Frequently Asked Questions

What does this Heart Disease Risk Estimate Using Artificial Intelligence Software do?

This heart attack and cardio vascular disease risk screening AI software provides you with an interactive absolute risk estimate (%) of a normal/healthy person who never had a heart attack or a heart diseas. This AI software calculates the risk of having a heart disease/cardiovascular event over a specific period of time and an idea of the potential benefit of treatment.

This calculator was designed to dynamically show how the value of specific risk factors impact cardiovascular risk. Results are shown using a representation of 100 "happy faces" to help illustrate risk. This format seems to be one of the better ways to present risks visually. Risks are also provided as absolute numbers and rounded off to one decimal point even though any cardiovascular estimates likely are +/- 5-10% at best.

There are TWO different views for this calculator. An ENHANCED VIEW which displays all the different types of interventions including non-drug, drugs for cholesterol, drugs for statins, drugs for blood pressure, drugs for glucose. The BASIC VIEW displays only non-drug, statins and ASA as interventions and all patient variables need to be inserted before any numbers are presented. In addition, it only displays three different faces - those with no event, those with an event, and those for benefit. You can toggle back and forth between the two different views.

How do I use this Heart Disease Risk Estimate Using Artificial Intelligence Software?

1) Enter the unmodifiable factors

These are the factors that really can't be changed - age, gender, ethnicity/race.

2) Enter the modifiable factors

These are factors that can mostly be modified with lifestyle changes or medications - smoking, blood pressure, cholesterol, diabetes.

a) smoking status - yes or no

b) diabetes - yes or no

c) systolic blood pressure in mmHg - the calculator uses 120mm/Hg to calculate baseline risk

d) total cholesterol in mmol/L - the calculator uses 3 mmol/L to calculate baseline risk

e) HDL cholesterol in mmol/L - the calculator uses 1.3 mmol/L to calculate baseline risk

HDL cholesterol in mmol/L - the calculator uses 1.3 mmol/L to calculate baseline risk

g) family history of CVD, chronic kidney disease, atrial fibrillation, rheumatoid arthritis - yes or no

h) height/weight - BMI is calculated automatically and if height and weight are not entered it imputes a value of 25kg/m2the calculator

3) Estimate benefit

Estimate of benefit - clicking on an intervention like Physical Activity, Mediterranean Diet vs Low Fat, Vitamin/Omega-3 supplements, (if SBP is >/= 140 mmHg) BP Meds, (regardless of cholesterol) Low-mod intensity statins, High intensity statins, Fibrates, Niacin, Ezetimibe, (if diabetic) Metformin, Sulfonylureas, Insulins, Glitazones, GLPs, DPP-4s, Meglitinides, SGLT2 and ASA will input a specific relative benefit and this will be applied to the absolute risk to calculate the % of people who could benefit from a therapy. Examples of relative benefits are taken from a synopsis of the best available evidence - see below for more details

ABSOLUTE NUMBERS - calculated risk and benefits

a) NO EVENT BLUE FACES - are the % of people who would NOT have an event over the time period

b) TOTAL WITH AN EVENT RED/PINK FACES - are the % of people who will have an event over the time period

d) NUMBER WHO BENEFIT  GREEN FACES - are the % of people who would not have an event because of "treatment" and is based on reducing the absolute risk of an event by the relative estimate of benefit

d) NNT is the number needed to treat and is infinity unless a treatment that has a benefit is added

b) BASELINE EVENTS RED FACES - are the % of people who would have an event if they had "baseline" risk factors (non-smoker, non-diabetic, SBP = 120, Total cholesterol = 3, HDL = 1.3)

c) ADDITIONAL EVENTS PINK FACES - are the % of people who would have an event OVER AND ABOVE the baseline rate and is that risk attributable to the persons specific risk factors

Where do the estimates of benefit come from?

The table below includes a reasonable estimate of the relative benefits of certain "therapies" and comes from a variety of meta-analyses and studies. They have been rounded off and some of the numbers have far more solid evidence behind them than others. Not all evidence shows the same numbers so there has been some neccesary personal judgement involved as well. One can apply these numbers to the individual patient's absolute risk by clicking on the specific treatment. Risk reductions for diabetes drugs only apply if the patient has been identified as a diabetic. For blood pressure, while somewhat arbitrary, risk reductions only apply if the patient has a blood pressure of 140mmHg or greater. A recent Cochrane review (CD006742) suggested that "Antihypertensive drugs used in the treatment of adults (primary prevention) with mild hypertension (systolic BP 140-159mmHg and/or diastolic BP 90-99mmHg) have not been shown to reduce mortality or morbidity in RCTs." In addition, the ACCORD study (type 2 diabetics) showed no benefit of lowering blood pressure below 140mmHg. However, a recent meta-analysis suggest that a benefit is achieved regardless of baseline systolic blood pressure. Please check out the scientfic study here :BMJ 2009;338:b1665doi:10.1136/bmj.b1665.

