"All those over 55 should be offered drugs to lower cholesterol and blood pressure, according to a new study," reported BBC News. It said the report suggests that when assessing the risk of heart problems, offering treatment to all over-55s had the same results as testing for cholesterol or blood pressure problems. The authors have also argued it would be simpler and more cost-effective.
The news piece is based on a good-quality modelling study that has a well-argued case for an “age-alone” screening strategy. Current guidelines recommend that decisions to prescribe statins or blood pressure treatments for people at risk from cardiovascular disease (CVD) are based on a combination of multiple risk factors including age, sex, smoking, diabetes, serum cholesterol and blood pressure. This study compared screening people for treatment using this method with treating all people over 55. The findings suggest that both approaches perform at about the same accuracy and are similarly useful in preventing heart disease and stroke.
These findings will prove useful for decision makers, but on their own are unlikely to be sufficient to change policy. Further real life testing of both approaches will be needed. Age screening for future cardiovascular disease is simpler than current assessments, and avoiding blood tests and medical examinations seems an advantage. However, some people consider this to be "over medicalisation", and fear that increased use of statins and low dose blood pressure-lowering drugs as preventive treatments might lead to more adverse effects. Further research and debate is needed.
Where did the story come from?
The study was carried out by researchers from the Wolfson Institute of Preventive Medicine in London. The authors have no support or funding to report. The study was published in the peer-reviewed medical journal PLoS ONE.
One of the authors, Professor Sir Nicholas Wald, reportedly holds patents for a combination pill for the prevention of cardiovascular disease.
Both the Daily Mail and The Daily Telegraph focus on the main implications of this modelling study, which is that if the authors’ proposals were implemented, screening by age would result in all over-55s being prescribed statins. Overall, the reports are generally accurate.
What kind of research was this?
In this study, the researchers created a model to compare the effect of different screening methods on the risk of future CVD events (such as heart attack or stroke).
The model was of a hypothetical population of 500,000 people up to 89 years of age, whose 10-year risk of having a CVD event was either estimated according to their age, or through a calculation known as the Framingham risk equation. This is a standard equation that is used to predict an individual’s chance of having an event based on a combination of risk factors (age, smoking, diabetes, blood pressure and cholesterol levels).
The researchers said that out of all of the factors taken into account in the Framingham risk equations, age has the greatest influence on a person’s subsequent risk of CVD. They suggested that using age alone may be a simpler screening strategy to decide which people need treatments that could prevent the onset of CVD.
The aim of this study was to compare the accuracy of various thresholds of age and risk level in predicting future CVD events, with screening using age and the other risk factors based on regular, five-year, Framingham risk assessments.
What did the research involve?
The researchers explain that screening for risk of future CVD events usually involves testing of the major risk factors, such as blood pressure and serum cholesterol, in combination with age, smoking and diabetes history. However, as age is the factor most closely linked to an individual’s chance of a heart attack or stroke, the policy of selecting people above a certain age is, in effect, already selecting people at high risk. As such, they propose that the other factors that are used in screening add little extra prognostic information.
In the modelling study, vascular risk was estimated using the Framingham equations in a theoretical sample population of 500,000 people aged less than 89 years. This sample population was generated using a computer simulation that ensured the population had the same age and sex distributions based on national statistics for England and Wales in 2007. The distribution of risk factors within this hypothetical population was made using data from a health survey of England. Each hypothetical person was allocated as either a smoker or non-smoker, diabetic or non-diabetic, and was assigned values for systolic blood pressure and total and HDL cholesterol.
The risk of a first CVD event was taken as the combined risks of the person experiencing heart disease death, a non-fatal heart attack or stroke. These risks had been estimated using data from the Framingham Heart Study, a large cohort study in which these three outcomes had been individually specified. These estimates were then used to identify people in the hypothetical population who would have a CVD event over the 10-year period that was modelled.
The researchers looked at the diagnostic accuracy and usefulness of two strategies:
- Screening using age alone (age screening) followed by preventive treatment to reduce risk from age 55 years onwards.
- Screening using multiple risk factors and age (Framingham screening) followed by preventive treatment to reduce risk.
