This is a plain English summary of an original research article. The views expressed are those of the author(s) and reviewer(s) at the time of publication.
Annual testing to monitor blood lipids (fat and cholesterol) was both effective and saved money long term in by reducing cardiovascular disease in people with or without known heart disease.
Given the relatively low cost of treating high lipids (typically by using statins) it was more cost effective to over-treat people than to under-treat them.
The findings of this detailed modelling study are broadly in line with guidance. They also add clarity to previous uncertainties around the frequency of lipid monitoring and its relationship to diagnosis, prognosis and treatment.
More information is needed on patient willingness to be treated when they might not be experiencing any symptoms and on the long-term safety profile of some statins.
Why was this study needed?
Cardiovascular disease causes almost a third of deaths in the UK, many of which are potentially preventable. Measuring blood lipids helps identify those at risk and guides treatment with statins to lower blood lipids. Along with measuring blood pressure and assessing other risk factors such as smoking, it is used routinely in the UK to assess a person’s risk of future heart attack or stroke. An Oxfordshire based study showed that lipid testing had increased more than 15-fold in the last 20 years in this region and UK expenditure on statin treatment was more than £800 million in 2011. Despite the popularity of lipid testing there is uncertainty about how often it should be tested and which lipid measure or combination of measures is best.
This study aimed to estimate the clinical and economic value of using different lipid measurements and different lipid testing intervals for preventing cardiovascular disease. In particular, it explored the impact of natural biological lipid level variation and measurement error – so called “noise” – on the value of different lipid-monitoring strategies.
What did this study do?
This modelling study incorporated data from two systematic reviews and patient level data from the UK and Japan to reach its predictions.
A systematic review of prospective studies investigated the ability of different lipid markers (and their ratios) to predict cardiovascular events (e.g. non-fatal heart attacks and strokes), deaths from cardiovascular disease and death from any cause.
A second systematic review pooled the results from trials to estimate the lipid lowering effects of a high-intensity statin – atorvastatin – at different doses.
A third data source, individual health records from people attending a GP in the UK and from a Japanese study, was used to study variability in lipids over time, including both natural biological variation and variation due to measurement imprecision (noise).
The final step was to estimate the likely cost of different lipid monitoring strategies, taking into account lipid variability, test accuracy (false positives and negatives) and quality of life.
What did it find?
- A range of lipid blood measures (see Definitions) showed similar ability to predict risk of cardiovascular events. In people not already taking statins, the most predictive measures were non-HDL cholesterol, ApoB and composite measures (ratios of LDL/HDL cholesterol and TC/HDL). Other associations were smaller.
- For all the different lipid tests, “noise” or biological variation and measurement error make up 25% of the variation in measurements on people who are not already taking statins or using lifestyle measures to reduce their lipid levels. This means that many observed changes in repeat rest results are not “true” differences.
- Overall, to prevent cardiovascular disease and be cost effective, annual lipid monitoring was more beneficial compared with less frequent approaches. This applied to primary prevention in people with 10%, 15% or 20% risk of cardiovascular disease over ten years according to the Qrisk score and those already with heart disease (secondary prevention).
- When measuring lipids more frequently, for example, annually instead of every five years, there is a trade-off between detecting more people who are at risk and incorrectly classifying (and treating) others who are not at risk.
What does current guidance say on this issue?
A 2014 NICE guideline recommends ongoing assessment of cardiovascular disease risk for people over 40 years old using risk scores. Targeted comprehensive assessment should only be carried out for those shown to have more than 10% risk of cardiovascular disease over ten years. It recommends against comprehensive assessment for everyone over 40 years old.
Before starting medication, such as statins, to reduce blood lipid levels the guideline recommends taking an initial blood test to include measuring total cholesterol, HDL cholesterol, non‑HDL cholesterol and triglyceride concentrations. Statins chosen for treatment should be high intensity statins, such as atorvastatin (see Definitions). For people already taking statins to reduce their blood lipids, the NICE guideline recommends annual review, which may be informed by fasting lipid measurements.
What are the implications?
This modelling study’s findings are broadly in line with guidance. The findings clarify issues on blood test monitoring and its relationship to diagnosis, prognosis and treatment that are more complex than they may first seem.
Whichever test is used, it can be difficult to determine a person’s cardiovascular disease risk exactly, as variation in their repeat test results may simply be due to “noise”. However, statin treatment and monitoring are relatively cheap, whereas the consequences of cardiovascular disease are severe and costly to individuals, society and health services. So it is more cost effective to over-treat than under-treat.
Further research is needed into whether patients are prepared to take statins in the long term, with potential side effects, to protect against future ill-health.
Citation and Funding
Perera R, McFadden E, McLellan J, et al. Optimal strategies for monitoring lipid levels in patients at risk or with cardiovascular disease: a systematic review with statistical and cost-effectiveness modelling. Health Technol Assess 2015;(19):1-402.
This project was funded by the National Institute for Health Research Health Technology Assessment programme (project number 10/97/01).
Heart UK. Cholesterol – know your numbers. Maidenhead: Heart UK; 2015.
NICE. Cardiovascular disease: risk assessment and reduction, including lipid modification. CG181. London: National Institute for Health and Care Excellence; 2014.
Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre