Beta-binomial analysis of variance model for network meta-analysis of diagnostic test accuracy data

There are several generalized linear mixed models to combine direct and indirect evidence on several diagnostic tests from related but independent diagnostic studies simultaneously also known as network meta-analysis. The popularity of these models is due to the attractive features of the normal distribution and the availability of statistical software to obtain parameter estimates. However, modeling the latent sensitivity and specificity using the normal distribution after transformation is neither natural nor computationally convenient. In this article, we develop a meta-analytic model based on the bivariate beta distribution, allowing to obtain improved and direct estimates for the global sensitivities and specificities of all tests involved, and taking into account simultaneously the intrinsic correlation between sensitivity and specificity and the overdispersion due to repeated measures. Using the beta distribution in regression has the following advantages, that the probabilities are modeled in their proper scale rather than a monotonic transform of the probabilities. Secondly, the model is flexible as it allows for asymmetry often present in the distribution of bounded variables such as proportions, which is the case with sparse data common in meta-analysis. Thirdly, the model provides parameters with direct meaningful interpretation since further integration is not necessary to obtain the meta-analytic estimates.

[1]  P. Albert,et al.  Models for longitudinal data: a generalized estimating equation approach. , 1988, Biometrics.

[2]  K. Cuschieri,et al.  Human papillomavirus testing versus repeat cytology for triage of minor cytological cervical lesions , 2013, Journal of Family Planning and Reproductive Health Care.

[3]  P N Valenstein,et al.  Evaluating diagnostic tests with imperfect standards. , 1990, American journal of clinical pathology.

[4]  Victoria N Nyaga.  Copula Based Bivariate Beta-Binomial Model for Diagnostic TestAccuracy Studies , 2016 .

[5]  Andrew GelmanyJanuary,et al.  Prior distribution , 2000 .

[6]  Marc Arbyn,et al.  Triage of women with minor abnormal cervical cytology: Meta‐analysis of the accuracy of an assay targeting messenger ribonucleic acid of 5 high‐risk human papillomavirus types , 2013, Cancer cytopathology.

[7]  R. Nelsen.  An Introduction to Copulas , 1998 .

[8]  Sophia Rabe-Hesketh,et al.  Avoiding zero between‐study variance estimates in random‐effects meta‐analysis , 2013, Statistics in medicine.

[9]  Yang Ning,et al.  Meta‐analysis of studies with bivariate binary outcomes: a marginal beta‐binomial model approach , 2016, Statistics in medicine.

[10]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data: Little/Statistical Analysis with Missing Data , 2002 .

[11]  Marc Aerts,et al.  ANOVA model for network meta-analysis of diagnostic test accuracy data , 2016, Statistical methods in medical research.

[12]  Jiqiang Guo,et al.  Stan: A Probabilistic Programming Language. , 2017, Journal of statistical software.

[13]  S D Walter,et al.  Meta-analysis of diagnostic tests with imperfect reference standards. , 1999, Journal of clinical epidemiology.

[14]  Simon Jackman,et al.  Hierarchical Statistical Models , 2009 .

[15]  Bradley P Carlin,et al.  Network meta-analysis of randomized clinical trials: Reporting the proper summaries , 2014, Clinical trials.

[16]  Susanne Rehm,et al.  Prospective evaluation of p16/Ki‐67 dual‐stained cytology for managing women with abnormal Papanicolaou cytology: PALMS study results , 2015, Cancer cytopathology.

[17]  Roderick J. A. Little,et al.  Mixed Normal and Non‐Normal Data with Missing Values, Ignoring the Missing‐Data Mechanism , 2014 .

[18]  Peter Dalgaard,et al.  R Development Core Team (2010): R: A language and environment for statistical computing , 2010 .

[19]  Paolo Giorgi-Rossi,et al.  Triage of women with minor abnormal cervical cytology: Meta‐analysis of the accuracy of an assay targeting messenger ribonucleic acid of 5 high‐risk human papillomavirus types , 2014, Cancer cytopathology.

[20]  Oliver Kuss,et al.  Nonparametric meta‐analysis for diagnostic accuracy studies , 2015, Statistics in medicine.

[21]  Kristian Thorlund,et al.  Attention should be given to multiplicity issues in systematic reviews. , 2008, Journal of clinical epidemiology.

