Guidance
Rationale and impact
- Full formal risk assessment
- Communication about risk assessment, lifestyle changes and treatment
- Aspirin for primary prevention of cardiovascular disease
- Cardioprotective diet
- Discussions and assessment before starting statins
- Statins for primary prevention of cardiovascular disease
- Statins for secondary prevention of cardiovascular disease
- Optimising treatment for people on statins
- Lipid target for secondary prevention of cardiovascular disease
- Assessing response to treatment
Rationale and impact
These sections briefly explain why the committee made the recommendations and how they might affect practice.
Full formal risk assessment
Recommendations 1.1.7 to 1.1.11
Why the committee made the recommendations
The committee agreed that the evidence suggested QRISK3 performed best among tools evaluated in a UK population to assess the risk of a person without established CVD having a CVD event within the next 10 years. They agreed that no tool is very good at accurately discriminating between who will and who will not have a CVD event, but that tools are useful for guiding decisions about interventions to prevent CVD based on estimated risk.
A small amount of evidence suggested that the additional fields included in QRISK3 (such as severe mental illness, regular corticosteroid use and atypical antipsychotic use) enabled the tool to perform better than QRISK2 at predicting CVD events for people with these risk factors. Use of QRISK3 should, therefore, result in more people within these groups being appropriately considered for risk reduction approaches including statin treatment.
It was noted that the group of people with severe mental illness used to develop and validate QRISK3 included a high proportion of people with severe and moderate depression. This is reflected in the definition of severe mental illness in QRISK3 but does not reflect the definition used in electronic clinical systems in primary care. The committee agreed, informed by their clinical experience and expert opinion, that people with moderate to severe depression are not considered to have as great an increased risk of CVD as people with schizophrenia, bipolar disorder and other psychoses. By using data that grouped these conditions together, QRISK3 may underestimate CVD risk for people with schizophrenia, bipolar disorder and other psychoses. Despite this the committee agreed to recommend use of the tool for people with severe mental illness (however defined), but clinical judgement should inform interpretation.
The committee was aware that the NHS Health Check best practice guidance states that gender should be recorded as reported by the individual. If the individual discloses gender reassignment, they should be provided with CVD risk calculations based on both genders and advised to discuss with their GP which calculation is most appropriate for them as an individual. They agreed that healthcare professionals are expected to follow this guidance when undertaking formal risk assessments.
An age range is given for QRISK3 because it is only intended for people aged between 25 and 84 years (inclusive).
The committee agreed that the use of a risk tool remains appropriate in people with type 2 diabetes to support shared decision making. They agreed, based on the evidence, that QRISK3 performed better than QRISK2 for the population as whole and so should be used for people with type 2 diabetes.
The committee agreed to remove a 2014 recommendation to complete as many fields of the risk assessment tool as possible because QRISK3 explains that the tool can overestimate risk if fields are left blank. They also noted that BMI, ethnicity and family history of CVD should be recorded in people's medical records, so the committee agreed that the 2014 recommendation on recording this information was no longer needed.
Evidence on the performance of QRISK3 was not considered sufficient to suggest changing the 2014 recommendations against using a CVD risk tool in people with type 1 diabetes or CKD as it had not been validated in a separate population from that in which the tool was developed. The committee agreed these groups should be considered high risk, as should people with familial hypercholesterolaemia.
Based on their clinical experience, the committee identified a list of factors for which all CVD risk tools underestimate risk. They highlighted the importance of using clinical judgement to interpret risk scores. As risk scores are used to guide decisions about interventions to prevent CVD, the committee agreed it was particularly important to ensure people are not incorrectly considered to be at low risk.
The 2014 recommendation to consider people aged 85 and older to be at increased risk of CVD was retained as the committee agreed with this statement and there are still no tools for this age group. The committee highlighted it is important that this group be considered for interventions to prevent CVD even though a formal risk assessment would not be carried out.
The evidence for lifetime risk tools was not considered sufficient to recommend their use instead of 10‑year risk tools. However, the committee agreed they can have value in communication of risk. See the section on communication about risk assessment, lifestyle changes and treatment.