Intervention Relative CVD Benefits Evidence
Physical Activity ~25% Eur J Cardiovasc Prev Rehabil 2008;15:239-46, Int J Environ Res Public Health 2012;9:391-407
Mediterranean Diet vs Low Fat ~30% N Engl J Med 2013 DOI:10.1056/NEJMoa1200303
Vitamins/Omega 3 supplements ~0% Lancet 1996;347:781-6, JAMA 2007;297:842-57, ACP Journal Club 2007;147(1):4
Blood pressure (only if SBP 140 or higher) ~30% (~50% if diabetic) J Htn 1993;11(suppl4):S61-73, Cochrane, Lancet 2004;364:1684–9, Ann Intern Med 2003;138:587-92, CD004349
Low-Moderate Intensity Statins ~25% Lancet 2008;371:117-25; Women-J Am Coll Cardiol 2012;59:572–82; Over age 65-J Am Coll Cardiol 2013 Aug 14 pii:S0735-1097(13)03880-1
High-Intensity Statins ~35%  
Fibrates ~0% NEJM 2010;362:1563-74, Amer J Med 2009;122:962.e1-962.e8, Int J Cardiol 2009 doi:10.1016/j.ijcard.2008.11.211
Niacin No studies
Ezetimibe ~0% NEJM 2008;358:1431-43
Sulfonylureas ~0% Diabet Med 2013;30:1160–71
Insulin ~0% N Engl J Med 2012; 367:319-328
Metformin ~35% UKPDS 34
Glitazones ~0% CD006063,CD006060
Dipeptidyl peptidase-4 inhibitors (DPP-4) ~0% SAVOR-TIMI 53, EXAMINE
ASA ~15%M ~10%F Arch Intern Med 2012;172:209-16
Combinations of the above interventions   Nobody really knows but hopefully there is some sort of additive benefit. If you have data on this or a way this can be shown or calculated please let me know.

Why aren't surrogate marker or treatment thresholds provided?

You will notice there are no designated target or treatment thresholds from guidelines. These were purposely not added as they are arbitrary and one should use a reasonable estimate of risks and benefits to help facilitate shared-informed decision-making.

Diabetes and smoking aren't really yes or no risk factors?


Smoking history is typically based on self-report. It is unclear exactly what should constitute a smoker. What seems reasonable is that if a person has been a smoker over the last 5 years they should be considered a smoker. If they stopped smoking 5-10 years ago they should likely be considered a non-smoker for the purposes of cardiovascular risk calculation. However, it is important to remember that smoking also increases the risks for cancers and lung disease and these risks may not be as reversible when a person becomes a non-smoker.


Many cardiovascular risk calculators identify diabetes as a yes or no variable. Framingham defines diabetes as a fasting glucose of > 7mmol/L (126 mg/dL) for the offspring cohort or 7.8 (140) for the original cohort. It is clear however that diabetes is not a yes or no risk factor. Based on the UKPDS risk engine for every 2% increase in A1c, over 6%CVD risk goes up 33%. Type-2 diabetes also increases the chance of microvascular disease. Compared to macrovascular risks the absolute risk of hard microvascular endpoints like renal failure or blindness are far less frequent. However incidence data is not as robust for these endpoints. In UKPDS 34 less than 1% of patients developed renal failure over 11 years and roughly 5% developed blindness in one eye. In STENO, in which patients also had microalbuminuria, 8% of patients in the standard treatment group went on dialysis and 9% became blind in at least 1 eye over 14 years. Vijan et al estimated a 60-year-old diabetic with an A1c of 9% has a lifetime chance of dialysis of around 1 in 100 and a similar lifetime chance of blindness.

What specific outcomes do the cardiovascular terms include and how are they defined?


A brief recurrent chest discomfort of up to 15 minutes duration, precipitated by exertion or emotion and relieved by rest or by nitroglycerine AND coronary insufficiency = prolonged (>15 minutes) chest pain but no indication of heart tissue damage.

Coronary heart disease (CHD)

CHD = fatal/non-fatal heart attacks + angina/coronary insufficiency.

Cardiovascular disease (CVD)

CVD = coronary heart disease (CHD) + heart failure + strokes + intermittent claudication.

Heart attack

Heart attacks = fatal or non-fatal myocardial infarctions.

Acute, recent, or old - an acute myocardial infarction is defined as the presence of at least two of three findings:

Heart failure

A minimum of two major or one major and two minor criteria to be present concurrently

Major Criteria:

Minor criteria:

Intermittent claudication

A cramping discomfort in the calf clearly provoked by walking some distance with the pain appearing sooner when walking quickly or uphill and being relieved within a few minutes by rest.


A sudden or rapid onset of a focal neurologic deficit persisting for greater than 24 hours AND Transient ischemic event (TIA) = a focal neurologic deficit of sudden or rapid onset that fully resolves in less than 24 hours.

Heart attacks and strokes.

Atherosclerotic cardiovascular disease = CHD death + nonfatal heart attacks + fatal/nonfatal strokes.

Scientific References for Heart Risk Equations used here