Regression modelling of weighted κ by using generalized estimating equations

被引:33
作者
Gonin, R
Lipsitz, SR
Fitzmaurice, GM
Molenberghs, G
机构
[1] Westat WB486, Rockville, MD 20850 USA
[2] Dana Farber Canc Inst, Boston, MA 02115 USA
[3] Harvard Univ, Boston, MA 02115 USA
[4] Limburgs Univ Ctr, Diepenbeek, Belgium
关键词
disease index; generalized estimating equations; intraclass correlation coefficient; measure of agreement; ordinal data; unbalanced data; weighted kappa;
D O I
10.1111/1467-9876.00175
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
In many clinical studies more than one observer may be rating a characteristic measured on an ordinal scale. For example, a study may involve a group of physicians rating a feature seen on a pathology specimen or a computer tomography scan. In clinical studies of this kind, the weighted it coefficient is a popular measure of agreement for ordinally scaled ratings. Our research stems from a study in which the severity of inflammatory skin disease was rated. The investigators wished to determine and evaluate the strength of agreement between a variable number of observers taking into account patient-specific (age and gender) as well as rater-specific (whether board certified in dermatology) characteristics. This suggested modelling kappa as a function of these covariates. We propose the use of generalized estimating equations to estimate the weighted kappa coefficient. This approach also accommodates unbalanced data which arise when some subjects are not judged by the same set of observers. Currently an estimate of overall kappa for a simple unbalanced data set without covariates involving more than two observers is unavailable. In the inflammatory skin disease study none of the covariates were significantly associated with kappa, thus enabling the calculation of an overall weighted a for this unbalanced data set. In the second motivating example (multiple sclerosis), geographic location was significantly associated with kappa. In addition we also compared the results of our method with current methods of testing for heterogeneity of weighted ii coefficients across strata (geographic location) that are available for balanced data sets.
引用
收藏
页码:1 / 18
页数:18
相关论文
共 27 条
[1]  
AGRASTI A, 1990, CATEGORICAL DATA ANA
[2]   NONOPTIMALLY WEIGHTED LEAST-SQUARES [J].
BLOCH, DA ;
MOSES, LE .
AMERICAN STATISTICIAN, 1988, 42 (01) :50-53
[4]   A COEFFICIENT OF AGREEMENT FOR NOMINAL SCALES [J].
COHEN, J .
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 1960, 20 (01) :37-46
[5]   COMPUTATIONAL EXPERIENCE WITH CONFIDENCE-REGIONS AND CONFIDENCE-INTERVALS FOR NONLINEAR LEAST-SQUARES [J].
DONALDSON, JR ;
SCHNABEL, RB .
TECHNOMETRICS, 1987, 29 (01) :67-82
[6]   The statistical analysis of kappa statistics in multiple samples [J].
Donner, A ;
Klar, N .
JOURNAL OF CLINICAL EPIDEMIOLOGY, 1996, 49 (09) :1053-1058
[7]   ESTIMATION OF INTRA-CLASS CORRELATION IN THE ANALYSIS OF FAMILY DATA [J].
DONNER, A ;
KOVAL, JJ .
BIOMETRICS, 1980, 36 (01) :19-25
[8]  
Efron B., 1982, SOC IND APPL MATH CB, V38, DOI [10.1137/1.9781611970319, DOI 10.1137/1.9781611970319]
[9]   Reliability testing of the dermatology index of disease severity (DIDS) - An index for staging the severity of cutaneous inflammatory disease [J].
Faust, HB ;
Gonin, R ;
Chuang, TY ;
Lewis, CW ;
Melfi, CA ;
Farmer, ER .
ARCHIVES OF DERMATOLOGY, 1997, 133 (11) :1443-1448
[10]   EQUIVALENCE OF WEIGHTED KAPPA AND INTRACLASS CORRELATION COEFFICIENT AS MEASURES OF RELIABILITY [J].
FLEISS, JL ;
COHEN, J .
EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 1973, 33 (03) :613-619