How the recommendations might affect practice
QRISK2 is currently integrated into electronic clinical systems, so 10‑year CVD risk assessments can be generated using data already available in a person's electronic records. At the time of development, the committee was aware of ongoing discussions about continuation of the inclusion of QRISK in electronic clinical systems.
Using QRISK3 instead of QRISK2 will require clinical systems to be updated by software developers for the impact on practice to be minimised. Public Health England issued guidance in August 2021 on using QRISK3 in NHS health checks and how to deal with the transition period (responsibility for the NHS Health Check programme has transferred to the Office for Health Improvement and Disparities, but the guidance produced by Public Health England remains current).
QRISK3 requires some additional clinical information that was not required for QRISK2. However, if integrated into electronic clinical systems, QRISK3 is not likely to require additional resources over QRISK2. There may be some implementation costs as healthcare professionals become familiar with the additional information included in QRISK3 and in managing the transition period.
Communication about risk assessment, lifestyle changes and treatment
Why the committee made the recommendation
The committee agreed that the evidence did not support using lifetime CVD risk assessment tools to guide decisions on the need for statin treatment because their accuracy could not be reliably assessed.
However, the committee noted that the usefulness of lifetime risk tools is primarily in communicating risk. They agreed by consensus that lifetime risk tools should be considered to help inform discussions about risk and motivate lifestyle changes. The committee highlighted that these tools may underestimate the ongoing benefit of lipid-lowering treatments as they do not predict risk reduction from taking medicines, and noted this should be considered when interpreting the results. They agreed lifetime risk calculation would not be necessary for everyone, but it may be particularly useful for people with a QRISK3 score less than 10% or under 40s who have CVD risk factors.
How the recommendation might affect practice
Lifetime risk tools are not routinely used in current clinical practice. The committee noted that there may be resource implications for calculating lifetime risk score estimates because lifetime risk tools are not currently embedded into electronic clinical systems, so scores are not automatically generated. There may also be implementation costs related to educating healthcare professionals about lifetime risk calculators. It is not clear if use of lifetime risk tools will result in longer consultations.
The committee agreed that the online calculators for lifetime risk tools such as QRISK-lifetime were easy to complete and provided some interpretation of the risk scores to aid discussions, but acknowledged that lifetime risk assessment would not be done for everyone.
The committee believe that using lifetime risk tools may have a long-term benefit in encouraging people to participate in lifestyle changes or engage in treatment, if appropriate. Given this, any additional time costs were considered likely to improve management of CVD risk and reduce future CVD events.
Aspirin for primary prevention of cardiovascular disease
Why NICE made the recommendation
NICE's surveillance team reviewed the evidence about aspirin for the primary prevention of CVD. Based on the review, NICE decided to add a do not routinely offer recommendation about this. For full details see the January 2023 exceptional surveillance report.
Cardioprotective diet
Why the committee made the recommendation
There was no available evidence comparing the effectiveness of dietary cholesterol strategies with normal diets for adults with and without CVD, so the committee updated the 2014 recommendation based on their clinical experience and expert opinion. They removed the reference to restricting dietary cholesterol intake.
Only evidence on dietary cholesterol was in scope for review and therefore the guidance on total fat intake and proportion of saturated fat versus unsaturated fat was not changed.
How the recommendation might affect practice
The committee agreed healthcare professionals already better understand the lack of a relationship between dietary cholesterol and CVD risk, so the new recommendation reflects current practice. The committee agreed there should be no change in practice or resource impact to the NHS because of this updated recommendation.
Discussions and assessment before starting statins
Recommendations 1.5.1 to 1.5.7
Why the committee made the recommendations
New evidence on adverse effects while on statins supported the 2014 recommendation. Evidence on the risk of muscle pain and rhabdomyolysis with statin use demonstrated a real effect, but the large body of evidence showed this was a very small increased risk when compared with similar populations not on statins; that is, when using high-intensity statins approximately 16% of people reported experiencing muscle pain, but of these cases only around 1 in 12 were likely to be due to the statin.
