| Popis: |
IntroductionThe previously developed LIFEtime-perspective CardioVascular Disease (LIFE-CVD) model can be used to predict lifetime cardiovascular disease risk, CVD-free life expectancy, and lifetime benefit from cardiovascular risk factor treatment in apparently healthy people aged 45 to 80 years. However, there was an unmet need to be able to apply the model in patients younger than 45 years, and to accurately estimate treatment effects in patients with a life expectancy exceeding 90 years.AimUpdate the LIFE-CVD model to enable application of the model in people aged 35 to 89 years, and to allow more accurate estimation of treatment effects in patients with a life expectancy exceeding 90 years.MethodsThe study was conducted using data from the same studies as were used for derivation and validation of the original model, including the Multi-Ethnic Study of Atherosclerosis (MESA) cohort, Atherosclerosis Risk in Communities Study (ARIC) cohort, and the European Prospective Investigation into Cancer-Netherlands (EPIC-NL) and EPIC-Norfolk cohort studies. Age-specific baseline survivals were smoothed by predicting the progression of baseline survivals with age, using a local polynomial regression function and a exponential function for CVD, and non-CVD mortality baseline survivals respectively. Using these functions, baseline survivals were then extrapolated to the age range of 35 to 100 years. External validation using the newly updated baseline survivals was performed.ResultsPerformance of the updated model was not dissimilar from the original model, with C-statistics for discrimination ranging from 0.70-0.76 in the external study populations. Calibration plots showed a good agreement between predicted and observed 10-year CVD risks. Estimation of treatment effects in patients with a life expectancy exceeding 90 years was improved.ConclusionThis update of the LIFE-CVD model improves the clinical usability of the model by increasing the age range and improving the method of estimation of lifetime treatment effects. |
