{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T01:36:07Z","timestamp":1790818567757,"version":"4.1.0"},"reference-count":40,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2021,1,14]],"date-time":"2021-01-14T00:00:00Z","timestamp":1610582400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2021,1,14]],"date-time":"2021-01-14T00:00:00Z","timestamp":1610582400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100001665","name":"Agence Nationale de la Recherche","doi-asserted-by":"publisher","award":["ANR-16-LCV1-0003-01"],"award-info":[{"award-number":["ANR-16-LCV1-0003-01"]}],"id":[{"id":"10.13039\/501100001665","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001665","name":"Agence Nationale de la Recherche","doi-asserted-by":"publisher","award":["ANR-16-LCV1-0003-01"],"award-info":[{"award-number":["ANR-16-LCV1-0003-01"]}],"id":[{"id":"10.13039\/501100001665","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001665","name":"Agence Nationale de la Recherche","doi-asserted-by":"publisher","award":["ANR-16-LCV1-0003-01"],"award-info":[{"award-number":["ANR-16-LCV1-0003-01"]}],"id":[{"id":"10.13039\/501100001665","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title><jats:p>In clinical research, there is a growing interest in the use of propensity score-based methods to estimate causal effects. G-computation is an alternative because of its high statistical power. Machine learning is also increasingly used because of its possible robustness to model misspecification. In this paper, we aimed to propose an approach that combines machine learning and G-computation when both the outcome and the exposure status are binary and is able to deal with small samples. We evaluated the performances of several methods, including penalized logistic regressions, a neural network, a support vector machine, boosted classification and regression trees, and a super learner through simulations. We proposed six different scenarios characterised by various sample sizes, numbers of covariates and relationships between covariates, exposure statuses, and outcomes. We have also illustrated the application of these methods, in which they were used to estimate the efficacy of barbiturates prescribed during the first 24\u00a0h of an episode of intracranial hypertension. In the context of GC, for estimating the individual outcome probabilities in two counterfactual worlds, we reported that the super learner tended to outperform the other approaches in terms of both bias and variance, especially for small sample sizes. The support vector machine performed well, but its mean bias was slightly higher than that of the super learner. In the investigated scenarios, G-computation associated with the super learner was a performant method for drawing causal inferences, even from small sample sizes.<\/jats:p>","DOI":"10.1038\/s41598-021-81110-0","type":"journal-article","created":{"date-parts":[[2021,1,14]],"date-time":"2021-01-14T11:11:08Z","timestamp":1610622668000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":23,"title":["G-computation and machine learning for estimating the causal effects of binary exposure statuses on binary outcomes"],"prefix":"10.1038","volume":"11","author":[{"given":"Florent","family":"Le Borgne","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Arthur","family":"Chatton","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Maxime","family":"L\u00e9ger","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"R\u00e9mi","family":"Lenain","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yohann","family":"Foucher","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2021,1,14]]},"reference":[{"key":"81110_CR1","doi-asserted-by":"publisher","unstructured":"Blakely, T., Lynch, J., Simons, K., Bentley, R. & Rose, S. Reflection