1
|
Liu YL, Zhang B, Chu H, Chen Y. Network meta-analysis made simple: a composite likelihood approach. MEDRXIV : THE PREPRINT SERVER FOR HEALTH SCIENCES 2024:2024.06.19.24309163. [PMID: 38947001 PMCID: PMC11213057 DOI: 10.1101/2024.06.19.24309163] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 07/02/2024]
Abstract
Network meta-analysis, also known as mixed treatments comparison meta-analysis or multiple treatments meta-analysis, extends conventional pairwise meta-analysis by simultaneously synthesizing multiple interventions in a single integrated analysis. Despite the growing popularity of network metaanalysis within comparative effectiveness research, it comes with potential challenges. For example, within-study correlations among treatment comparisons are rarely reported in the published literature. Yet, these correlations are pivotal for valid statistical inference. As demonstrated in earlier studies, ignoring these correlations can inflate mean squared errors of the resulting point estimates and lead to inaccurate standard error estimates. This paper introduces a composite likelihood-based approach that ensures accurate statistical inference without requiring knowledge of the within-study correlations. The proposed method is computationally robust and efficient, with substantially reduced computational time compared to the state-of-the-science methods implemented in R packages. The proposed method was evaluated through extensive simulations and applied to two important applications including a network meta-analysis comparing interventions for primary open-angle glaucoma, and another comparing treatments for chronic prostatitis and chronic pelvic pain syndrome.
Collapse
Affiliation(s)
- Yu-Lun Liu
- Peter O'Donnell Jr. School of Public Health, University of Texas Southwestern Medical Center, Dallas, TX, USA
| | - Bingyu Zhang
- Center for Health AI and Synthesis of Evidence, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
- Applied Mathematics and Computational Science, University of Pennsylvania, Philadelphia, PA, USA
| | - Haitao Chu
- Statistical Research and Data Science, Pfizer Inc., New York, NY, USA
| | - Yong Chen
- Center for Health AI and Synthesis of Evidence, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
- Applied Mathematics and Computational Science, University of Pennsylvania, Philadelphia, PA, USA
- Leonard Davis Institute of Health Economics, Philadelphia, PA, USA
- Penn Medicine Center for Evidence-based Practice, Philadelphia, PA, USA
- Penn Institute for Biomedical Informatics, Philadelphia, PA, USA
| |
Collapse
|
2
|
Liu YL, Ying GS, Quinn GE, Zhou XH, Chen Y. Extending Hui-Walter framework to correlated outcomes with application to diagnosis tests of an eye disease among premature infants. Stat Med 2022; 41:433-448. [PMID: 34859902 PMCID: PMC8884176 DOI: 10.1002/sim.9269] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/22/2019] [Revised: 08/28/2021] [Accepted: 11/05/2021] [Indexed: 11/08/2022]
Abstract
Diagnostic accuracy, a measure of diagnostic tests for correctly identifying patients with or without a target disease, plays an important role in evidence-based medicine. Diagnostic accuracy of a new test ideally should be evaluated by comparing to a gold standard; however, in many medical applications it may be invasive, costly, or even unethical to obtain a gold standard for particular diseases. When the accuracy of a new candidate test under evaluation is assessed by comparison to an imperfect reference test, bias is expected to occur and result in either overestimates or underestimates of its true accuracy. In addition, diagnostic test studies often involve repeated measurements of the same patient, such as the paired eyes or multiple teeth, and generally lead to correlated and clustered data. Using the conventional statistical methods to estimate diagnostic accuracy can be biased by ignoring the within-cluster correlations. Despite numerous statistical approaches have been proposed to tackle this problem, the methodology to deal with correlated and clustered data in the absence of a gold standard is limited. In this article, we propose a method based on the composite likelihood function to derive simple and intuitive closed-form solutions for estimates of diagnostic accuracy, in terms of sensitivity and specificity. Through simulation studies, we illustrate the relative advantages of the proposed method over the existing methods that simply treat an imperfect reference test as a gold standard in correlated and clustered data. Compared with the existing methods, the proposed method can reduce not only substantial bias, but also the computational burden. Moreover, to demonstrate the utility of this approach, we apply the proposed method to the study of National-Eye-Institute-funded Telemedicine Approaches to Evaluating of Acute-Phase Retinopathy of Prematurity (e-ROP), for estimating accuracies of both the ophthalmologist examination and the image evaluation.
