Re-calibration of coronary risk prediction: an example of the Seven Countries Study.
Re-calibration of coronary risk prediction: an example of the Seven Countries Study.
复制标题
DOI：
10.1038/s41598-017-17784-2
复制
发表时间：
2017-12-14
影响因子：
4.6
通讯作者：
Menotti A
中科院分区：
文献类型：
--
作者：
Puddu PE;Piras P;Kromhout D;Tolonen H;Kafatos A;Menotti A
我们的目标是使用来自七国研究的北欧（NE, N = 2360）或南欧（SE, N = 2789）中年男性的六个标准危险因素进行校准和重新校准过程，其参数和数据完全已知，以确定重新校准是否给出了正确的答案。Demler （GNDD）在2015年修改的Greenwood-Nam-D 'Agostino技术使用观察/预期冠心病死亡风险的10十分位数产生了卡方统计，对应于使用二元数据的多逻辑方程中使用的Hosmer-Lemeshaw卡方。GNDD测试使用观察和预测事件的生存概率，而不是事件的数量。该练习以五种不同的方式将ne -预测模型的参数应用于SE（反之亦然），并将模拟重新校准的结果与实际数据进行比较。只有替换风险因素系数才能获得良好的重新校准，其量级相似，NE-SE之间无显著差异。在所有其他方法中，都无法获得良好的重新校准。这足以赞扬重新评价大多数调查的总体需要，这些调查在没有GNDD或其他适当的技术来统计评估潜在差异的情况下，得出的结论是重新校准是一种公平的方法，因此可以使用，而无需特别谨慎。
We aimed at performing a calibration and re-calibration process using six standard risk factors from Northern (NE, N = 2360) or Southern European (SE, N = 2789) middle-aged men of the Seven Countries Study, whose parameters and data were fully known, to establish whether re-calibration gave the right answer. Greenwood-Nam-D’Agostino technique as modified by Demler (GNDD) in 2015 produced chi-squared statistics using 10 deciles of observed/expected CHD mortality risk, corresponding to Hosmer-Lemeshaw chi-squared employed for multiple logistic equations whereby binary data are used. Instead of the number of events, the GNDD test uses survival probabilities of observed and predicted events. The exercise applied, in five different ways, the parameters of the NE-predictive model to SE (and vice-versa) and compared the outcome of the simulated re-calibration with the real data. Good re-calibration could be obtained only when risk factor coefficients were substituted, being similar in magnitude and not significantly different between NE-SE. In all other ways, a good re-calibration could not be obtained. This is enough to praise for an overall need of re-evaluation of most investigations that, without GNDD or another proper technique for statistically assessing the potential differences, concluded that re-calibration is a fair method and might therefore be used, with no specific caution.
登录
查看更多内容
DOI：
10.5114/aoms.2017.68237
发表时间：
2017-06
期刊：
Archives of medical science : AMS
影响因子：
--
作者：
Faridi KF;Lupton JR;Martin SS;Banach M;Quispe R;Kulkarni K;Jones SR;Michos ED
通讯作者：
Michos ED
影响因子：
5.7
作者：
Menotti, A;Lanti, H;Kromhout, D
通讯作者：
Kromhout, D
影响因子：
3.5
作者：
De Bacquer, Dirk;De Backer, Guy
通讯作者：
De Backer, Guy
影响因子：
120.7
作者：
Liu, J;Hong, YL;Zhao, D
通讯作者：
Zhao, D
DOI：
10.1097/hjr.0b013e3282fb040f
发表时间：
2008-08-01
影响因子：
--
作者：
Marques-Vidal, Pedro;Rodondi, Nicolas;Vollenweider, Peter
通讯作者：
Vollenweider, Peter