Validation of a new predictive risk model: measuring the impact of the major modifiable risks of death for patients and populations

• Published in: Population health metrics 2015-10, Vol.13 (1), p.27-27, Article 27
• Main Authors: Lim, Stephen S, Carnahan, Emily, Nelson, Eugene C, Gillespie, Catherine W, Mokdad, Ali H, Murray, Christopher J L, Fisher, Elliott S
• Format: Article
• Language: English
• Subjects: Health aspects; Measurement; Mediation; Medical research; Medicine, Experimental; Methods; Mortality; Patient education; Risk assessment
• ISSN: 1478-7954; 1478-7954
• DOI: 10.1186/s12963-015-0059-8

Modifiable risks account for a large fraction of disease and death, but clinicians and patients lack tools to identify high risk populations or compare the possible benefit of different interventions. We used data on the distribution of exposure to 12 major behavioral and biometric risk factors inthe US population, mortality rates by cause, and estimates of the proportional hazards of risk factor exposure from published systematic reviews to develop a risk prediction model that estimates an adult's 10 year mortality risk compared to a population with optimum risk factors. We compared predicted risk to observed mortality in 8,241 respondents in NHANES 1988-1994 and NHANES 1999-2004 with linked mortality data up to the end of 2006. Predicted risk showed good discrimination with an area under the receiver operating characteristic (ROC) curve of 0.84 (standard error 0.01) for women and 0.84 (SE 0.01) for men. Across deciles of predicted risk, mortality was accurately predicted in men ((Χ (2) statistic = 12.3 for men, p=0.196) but slightly overpredicted in the highest decile among women (Χ (2) statistic = 22.8, p=0.002). Mortality risk was highly concentrated; for example, among those age 30-44 years, 5.1 % (95 % CI 4.1 % - 6.0 %) of the male and 5.9 % (95 % CI 4.8 % - 6.9 %) of the female population accounted for 25 % of the risk of death. The risk model accurately predicted mortality in a representative sample of the US population and could be used to help inform patient and provider decision-making, identify high risk groups, and monitor the impact of efforts to improve population health.