Collapse
Affiliation(s)
- Yu-Lun Liu
- Department of Population and Data Sciences, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA.,Correspondence to: Yong Chen, Department of Biostatistics, Epidemiology and Informatics, University of Pennsylvania, Philadelphia, PA 19104, USA or Yu-Lun Liu, Department of Population and Data Sciences, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA. or
| | - Gui-Shuang Ying
- Department of Ophthalmology, University of Pennsylvania, Philadelphia, PA 19104, USA
| | - Graham E. Quinn
- Division of Pediatric Ophthalmology, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, PA 19104, USA
| | - Xiao-Hua Zhou
- Department of Biostatistics, School of Public Health, Peking University, China.,Beijing International Center for Mathematical Research, Peking University, China
| | - Yong Chen
- Department of Biostatistics, Epidemiology and Informatics, University of Pennsylvania, Philadelphia, PA 19104, USA.,Correspondence to: Yong Chen, Department of Biostatistics, Epidemiology and Informatics, University of Pennsylvania, Philadelphia, PA 19104, USA or Yu-Lun Liu, Department of Population and Data Sciences, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA. or
| |
Collapse
|
3
|
Hong C, Salanti G, Morton SC, Riley RD, Chu H, Kimmel SE, Chen Y. Testing small study effects in multivariate meta-analysis. Biometrics 2020; 76:1240-1250. [PMID: 32720712 DOI: 10.1111/biom.13342] [Citation(s) in RCA: 14] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/23/2018] [Revised: 06/06/2019] [Accepted: 09/10/2019] [Indexed: 01/10/2023]
Abstract
Small study effects occur when smaller studies show different, often larger, treatment effects than large ones, which may threaten the validity of systematic reviews and meta-analyses. The most well-known reasons for small study effects include publication bias, outcome reporting bias, and clinical heterogeneity. Methods to account for small study effects in univariate meta-analysis have been extensively studied. However, detecting small study effects in a multivariate meta-analysis setting remains an untouched research area. One of the complications is that different types of selection processes can be involved in the reporting of multivariate outcomes. For example, some studies may be completely unpublished while others may selectively report multiple outcomes. In this paper, we propose a score test as an overall test of small study effects in multivariate meta-analysis. Two detailed case studies are given to demonstrate the advantage of the proposed test over various naive applications of univariate tests in practice. Through simulation studies, the proposed test is found to retain nominal Type I error rates with considerable power in moderate sample size settings. Finally, we also evaluate the concordance between the proposed tests with the naive application of univariate tests by evaluating 44 systematic reviews with multiple outcomes from the Cochrane Database.
Collapse
Affiliation(s)
- Chuan Hong
- Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
| | - Georgia Salanti
- Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland
| | - Sally C Morton
- Department of Statistics, Virginia Tech, Blacksburg, Virginia
| | - Richard D Riley
- Centre for Prognosis Research, School of Medicine, Keele University, Staffordshire, UK
| | - Haitao Chu
- Division of Biostatistics, University of Minnesota, Minneapolis, Minnesota
| | - Stephen E Kimmel
- Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania.,Department of Biostatistics, Epidemiology & Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
| | - Yong Chen
- Department of Biostatistics, Epidemiology & Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
| |
Collapse
|
4
|
Chatelain M, Drobniak SM, Szulkin M. The association between stressors and telomeres in non‐human vertebrates: a meta‐analysis. Ecol Lett 2019; 23:381-398. [DOI: 10.1111/ele.13426] [Citation(s) in RCA: 99] [Impact Index Per Article: 19.8] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/05/2019] [Revised: 10/10/2019] [Accepted: 10/16/2019] [Indexed: 12/21/2022]
Affiliation(s)
- Marion Chatelain
- Centre of New Technologies University of Warsaw Banacha 2C 02‐097 Warszawa Poland
| | - Szymon M. Drobniak
- Institute of Environmental Sciences Jagiellonian University Gronostajowa 7 30‐387 Kraków Poland
- Ecology & Evolution Research Centre School of Biological, Environmental and Earth Sciences University of New South Wales Sydney Australia