on modern methods: when worlds collide: prediction, machine learning and causal inference. Int. J. Epidemiol. https:\/\/doi.org\/10.1093\/ije\/dyz132.","DOI":"10.1093\/ije\/dyz132"},{"key":"81110_CR2","doi-asserted-by":"publisher","first-page":"41","DOI":"10.1093\/biomet\/70.1.41","volume":"70","author":"PR Rosenbaum","year":"1983","unstructured":"Rosenbaum, P. R. & Rubin, D. B. The central role of the propensity score in observational studies for causal effects. Biometrika 70, 41\u201355 (1983).","journal-title":"Biometrika"},{"key":"81110_CR3","doi-asserted-by":"publisher","first-page":"826","DOI":"10.1016\/j.jclinepi.2009.11.020","volume":"63","author":"D Westreich","year":"2010","unstructured":"Westreich, D., Lessler, J. & Funk, M. J. Propensity score estimation: neural networks, support vector machines, decision trees (CART), and meta-classifiers as alternatives to logistic regression. J. Clin. Epidemiol. 63, 826\u2013833 (2010).","journal-title":"J. Clin. Epidemiol."},{"key":"81110_CR4","doi-asserted-by":"publisher","first-page":"1393","DOI":"10.1016\/0270-0255(86)90088-6","volume":"7","author":"JM Robins","year":"1986","unstructured":"Robins, J. M. A new approach to causal inference in mortality studies with a sustained exposure period\u2014application to control of the healthy worker survivor effect. Math. Model. 7, 1393\u20131512 (1986).","journal-title":"Math. Model."},{"key":"81110_CR5","doi-asserted-by":"publisher","first-page":"731","DOI":"10.1093\/aje\/kwq472","volume":"173","author":"JM Snowden","year":"2011","unstructured":"Snowden, J. M., Rose, S. & Mortimer, K. M. Implementation of G-computation on a simulated data set: demonstration of a causal inference technique. Am. J. Epidemiol. 173, 731\u2013738 (2011).","journal-title":"Am. J. Epidemiol."},{"key":"81110_CR6","doi-asserted-by":"publisher","first-page":"546","DOI":"10.1002\/pds.1555","volume":"17","author":"S Setoguchi","year":"2008","unstructured":"Setoguchi, S., Schneeweiss, S., Brookhart, M. A., Glynn, R. J. & Cook, E. F. Evaluating uses of data mining techniques in propensity score estimation: a simulation study. Pharmacoepidemiol. Drug Saf. 17, 546\u2013555 (2008).","journal-title":"Pharmacoepidemiol. Drug Saf."},{"key":"81110_CR7","doi-asserted-by":"publisher","first-page":"337","DOI":"10.1002\/sim.3782","volume":"29","author":"BK Lee","year":"2010","unstructured":"Lee, B. K., Lessler, J. & Stuart, E. A. Improving propensity score weighting using machine learning. Stat. Med. 29, 337\u2013346 (2010).","journal-title":"Stat. Med."},{"key":"81110_CR8","doi-asserted-by":"publisher","first-page":"106","DOI":"10.1002\/sim.6322","volume":"34","author":"S Gruber","year":"2015","unstructured":"Gruber, S., Logan, R. W., Jarr\u00edn, I., Monge, S. & Hern\u00e1n, M. A. Ensemble learning of inverse probability weights for marginal structural modeling in large observational datasets. Stat. Med. 34, 106\u2013117 (2015).","journal-title":"Stat. Med."},{"key":"81110_CR9","doi-asserted-by":"publisher","first-page":"108","DOI":"10.1093\/aje\/kwu253","volume":"181","author":"R Pirracchio","year":"2015","unstructured":"Pirracchio, R., Petersen, M. L. & van der Laan, M. Improving propensity score estimators\u2019 robustness to model misspecification using super learner. Am. J. Epidemiol. 181, 108\u2013119 (2015).","journal-title":"Am. J. Epidemiol."},{"key":"81110_CR10","doi-asserted-by":"crossref","first-page":"1049","DOI":"10.1002\/bimj.201800132","volume":"61","author":"M Cannas","year":"2019","unstructured":"Cannas, M. & Arpino, B. A comparison of machine learning algorithms and covariate balance measures for propensity score matching and weighting. Biom. J. 61, 1049\u20131072 (2019).","journal-title":"Biom. J."