Our dash to bring relief to tsunami survivors
When striving for quality points sells patients short
Dr Eugene Hughes pleads for his GP colleagues to view the quality framework targets for diabetes as no more than a bare minimum
Never has there been so much activity in the field of diabetes. In practices up and down the country people with diabetes are being weighed, told to stop smoking and having their blood pressures taken. Laboratories are overwhelmed with requests for HbA1c and lipid profiles. Gallons of early-morning urine await microalbumin testing.
And what exactly is behind this new-found enthusiasm? An altruistic desire to improve metabolic outcomes in people with diabetes? The rediscovery of a set of guidelines on a dusty shelf? Of course not it's the quality and outcomes framework of the new GMS contract. It's about points. It's about money.
Not that I have any problem with this in itself. For many years diabetes enthusiasts among GPs have slaved away at such endeavours for the miserable pittance offered by the chronic disease management payments. Now at last there is a reasonably realistic payment involved.
And, of course, it will benefit people with diabetes won't it? Surely my increasingly elderly patients with long-standing type 2 diabetes will enjoy an improved quality of life as a result of being bombarded with tests and having to eat a bowlful of medication each morning? After all, the quality framework is evidence-based and that's what the evidence shows....er, doesn't it?
Two fundamental flaws
I have two fundamental problems with the quality framework. The first and most worrying is that the outcome indicators for blood pressure, HbA1c and cholesterol are not evidence-based at all. They are merely convenient stages on the payment scale that awards points to practices according to the proportion of eligible patients hitting the relevant indicator level.
The real danger is that these indicator levels will come to be accepted as appropriate targets for individual patients. They are not. They are merely the absolute bare minimum we should be aiming for.
The second flaw in the framework is that, despite its title, it has surprisingly little to do with outcomes. In relation to diabetes, 13 of the 18 indicators relate solely to processes of care.
Crude payment indicators not evidence-based targets
Much has been made of the claim that everything in the quality and outcomes framework is 'evidence-based'.
At the risk of teaching grandma to suck eggs, it may be worth recapping on a couple of basic principles.
Evidence-based medicine is the application of quality research findings to individuals and populations in the belief that the intervention applied is likely to bring benefit.
Targets are levels of control that may be appropriate to populations or individuals. At a population level, a suggested target for example for glycaemic control is likely to be based on available scientific evidence.
But clinical trials are conducted on specific groups of subjects that may not represent a 'normal' practice population. Most trials exclude patients who are very elderly, those who have co-morbidities and the non-compliant.
At an individual level, the population-wide target may therefore be adjusted to reflect the particular situation of the patient. So, whereas a population target for HbA1c of 7 per cent may be appropriate, an individual currently at
10 per cent may work towards a target of 9 per cent, if that is deemed sensible and achievable.
Likewise, an 85-year-old driver with an HbA1c of 8 per cent may derive
no benefit from getting down to
7 per cent, but may be at risk of complications such as hypoglycaemia.
Let's return to the Q&O. Any claim that it is evidence-based begins to unravel as soon as you scrutinise the indicator levels set for HbA1c, blood pressure and cholesterol.
The evidence for the benefit of tight control of blood glucose comes from trials such as the UK Prospective Diabetes Study. Based on this evidence, Diabetes UK suggests the following targets for HbA1c:
<6.5 per="" cent="" optimal="">6.5>
6.5 - 7.5 per cent acceptable control
>7.5 per cent poor control
The American Diabetes Association (ADA) recommends a target of <7.0 per="">7.0>
In contrast, to earn the maximum 16 points available for indicator DM6 of the Q&O, practices have to get 50 per cent of their diabetic patients to achieve a last recorded HbA1c of 7.4 per cent or less. There is some latitude built in to reflect how difficult it is to achieve this level.
The 7.4 per cent figure is not an evidence-based target, it is a convenient payment stage. There is a danger that health professionals will accept 7.4 per cent as a metabolic target and will consider that, having achieved this level, there is no point in attempting further improvement. But the evidence suggests <6.5 per="" cent="" is="" optimal,="" and="" brings="" additional="" benefits="" in="" terms="" of="" microvascular="" and="" macrovascular="">6.5>
The evidence for the benefits of good blood pressure control in patients with diabetes comes from a host of trials including HOT, HOPE, ALLHAT and UKPDS.
