AN ALTERNATIVE TO THE COCHRAN- STATISTIC FOR ANALYSIS OF HETEROGENEITY IN META-ANALYSIS OF DIAGNOSTIC TESTS BASED ON HJ BIPLOT

The possible heterogeneity among individual studies constituting a meta-analysis is traditionally evaluated using Cochran Q and Higgins [I.sup.2] statistics. However, both indices have deficiencies: The Q statistic detects heterogeneity but does not allow its quantification, whereas the [I.sup.2] index allows for quantification of heterogeneity but does not indicate which studies are responsible for it. This problem is solved by additionally using the HJ biplot of the matrix containing the information about true positives, true negatives, false positives, and false negatives for each study. This means that the information contained in such a tetrachoric table contains the joint frequency distribution of the true classification of the disease and that provided by the diagnostic test. MSC: 62H30 KEYWORDS: Meta-analysis, effect size, heterogeneity, Q-statistic, [I.sup.2] index, HJ biplot` La forma tradicional de evaluar la posible heterogeneidad de los estudios individuales que conforman un meta-analisis, es utilizando los estadisticos Q de Cochran e [I.sup.2] de Higgins. Sin embargo, ambos indices presentan deficiencias: el Q detecta heterogeneidad, pero no permite cuantificarla y el [I.sup.2] permite cuantificarla, pero no conocer cuales son los estudios responsables de la heterogeneidad. Este problema se resuelve utilizando, ademas, el HJ-Biplot de la matriz que contiene la informacion sobre VP, VN, FP y FN, para todos y cada uno de los estudios; es decir la informacion contenida en la tabla tetracorica que contiene la distribucion de frecuencias conjunta de la verdadera clasificacion de la enfermedad y la proporcionada por el test diagnostico. PALABRAS CLAVES: Meta-analisis, tamano del efecto, heterogeneidad, estadistico Q, indice I2, HJ biplot