Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data
Hierarchical models such as the bivariate and hierarchical summary receiver operating characteristic (HSROC) models are recommended for meta-analysis of test accuracy studies. These models are challenging to fit when there are few studies and/or sparse data (for example zero cells in contingency tab...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Published: |
Sage
2015
|
| Subjects: | |
| Online Access: | https://eprints.nottingham.ac.uk/31489/ |
| _version_ | 1848794213513166848 |
|---|---|
| author | Takwoingi, Yemisi Guo, Boliang Riley, Richard D. Deeks, Jonthan J. |
| author_facet | Takwoingi, Yemisi Guo, Boliang Riley, Richard D. Deeks, Jonthan J. |
| author_sort | Takwoingi, Yemisi |
| building | Nottingham Research Data Repository |
| collection | Online Access |
| description | Hierarchical models such as the bivariate and hierarchical summary receiver operating characteristic (HSROC) models are recommended for meta-analysis of test accuracy studies. These models are challenging to fit when there are few studies and/or sparse data (for example zero cells in contingency tables due to studies reporting 100% sensitivity or specificity); the models may not converge, or give unreliable parameter estimates. Using simulation, we investigated the performance of seven hierarchical models incorporating increasing simplifications in scenarios designed to replicate realistic situations for meta-analysis of test accuracy studies. Performance of the models was assessed in terms of estimability (percentage of meta-analyses that successfully converged and percentage where the between study correlation was estimable), bias, mean square error and coverage of the 95% confidence intervals. Our results indicate that simpler hierarchical models are valid in situations with few studies or sparse data. For synthesis of sensitivity and specificity, univariate random effects logistic regression models are appropriate when a bivariate model cannot be fitted. Alternatively, an HSROC model that assumes a symmetric SROC curve (by excluding the shape parameter) can be used if the HSROC model is the chosen meta-analytic approach. In the absence of heterogeneity, fixed effect equivalent of the models can be applied. |
| first_indexed | 2025-11-14T19:12:37Z |
| format | Article |
| id | nottingham-31489 |
| institution | University of Nottingham Malaysia Campus |
| institution_category | Local University |
| last_indexed | 2025-11-14T19:12:37Z |
| publishDate | 2015 |
| publisher | Sage |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | nottingham-314892020-05-04T17:09:58Z https://eprints.nottingham.ac.uk/31489/ Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data Takwoingi, Yemisi Guo, Boliang Riley, Richard D. Deeks, Jonthan J. Hierarchical models such as the bivariate and hierarchical summary receiver operating characteristic (HSROC) models are recommended for meta-analysis of test accuracy studies. These models are challenging to fit when there are few studies and/or sparse data (for example zero cells in contingency tables due to studies reporting 100% sensitivity or specificity); the models may not converge, or give unreliable parameter estimates. Using simulation, we investigated the performance of seven hierarchical models incorporating increasing simplifications in scenarios designed to replicate realistic situations for meta-analysis of test accuracy studies. Performance of the models was assessed in terms of estimability (percentage of meta-analyses that successfully converged and percentage where the between study correlation was estimable), bias, mean square error and coverage of the 95% confidence intervals. Our results indicate that simpler hierarchical models are valid in situations with few studies or sparse data. For synthesis of sensitivity and specificity, univariate random effects logistic regression models are appropriate when a bivariate model cannot be fitted. Alternatively, an HSROC model that assumes a symmetric SROC curve (by excluding the shape parameter) can be used if the HSROC model is the chosen meta-analytic approach. In the absence of heterogeneity, fixed effect equivalent of the models can be applied. Sage 2015-06-26 Article PeerReviewed Takwoingi, Yemisi, Guo, Boliang, Riley, Richard D. and Deeks, Jonthan J. (2015) Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data. Statistical Methods in Medical Research . ISSN 1477-0334 diagnostic accuracy meta-analysis hierarchical models HSROC model bivariate model sensitivity specificity diagnostic odd ration sparse data random effects http://smm.sagepub.com/content/early/2015/06/25/0962280215592269 doi:10.1177/0962280215592269 doi:10.1177/0962280215592269 |
| spellingShingle | diagnostic accuracy meta-analysis hierarchical models HSROC model bivariate model sensitivity specificity diagnostic odd ration sparse data random effects Takwoingi, Yemisi Guo, Boliang Riley, Richard D. Deeks, Jonthan J. Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| title | Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| title_full | Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| title_fullStr | Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| title_full_unstemmed | Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| title_short | Performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| title_sort | performance of methods for meta-analysis of diagnostic test accuracy with few studies or sparse data |
| topic | diagnostic accuracy meta-analysis hierarchical models HSROC model bivariate model sensitivity specificity diagnostic odd ration sparse data random effects |
| url | https://eprints.nottingham.ac.uk/31489/ https://eprints.nottingham.ac.uk/31489/ https://eprints.nottingham.ac.uk/31489/ |