| | - Marta Szulkin
- Centre of New Technologies University of Warsaw Banacha 2C 02‐097 Warszawa Poland
| |
Collapse
|
5
|
Generating the evidence for risk reduction: a contribution to the future of food-based dietary guidelines. Proc Nutr Soc 2018; 77:432-444. [PMID: 29708078 DOI: 10.1017/s0029665118000125] [Citation(s) in RCA: 20] [Impact Index Per Article: 3.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/07/2023]
Abstract
A major advantage of analyses on the food group level is that the results are better interpretable compared with nutrients or complex dietary patterns. Such results are also easier to transfer into recommendations on primary prevention of non-communicable diseases. As a consequence, food-based dietary guidelines (FBDG) are now the preferred approach to guide the population regarding their dietary habits. However, such guidelines should be based on a high grade of evidence as requested in many other areas of public health practice. The most straightforward approach to generate evidence is meta-analysing published data based on a careful definition of the research question. Explicit definitions of study questions should include participants, interventions/exposure, comparisons, outcomes and study design. Such type of meta-analyses should not only focus on categorical comparisons, but also on linear and non-linear dose-response associations. Risk of bias of the individual studies of the meta-analysis should be assessed, rated and the overall credibility of the results scored (e.g. using NutriGrade). Tools such as a measurement tool to assess systematic reviews or ROBIS are available to evaluate the methodological quality/risk of bias of meta-analyses. To further evaluate the complete picture of evidence, we propose conducting network meta-analyses (NMA) of intervention trials, mostly on intermediate disease markers. To rank food groups according to their impact, disability-adjusted life years can be used for the various clinical outcomes and the overall results can be compared across the food groups. For future FBDG, we recommend to implement evidence from pairwise and NMA and to quantify the health impact of diet-disease relationships.
Collapse
|
6
|
Chen Y, Cai Y, Hong C, Jackson D. Inference for correlated effect sizes using multiple univariate meta-analyses. Stat Med 2016; 35:1405-22. [PMID: 26537017 PMCID: PMC4821787 DOI: 10.1002/sim.6789] [Citation(s) in RCA: 7] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/24/2014] [Accepted: 10/13/2015] [Indexed: 12/17/2022]
Abstract
Multivariate meta-analysis, which involves jointly analyzing multiple and correlated outcomes from separate studies, has received a great deal of attention. One reason to prefer the multivariate approach is its ability to account for the dependence between multiple estimates from the same study. However, nearly all the existing methods for analyzing multivariate meta-analytic data require the knowledge of the within-study correlations, which are usually unavailable in practice. We propose a simple non-iterative method that can be used for the analysis of multivariate meta-analysis datasets, that has no convergence problems, and does not require the use of within-study correlations. Our approach uses standard univariate methods for the marginal effects but also provides valid joint inference for multiple parameters. The proposed method can directly handle missing outcomes under missing completely at random assumption. Simulation studies show that the proposed method provides unbiased estimates, well-estimated standard errors, and confidence intervals with good coverage probability. Furthermore, the proposed method is found to maintain high relative efficiency compared with conventional multivariate meta-analyses where the within-study correlations are known. We illustrate the proposed method through two real meta-analyses where functions of the estimated effects are of interest.
Collapse
Affiliation(s)
- Yong Chen
- Department of Biostatistics and EpidemiologyUniversity of PennsylvaniaPhiladelphiaPennsylvania19104U.S.A.
| | - Yi Cai
- Division of BiostatisticsUniversity of Texas School of Public Health1200 Pressler St, HoustonTexas 77030U.S.A.
| | - Chuan Hong
- Division of BiostatisticsUniversity of Texas School of Public Health1200 Pressler St, HoustonTexas 77030U.S.A.
| | - Dan Jackson
- MRC Biostatistics Unit, CambridgeCambridge Institute of Public HealthForvie Site, Robinson Way, Cambridge CB2 0SRU.K.
| |
Collapse
|