased by two units at a time until blood sugar comes under control.
In a viral illness sugars may have increased from 7-8mmol/l, up to 11 or 12mmol/l, and then come down when the patient improves. In an elderly patient who feels unwell during an acute illness, check blood glucose.
If it is around 13 or 14mmol/l it is reasonable to do nothing for a few days apart from monitor and wait for it to settle.
I feel there is a danger of treating elderly disabled type 2 diabetics too aggressively by focusing on HbA1C and forgetting about quality of life. What is your view?
The reason for reducing blood glucose to 6.5mmol/l is to reduce cardiovascular, retinal and renal complications. Nevertheless, for elderly non-insulin dependent patients, one must consider quality of life and risk of hypoglycaemia with overly aggressive therapy.
In the older patient, uncontrolled blood pressure actually has more impact in terms of stroke risk than a sugar of 11-12mmol/l.
In general, the decision about initiating insulin should be based on the presence of symptoms such as nocturia, thirst or weight loss. For asymptomatic patients with co-morbidity such as rheumatoid arthritis or visual impairment I would pragmatically accept an HbA1C of 8.5-9 per cent after discussing the pros and cons with them.
How often should patients on once-daily overnight insulin monitor blood sugar?
These patients should check their early-morning glucose to prevent hypos. In patients over 80 without complications a target of an HbA1C under 7 per cent will probably lead to hypoglycaemic attacks and affect quality of life so I would accept a slightly higher HbA1c in these patients.
Is there any way back from insulin, perhaps for patients who have been on it for a while and have never been tried on a glitazone?
If a patient is well controlled with metformin and insulin there is little reason to change. The reason we are reluctant to start insulin in type 2 diabetics, apart from the practical difficulties, is that it causes massive weight gain. So patients with metabolic syndrome need ever-increasing doses of insulin often more than 150 units because of insulin resistance and they keep gaining weight. If they still have an HbA1c of 8-9 per cent or more one could argue about stopping the insulin completely.
In patients who are using insulin as an adjunct overnight (rather than those who have been converted to four times a day) it is acceptable to stop insulin suddenly and they will not develop ketoacidosis or hyperosmolar hyperglycaemia. Alternatively reduce by 10 units a week.
However, these patients must continue to monitor their blood sugar because there are type 2 diabetics who have depleted their own insulin stores and are now 'insulin dependent'.
of newly diagnosed type 2 diabetics~
How diabetes care is set to change with new contract
is set to change
with new contract
Indicator Points Maximum
Practice register 6
Record of BMI* 3 90%
Record of smoking status* 3 90%
Smokers offered cessation advice* 5 90%
Patients with record of HbA1c* 3 90%
with record of retinal screening* 5 90%
with record of neuropathy test* 3 90%
with record of BP* 3 90%
with record of microalbuminuria* 3 90%
with record of serum creatinine* 3 90%
with record of total cholesterol* 3 90%
who have had influenza immunisation 3 85%
during preceding September-March
Indicator Points Maximum
Note: All minimum thresholds are 25%
*within previous 15 months
Points for achieving surrogate outcomes
with last HbA1c Ã7.4 per cent* 16 50%
with last HbA1c Ã10 per cent* 11 85%
with record of peripheral pulses* 3 90%
with last BP Ã140/85 17 55%
with proteinuria or microalbuminuria
treated with ACEI or ARBs 3 70%
with last total cholesterol Ã5mmol/l* 6 60%
Note: all minimum thresholds are 25 per cent
*within previous 15 months1.0>