The following text from the article by Macleod et al (Does consideration of either psychological or material disadvantage improve coronary risk prediction? Prospective observational study of Scottish men. J Epidemiol Community Health 2007;61:833–837) was omitted form the printed article and should have appeared at the end of the results section as the last paragraph before the discussion:

Prediction of coronary heart disease mortality requires a combination of risk factors which can discriminate those individuals at future risk, as investigated above, and calibration to ensure accuracy in the magnitude of risk predicted for those individuals and the cohort as a whole. The standard Framingham risk equation under‐predicts the risk in each occupational class in the current cohort, and consequently has high specificity but very low sensitivity (Table 3). Recalibration improves matters, but sensitivity is still unacceptably low. Adding lifetime socio‐economic position to the recalibrated equation has little additional influence.