},{"key":"81110_CR11","doi-asserted-by":"crossref","unstructured":"Chatton, A. et al. G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study. Sci. Rep. (in press) (2020).","DOI":"10.1038\/s41598-020-65917-x"},{"key":"81110_CR12","doi-asserted-by":"publisher","first-page":"S91","DOI":"10.1016\/j.jclinepi.2013.02.017","volume":"66","author":"SD Lendle","year":"2013","unstructured":"Lendle, S. D., Fireman, B. & van der Laan, M. J. Targeted maximum likelihood estimation in safety analysis. J. Clin. Epidemiol. 66, S91-98 (2013).","journal-title":"J. Clin. Epidemiol."},{"key":"81110_CR13","doi-asserted-by":"publisher","first-page":"23222","DOI":"10.1038\/srep23222","volume":"6","author":"KE Colson","year":"2016","unstructured":"Colson, K. E. et al. Optimizing matching and analysis combinations for estimating causal effects. Sci. Rep. 6, 23222 (2016).","journal-title":"Sci. Rep."},{"key":"81110_CR14","doi-asserted-by":"publisher","first-page":"523","DOI":"10.1214\/07-STS227","volume":"22","author":"JDY Kang","year":"2007","unstructured":"Kang, J. D. Y. & Schafer, J. L. Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data. Stat. Sci. 22, 523\u2013539 (2007).","journal-title":"Stat. Sci."},{"key":"81110_CR15","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1080\/00273171.2012.640600","volume":"47","author":"PC Austin","year":"2012","unstructured":"Austin, P. C. Using ensemble-based methods for directly estimating causal effects: an investigation of tree-based G-computation. Multivar. Behav. Res. 47, 115\u2013135 (2012).","journal-title":"Multivar. Behav. Res."},{"key":"81110_CR16","doi-asserted-by":"publisher","first-page":"688","DOI":"10.1037\/h0037350","volume":"66","author":"DB Rubin","year":"1974","unstructured":"Rubin, D. B. Estimating causal effects of treatments in randomized and nonrandomized studies. J. Educ. Psychol. 66, 688\u2013701 (1974).","journal-title":"J. Educ. Psychol."},{"key":"81110_CR17","doi-asserted-by":"publisher","first-page":"265","DOI":"10.1136\/jech.2002.006361","volume":"58","author":"MA Hern\u00e1n","year":"2004","unstructured":"Hern\u00e1n, M. A. A definition of causal effect for epidemiological research. J. Epidemiol. Commun. Health 58, 265\u2013271 (2004).","journal-title":"J. Epidemiol. Commun. Health"},{"key":"81110_CR18","doi-asserted-by":"publisher","unstructured":"Lin, S.-H. & Ikram, M. A. On the relationship of machine learning with causal inference. Eur. J. Epidemiol. https:\/\/doi.org\/10.1007\/s10654-019-00564-9 (2019).","DOI":"10.1007\/s10654-019-00564-9"},{"key":"81110_CR19","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1007\/s10654-019-00494-6","volume":"34","author":"TJ VanderWeele","year":"2019","unstructured":"VanderWeele, T. J. Principles of confounder selection. Eur. J. Epidemiol. 34, 211\u2013219 (2019).","journal-title":"Eur. J. Epidemiol."},{"key":"81110_CR20","doi-asserted-by":"publisher","first-page":"1213","DOI":"10.1093\/aje\/kwr364","volume":"174","author":"JA Myers","year":"2011","unstructured":"Myers, J. A. et al. Effects of adjusting for instrumental variables on bias and precision of effect estimates. Am. J. Epidemiol. 174, 1213\u20131222 (2011).","journal-title":"Am. J. Epidemiol."},{"key":"81110_CR21","doi-asserted-by":"publisher","first-page":"1149","DOI":"10.1093\/aje\/kwj149","volume":"163","author":"MA Brookhart","year":"2006","unstructured":"Brookhart, M. A. et al. Variable selection for propensity score models. Am. J. Epidemiol. 163, 1149\u20131156 (2006).","journal-title":"Am. J. Epidemiol."