zbMATH — the first resource for mathematics

An application of targeted maximum likelihood estimation to the meta-analysis of safety data. (English) Zbl 1271.62225
Summary: Safety analysis to estimate the effect of a treatment on an adverse event poses a challenging statistical problem even in randomized controlled trials because these events are typically rare, so studies originally powered for efficacy are underpowered for safety outcomes. A meta-analysis of data pooled across multiple studies may increase power, but missingness in the outcome or failed randomization can introduce bias. This article illustrates how targeted maximum likelihood estimation (TMLE) can be applied in a meta-analysis to reduce bias in causal effect estimates, and compares performance with other estimators in the literature. A simulation study in which missingness in the outcome is at random or completely at random highlights the differences in estimators with respect to the potential gains in bias and efficiency. Risk difference, relative risk, and odds ratio of the effect of treatment on 30-day mortality are estimated from data from eight randomized controlled trials. When an outcome event is rare there may be little opportunity to improve efficiency, and associations between covariates and the outcome may be hard to detect. TMLE attempts to exploit the available information to either meet or exceed the performance of a less sophisticated estimator.

62N02 Estimation in survival analysis and censored data
62P10 Applications of statistics to biology and medical sciences; meta analysis
62F10 Point estimation
Hmisc; R; SuperLearner; tmle
Full Text: DOI
[1] Bang, Doubly robust estimation in missing data and causal inference models, Biometrics 61 pp 962– (2005) · Zbl 1087.62121 · doi:10.1111/j.1541-0420.2005.00377.x
[2] Bickel, Efficient and Adaptive Estimation for Semiparametric Models (1997)
[3] Gruber, TMLE: Targeted Maximum Likelihood Estimation (2012)
[4] Gruber, A targeted maximum likelihood estimator of a causal effect on a bounded continuous outcome, The International Journal of Biostatistics 6 pp 1– (2010) · doi:10.2202/1557-4679.1260
[5] Gruber, TMLE: An R package for targeted maximum likelihood estimation (2012)
[6] Hampel, The influence curve and its role in robust estimation, Journal of the American Statistical Association 69 pp 383– (1974) · Zbl 0305.62031 · doi:10.1080/01621459.1974.10482962
[7] Harrell, HMISC: Harrell Miscellaneous (2010)
[8] Heitjan, Ignorability and coarse data, The Annals of Statistics 19 pp 2244– (1991) · Zbl 0745.62004 · doi:10.1214/aos/1176348396
[9] Hernan, Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men, Epidemiology 11 pp 561– (2000) · doi:10.1097/00001648-200009000-00012
[10] Kang, Demystifying double robustness: A comparison of alternative strategies for estimating a population mean from incomplete data (with discussion), Statistical Science 22 pp 523– (2007) · Zbl 1246.62073 · doi:10.1214/07-STS227
[11] Little, Statistical analysis with missing data (1987) · Zbl 0665.62004
[12] Moore, Covariate adjustment in randomized trials with binary outcomes: Targeted maximum likelihood estimation, Statistics in Medicine 28 pp 39– (2009) · doi:10.1002/sim.3445
[13] Neugebauer, DSA: Deletion/Substitution/Addition Algorithm (2010)
[14] Polley, SuperLearner: Super Learner Prediction (2010)
[15] Porter, The relative performance of targeted maximum likelihood estimators, The International Journal of Biostatistics 7 pp 1– (2011) · doi:10.2202/1557-4679.1308
[16] R Development Core Team 2012 R: A Language and Environment for Statistical Computing Vienna, Austria R Foundation for Statistical Computing
[17] Robins, A new approach to causal inference in mortality studies with sustained exposure periods-Application to control of the healthy worker survivor effect, Mathematical Modelling 7 pp 1393– (1986) · Zbl 0614.62136 · doi:10.1016/0270-0255(86)90088-6
[18] Robins, Marginal structural models versus structural nested models as tools for causal inference, in: Statistical Models in Epidemiology, the Environment, and Clinical Trials (Minneapolis, MN, 1997) pp 95– (2000a) · doi:10.1007/978-1-4612-1284-3_2
[19] Robins, Robust estimation in sequentially ignorable missing data and causal inference models, in: Proceedings of the American Statistical Association: Section on Bayesian Statistical Science pp 6– (2000b)
[20] Robins, Recovery of information and adjustment for dependent censoring using surrogate markers, AIDS Epidemiology (1992) · doi:10.1007/978-1-4757-1229-2_14
[21] Robins, Comment on the Bickel and Kwon article, ’Inference for semiparametric models: Some questions and an answer’, Statistica Sinica 11 pp 920– (2001)
[22] Robins, Comment on ”On Profile Likelihood” by S.A. Murphy and A.W. van der Vaart, Journal of the American Statistical Association-Theory and Methods 450 pp 431– (2000) · Zbl 0995.62033 · doi:10.2307/2669381
[23] Robins, Comment: Performance of double-robust estimators when ”inverse probability” weights are highly variable, Statistical Science 22 pp 544– (2007) · Zbl 1246.62076 · doi:10.1214/07-STS227D
[24] Scharfstein, Adjusting for non-ignorable drop-out using semiparametric nonresponse models, (with discussion and rejoinder), Journal of the American Statistical Association 94 pp 1096– (1999) · doi:10.1080/01621459.1999.10473862
[25] Sinisi, The Deletion/Substitution/Addition algorithm in loss function based estimation: Applications in genomics, Journal of Statistical Methods in Molecular Biology 3 pp 1– (2004) · Zbl 1166.62368
[26] Stitelman , O. M. Gruttola , V. D. van der Laan , M. J. 2011 A general implementation of TMLE for longitudinal data applied to causal inference in survival analysis University of California Berkeley
[27] Tan, Bounded, efficient and doubly robust estimation with inverse weighting, Biometrika 97 pp 661– (2010) · Zbl 1195.62037 · doi:10.1093/biomet/asq035
[28] van der Laan, The cross-validated adaptive epsilon-net estimator, Statistics and Decisions 24 pp 373– (2006) · Zbl 1111.62003
[29] van der Laan, Targeted minimum loss based estimation of causal effects of multiple time point interventions, The International Journal of Biostatistics 8 pp 1– (2012) · doi:10.1515/1557-4679.1370
[30] van der Laan, Super Learner, Statistical Applications in Genetics and Molecular Biology 6 pp 1– (2007) · Zbl 1166.62387 · doi:10.2202/1544-6115.1309
[31] van der Laan, Unified Methods for Censored Longitudinal Data and Causality (2003) · Zbl 1013.62034 · doi:10.1007/978-0-387-21700-0
[32] van der Laan, Targeted Learning: Prediction and Causal Inference for Observational and Experimental Data (2011) · Zbl 05911753 · doi:10.1007/978-1-4419-9782-1
[33] van der Laan, Targeted maximum likelihood learning, The International Journal of Biostatistics 2 pp 1– (2006) · doi:10.2202/1557-4679.1043
[34] Venables, Modern Applied Statistics with S (2002) · doi:10.1007/978-0-387-21706-2
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.