[22]  Jack Cuzick,et al.  Genotyping for Human Papillomavirus Types 16 and 18 in Women With Minor Cervical Lesions: A Systematic Review and Meta-analysis. , 2016, Annals of internal medicine.

[23]  Johannes B Reitsma,et al.  Problems in detecting misfit of latent class models in diagnostic research without a gold standard were shown. , 2016, Journal of clinical epidemiology.

[24]  Geert Molenberghs,et al.  The Direct Likelihood Method , 2007 .

[25]  M. Sklar.  Fonctions de repartition a n dimensions et leurs marges , 1959 .

[26]  M. Aitkin,et al.  Meta-analysis by random effect modelling in generalized linear models. , 1999, Statistics in medicine.

[27]  D. Rubin.  INFERENCE AND MISSING DATA , 1975 .

[28]  Maria Adam,et al.  A multivariate method for meta‐analysis and comparison of diagnostic tests , 2016, Statistics in medicine.

[29]  Carol Jagger,et al.  Assessing the validity of the Global Activity Limitation Indicator in fourteen European countries , 2015, BMC Medical Research Methodology.

[30]  Johannes B Reitsma,et al.  Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. , 2005, Journal of clinical epidemiology.

[31]  Nicola J Cooper,et al.  Network meta-analysis of multiple outcome measures accounting for borrowing of information across outcomes , 2014, BMC Medical Research Methodology.

[32]  Haitao Chu,et al.  Bivariate random effects models for meta-analysis of comparative studies with binary outcomes: Methods for the absolute risk difference and relative risk , 2012, Statistical methods in medical research.

[33]  Mir Said Siadaty,et al.  Repeated-measures modeling improved comparison of diagnostic tests in meta-analysis of dependent studies. , 2004, Journal of clinical epidemiology.

[34]  K. Atkins,et al.  p16/Ki‐67 dual‐stain cytology in the triage of ASCUS and LSIL Papanicolaou cytology , 2011, Cancer cytopathology.

[35]  Jack Cuzick,et al.  The APTIMA HPV assay versus the hybrid capture 2 test in triage of women with ASC‐US or LSIL cervical cytology: A meta‐analysis of the diagnostic accuracy , 2013, International journal of cancer.

[36]  Johannes B. Reitsma,et al.  A review of solutions for diagnostic accuracy studies with an imperfect or missing reference standard. , 2009, Journal of clinical epidemiology.

[37]  Julian Peto,et al.  Evidence regarding human papillomavirus testing in secondary prevention of cervical cancer. , 2012, Vaccine.

[38]  Jing Ning,et al.  A composite likelihood method for bivariate meta-analysis in diagnostic systematic reviews , 2017, Statistical methods in medical research.

[39]  Alan Agresti,et al.  Examples in which misspecification of a random effects distribution reduces efficiency, and possible remedies , 2004, Comput. Stat. Data Anal..

[40]  G. Barrie Wetherill.  The generalized linear model , 1981 .

[41]  Aristidis K. Nikoloulopoulos,et al.  A mixed effect model for bivariate meta‐analysis of diagnostic test accuracy studies using a copula representation of the random effects distribution , 2015, Statistics in medicine.

[42]  David J Spiegelhalter,et al.  A re-evaluation of random-effects meta-analysis , 2009, Journal of the Royal Statistical Society. Series A,.

[43]  S D Walter,et al.  Estimation of test error rates, disease prevalence and relative risk from misclassified data: a review. , 1988, Journal of clinical epidemiology.

[44]  Marc Arbyn,et al.  p16INK4a immunocytochemistry versus human papillomavirus testing for triage of women with minor cytologic abnormalities , 2012, Cancer cytopathology.

[45]  Margaret Sullivan Pepe,et al.  Insights into latent class analysis of diagnostic test performance. , 2007, Biostatistics.

[46]  Emmanuel Lesaffre,et al.  A general framework for comparative Bayesian meta-analysis of diagnostic studies , 2015, BMC Medical Research Methodology.

[47]  Geert Molenberghs,et al.  Direct likelihood analysis versus simple forms of imputation for missing data in randomized clinical trials , 2005, Clinical trials.

[48]  M. J. Frank.  On the simultaneous associativity ofF(x,y) andx +y -F(x,y) , 1979 .