},{"key":"81110_CR22","doi-asserted-by":"publisher","first-page":"459","DOI":"10.1007\/s10654-018-0390-z","volume":"33","author":"AI Naimi","year":"2018","unstructured":"Naimi, A. I. & Balzer, L. B. Stacked generalization: an introduction to super learning. Eur. J. Epidemiol. 33, 459\u2013464 (2018).","journal-title":"Eur. J. Epidemiol."},{"key":"81110_CR23","doi-asserted-by":"publisher","first-page":"437","DOI":"10.1007\/s10654-018-0405-9","volume":"33","author":"AP Keil","year":"2018","unstructured":"Keil, A. P. & Edwards, J. K. You are smarter than you think: (super) machine learning in context. Eur. J. Epidemiol. 33, 437\u2013440 (2018).","journal-title":"Eur. J. Epidemiol."},{"key":"81110_CR24","doi-asserted-by":"publisher","first-page":"1406","DOI":"10.1111\/j.1541-0420.2011.01619.x","volume":"67","author":"TJ VanderWeele","year":"2011","unstructured":"VanderWeele, T. J. & Shpitser, I. A new criterion for confounder selection. Biometrics 67, 1406\u20131413 (2011).","journal-title":"Biometrics"},{"key":"81110_CR25","doi-asserted-by":"publisher","first-page":"3629","DOI":"10.1002\/sim.3200","volume":"27","author":"G Lefebvre","year":"2008","unstructured":"Lefebvre, G., Delaney, J. A. C. & Platt, R. W. Impact of mis-specification of the treatment model on estimates from a marginal structural model. Stat. Med. 27, 3629\u20133642 (2008).","journal-title":"Stat. Med."},{"key":"81110_CR26","doi-asserted-by":"publisher","first-page":"471","DOI":"10.1080\/00273171.2015.1036965","volume":"50","author":"DM McNeish","year":"2015","unstructured":"McNeish, D. M. Using lasso for predictor selection and to assuage overfitting: a method long overlooked in behavioral sciences. Multivar. Behav. Res. 50, 471\u2013484 (2015).","journal-title":"Multivar. Behav. Res."},{"key":"81110_CR27","doi-asserted-by":"publisher","DOI":"10.1093\/aje\/kwz189","author":"Q Bi","year":"2019","unstructured":"Bi, Q., Goodman, K. E., Kaminsky, J. & Lessler, J. What is machine learning? A primer for the epidemiologist. Am. J. Epidemiol. https:\/\/doi.org\/10.1093\/aje\/kwz189 (2019).","journal-title":"Am. J. Epidemiol."},{"key":"81110_CR28","doi-asserted-by":"crossref","unstructured":"van der Laan, M. J., Polley, E. C. & Hubbard, A. E. Super learner. Stat. Appl. Genet. Mol. Biol. 6, Article25 (2007).","DOI":"10.2202\/1544-6115.1309"},{"key":"81110_CR29","doi-asserted-by":"publisher","first-page":"991","DOI":"10.1080\/01621459.2013.823775","volume":"109","author":"B Efron","year":"2014","unstructured":"Efron, B. Estimation and accuracy after model selection. J. Am. Stat. Assoc. 109, 991\u20131007 (2014).","journal-title":"J. Am. Stat. Assoc."},{"key":"81110_CR30","doi-asserted-by":"publisher","first-page":"1768","DOI":"10.1093\/bioinformatics\/btm232","volume":"23","author":"M Schumacher","year":"2007","unstructured":"Schumacher, M., Binder, H. & Gerds, T. Assessment of survival prediction models based on microarray data. Bioinformatics 23, 1768\u20131774 (2007).","journal-title":"Bioinformatics"},{"key":"81110_CR31","doi-asserted-by":"crossref","unstructured":"Foucher, Y. & Danger, R. Time dependent ROC curves for the estimation of true prognostic capacity of microarray data. Stat. Appl. Genet. Mol. Biol. 11, Article 1 (2012).","DOI":"10.1515\/1544-6115.1815"},{"key":"81110_CR32","doi-asserted-by":"publisher","first-page":"222","DOI":"10.1002\/pst.537","volume":"11","author":"E Gayat","year":"2012","unstructured":"Gayat, E., Resche-Rigon, M., Mary, J.-Y. & Porcher, R. Propensity score applied to survival data analysis through proportional hazards models: a Monte Carlo study. Pharm. Stat. 11, 222\u2013229 (2012).","journal-title":"Pharm. Stat."