Estimating 5-year absolute risk of cardiovascular disease using routinely collected electronic medical records from Australian general practices
Authors' conclusion
CVD risk equations based on routine Australian primary care data performed strongly and generalised well across diverse settings. These models offer potential for automated integration into general practice software to support real-time CVD risk assessment.
Study context
- Population
- Individuals aged 30–74 years on 1 January 2017 with no prior cardiovascular disease (CVD) history and at least one record for an anthropometric measurement or pathology test, from primary care EMR data of the New South Wales Health Lumos programmepatients, 680 general practices
- Participants
- 850,216 individualsEvents: 33,578
- Predictors
- Demographics, smoking, chronic conditions, clinical variables and medications
- Predicted outcome
- fatal or non-fatal CVD event
Methodological details5×2 cross-validation, linked with hospital and death records, Geographic (internal-external) validation across 10 Primary Health Networks
- 5×2 cross-validation
- Linked with hospital and death records
- Geographic (internal-external) validation across 10 Primary Health Networks
Model performance
As reported in the abstract
5-year risk
| Test or model | Harrell’s C |
|---|---|
| Full models with 28 predictors, females | 0.803 (95% CI 0.801 to 0.804) |
| Full models with 28 predictors, males | 0.772 (95% CI 0.770 to 0.773) |
Harrell’s C
Full models with 28 predictors, females0.803 (95% CI 0.801 to 0.804)
Full models with 28 predictors, males0.772 (95% CI 0.770 to 0.773)
Further findings
| Outcome | Males | Females |
|---|---|---|
| Geographic (internal-external) validation across 10 Primary Health Networks | ||
| C-index range | 0.747-0.788 | 0.772-0.827 |
Least absolute shrinkage and selection operator models with 12-15 predictors
- Performed similarly
Calibration
- Well calibrated across age, socioeconomic and smoking strata
Percentile-based net reclassification improvement vs model based on AusCVDRisk variables
- event-Net Reclassification Improvement up to
- 0.185
Original publication
Estimating 5-year absolute risk of cardiovascular disease using routinely collected electronic medical records from Australian general practices. Heart. Published 17 November 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →