Advice on using the cardiovascular risk charts and calculator

This cardiovascular risk calculator can be used to estimate the risk of a cardiovascular event occuring over a given time period, with corresponding graphical displays including charts corresponding to those published by the Joint British Societies (JBS) and found in the British National Formulary (BNF). The site will also produce "smiley face" charts which some individuals may find useful, and bar graphs allowing comparison with patients with modified risk factors.

Contents


Different risk equations

This calculator can compute risk values using the Framingham equation, the formula used by the Joint British Societies (JBS) which is derived from Framingham, and the Scottish ASSIGN score.

Current UK guidelines recommend risk stratification based on the probability of cardiovascular disease (as opposed to coronary heart disease or other outcomes). NICE currently recommend use of the JBS formula. ASSIGN is tailored to the Scottish population.

Framingham

Until recently, the majority of people probably used the Framingham equation for calculating cardiovascular risk. The Framingham equation has the advantage of allowing calculations over various time periods (4 to 12 years) and for different outcomes: cardiovascular disease, stroke, coronary disease, myocardial infarction and death from either coronary or cardiovascular disease.

Joint British Societies (JBS)

The official cardiovascular risk charts published in the British National Formulary (BNF) are based on those given in the JBS guidelines. The BNF calculation option provided on this site is designed to correspond as closely as possible to the official published charts. As such, the LVH and diabetes risk factors originally incorporated in Framingham, are not included in the JBS charts as these patients are automatically considered high risk. JBS/BNF calculates cardiovascular disease risk, based on the sum of the coronary disease and stroke risks given by Framingham.

ASSIGN

The ASSIGN score was developed in conjunction with the Scottish Intercollegiate Guidelines Network, to address the issue that the Framingham equation may not accurately calculate risk in other populations. ASSIGN is therefore tailored to the Scottish population. It, too, calculates risk of cardiovascular disease, and also includes additional risk factors not used by Framingham (family history and social deprivation). The original ASSIGN score uses SIMD deprivation scores from 2004, but this website uses SIMD scores from 2006 in line with the official ASSIGN website.

top


Outcomes

The following outcomes can be calculated by this site:

  • CVD (BNF) - risk of developing cardiovascular disease, based on the calculation used by JBS/BNF. This option may be used to produce charts similar to those produced by the JBS.
  • CVD (ASSIGN) - risk of developing cardiovascular disease, based on the ASSIGN calculation.
  • CVD (Framingham) - risk of developing cardiovascular disease, based on the original Framingham equation.
  • CHD - risk of developing coronary heart disease (MI, CHD death, angina, coronary insufficiency), based on Framingham.
  • MI - risk of suffering a myocardial infarction (a "heart attack"), based on Framingham.
  • Stroke - risk of suffering a stroke (including transient ischaemic events (TIAs), based on Framingham.
  • CVD death - risk of dying from cardiovascular disease, based on Framingham.
  • CHD death - risk of dying from coronary heart disease, based on Framingham.

The definitions of "cardiovascular disease" vary depending on the equation used, and are detailed below.

top


Differences between equations

There are various differences between the different calculations, including the definition of cardiovascular disease used, and the risk factors included in the calculation. These differences are summarised below.

Definition of cardiovascular disease

  • Original Framingham definition - CHD (MI, CHD death, angina, coronary insufficiency) plus stroke, congestive heart failure and peripheral vascular disease
  • JBS definition - sum of the CHD (as defined above) and stroke risks. This definition is the one used in the printed charts.
  • ASSIGN definition - any cardiovascular death, CHD (ICD-9 410414, ICD-10 I20-I25) including angioplasty and bypass grafting, cerebrovascular disease.

Time period to calculate risk over

This can be varied between 4 and 12 years for any of the Framingham calculations, but is fixed at 10 years for the BNF or ASSIGN scores.

Age range

The printed BNF charts are restricted to 3 age groups (<50 years, 50 to 59 years, 60+ years). This site allows any age in the range 30 to 75 years. Strictly speaking, the age range for Framingham/BNF is 35 to 75, and for ASSIGN, 30 to 74 years, but these limits have not been enforced in order to facilitate comparisons between equations.

Smoking

The Framingham and BNF calculations treat smoking as a binary variable (yes/no), whereas ASSIGN uses number of cigarettes per day (this defaults to 10 in smokers).

Diabetes

The ASSIGN and Framingham calculations allow incorpation of diabetes as a risk factor. The BNF calculation assumes that the person does not have diabetes.

Family history

ASSIGN includes premature family history of cardiovascular disease as a risk factor, whereas the Framingham and BNF calculations do not. The Joint British Societies advise that in individuals with a positive family history, the risk can corrected be multiplying by 1.5. This feature is not included on this website.

Left ventricular hypertrophy (LVH)

The Framingham calculations include LVH as a risk factor. LVH refers to the original Framingham ECG criteria. It is important to distinguish this from echocardiographic findings of LVH, which are more common, or other ECG criteria such as the Sokolow-Lyon criteria or those employed by automated ECG reporting. ASSIGN does not incorporate LVH as a risk factor (it was only found to be significant in the ASSIGN model for women). The BNF calculation assumes LVH is absent.

Scottish Index of Multiple Deprivation (SIMD)

ASSIGN incorporates SIMD as a risk factor. SIMD can range between 0.54 (least deprived) to 87.6 (most deprived), and is based on the 2006 score (which differs only slightly from the original score in 2004). SIMD is assumed to be 20 if left blank and no postcode is given. Non-Scottish postcodes will generate an error.

