Chen CC, Barnhart HX. Comparison between ICC and CCC to evaluate agreement for data without and with replications. Comput Stat Data Anal 2008;53:554-64. Finally, the correlation coefficient (CCC) proposed by Lin9 must assess the agreement between two measures without adopting an underlying ANOVA model. The CCC is defined as Chen CC, Barnhart HX. Evaluation of compliance with intraclassade correlation coefficient and correlation coefficient for data with repeated measurements. Comput Stat Data Anal 2013;60:132-45. where μ j and σ 2 j are the average and variance of the Jth test. Note that C b depends in part on “bias” if the interest is to estimate the difference between the means of the two tests, i.e. μ 1 x μ 2. C b is also called the “bia correction factor.” 9 The CCC can therefore be designed as the product of a consistency measure (i.e.

the Pearson correlation coefficient) and a distortion measure. In other words, the CCC quantifies not only how the observations fall on the regression line (by B), but also how close this regression line is to the 45-degree line of perfect concordance (above C b). These two themes – knowledge of objectives and consideration of theory – are the main keys to a successful analysis of contractual data. Here are some other, more specific questions about choosing the appropriate methods for a given study. The subsampling can be laminated using all available information on all samples, such as. B other diagnostic results or demographic variables. In a second example, we present data to compare two human papillomavirus (HPV) DNA tests that we categorized into five ordinal categories. We scanned 25% of the final deposit in 48 layers of the sample. Stratification allowed for subsampling of rare subgroups, oversampling informational samples, and subsampling non-experimental samples.

Another important area of research is to propose effective samples to compare the diagnostic test agreement. The selection of sample size and the allocation of sample size to strata require consideration, particularly for strata with rare events where a large number of samples need to be reinted to find a single positive specimen. Another important avenue of research is the effective use of the auxiliary information available for each stratification specimen. The example of an HPV test showed the value of several previous test results to highlight the most informative samples: those that probably do not have consistent results. Such effective sample samples could achieve most of the statistical effectiveness of re-examining all samples, with significant reductions in the cost of studies and consumption of valuable specimens. We believe that effective sampling designs should be used more broadly to improve the efficiency of the biomarker pipeline. We approved a hybrid capture test 2 (HC2; Qiagen Corporation, Gaithersburg, MD, USA) and C. trachomatis Detection and genoTyping Assay (Ct-DT); DDL Diagnostic Laboratory, Voorburg, Netherlands) for the detection of chlamydia infection [9]. HC2 had previously undergone a full pelvic examination on all women in the Cervarix Vaccine Trial (CVT) in Costa Rica, resulting in 827 women who tested positive for HC2 Chlamydia and 4998 women who were negative for HC2 chlamydia.

The 827 HC2-positives were re-tested by Ct-DT, resulting in 27 negative and 800 positive results. To minimize the cost of studying and the consumption of samples: 402 (8%) negative samples of HC2 were randomly selected for a new trial by Ct-DT. Of which 402, 6 tested positive ct-DT and 396 ct-DT negative. Weighting of 402 out of the total of 4998 negative HC2, 4923 ct-DT negative and 75 ct-DT considered positive. The problem of subsampling can also be formulated as a two-step laminated sample (also in double sampling [22, Ch.