Stroke Risk Assessment

Tools for assessment of stroke risk

Determining the risk of first a first stroke is useful for optimal patient management.10

Guidelines recommend using validated stroke risk-assessment tools for individuals who have not had a primary stroke event or transient ischaemic attack (TIA).9 Different stroke risk calculators are summarized in the table below.  

ABCD risk assessments for TIA

The ABCD or ABCD2 score is a simple scoring system to identify individuals at a high early risk of a stroke after a TIA. The “2” represents the presence of diabetes. The assessment predicts a 7-day risk of stroke based on:43-45

  • Age (1 point for patients 60 years of age or older),
  • Blood pressure (1 point for a blood pressure greater than 140/90 mm Hg),
  • Clinical features (2 points for unilateral weakness with or without speech impairment or 1 point for speech impairment without weakness)
  • Duration (1 point for 10 to 59 minutes, 2 points for greater than 59 minutes).

Both ABCD and ABCD2 scores are highly predictive of stroke risk among TIA clinic outpatients. Scores of 3 or greater may be clinically useful in identifying TIA outpatients at raised risk of stroke in the medium to long term.45

Benefits of stroke risk assessment

  • Identify patients who are unaware of their elevated risk
  • Assess the total risk of multiple factors
  • Discern the utility of modifications and treatments
  • Project a quantitative risk to select appropriate treatments or stratification in clinical trials
  • Guide appropriate use of diagnostic tests

Limitations of stroke risk assessment

Although most stroke risk calculators were designed for cardiovascular disease, the AHA/ASA guidelines suggest that cardiovascular disease risk calculators can also be used to estimate stroke risk.46 Assessment of stroke risk using these risk assessment calculators is not perfect because they can over- or under-estimate stroke risk. Therefore, the values should be interpreted with caution.

Stroke risk assessment is not indicated in the following patients as they are considered to be at high risk, requiring lifestyle changes and medical interventions:

  • Patients age 85 years and older
  • Patients with pre-existing atherosclerotic cardiovascular disease (ASCVD)
  • Patients with familial hypercholesterolemia or inherited disorders of lipid metabolism
  • Patients with type 1 diabetes mellitus
  • Patients with chronic kidney disease stage 3, 4, or 5

Commonly Used Stroke Risk Calculators

Framingham Risk Score (FRS)45Includes low-density lipoprotein cholesterol (LDL-C) with high-sensitivity CRP (hs-CRP) and family history to select patients for statin therapy
  • Validated predictor of CVD and thus of stroke
  • Several elderlies become candidates for beneficial statin therapy
  • Requires repeated testing which can lower compliance and accuracy of risk assessment
The United Kingdom Prospective Diabetes Study—UKPDS56, UKPDS6045Risk of new coronary heart disease in type II diabetic patients
  • Model includes systolic BP, lipid levels, glycaemic control, and factors like age, gender, ethnicity, smoking status, and duration of diabetes
  • Tool is specific for patients with type 2 diabetes mellitus
Reynolds Risk Score (RRS)46Includes gender, age, systolic BP, smoking, hs-CRP, total cholesterol, HDL-C, family history of premature MI, and haemoglobin A1C (if diabetic)
  • Validated, reliable, easy to use tool to estimate risk separately for men and women
  • Initially devised for patients without diabetes, so potentially less reliable in patients with diabetes
Pooled Cohort Risk Score (PCRS or ASCVD)47Based on pooled cohort risk equations, is valid and reliable, and provides 10-year risk for hard ASCVD incidents
  • Provides 10-year estimate of patient’s risk for ASCVD including baseline assessment at 1st visit and impact of future interventions
  • Useful for prediction of secondary stroke
  • Shows change in risk factors over time and thus needs initial and follow-up values (not useful to show immediate risk)
Systematic Coronary Risk Evaluation (SCORE)48Database covers 12 large cohort studies
  • Calculations based on extremely large data sets from the European population and estimate the risk of death from CVD or stroke
  • Used extensively in all ethnic groups for calibrating each country’s mortality statistics low- and high-risk charts to be used for different geographies (validity of chart is limited by reliable mortality statistics for the region/country)
QRISK-345Large cohort data found useful for primary incidence of CVD and stroke and adjusted for several conditions while taking into consideration multiple parameters which make calculations more reliable
  • Recommended risk assessment tool for people of ages 30 to 84 years
  • Updated every year for variations in population trends and changes in national guidelines
  • More appropriate for the UK population as it is validated only in this population
World Health Organization (WHO)45Risk prediction is based on age, gender, total cholesterol, systolic BP, and diabetes type 2 status
  • Useful in primary health care setting with limited resources
  • Designed for low- to middle-income countries where risk prediction charts are unavailable
  • Easy to use with color-coded zones
  • Uses fewer variables for practical use in low-resource settings
Prospective Cardiovascular Munster Study (PROCAM)45Predicts 10-year CVD risk based on history of MI, high cholesterol, diabetes, triglycerides, systolic BP, smoking, and age
  • Conducted on cohort of German population
  • Tends to overestimate risk for western European population such as the UK
MyRisk Stroke Calculator49Population-based longitudinal study on stroke incidence
  • Provides 10-year estimate of stroke risk based on age, gender, education, renal disease, diabetes, CHF, peripheral artery disease, IHD, smoking, alcoholism, physical activity, anger, depression, and anxiety using point-based calculations
  • Is a simple method for use for population information dissemination
AtheroEdge Composite Risk Score (AECRS1.010yr)5010-year risk calculator incorporating carotid artery image phenotypes with conventional cardiovascular risk factors
  • Shows great potential as a comprehensive predictor with higher accuracy
  • Comparison with current predictors and validity testing further studies are required