The committee agreed to strengthen the recommendation to reassure people that the risk of these adverse effects occurring is low.
The evidence supported the other 2014 recommendations, so they were retained. The committee agreed a full lipid profile should be provided before starting statins and amended the recommendation on lipid measures accordingly. See the section on assessing response to treatment for further details.
How the recommendations might affect practice
The recommendation on adverse effects has been strengthened to emphasise the low risk of experiencing severe muscle adverse effects because of statin treatment. It is not expected to have an impact on resource use as discussions on adverse effects are already an important part of current practice in prescribing and monitoring statins.
Statins for primary prevention of cardiovascular disease
Recommendations 1.6.1 to 1.6.12
Why the committee made the recommendations
Evidence on both the effectiveness and adverse effects of statins showed high-intensity statins are clinically effective and cost effective compared to no statins, low-intensity statins, or medium-intensity statins for preventing CVD in people without CVD.
The committee agreed to retain 20 mg as the recommended starting dose for all people starting atorvastatin for primary prevention of CVD. Although there was committee consensus that higher doses have a greater effect, they agreed that starting at the lowest effective dose was likely to be preferable to people, but that up-titration of the dose should be considered as appropriate, following recommendation 1.9.1.
The evidence supported the 2014 recommendations on optimising lifestyle changes and treating comorbidities and secondary causes of dyslipidaemia before starting statins, so they were retained.
The committee agreed to retain the following recommendations for research because there is still a lack of direct evidence in these areas:
Statins and QRISK score
Evidence showed that statins are cost effective for people with 10‑year CVD risk scores less than 10%.
The committee agreed that if more people took statins there would be a greater reduction in CVD events. However, they also recognised that practical considerations needed to be taken into account.
They agreed that risk scores are an important aid to shared decision making on statins. National audit data (CVDPREVENT) suggests that 60% of people without CVD and a QRISK score of 20% or more are prescribed lipid-lowering treatment, compared with 50% for people with scores of 10% or more. Therefore, the committee consensus was that an even smaller proportion of people with scores less than 10% may choose to take statins.
The committee agreed that focusing on increasing uptake among people with the most potential to benefit would have more impact than lowering the statin treatment threshold. The 10% 10‑year QRISK score was therefore retained as the threshold for offering statins. Although QRISK3 is specified in the recommendations, it is acknowledged that QRISK2 may be used in some circumstances until QRISK3 is embedded in electronic clinical systems (see the panel after recommendation 1.1.8 for details). The 10% threshold applies whether QRISK2 or 3 is used.
Despite this, the committee agreed that a more person-centred approach should be adopted and recommended atorvastatin 20 mg as an option for people who want to take statins, irrespective of their QRISK3 score, or where clinical judgement suggests the person may be at high risk of CVD (for example, if the person has CVD risk factors not covered by QRISK3).
How the recommendations might affect practice
Most recommendations about statin treatment have been retained from the 2014 update of the guideline and so should not require a change in practice.
National audit data suggests that about half of people with a QRISK score of 10% or more are on lipid-lowering treatment. It is unclear if people are not being offered treatment or if they are declining or stopping treatment.
The recommendation to consider starting atorvastatin 20 mg for people with QRISK3 scores less than 10% is a change in practice. The impact on medication and monitoring costs and workload will depend on the level of uptake. For details of the impact of an increase in statin use, see the impact section on lipid target for secondary prevention of CVD.
Statins for secondary prevention of cardiovascular disease
Recommendations 1.7.2 to 1.7.5
Why the committee made the recommendations
Evidence on both the effectiveness and adverse effects of statins showed high-intensity statins are clinically effective and cost effective compared to no statins, low-intensity statins, or medium-intensity statins for preventing CVD in people with CVD.
The evidence supported the 2014 recommendations on initial treatment with statins and so they were retained. The recommendation on acute coronary syndromes was amended to clarify that a full lipid profile is needed on admission to hospital and 2 to 3 months after starting treatment. See the section on assessing response to treatment for further details.
