Distribution of Cardiovascular Disease Risk Based on the Updated 2023 Guideline-Recommended Australian Cardiovascular Disease Risk Algorithm and Comparison With the 2012 Algorithm: An Observational Study.
Lazarevic N, Bhat M, et al. • The Medical journal of Australia • 2026
Using the 2023 Australian CVD risk algorithm, 9.7% of people aged 45–74 without existing CVD were at high risk, 26.4% at intermediate risk, and 63.9% at low risk, compared with 17.6%, 11.6%, and 70.8% respectively under the 2012 algorithm, with differences largely driven by changes to clinically determined high-risk criteria.
Key Findings
Results
Under the 2023 Australian CVD risk algorithm, 9.7% of participants were classified as high CVD risk, 26.4% as intermediate, and 63.9% as low risk.
High risk was defined as ≥10% 5-year risk or clinically determined high risk under the 2023 algorithm.
Intermediate risk was defined as 5% to <10% 5-year risk.
Low risk was defined as <5% 5-year risk.
Sample consisted of 115,873 people aged 45–74 years without existing CVD from the MedicineInsight primary care database.
Data were drawn from clinical encounters between September 2020 and August 2022.
Results
Under the 2012 Australian CVD risk algorithm, 17.6% of participants were classified as high CVD risk, 11.6% as intermediate, and 70.8% as low risk.
High risk under the 2012 algorithm was defined as >15% 5-year risk or clinically determined high risk.
Intermediate risk was defined as 10%–15% 5-year risk.
Low risk was defined as <10% 5-year risk.
The same study population of 115,873 participants was used for the 2012 algorithm comparison.
The proportion at high risk dropped from 17.6% (2012) to 9.7% (2023), while the intermediate-risk group grew from 11.6% to 26.4%.
Results
The difference in proportions classified at high risk between the two algorithms was largely driven by changes to clinically determined criteria for high risk.
The 2023 algorithm revised which clinical conditions automatically confer high-risk status independent of calculated score.
Changes to these clinically determined criteria, rather than the underlying risk score formula alone, accounted for most of the shift in high-risk proportions.
The authors note this 'likely reflects more accurate updated CVD risk estimation for the contemporary Australian population.'
Results
There was moderate-to-substantial agreement and concordance between the 2023 and 2012 CVD risk algorithms.
Linear-weighted kappa = 0.62, indicating moderate-to-substantial agreement in risk categorisation between the two algorithms.
Kendall's tau-b = 0.74, indicating substantial concordance between continuous risk estimates from the two algorithms.
Bland-Altman plots were used alongside Cohen's kappa to assess agreement and concordance.
Despite overall agreement, meaningful differences existed in the proportions assigned to each risk category.
Results
The proportion of people estimated at low risk and not routinely recommended pharmacotherapy was similar between the 2023 and 2012 guidelines and aligned with international standards.
63.9% were at low risk under 2023 guidelines versus 70.8% under 2012 guidelines.
The authors state that 'proportions estimated at low risk and not routinely recommended pharmacotherapy align with international standards and were similar between guidelines.'
Fewer people would be recommended pharmacotherapy due to high-risk classification under the 2023 versus 2012 guidelines.
Methods
The study used MedicineInsight, a longitudinal primary care database covering approximately 8% of Australian general practices, as its data source.
115,873 participants aged 45–74 years without existing CVD were included.
Participants had a clinical encounter recorded between September 2020 and August 2022.
The database is a longitudinal primary care dataset covering 8% of Australian general practices.
The 2023 algorithm's discretionary reclassification step based on additional risk-enhancing factors was not tested in this study.
Discussion
The 2023 guidelines include a discretionary reclassification step allowing clinicians to adjust risk based on additional factors, which was not evaluated in this study.
The authors describe this as an 'untested in our study' component of the 2023 guidelines.
The authors emphasise 'the use of this reclassification step by clinicians and consideration of the benefits of pharmacotherapy for those at intermediate risk.'
This step could potentially move some individuals from intermediate to high risk, increasing recommended pharmacotherapy use.
The study's findings on intermediate-risk proportions (26.4%) should be interpreted in the context of this untested discretionary adjustment.
What This Means
This Australian study compared two sets of guidelines — from 2012 and 2023 — used by doctors to estimate a person's risk of having a heart attack or stroke in the next five years. Using health records from nearly 116,000 people aged 45 to 74 who visited a GP between 2020 and 2022 and had no prior heart disease, the researchers applied both sets of guidelines to see how many people would fall into high, intermediate, or low risk categories under each approach.
The results showed meaningful differences between the two guidelines. Under the older 2012 guidelines, about 17.6% of people were classified as high risk (warranting preventive medication), while under the updated 2023 guidelines, only 9.7% were classified as high risk. At the same time, the intermediate-risk group grew substantially — from 11.6% under the 2012 guidelines to 26.4% under the 2023 guidelines. The main reason fewer people were classified as high risk under the new guidelines was changes to which medical conditions automatically place someone in the high-risk category, rather than changes to the underlying risk score formula. The two algorithms showed moderate-to-substantial overall agreement in how they ranked people.
This research suggests that the updated 2023 guidelines, by recalibrating risk thresholds to better reflect the current Australian population, may result in fewer people being automatically recommended preventive medications like statins or blood pressure drugs based on high-risk classification alone. However, the authors highlight that a large proportion of people — more than one in four — now fall into the intermediate-risk category where treatment should still be actively considered, especially for those with additional risk-enhancing factors. Doctors are encouraged to use an optional step in the 2023 guidelines that allows them to adjust a person's risk category based on these additional factors, which could help ensure people who would benefit from preventive treatment are not missed.
Lazarevic N, Bhat M, Joshy G, Butler D, Woodward M, Patel A, et al.. (2026). Distribution of Cardiovascular Disease Risk Based on the Updated 2023 Guideline-Recommended Australian Cardiovascular Disease Risk Algorithm and Comparison With the 2012 Algorithm: An Observational Study.. The Medical journal of Australia. https://doi.org/10.5694/mja2.70280