Publication: A comparison of bivariate, multivariate random-effects, and Poisson correlated gamma-frailty models to meta-analyze individual patient data of ordinal scale diagnostic tests
Issued Date
2017-11-01
Resource Type
ISSN
15214036
03233847
03233847
Other identifier(s)
2-s2.0-85022328420
Rights
Mahidol University
Rights Holder(s)
SCOPUS
Bibliographic Citation
Biometrical Journal. Vol.59, No.6 (2017), 1317-1338
Suggested Citation
Gabrielle Simoneau, Brooke Levis, Pim Cuijpers, John P.A. Ioannidis, Scott B. Patten, Ian Shrier, Charles H. Bombardier, Flavia de Lima Osório, Jesse R. Fann, Dwenda Gjerdingen, Femke Lamers, Manote Lotrakul, Bernd Löwe, Juwita Shaaban, Lesley Stafford, Henk C.P.M. van Weert, Mary A. Whooley, Karin A. Wittkampf, Albert S. Yeung, Brett D. Thombs, Andrea Benedetti A comparison of bivariate, multivariate random-effects, and Poisson correlated gamma-frailty models to meta-analyze individual patient data of ordinal scale diagnostic tests. Biometrical Journal. Vol.59, No.6 (2017), 1317-1338. doi:10.1002/bimj.201600184 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/42415
Research Projects
Organizational Units
Authors
Journal Issue
Thesis
Title
A comparison of bivariate, multivariate random-effects, and Poisson correlated gamma-frailty models to meta-analyze individual patient data of ordinal scale diagnostic tests
Author(s)
Gabrielle Simoneau
Brooke Levis
Pim Cuijpers
John P.A. Ioannidis
Scott B. Patten
Ian Shrier
Charles H. Bombardier
Flavia de Lima Osório
Jesse R. Fann
Dwenda Gjerdingen
Femke Lamers
Manote Lotrakul
Bernd Löwe
Juwita Shaaban
Lesley Stafford
Henk C.P.M. van Weert
Mary A. Whooley
Karin A. Wittkampf
Albert S. Yeung
Brett D. Thombs
Andrea Benedetti
Brooke Levis
Pim Cuijpers
John P.A. Ioannidis
Scott B. Patten
Ian Shrier
Charles H. Bombardier
Flavia de Lima Osório
Jesse R. Fann
Dwenda Gjerdingen
Femke Lamers
Manote Lotrakul
Bernd Löwe
Juwita Shaaban
Lesley Stafford
Henk C.P.M. van Weert
Mary A. Whooley
Karin A. Wittkampf
Albert S. Yeung
Brett D. Thombs
Andrea Benedetti
Other Contributor(s)
McGill University
Lady Davis Institute for Medical Research
Vrije Universiteit Amsterdam
Stanford University
University of Calgary
University of Washington, Seattle
Universidade de Sao Paulo - USP
University of Minnesota Twin Cities
VU University Medical Center
Mahidol University
Universitätsklinikum Hamburg-Eppendorf und Medizinische Fakultät
School of Medical Sciences - Universiti Sains Malaysia
Royal Women's Hospital, Carlton
Academic Medical Centre, University of Amsterdam
VA Medical Center
Massachusetts General Hospital
Centre universitaire de santé McGill
Lady Davis Institute for Medical Research
Vrije Universiteit Amsterdam
Stanford University
University of Calgary
University of Washington, Seattle
Universidade de Sao Paulo - USP
University of Minnesota Twin Cities
VU University Medical Center
Mahidol University
Universitätsklinikum Hamburg-Eppendorf und Medizinische Fakultät
School of Medical Sciences - Universiti Sains Malaysia
Royal Women's Hospital, Carlton
Academic Medical Centre, University of Amsterdam
VA Medical Center
Massachusetts General Hospital
Centre universitaire de santé McGill
Abstract
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Individual patient data (IPD) meta-analyses are increasingly common in the literature. In the context of estimating the diagnostic accuracy of ordinal or semi-continuous scale tests, sensitivity and specificity are often reported for a given threshold or a small set of thresholds, and a meta-analysis is conducted via a bivariate approach to account for their correlation. When IPD are available, sensitivity and specificity can be pooled for every possible threshold. Our objective was to compare the bivariate approach, which can be applied separately at every threshold, to two multivariate methods: the ordinal multivariate random-effects model and the Poisson correlated gamma-frailty model. Our comparison was empirical, using IPD from 13 studies that evaluated the diagnostic accuracy of the 9-item Patient Health Questionnaire depression screening tool, and included simulations. The empirical comparison showed that the implementation of the two multivariate methods is more laborious in terms of computational time and sensitivity to user-supplied values compared to the bivariate approach. Simulations showed that ignoring the within-study correlation of sensitivity and specificity across thresholds did not worsen inferences with the bivariate approach compared to the Poisson model. The ordinal approach was not suitable for simulations because the model was highly sensitive to user-supplied starting values. We tentatively recommend the bivariate approach rather than more complex multivariate methods for IPD diagnostic accuracy meta-analyses of ordinal scale tests, although the limited type of diagnostic data considered in the simulation study restricts the generalization of our findings.