Optimising treatment for people on statins
Why the committee made the recommendation
The committee amended the 2014 consensus recommendation on what to do if someone taking statins does not reach their lipid target to apply to both the existing primary prevention target and the new secondary prevention target. The recommended actions remain the same.
Lipid target for secondary prevention of cardiovascular disease
Recommendation 1.7.1, recommendations 1.7.8 to 1.7.11 and recommendations 1.10.1 to 1.10.2
Why the committee made the recommendations
Lipid target
The committee agreed LDL cholesterol and non-HDL cholesterol levels should be reduced as much as possible in people with CVD. However, people respond differently to statins and other lipid-lowering treatments, and it is not cost effective to offer the full range of treatments to everyone with CVD.
The clinical evidence consisted of 34 randomised control trials (RCTs). Clinically significant reductions in LDL cholesterol and non-HDL cholesterol levels compared to placebo were seen for all 4 lipid-lowering treatments covered by the clinical trials: alirocumab, evolocumab, ezetimibe and inclisiran. The majority of people in the trials were also taking statins.
Modest reductions in major CVD events such as myocardial infarction, stroke and related deaths were also seen for all 4 medicines. The committee recognised that some of the trials involved short follow-up periods of 1 year or less, so these medicines are likely to have a bigger impact on CVD events over the long term. There was no clinically important increased risk of adverse events. Injection site reactions were more frequent with alirocumab, evolocumab and inclisiran than with placebo but these were mild and not persistent.
An economic model was developed using estimates of the impact of lipid-lowering treatments on LDL cholesterol (from a network meta-analysis of the 34 RCTs), combined with estimates of the impact of LDL cholesterol reduction on major cardiovascular events (from a published meta-analysis of statin RCTs). The economic model calculated, for all possible baseline LDL cholesterol levels, the reduced admissions to hospital for stroke, myocardial infarction and cardiovascular procedures and the associated life expectancy increases, quality of life improvements and treatment cost savings as a result of taking lipid-lowering treatments. This was offset against the cost of lipid-lowering treatments and associated monitoring costs. The data for the 2 PCSK9 inhibitors (alirocumab and evolocumab) were combined and were not analysed separately because treatment effects were found to be the same in the network meta-analysis.
The model estimated the absolute LDL cholesterol target at which it was cost effective to escalate treatment for people on a high-intensity statin. The analysis explored different combinations of lipid-lowering treatments.
Escalation of treatment was cost effective for people on statins with LDL cholesterol levels of more than 2.2 mmol per litre. There was a little more uncertainty about the cost effectiveness of escalating treatment for people with LDL cholesterol levels between 2.0 mmol and 2.2 mmol per litre. An LDL cholesterol target of 1.8 mmol per litre was not cost effective. The committee decided that 2.0 mmol per litre was likely to be cost effective and would allow more people to be treated than 2.2 mmol per litre.
Even though the main economic analysis was based on the impact of lipid-lowering treatment on LDL cholesterol levels, the committee recognised the need to identify a non-HDL cholesterol target for use when LDL cholesterol levels have not been requested or calculated.
Using the distribution of LDL cholesterol levels for the population with CVD and on a statin in the clinical practice research datalink (CPRD) dataset, 42% of people had LDL cholesterol levels of 2.0 mmol per litre or more. Using the same data, the threshold for non-HDL cholesterol that would produce an identical number of people being escalated for treatment was 2.6 mmol per litre. An alternative approach would be to use the Friedewald equation and insert the mean triglyceride level of 1.4 mmol per litre along with the LDL cholesterol of 2.0 mmol per litre. This approach also indicates a non-HDL cholesterol target of 2.6 mmol per litre.
Both target measures are slightly higher than other national and international targets because, unlike other targets, the LDL cholesterol target is based on the cost effectiveness of treatment escalation. However, the committee thought it was sufficiently similar and, because it was more affordable, was more likely to be implemented.