},{"key":"81110_CR33","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1080\/09332480.2019.1579578","volume":"32","author":"MA Hern\u00e1n","year":"2019","unstructured":"Hern\u00e1n, M. A., Hsu, J. & Healy, B. A second chance to get causal inference right: a classification of data science tasks. Chance 32, 42\u201349 (2019).","journal-title":"Chance"},{"key":"81110_CR34","doi-asserted-by":"publisher","first-page":"S8","DOI":"10.1038\/ijo.2008.82","volume":"32","author":"MA Hern\u00e1n","year":"2008","unstructured":"Hern\u00e1n, M. A. & Taubman, S. L. Does obesity shorten life? The importance of well-defined interventions to answer causal questions. Int. J. Obes. 32, S8\u2013S14 (2008).","journal-title":"Int. J. Obes."},{"key":"81110_CR35","first-page":"353","volume":"21","author":"I D\u00edaz","year":"2020","unstructured":"D\u00edaz, I. Machine learning in the estimation of causal effects: targeted minimum loss-based estimation and double\/debiased machine learning. Biostatistics 21, 353\u2013358 (2020).","journal-title":"Biostatistics"},{"key":"81110_CR36","doi-asserted-by":"publisher","first-page":"802","DOI":"10.1214\/12-AOS1077","volume":"41","author":"R Berk","year":"2013","unstructured":"Berk, R., Brown, L., Buja, A., Zhang, K. & Zhao, L. Valid post-selection inference. Ann. Stat. 41, 802\u2013837 (2013).","journal-title":"Ann. Stat."},{"key":"81110_CR37","doi-asserted-by":"publisher","first-page":"96","DOI":"10.1097\/EDE.0000000000000762","volume":"29","author":"R Wyss","year":"2018","unstructured":"Wyss, R. et al. Using super learner prediction modeling to improve high-dimensional propensity score estimation. Epidemiology 29, 96\u2013106 (2018).","journal-title":"Epidemiology"},{"key":"81110_CR38","doi-asserted-by":"publisher","first-page":"191","DOI":"10.1097\/EDE.0000000000000787","volume":"29","author":"ME Karim","year":"2018","unstructured":"Karim, M. E., Pang, M. & Platt, R. W. Can we train machine learning methods to outperform the high-dimensional propensity score algorithm?. Epidemiology 29, 191\u2013198 (2018).","journal-title":"Epidemiology"},{"key":"81110_CR39","doi-asserted-by":"publisher","first-page":"891","DOI":"10.1093\/aje\/kwx385","volume":"187","author":"AP Keil","year":"2018","unstructured":"Keil, A. P. et al. Resolving an apparent paradox in doubly robust estimators. Am. J. Epidemiol. 187, 891\u2013892 (2018).","journal-title":"Am. J. Epidemiol."},{"key":"81110_CR40","doi-asserted-by":"publisher","first-page":"1938","DOI":"10.1177\/0962280213507034","volume":"25","author":"R Pirracchio","year":"2016","unstructured":"Pirracchio, R. et al. Propensity score estimators for the average treatment effect and the average treatment effect on the treated may yield very different estimates. Stat. Methods Med. Res. 25, 1938\u20131954 (2016).","journal-title":"Stat. Methods Med. Res."}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-021-81110-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-021-81110-0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-021-81110-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,10,17]],"date-time":"2023-10-17T17:33:15Z","timestamp":1697563995000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-021-81110-0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,1,14]]},"references-count":40,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2021,12]]}},"alternative-id":["81110"],"URL":"https:\/\/doi.org\/10.1038\/s41598-021-81110-0","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,1,14]]},"assertion":[{"value":"6 July 2020","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"24 December 2020","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 January 2021","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"1435"}}