The most recent evidence-based guidelines come from the British Hypertension Society, which set a target of 130/80mmHg for most people with diabetes. The ADA suggests 130/85mmHg.
Yet under indicator DM12 of the Q&O, practices can earn the maximum 17 points for getting the last recorded blood pressure down to 145/85mmHg or below in 55 per cent of their diabetic patients.
But 145/85mmHg is not an evidence-based level. It is a convenient payment stage.
There is a danger that health professionals will accept 145/85mmHg as a blood pressure target and will consider that, having reached this level, there is no point in attempting further improvement. However, the evidence suggests the lower the blood pressure, the better.
The evidence for secondary prevention comes from studies such as 4S, CARE, and LIPID. The evidence for primary prevention comes from the HPS and CARDS. Most authorities agree that diabetes is itself a cardiovascular risk factor, so all prevention is in fact secondary. This suggests that most, if not all, patients should be on lipid-lowering therapy.
The evidence-based targets set out in key guidelines are under revision, but it seems likely that the Joint British Societies' updated guidelines, due out later this year, will recommend targets of:
Total cholesterol 4mmol/l
LDL cholesterol 2mmol/l
Yet under DM17 a total cholesterol
of 5mmol/l or less must be achieved in 60 per cent of patients to earn the practice six points.
The 5mmol/l level is not an evidence- based target. It is a convenient payment stage. There is a danger that health professionals will accept 5mmol/l as a metabolic target and will consider that, having reached this level, there is no point in attempting further improvement. But the evidence suggests the lower the better. Moreover, there is also no reference in the Q&O to LDL levels, despite the fact there is increasing recognition that they may have greater prognostic significance than total cholesterol levels.
Outcomes are neglected
A subsidiary problem is the preoccupation of the Q&O with process measures.
Returning for a minute to the importance of definitions, if I measure a patient's BMI, that is a process. If I then suggest to the patient that a lower BMI might be beneficial to health, and offer assistance in terms of dietary advice or medication, that is an intervention. If the patient subsequently loses weight (or even if they don't) that is an outcome.
If the Q&O as its name suggests were genuinely about outcomes, it would do more to reward improved outcomes and the interventions that underpin them. To illustrate this, take the extreme example of two fictitious practices, each with the same number of diabetic patients on its list (see table, above right).
Practice 2 will have undertaken an immense amount of extra work compared with practice 1. There will have been significant improvements in microvascular and macrovascular outcomes in the practice population. The costs in terms of human resources and appropriate prescribing of medication will not be met by the extra Q&O income provided by just three additional points.
Chase the points but
remember the evidence
To conclude, the Q&O has far too little to do with either quality or outcomes. If it were about quality, it would set disease outcome targets that can be supported by reference to an evidence base. If it were about outcomes, it would do much more to reward practices that go beyond the processes of data collection and actually strive to achieve better disease control for their diabetes patients.
My message is to go for the points, but don't lose your soul. Never forget the real outcome targets that we should strive for nor the evidence underpinning them.
How the quality framework rewards process more than outcome
Practice 1 has recorded data on all its diabetic patients, showing that:
·100% have a BMI >30
·100% are smokers and have been told to stop but haven't
·100% have a BP >140/85mmHg
·100% have a total cholesterol >5mmol/l
·100% have HbA1c values >7.4 but <10 per="">10>
·100% have a record of absent peripheral pulses
·100% have a record of abnormal neurological findings
·100% have abnormal creatinine and microalbumin levels, but none are on ACE inhibitor
The practice opts to take no action. It has achieved 49 points almost 50 per cent of the total for diabetes but it is not providing quality care.
Practice 2 has:
·Reduced the average BMI of all patients from 30 to 25
·Stopped all patients smoking
·Reduced the average BP from 160/100 to 150/90mmHg
·Reduced average cholesterol values from
6 to 5.1mmol/l
·Referred all patients with at-risk feet for further evaluation
·Reduced the average HbA1c from 9 to 7.5 per cent
·Acted on abnormal microalbuminuria testing by appropriate prescribing of ACE inhibitor
This practice will achieve 52 points just
three more than practice 1.
Eugene Hughes is a GP in the Isle of Wight, a member of Primary Care Diabetes Europe and editor of Diabetes and Primary Care journal