Statin plus ezetimibe
In a separate analysis, escalation with statin plus ezetimbe (but no injectable treatment) was evaluated at different LDL cholesterol levels. Ezetimibe was cost effective regardless of LDL cholesterol, so the committee agreed that it could be considered for people with lipid levels below the agreed target. They noted that the trade-off between further reducing risk and increasing medication should be taken into account. These should be considered and fully discussed with the person as part of informed shared decision making. Furthermore, the committee agreed that adherence may be lower for people on 2 pills rather than 1, especially if they are below the target.
Statins are contraindicated or not tolerated
The committee did not review the evidence for the clinical effectiveness of lipid-lowering treatments in people who are statin intolerant or for whom statins are contraindicated but based their recommendations on NICE's technology appraisal guidance on alirocumab, bempedoic acid, evolocumab, ezetimibe and inclisiran.
The committee emphasised that statin treatment is known to be the most effective method of reducing the risk of CVD events and that this should be the main treatment for most people. They highlighted the importance of reviewing statin medication in response to adverse effects before deciding someone is statin intolerant.
An economic analysis estimated the absolute LDL cholesterol target at which it was cost effective to escalate treatment for people on ezetimibe who are statin intolerant. The analysis explored different combinations of lipid-lowering treatments.
The committee discussed whether the lipid target should be different because of the different treatment options and associated costs. However, it was noted that this may introduce inequality regarding access to lipid-lowering treatment. Also, the target at which escalation of lipid-lowering treatment is cost effective did not change when the statin intolerant population was included in the economic model alongside the statin tolerant population, largely because the prevalence of statin intolerance is relatively low. Therefore, the committee agreed that the target for people who cannot take statins should be the same as for those who can take them.
They recommended offering ezetimibe to people who cannot take statins (in line with NICE's technology appraisal guidance on ezetimibe) and, if this does not achieve the lipid target in this guideline, to offer alternative or additional lipid-lowering treatments (in line with other technology appraisal guidance).
How the recommendations might affect practice
It is expected that recommending a specific lipid target for secondary prevention of CVD will lead to an increased use of lipid-lowering treatments. The committee was aware that NHS England had recently introduced a target as part of the Quality and Outcomes Framework (QOF). The target recommended in this guideline is similar to the 2023/24 QOF, although data showed that, in many people, the QOF target is not being met. In June 2023, the CVDPREVENT audit reported that 28.7% were meeting the target.
Increased uptake of statins, ezetimibe and other lipid-lowering treatments will result in higher medication and monitoring costs to the NHS. It will also contribute to an increased workload in primary care, including for GP practices and pharmacies, and in laboratories that process lipid profile and liver function tests. The committee agreed that increased uptake of lipid-lowering treatments is necessary for an overall improvement in population health, but that the extra cost of lipid-lowering treatment would be partly offset by savings due to a reduction in CVD events (including hospital admissions for stroke, heart disease and cardiovascular procedures).
Assessing response to treatment
Recommendations 1.11.1 to 1.11.12
Why the committee made the recommendations
The committee agreed that more flexibility in the timing of blood tests to measure lipid levels after starting high-intensity statins was reflective of actual clinical practice and recommended a timeframe of 2 to 3 months, rather than at 3 months of treatment as recommended in the 2014 guideline. They also recommended blood tests should be done 2 to 3 months after changing treatment.
They agreed that blood tests should provide a full lipid profile and that LDL cholesterol levels can be calculated rather than measured directly. They recommended that a full lipid profile should be done to inform annual medication reviews about secondary prevention of CVD, and considered for reviews about primary prevention of CVD.
These changes were also applied to the recommendations on baseline blood tests before starting statins and statin treatment after an acute coronary syndrome.
The evidence supported the 2008 and 2014 recommendations on when to measure creatine kinase, not stopping statins because of an increase in blood glucose or HbA1c and restarting statins if stopped because of drug interactions or illness, so they were retained.
How the recommendations might affect practice
The requirement to check lipid levels after changing lipid-lowering treatment, as well as after starting treatment, will result in higher monitoring costs. However, calculating LDL cholesterol as well as non-HDL cholesterol should not add to the cost per test.