Methods for Calculating Sensitivity and Specificity of Clustered Data: A Tutorial

被引:214
作者
Genders, Tessa S. S. [1 ,2 ]
Spronk, Sandra [1 ,2 ]
Stijnen, Theo [5 ]
Steyerberg, Ewout W. [3 ]
Lesaffre, Emmanuel [4 ]
Hunink, M. G. Myriam [1 ,2 ,6 ]
机构
[1] Erasmus Univ, Med Ctr Rotterdam, Dept Epidemiol, NL-3000 CA Rotterdam, Netherlands
[2] Erasmus Univ, Med Ctr Rotterdam, Dept Radiol, NL-3000 CA Rotterdam, Netherlands
[3] Erasmus Univ, Med Ctr Rotterdam, Dept Publ Hlth, NL-3000 CA Rotterdam, Netherlands
[4] Erasmus Univ, Med Ctr Rotterdam, Dept Biostat, NL-3000 CA Rotterdam, Netherlands
[5] Leiden Univ, Med Ctr, Dept Med Stat, Leiden, Netherlands
[6] Harvard Univ, Sch Publ Hlth, Dept Hlth Policy & Management, Boston, MA 02115 USA
关键词
CORRELATED DATA; ROC CURVES; TESTS;
D O I
10.1148/radiol.12120509
中图分类号
R8 [特种医学]; R445 [影像诊断学];
学科分类号
1002 ; 100207 ; 1009 ;
摘要
The performance of a diagnostic test is often expressed in terms of sensitivity and specificity compared with the reference standard. Calculations of sensitivity and specificity commonly involve multiple observations per patient, which implies that the data are clustered. Whether analysis of sensitivity and specificity per patient or using multiple observations per patient is preferable depends on the clinical context and consequences. The purpose of this article was to discuss and illustrate the most common statistical methods that calculate sensitivity and specificity of clustered data, adjusting for the possible correlation between observations within each patient. This tutorial presents and illustrates the following methods: (a) analysis at different levels ignoring correlation, (b) variance adjustment, (c) logistic random-effects models, and (d) generalized estimating equations. The choice of method and the level of reporting should correspond with the clinical decision problem. If multiple observations per patient are relevant to the clinical decision problem, the potential correlation between observations should be explored and taken into account in the statistical analysis. (C) RSNA, 2012
引用
收藏
页码:910 / 916
页数:7
相关论文
共 31 条
[1]  
[Anonymous], 2002, Statistical Methods in Diagnostic Medicine
[2]  
[Anonymous], 2003, Statistical Methods for Rates and Proportions
[3]  
Beam C A, 1998, Stat Methods Med Res, V7, P324
[4]  
Beam C A, 1995, Acad Radiol, V2 Suppl 1, pS4
[5]  
Beam CA, 1995, ACAD RADIOL S, V2, pS68
[6]  
Beam CA, 1995, ACAD RADIOL S, V2, pS57
[7]  
Gatsonis C A, 1995, Acad Radiol, V2 Suppl 1, pS14
[8]  
GATSONIS CA, 1995, ACAD RADIOL S, V2, pS61
[9]  
GATSONIS CA, 1995, ACAD RADIOL S, V2, pS57
[10]  
Kirkwood BR, 2010, Essential medical statistics