Science, Medicine, and the Anesthesiologist

Cardiovascular disease is a leading cause of death worldwide. Risk prediction tools facilitate interventions, but currently used prediction algorithms may underestimate risk in emerging high-risk populations. This study analyzed two established research databases containing anonymized health data from more than 16 million patients, aged 18 to 84 yr, receiving routine care through the United Kingdom’s National Health Service since 1989. The primary outcome was incidence of cardiovascular disease defined as fatal or nonfatal ischemic heart disease, cerebrovascular disease, or transient ischemic attack. The new algorithm (QR4) estimates the 10-yr risk using cause-specific proportional hazard models derived and validated using data from 9.98 and 6.79 million patients, respectively. The QR4, adjusted for age and body mass index, identified seven new predictors in men and women: chronic obstructive pulmonary disease (pooled hazard ratio 1.38 [95% CI, 1.33 to 1.42]), learning disability (1.25 [1.16 to 1.34]), Down syndrome (2.59 [2.10 to 3.07]), brain cancers (5.05 [3.16 to 6.93]), blood cancers (2.08 [1.82 to 2.34]), lung cancers (1.67 [1.43 to 1.90]), and oral cancers (1.51 [1.34 to 1.68]), in addition to two new predictors in women: pre-eclampsia (1.56 [1.36 to 1.78]) and postpartum depression (1.18 [1.11 to 1.26]).