Are the NICE guidelines designed for primary or secondary prevention of CHD?
Both. These guidelines update the relevant sections of the NSF for CHD and are the latest formal guidance on CHD for the NHS. The recommendations for primary prevention include the identification and assessment of cardiovascular risk to include a determination of an individual's absolute risk of developing cardiovascular disease (CVD).
Recommendations for lipid modification are made for patients identified as high risk and for those with established CVD (acute coronary syndrome, stroke and peripheral artery disease). Interventions to modify lipids were only considered when there was evidence that they could reduce the risk of CVD events. Pharmacological and non-pharmacological risk factors were considered.
The draft guidance identified QRISK as a more accurate method of estimating cardiovascular risk. What was behind the decision to continue using the older, Framingham risk assessment?
The guideline development group (GDG) recommended that healthcare professionals should be aware that risk tools give only approximate results, that they should be used as the starting point for a discussion between clinicians and patients.
The strengths and weaknesses of Framingham are known and it is already widely used. Although there is strong demand for a risk equation developed and validated in a UK population, the GDG decided that a number of issues needed clarification. A recommendation was made for urgent research into the development of a risk tool for the contemporary UK population.
The GDG went to some length to assess QRISK. When the revised QRISK paper was published, because of the particularly advanced statistical techniques used to account for missing data, the GDG members requested specialist statistical advice from three leading experts. Having received generally positive recommendations, the GDG provisionally recommended QRISK, subject to stakeholder consultation.
Many broader issues were raised by the general medical community than those identified in the specialist reviews. The GDG reviewed issues such as the ascertainment of outcomes, the importance of independent validation, validation of QRISK other than in general practice records, use across different healthcare settings and comparisons with other risk tools.
Other considerations included over- and underestimation of risk with specific tools. The GDG considered that because statins were known to have benefits below the thresholds currently used, underestimation might have greater morbidity consequences than overestimation. The stakeholders also questioned the accuracy of recorded risk factor data used in GP datasets, but this was not considered relevant by the GDG.
Like Framingham, QRISK1 did not include ethnicity, but it did not include any adjustment to take this into account and therefore could have led to worsening health inequalities for this high-risk group. QRISK has now been developed to address this omission, although the authors indicate that it is important that it is validated by another team on independent populations.
How does this fit in with the government's cardiovascular screening programme, announced earlier this year?
This guideline describes cardiovascular risk assessment and the modification of blood lipids; ultimately, it is a guideline, not a programme. The details of the cardiovascular risk programme are undergoing development and stakeholder consultation, but what seems apparent at this stage is that it may target a much wider group of conditions, including diabetes, which obviously are not dealt with by this NICE guideline.
How should GPs estimate left ventricular hypertrophy (LVH)? Are all potential at-risk patients expected to have an ECG/echo?
Presence of LVH should be factored into the equation where this is known. This does not mean there should be general screening of LVH for patients undergoing risk assessment.
The systematic approach recommended in the guidelines promises to identify an additional 1.5 million patients at risk, who should be taking statins. How should GPs support these previously 'well' patients?
First, it is important to use every day, jargon-free language to communicate information on risk. Adequate time should be set aside during the consultation to provide information on risk assessment and to allow any questions to be answered.
Patients should be offered information about their absolute risk of CVD and about the absolute benefits and harms of an intervention over a 10-year period, in a form that presents individualised risk/benefit scenarios, the absolute risk of events numerically, and appropriate diagrams and text.
Clinicians should also find out what, if anything, the patient has already been told about their CVD risk and how they feel about it, and explore their beliefs about what determines future health because this may affect their attitude to changing risk.
Many people take OTC aspirin for primary prevention of CHD and I think this provides a good analogy for the aftercare of patients receiving statins for primary prevention of CVD.
The 1.5 million at-risk patients eligible for statins equates to approximately 45 patients per GP, notwithstanding the likelihood that practice nurses will take up a substantial part of this workload. It should also be remembered that some of our patients are at considerable risk, with risks of 30 and 40 per cent over the next 10 years.
What is implementation of these new guidelines going to cost the average practice? Is extra funding available for GPs to implement a primary prevention strategy?
GPs routinely ration the delivery of primary healthcare to manage an increasing workload, but we have been performing risk assessment ever since the NSF for CHD was published in March 2000. In our locality, an audit showed that 60 per cent of statin prescribing was for primary prevention, so it has already been going on for quite a while.
