23 June 2008 | Julia Hippisley-Cox, Yana Vinogradova, Carol Coupland, John Robson, Rubin Minhas, Aziz Sheikh, Peter Brindle
The QRISK2 algorithm was developed and validated to estimate cardiovascular disease risk in England and Wales, incorporating ethnicity, deprivation, and other clinical factors. It outperformed the modified Framingham score in discrimination and calibration. QRISK2 explained 43% of variation in women and 38% in men, compared to 39% and 35% for the Framingham score. It reclassified 41.1% of high-risk patients (≥20% 10-year risk) to low risk, with observed risks below 20%, while 15.3% of QRISK2 high-risk patients were reclassified to low risk by Framingham, with observed risks above 20%. QRISK2 identified higher risk in populations with higher cardiovascular event rates, particularly among south Asians. The algorithm was validated using data from 531 general practices, with 2.3 million patients aged 35-74. QRISK2 incorporated factors like self-assigned ethnicity, age, sex, smoking, blood pressure, cholesterol, BMI, family history, and deprivation. It showed better performance in predicting cardiovascular events, especially among south Asians. The algorithm was developed using longitudinal data from the QRESEARCH database, which includes electronic health records and linked death certificates. QRISK2 was validated in a similar population to the one from which it was derived, suggesting a "home advantage." Further validation in other populations is recommended. QRISK2 is a more accurate and equitable tool for cardiovascular disease prevention, particularly for ethnic minorities. It is designed to be integrated into primary care systems for routine use, with potential for future updates as data quality improves. The algorithm is not a replacement for clinical judgment but aids in risk assessment. QRISK2 has been implemented in primary care, with plans to integrate it into electronic health records and computerized decision support tools.The QRISK2 algorithm was developed and validated to estimate cardiovascular disease risk in England and Wales, incorporating ethnicity, deprivation, and other clinical factors. It outperformed the modified Framingham score in discrimination and calibration. QRISK2 explained 43% of variation in women and 38% in men, compared to 39% and 35% for the Framingham score. It reclassified 41.1% of high-risk patients (≥20% 10-year risk) to low risk, with observed risks below 20%, while 15.3% of QRISK2 high-risk patients were reclassified to low risk by Framingham, with observed risks above 20%. QRISK2 identified higher risk in populations with higher cardiovascular event rates, particularly among south Asians. The algorithm was validated using data from 531 general practices, with 2.3 million patients aged 35-74. QRISK2 incorporated factors like self-assigned ethnicity, age, sex, smoking, blood pressure, cholesterol, BMI, family history, and deprivation. It showed better performance in predicting cardiovascular events, especially among south Asians. The algorithm was developed using longitudinal data from the QRESEARCH database, which includes electronic health records and linked death certificates. QRISK2 was validated in a similar population to the one from which it was derived, suggesting a "home advantage." Further validation in other populations is recommended. QRISK2 is a more accurate and equitable tool for cardiovascular disease prevention, particularly for ethnic minorities. It is designed to be integrated into primary care systems for routine use, with potential for future updates as data quality improves. The algorithm is not a replacement for clinical judgment but aids in risk assessment. QRISK2 has been implemented in primary care, with plans to integrate it into electronic health records and computerized decision support tools.