Based on this, the researchers could estimate the accuracy and usefulness of the screening strategies according to four measures:
- detection rate (sensitivity)
- false-positive rate
- the proportion of CVD-free years of life lost in affected individuals with positive results (person-years detection rate)
- the cost per CVD-free life year gained from preventive treatment. (the cost of extending one person’s life by one year – free of heart disease or stroke – by use of the medication)
The results varied depending on which thresholds of age or risk were used. As with most tests, along with a lowering of the thresholds the detection rate improved (more people were detected), however the false-positive rate also rose (more people were picked up as positive who did not in fact go on to have an event). The researchers used graphs to show how well their strategies performed against each other in terms of an ideal threshold for separating those who would go on to develop a vascular event from those who would not.
The researchers also validated the methods they had used by testing the rates of CVD events in their model against those actually observed in UK registry data.
What were the basic results?
Both strategies had a maximum 84% detection rate, meaning that among those people who went on to develop heart disease or stroke over 10 years, 84% were correctly identified.
The two methods also had similar false-positive rates – the proportion of people who would not have gone on to have a CVD event, but who were falsely identified as at risk by screening. Using age alone (up to 55 years of age) 24% of those identified as being at risk would not go on to develop heart problems. In comparison, the existing vascular risk assessment using Framingham screening with assessments every five years using the widely adopted 20% 10-year CVD risk cut-off, would identify 21% of people as false-positives.
The researchers say that offering everyone preventive treatment at the age of 55 would also be more cost effective. The estimated cost for each year of life free of heart disease or stroke gained was £2,000 for age screening and £2,200 for Framingham screening. These results were calculated assuming that a Framingham screen costs £150 and the annual cost of preventive treatment is £200.
Age screening using a cut-off of 55 years detected 86% of all first CVD events arising in the population every year for a 24% false-positive rate. In comparison, five-yearly Framingham screening produced a false-positive rate of 21% for the same 86% detection rate.
How did the researchers interpret the results?
The researchers say that vascular disease is common and serious. They call for a proactive cost-effective public health policy to reduce the rates of the disease and say that it should be designed to prevent most events and make access to preventive treatment simpler without making people become patients.
They conclude that age screening for future heart disease or stroke is simpler than Framingham screening as it avoids the blood tests and medical examinations that the Framingham assessment requires.
The researchers also say that the age cut-off of 55 could be set lower for people with diabetes as they have an especially high vascular risk and will already be aware of this.
This is a well-argued case for an “age-alone” screening strategy supported by a well-conducted modelling study.
Most of the concerns reported by the media concern the implications of using preventive treatments based on age rather than the accuracy or otherwise of the vascular assessment. For example, the thought of taking medication for life beyond a certain age is seen as over medicalisation by some, while others put more emphasis on the adverse effects of medication. Although these concerns are valid, they are not directly addressed by this research.
There are several other points raised by this study and that need to be taken into account when interpreting its findings:
- It would have been possible to model the cost of side effects or monitoring for these in people taking statins or blood pressure drugs, however these are usually minor and would have affected both arms of this study similarly. Minor symptoms such as muscle aches are common, but serious complications appear to be rare. Adverse effects and the number of people who subsequently stop treatment could be investigated further in other studies.
- Some experts have also expressed concerns that taking the pills might cause people to ignore other aspects of their health such as diet and exercise.
- This is a modelling or simulation study and so does not include real people. It is probable that trials of this strategy would be needed in a real population before a change in policy could be advocated.
- The Framingham risk score is the oldest and most studied prediction tool for cardiovascular disease, but there are others that may be more accurate.
Overall this study has been well conducted and will add to the debate about the best policies for preventing vascular disease. The age screening policy advocated here would result in a very large number of people receiving treatment (everyone over 55) and so even a 1% improvement in prognostic performance might be worthwhile. However, the BBC reports that the Department of Health and the British Heart Foundation (BHF) had suggested these findings be treated with some caution.
Natasha Stewart, senior cardiac nurse at the BHF, said: “There’s not enough evidence that everyone of a certain age should be offered treatment, such as statins, without taking other risk factors into account. Also, it's essential that we continue full risk assessments for younger people who may be at significant risk of getting heart and circulatory disease.”
Analysis by Bazian
Edited by NHS Website
Links to the headlines
BBC News, 5 May 2011
Daily Mail, 5 May 2011
The Daily Telegraph, 5 May 2011
Links to the science
PLoS ONE 6(5): e18742
Cochrane Database of Systematic Reviews 2011, Issue 1