This guideline informs the process of continuing care that was set in motion by the NSF for CHD, in addition to updating those aspects relevant to identification of high-risk primary prevention patients, risk assessment and lipid modification.
Some PCTs have set in place local enhanced services for the primary prevention of CVD and there is always the possibility that these matters may be raised within a future quality and outcomes framework.
The new guidelines do not give target cholesterol for primary prevention. Why is this?
The clinical effectiveness of higher-intensity statins and combining statins with other drugs has not been shown for primary prevention. Owing to this lack of evidence, the GDG agreed that this guideline would not recommend the use of target levels of cholesterol for primary prevention of CVD.
Also, the absolute benefits of statins in primary prevention are considerably less than the benefits in secondary prevention - benefit is proportional to risk and the lower the risk of the adverse prognosis, the lower the potential benefit. I think of the target as being the receipt of treatment by the patient because this is what the evidence base shows. Think of statins as the new aspirin; take one a day for primary prevention.
Why should we measure LFTs with simvastatin, when the Heart Protection Study did not endorse it?
The patients we see in general practice are undifferentiated and the patients clinical triallists enrol in research studies are not representative of the patients we see.
Patients in trials tend to be male, younger and have less comorbid conditions, which is probably why I and many other GPs find our perspective on drug treatments often differs from the specialists who perform clinical trials. The most dramatic example of this is in heart failure, where it has been suggested that less than 10 per cent of real-life patients would ever be suitable for entry into a clinical trial.
The patients in the Heart Protection Study and many other statin studies were triaged for statin tolerance through tolerability testing with a small dose of the statin before the real trial began. These 'run-in periods' result in a statin-tolerant population that does not provide a realistic basis for making decisions on safety.
The 'lower is better' philosophy is based on studies like these, but disregards the safety issues that these studies obscure. The licence for simvastatin still includes the need for measuring LFTs and this seems quite sensible, given that GPs do not have the luxury of knowing who will tolerate a drug in advance.
This guideline recommends that LFTs are measured at three months and one year and states that they do not need to be measured again if they are normal.
The Framingham adjustment factors increase risk by a factor of 1.4 if the patient is a South Asian man and by a factor of 1.5 to 2.0 if the patient has affected relatives. What is the basis for these figures?
The 1.4 inflation factor for South Asian men is based on the standardised mortality ratio, which is the age- and sex-adjusted death rate compared with the reference population. This is where the frequently quoted statistic that South Asians have a 40 per cent increased risk of MI comes from. It has traditionally been used to adjust the risk calculated in the white reference population to a level that corresponds with the South Asian population. It is imperfect, but has been the best there is.
The inflation factors of 1.5 to 2.0 for family history come from a variety of studies that explored risk of CVD, such as the MALMO Preventive Project, the Women's Health Study and several others. The studies gave different figures, based on the exact relationships between relatives (siblings versus parents and children, for example) and the age at which the event occurred. Generally speaking, the younger the family member experiencing the event, the greater the risk.
The NICE guidelines advise calculating BMI, but aren't we all supposed to be measuring waist circumference these days?
The NICE clinical guideline on obesity recommends that BMI should be used in adults, supplemented with measurement of waist circumference. BMI is height standardised and more reproducible than waist circumference. Waist circumference adds some value as a direct measure of adiposity. Where one is weak, the other is strong. Remember also that the indications for weight loss drugs are all specified in terms of BMI.
Secondary prevention targets are tighter than those set out in the quality and outcomes framework. What should GPs aim for, given that it is difficult and costly to hit the NICE target? It is not clear from the guideline which is acceptable, audit or target.
Up-titration of initial statin treatment to achieve recommended cholesterol levels will only be successful in 37 per cent of patients. Therefore, cholesterol levels of 5mmol/L for total cholesterol and 3mmol/L for LDL have been adopted for audit purposes, because although all patients should be considered for optimisation of recommended treatment, the reality is that the majority of these patients will not achieve total cholesterol of 4mmol/L or LDL of 2mmol/L. The audit standard is the minimum level of acceptable performance.
In a similar way, the audit targets for BP in the quality and outcomes framework are the minimum level of performance and in clinical practice, I would consider reducing BP below the audit figure, but only where this is appropriate.
- Dr Rubin Minhas is a GPSI in cardiology in Gillingham, Kent, and a member of the NICE GDG. The guideline was developed by the National Collaborating Centre for Primary Care (NCC-PC). Any views expressed in this article are those of the author and not necessarily those of NICE or the NCC-PC.