HDL cholesterol

ASSIGN incorporates HDL as a separate risk factor to total cholesterol. In contrast, Framingham and BNF use the total:HDL cholesterol ratio. The charts for ASSIGN therefore use total cholesterol on the x-axis, whereas Framingham/BNF use the total:HDL cholesterol ratio. If HDL is left blank, a value of 1.4 is assumed.

top


Different graphical styles

The site will produce 4 different graphical displays, selectable from the menu at the bottom left labelled "chart style".

BNF charts

These charts are designed to replicate the original printed charts in the BNF. The original printed charts were used for estimation of coronary heart disease (CHD). These have now been superceded by charts for estimation of cardiovascular disease (CVD). The traffic-light colour coding used by this website corresponds to the same system used by the printed charts. There are differences in the green/orange/red cut-offs used for CHD and CVD.

An alternative colour coding system has been employed on the website for the other Framingham outcomes (e.g. MI, death). It should be noted that the risk ranges portrayed by these colours have been chosen simply for clarity and ease-of-use, and are not based on scientific evidence or published guidelines.

Where blood pressure and/or cholesterol lie outwith the ranges on the chart axes, the blue circular marker will not be visible.

Smiley faces

This display is a simple way of expressing percentage risk, using a representation of 100 individual patients with the same characteristics. Green (happy) faces represent those patients who will not have a cardiovascular event in the given time period. Red (sad) faces represent individuals who will suffer an event. It is important to remember that there is no way of determining whether a given patient will fall into the green or red category.

Comparison bars

The bar graph allows you to compare the risk score based on the current risk factors with an otherwise identical individual who:

  • has lower blood pressure
  • takes a statin tablet
  • is a non-smoker
  • has lower blood pressure, takes a statin and does not smoker

The reduction with statin therapy is based on the average reductions within the Cholesterol Treatment Trialists Collaboration meta-analysis.

It is important to emphasise that "lower blood pressure" risks are not the same as those achieved by taking antihypertensives. For instance, a person treated with antihypertensives will have a higher risk for a given blood pressure than an individual with the same blood pressure who is not on treatment. The comparison with a lower blood pressure is made with 140mmHg (the standard threshold used to define hypertension) or a 10mmHg reduction (whichever is lower).

Similarly, the "non-smoker" risks are not the same as those achieved by simply stopping smoking. A non-smoker will be at lower risk than an ex-smoker (especially in the early years following smoking cessation). The "non-smoker" bar is only visible if the current patient is a smoker.

Thermometer

The thermometer provides a similar means of comparing risk values to that of the comparison bar graph described above, with similar caveats. The colour coding corresponds to that used in the BNF-style charts. "Non-smoker" will only appear if the patient is a smoker. The effects of a lower BP and statin treatment are only shown if the current patient risk is out-with the lowest (green) risk area.

top


Important notes on using the charts

These charts are designed as an aid to making clinical decisions, with respect to the use of lifestyle and drug interventions for modifying risk. They should NOT replace clinical judgement. For a detailed overview, users are advised to read the relevant section of the BNF. Users of the ASSIGN score are also referred to the SIGN-97 guidelines.

  • Use of charts is inappropriate in patients with existing disease (e.g. renal dysfunction, CHD, etc.)
  • Treatment of persistently/marked elevated BP or TC:HDL ratio is generally indicated regardless of estimated risk
  • Smoking status should reflect lifetime tobacco exposure, rather than current use
  • LVH refers to the original Framingham ECG criteria. It is important to distinguish this from echocardiographic findings of LVH, which are more common, or other ECG criteria such as the Sokolow-Lyon criteria or those employed by automated ECG reporting systems
  • Decision to treat with drugs should be based on repeated assessment of risk factors rather than a single measurement
  • Risk estimates are based on untreated levels of BP and cholesterol, and can only be used as a guide in persons already receiving treatment
  • Risk will be underestimated in the following groups
    • Elevated triglycerides (>1.7mmol/L)
    • Premature menopause
    • Impaired glycaemic function, despite no overt diabetes
  • For Framingham/BNF calculations, a family history of premature CVD will also result in underestimation of risk
  • Risk estimates have not been validated in ethnic minorities

top


Printing charts

If you try to print the chart, you may find that, although the text is printed, the graph itself is not. This is because the charts are drawn onto the background of the browser window, and unfortunately most browsers will not print background images.

To get around this problem, a "printer friendly" page is available which should print normally, although is currently restricted to the BNF-style charts only.

top


Notes and disclaimer

  • v3.0, 23/2/10: addition of "smiley face graphs", minor error checking corrections
  • v2.0, 24/9/08: incorporation of the different calculators into one single page, addition of the ASSIGN calculator
  • v1.2, 31/3/08: modified BNF calculation for CVD risk, correction of bug affecting accuracy of graphs, addition of marker showing BP and cholesterol on chart, automatic calculation updating, printer-friendly page option.

Written by Dr Rupert Payne using JavaScript. With thanks to Walter Zorn for his excellent JS graphics library. Tested using Firefox 3.6 and Internet Explorer 7.0.

This calculator has been published with the hope that it will be useful, but without warranty, implied or explicit, of fitness for a particular purpose. The author will not accept responsibility for any injury, damages or other outcomes arising as a result, direct or indirect, of using this calculator.

top