A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes
Abstract Background Individual patient data meta-analyses (IPD-MA) are often performed using a one-stage approach-- a form of generalized linear mixed model (GLMM) for binary outcomes. We compare (i) one-stage to two-stage approaches (ii) the performance of two estimation procedures (Penalized Quasi...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2017-02-01
|
| Series: | BMC Medical Research Methodology |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s12874-017-0307-7 |
| _version_ | 1818486257216389120 |
|---|---|
| author | Doneal Thomas Robert Platt Andrea Benedetti |
| author_facet | Doneal Thomas Robert Platt Andrea Benedetti |
| author_sort | Doneal Thomas |
| collection | DOAJ |
| description | Abstract Background Individual patient data meta-analyses (IPD-MA) are often performed using a one-stage approach-- a form of generalized linear mixed model (GLMM) for binary outcomes. We compare (i) one-stage to two-stage approaches (ii) the performance of two estimation procedures (Penalized Quasi-likelihood-PQL and Adaptive Gaussian Hermite Quadrature-AGHQ) for GLMMs with binary outcomes within the one-stage approach and (iii) using stratified study-effect or random study-effects. Methods We compare the different approaches via a simulation study, in terms of bias, mean-squared error (MSE), coverage and numerical convergence, of the pooled treatment effect (β 1) and between-study heterogeneity of the treatment effect (τ 1 2 ). We varied the prevalence of the outcome, sample size, number of studies and variances and correlation of the random effects. Results The two-stage and one-stage methods produced approximately unbiased β 1 estimates. PQL performed better than AGHQ for estimating τ 1 2 with respect to MSE, but performed comparably with AGHQ in estimating the bias of β 1 and of τ 1 2 . The random study-effects model outperformed the stratified study-effects model in small size MA. Conclusion The one-stage approach is recommended over the two-stage method for small size MA. There was no meaningful difference between the PQL and AGHQ procedures. Though the random-intercept and stratified-intercept approaches can suffer from their underlining assumptions, fitting GLMM with a random-intercept are less prone to misfit and has good convergence rate. |
| first_indexed | 2024-12-10T16:20:31Z |
| format | Article |
| id | doaj.art-e348fd12a89f434d805a281d2cc0959e |
| institution | Directory Open Access Journal |
| issn | 1471-2288 |
| language | English |
| last_indexed | 2024-12-10T16:20:31Z |
| publishDate | 2017-02-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Research Methodology |
| spelling | doaj.art-e348fd12a89f434d805a281d2cc0959e2022-12-22T01:41:50ZengBMCBMC Medical Research Methodology1471-22882017-02-0117111210.1186/s12874-017-0307-7A comparison of analytic approaches for individual patient data meta-analyses with binary outcomesDoneal Thomas0Robert Platt1Andrea Benedetti2Department of Epidemiology, Biostatistics & Occupational Health, McGill UniversityDepartment of Epidemiology, Biostatistics & Occupational Health, McGill UniversityDepartment of Epidemiology, Biostatistics & Occupational Health, McGill UniversityAbstract Background Individual patient data meta-analyses (IPD-MA) are often performed using a one-stage approach-- a form of generalized linear mixed model (GLMM) for binary outcomes. We compare (i) one-stage to two-stage approaches (ii) the performance of two estimation procedures (Penalized Quasi-likelihood-PQL and Adaptive Gaussian Hermite Quadrature-AGHQ) for GLMMs with binary outcomes within the one-stage approach and (iii) using stratified study-effect or random study-effects. Methods We compare the different approaches via a simulation study, in terms of bias, mean-squared error (MSE), coverage and numerical convergence, of the pooled treatment effect (β 1) and between-study heterogeneity of the treatment effect (τ 1 2 ). We varied the prevalence of the outcome, sample size, number of studies and variances and correlation of the random effects. Results The two-stage and one-stage methods produced approximately unbiased β 1 estimates. PQL performed better than AGHQ for estimating τ 1 2 with respect to MSE, but performed comparably with AGHQ in estimating the bias of β 1 and of τ 1 2 . The random study-effects model outperformed the stratified study-effects model in small size MA. Conclusion The one-stage approach is recommended over the two-stage method for small size MA. There was no meaningful difference between the PQL and AGHQ procedures. Though the random-intercept and stratified-intercept approaches can suffer from their underlining assumptions, fitting GLMM with a random-intercept are less prone to misfit and has good convergence rate.http://link.springer.com/article/10.1186/s12874-017-0307-7Individual patient data meta-analysesOne- and two-stage modelsGeneralized linear mixed modelsPenalized quasi-likelihoodAdaptive gauss-hermite quadratureFixed and random study-effects |
| spellingShingle | Doneal Thomas Robert Platt Andrea Benedetti A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes BMC Medical Research Methodology Individual patient data meta-analyses One- and two-stage models Generalized linear mixed models Penalized quasi-likelihood Adaptive gauss-hermite quadrature Fixed and random study-effects |
| title | A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes |
| title_full | A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes |
| title_fullStr | A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes |
| title_full_unstemmed | A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes |
| title_short | A comparison of analytic approaches for individual patient data meta-analyses with binary outcomes |
| title_sort | comparison of analytic approaches for individual patient data meta analyses with binary outcomes |
| topic | Individual patient data meta-analyses One- and two-stage models Generalized linear mixed models Penalized quasi-likelihood Adaptive gauss-hermite quadrature Fixed and random study-effects |
| url | http://link.springer.com/article/10.1186/s12874-017-0307-7 |
| work_keys_str_mv | AT donealthomas acomparisonofanalyticapproachesforindividualpatientdatametaanalyseswithbinaryoutcomes AT robertplatt acomparisonofanalyticapproachesforindividualpatientdatametaanalyseswithbinaryoutcomes AT andreabenedetti acomparisonofanalyticapproachesforindividualpatientdatametaanalyseswithbinaryoutcomes AT donealthomas comparisonofanalyticapproachesforindividualpatientdatametaanalyseswithbinaryoutcomes AT robertplatt comparisonofanalyticapproachesforindividualpatientdatametaanalyseswithbinaryoutcomes AT andreabenedetti comparisonofanalyticapproachesforindividualpatientdatametaanalyseswithbinaryoutcomes |