Evaluation of the Effect of a Continuous Treatment: A Machine Learning Approach with an Application to Treatment for Traumatic Brain Injury

Summary For a continuous treatment, the generalised propensity score (GPS) is defined as the conditional density of the treatment, given covariates. GPS adjustment may be implemented by including it as a covariate in an outcome regression. Here, the unbiased estimation of the dose–response function assumes correct specification of both the GPS and the outcome‐treatment relationship. This paper introduces a machine learning method, the ‘Super Learner’, to address model selection in this context. In the two‐stage estimation approach proposed, the Super Learner selects a GPS and then a dose–response function conditional on the GPS, as the convex combination of candidate prediction algorithms. We compare this approach with parametric implementations of the GPS and to regression methods. We contrast the methods in the Risk Adjustment in Neurocritical care cohort study, in which we estimate the marginal effects of increasing transfer time from emergency departments to specialised neuroscience centres, for patients with acute traumatic brain injury. With parametric models for the outcome, we find that dose–response curves differ according to choice of specification. With the Super Learner approach to both regression and the GPS, we find that transfer time does not have a statistically significant marginal effect on the outcomes. © 2015 The Authors. Health Economics Published by John Wiley & Sons Ltd.

[1]  S. Keleş,et al.  Statistical Applications in Genetics and Molecular Biology Asymptotic Optimality of Likelihood-Based Cross-Validation , 2011 .

[2]  Z Kadziola,et al.  Propensity Score Matching and Subclassification With Multi-Level Treatments. , 2014, Value in health : the journal of the International Society for Pharmacoeconomics and Outcomes Research.

[3]  P. Royston,et al.  Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling. , 1994 .

[4]  Alessandra Mattei,et al.  Nonparametric Estimators of Dose-Response Functions , 2011 .

[5]  Douglas G Altman,et al.  Dichotomizing continuous predictors in multiple regression: a bad idea , 2006, Statistics in medicine.

[6]  Alan R. Ellis,et al.  The role of prediction modeling in propensity score estimation: an evaluation of logistic regression, bCART, and the covariate-balancing propensity score. , 2014, American journal of epidemiology.

[7]  S. Rose.  Mortality risk score prediction in an elderly population using machine learning. , 2013, American journal of epidemiology.

[8]  Mark J van der Laan,et al.  Long-term consequences of the delay between virologic failure of highly active antiretroviral therapy and regimen modification , 2008, AIDS.

[9]  M. J. van der Laan,et al.  Mortality prediction in intensive care units with the Super ICU Learner Algorithm (SICULA): a population-based study. , 2015, The Lancet. Respiratory medicine.

[10]  M. Bullock,et al.  Surgical Management of Traumatic Brain Injury , 2011 .

[11]  Elizabeth A Stuart,et al.  Improving propensity score weighting using machine learning , 2010, Statistics in medicine.

[12]  D K Menon,et al.  Risk Adjustment In Neurocritical care (RAIN)--prospective validation of risk prediction models for adult patients with acute traumatic brain injury to use to evaluate the optimum location and comparative costs of neurocritical care: a cohort study. , 2013, Health technology assessment.

[13]  S. Vansteelandt,et al.  On regression adjustment for the propensity score , 2014, Statistics in medicine.

[14]  Zhong Zhao,et al.  Evaluating continuous training programmes by using the generalized propensity score , 2007 .

[15]  J. Robins,et al.  Marginal Structural Models and Causal Inference in Epidemiology , 2000, Epidemiology.

[16]  Stef van Buuren,et al.  MICE: Multivariate Imputation by Chained Equations in R , 2011 .

[17]  Peter C. Austin,et al.  Using Ensemble-Based Methods for Directly Estimating Causal Effects: An Investigation of Tree-Based G-Computation , 2012, Multivariate behavioral research.

[18]  Alessandra Mattei,et al.  A Stata Package for the Estimation of the Dose-response Function through Adjustment for the Generalized Propensity Score , 2008 .

[19]  Jamshid Ghajar,et al.  Guidelines for the Surgical Management of Traumatic Brain Injury Author Group , 2006 .

[20]  Iván Díaz,et al.  Targeted Data Adaptive Estimation of the Causal Dose–Response Curve , 2013 .

[21]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[22]  D. Rubin.  Causal Inference Using Potential Outcomes , 2005 .

[23]  Kosuke Imai,et al.  Causal Inference With General Treatment Regimes , 2004 .

