Mohsen Aarabi1, Peter R Jackson. 1. Unit of Clinical Pharmacology, Royal Hallamshire Hospital, Division of Clinical Sciences, University of Sheffield, Sheffield S10 2JF, UK.
Abstract
BACKGROUND: Although people of South Asian descent appear to be at increased risk of cardiovascular disease (CVD) compared to the general population there is no validated tool for predicting risk in this group. DESIGN: This study was based on an analysis of data from the Health Survey for England (HSE) 1998 and 1999 to identify a simple method for adjusting the Framingham equation to estimate coronary risk in South Asians which could be then applied to existing simple paper-based tools. METHODS: The different adjustments investigated were fixed increments to the age, total cholesterol (TC), and TC:high density lipoprotein (HDL) cholesterol ratio and multipliers for the TC:HDL cholesterol ratio. Framingham risk modified according to a factor derived from an earlier overview of prospective studies was used as a standard to estimate sensitivity and specificity of adjustment methods. The receiver-operating characteristic (ROC) plot was used to compare the different adjustments in the primary role of identifying individuals above or below a given risk threshold. RESULTS: All adjustment methods produced a graded monotonic increase of coronary risk in 4497 eligible subjects from the HSE 1998 dataset. Multiplying TC:HDL cholesterol ratio gave the largest area under the ROC curve. However adding 10 years to the age of South Asian people was the simplest way of calculating coronary heart disease risk using paper-based methods and still provided acceptable accuracy. CONCLUSION: Our result should be used to achieve the systematic evaluation of each individual for primary prevention of CHD in South Asians in primary care.
BACKGROUND: Although people of South Asian descent appear to be at increased risk of cardiovascular disease (CVD) compared to the general population there is no validated tool for predicting risk in this group. DESIGN: This study was based on an analysis of data from the Health Survey for England (HSE) 1998 and 1999 to identify a simple method for adjusting the Framingham equation to estimate coronary risk in South Asians which could be then applied to existing simple paper-based tools. METHODS: The different adjustments investigated were fixed increments to the age, total cholesterol (TC), and TC:high density lipoprotein (HDL) cholesterol ratio and multipliers for the TC:HDL cholesterol ratio. Framingham risk modified according to a factor derived from an earlier overview of prospective studies was used as a standard to estimate sensitivity and specificity of adjustment methods. The receiver-operating characteristic (ROC) plot was used to compare the different adjustments in the primary role of identifying individuals above or below a given risk threshold. RESULTS: All adjustment methods produced a graded monotonic increase of coronary risk in 4497 eligible subjects from the HSE 1998 dataset. Multiplying TC:HDL cholesterol ratio gave the largest area under the ROC curve. However adding 10 years to the age of South Asian people was the simplest way of calculating coronary heart disease risk using paper-based methods and still provided acceptable accuracy. CONCLUSION: Our result should be used to achieve the systematic evaluation of each individual for primary prevention of CHD in South Asians in primary care.
Authors: Latha P Palaniappan; Maria Rosario G Araneta; Themistocles L Assimes; Elizabeth L Barrett-Connor; Mercedes R Carnethon; Michael H Criqui; Gordon L Fung; K M Venkat Narayan; Hamang Patel; Ruth E Taylor-Piliae; Peter W F Wilson; Nathan D Wong Journal: Circulation Date: 2010-08-23 Impact factor: 29.690
Authors: S S Liem; P V Oemrawsingh; S C Cannegieter; S Le Cessie; J Schreur; F R Rosendaal; M J Schalij Journal: Neth Heart J Date: 2009-04 Impact factor: 2.380
Authors: Charlotte A Jones; Shefina Mawani; Kathryn M King; Selina Omar Allu; Megan Smith; Sailesh Mohan; Norman R C Campbell Journal: BMC Public Health Date: 2011-01-11 Impact factor: 3.295