[24]  Mark J van der Laan,et al.  An Application of Collaborative Targeted Maximum Likelihood Estimation in Causal Inference and Genomics , 2010, The international journal of biostatistics.

[25]  G. Imbens.  The Role of the Propensity Score in Estimating Dose-Response Functions , 1999 .

[26]  Lane F Burgette,et al.  A tutorial on propensity score estimation for multiple treatments using generalized boosted models , 2013, Statistics in medicine.

[27]  Romain Neugebauer,et al.  Targeted learning in real‐world comparative effectiveness research with time‐varying interventions , 2014, Statistics in medicine.

[28]  M. J. Laan,et al.  Targeted Learning of an Optimal Dynamic Treatment, and Statistical Inference for its Mean Outcome , 2014 .

[29]  Ekaterina Eliseeva,et al.  An Application Of Machine Learning Methods To The Derivation Of Exposure-Response Curves For Respiratory Outcomes , 2013 .

[30]  M. J. Laan,et al.  Targeted Learning: Causal Inference for Observational and Experimental Data , 2011 .

[31]  Elizabeth A Stuart,et al.  Matching methods for causal inference: A review and a look forward. , 2010, Statistical science : a review journal of the Institute of Mathematical Statistics.

[32]  Alfonso Flores-Lagunes,et al.  Estimating the Effects of Length of Exposure to a Training Program: The Case of Job Corps , 2007, SSRN Electronic Journal.

[33]  Keying Ye,et al.  Applied Bayesian Modeling and Causal Inference From Incomplete-Data Perspectives , 2005, Technometrics.

[34]  Mark J. van der Laan,et al.  Super Learner In Prediction , 2010 .

[35]  Andrew Gelman,et al.  Data Analysis Using Regression and Multilevel/Hierarchical Models , 2006 .

[36]  G. Imbens,et al.  The Propensity Score with Continuous Treatments , 2005 .

[37]  K. Imai,et al.  Covariate balancing propensity score , 2014 .

[38]  Juan Lu,et al.  Predicting Outcome after Traumatic Brain Injury: Development and International Validation of Prognostic Scores Based on Admission Characteristics , 2008, PLoS medicine.

[39]  P. Rosenbaum.  Sensitivity analysis for certain permutation inferences in matched observational studies , 1987 .

[40]  Ramon Diaz-Arrastia,et al.  Intracranial pressure monitoring in brain-injured patients is associated with worsening of survival. , 2008, The Journal of trauma.

[41]  K. Barlow.  Traumatic brain injury. , 2013, Handbook of clinical neurology.

[42]  M. J. van der Laan,et al.  Statistical Applications in Genetics and Molecular Biology Super Learner , 2010 .

[43]  M. J. van der Laan,et al.  Practice of Epidemiology Improving Propensity Score Estimators ’ Robustness to Model Misspecification Using Super Learner , 2015 .

[44]  Debashis Ghosh,et al.  A Boosting Algorithm for Estimating Generalized Propensity Scores with Continuous Treatments , 2015, Journal of causal inference.

[45]  Joseph Kang,et al.  Demystifying Double Robustness: A Comparison of Alternative Strategies for Estimating a Population Mean from Incomplete Data , 2007, 0804.2958.

[46]  G. Andrew,et al.  arm: Data Analysis Using Regression and Multilevel/Hierarchical Models , 2014 .

[47]  H Muskett,et al.  Development and validation of a risk model for identification of non-neutropenic, critically ill adult patients at high risk of invasive Candida infection: the Fungal Infection Risk Evaluation (FIRE) Study. , 2013, Health technology assessment.

[48]  Hilmar Schneider,et al.  Evaluating continuous training programmes by using the generalized propensity score , 2007 .

[49]  J. Sekhon,et al.  Evaluating treatment effectiveness under model misspecification: A comparison of targeted maximum likelihood estimation with bias-corrected matching , 2014, Statistical methods in medical research.

[50]  Kristin E. Porter,et al.  The Relative Performance of Targeted Maximum Likelihood Estimators , 2011, The international journal of biostatistics.

[51]  Alfonso Flores-Lagunes,et al.  Estimating the Effects of Length of Exposure to Instruction in a Training Program: The Case of Job Corps , 2012, Review of Economics and Statistics.

[52]  S. Dudoit,et al.  Unified Cross-Validation Methodology For Selection Among Estimators and a General Cross-Validated Adaptive Epsilon-Net Estimator: Finite Sample Oracle Inequalities and Examples , 2003 .