Search

While machine learning (ML) methods have received a lot of attention in recent years, these methods are primarily for prediction. Empirical researchers conducting policy evaluations are, on the other hand, pre-occupied with causal problems, trying to answer counterfactual questions: what would have happened in the absence of a policy? Because these counterfactuals can never be directly observed (described as the "fundamental problem of causal inference") prediction tools from the ML literature cannot be readily used for causal inference. In the last decade, major innovations have taken place incorporating supervised ML tools into estimators for causal parameters such as the average treatment effect (ATE). This holds the promise of attenuating model misspecification issues, and increasing of transparency in model selection. One particularly mature strand of the literature include approaches that incorporate supervised ML approaches in the estimation of the ATE of a binary treatment, under the \textit{unconfoundedness} and positivity assumptions (also known as exchangeability and overlap assumptions). This article reviews popular supervised machine learning algorithms, including the Super Learner. Then, some specific uses of machine learning for treatment effect estimation are introduced and illustrated, namely (1) to create balance among treated and control groups, (2) to estimate so-called nuisance models (e.g. the propensity score, or conditional expectations of the outcome) in semi-parametric estimators that target causal parameters (e.g. targeted maximum likelihood estimation or the double ML estimator), and (3) the use of machine learning for variable selection in situations with a high number of covariates., Comment: 40 pages 3 Figures

Health economic evaluations face the issues of non-compliance and missing data. Here, non-compliance is defined as non-adherence to a specific treatment, and occurs within randomised controlled trials (RCTs) when participants depart from their random assignment. Missing data arises if, for example, there is loss to follow-up, survey non-response, or the information available from routine data sources is incomplete. Appropriate statistical methods for handling non-compliance and missing data have been developed, but they have rarely been applied in health economics studies. Here, we illustrate the issues and outline some of the appropriate methods to handle these with an application to a health economic evaluation that uses data from an RCT. In an RCT the random assignment can be used as an instrument for treatment receipt, to obtain consistent estimates of the complier average causal effect, provided the underlying assumptions are met. Instrumental variable methods can accommodate essential features of the health economic context such as the correlation between individuals' costs and outcomes in cost-effectiveness studies. Methodological guidance for handling missing data encourages approaches such as multiple imputation or inverse probability weighting, that assume the data are Missing At Random, but also sensitivity analyses that recognise the data may be missing according to the true, unobserved values, that is, Missing Not at Random. Future studies should subject the assumptions behind methods for handling non-compliance and missing data to thorough sensitivity analyses. Modern machine learning methods can help reduce reliance on correct model specification. Further research is required to develop flexible methods for handling more complex forms of non-compliance and missing data., Comment: 41 pages

Non-adherence to assigned treatment is common in randomised controlled trials (RCTs). Recently, there has been an increased interest in estimating causal effects of treatment received, for example the so-called local average treatment effect (LATE). Instrumental variables (IV) methods can be used for identification, with estimation proceeding either via fully parametric mixture models or two-stage least squares (TSLS). TSLS is popular but can be problematic for binary outcomes where the estimand of interest is a causal odds ratio. Mixture models are rarely used in practice, perhaps because of their perceived complexity and need for specialist software. Here, we propose using multiple imputation (MI) to impute the latent compliance class appearing in the mixture models. Since such models include an interaction term between compliance class and randomised treatment, we use `substantive model compatible' MI (SMC MIC), which can also address other missing data, before fitting the mixture models via maximum likelihood to the MI datasets and combining results via Rubin's rules. We use simulations to compare the performance of SMC MIC to existing approaches and also illustrate the methods by re-analysing a RCT in UK primary health. We show that SMC MIC can be more efficient than full Bayesian estimation when auxiliary variables are incorporated, and is superior to two-stage methods, especially for binary outcomes., Comment: 24 pages, 3 Figures

Non-adherence to assigned treatment is a common issue in cluster randomised trials (CRTs). In these settings, the efficacy estimand may be also of interest. Many methodological contributions in recent years have advocated using instrumental variables to identify and estimate the local average treatment effect (LATE). However, the clustered nature of randomisation in CRTs adds to the complexity of such analyses. In this paper, we show that under certain assumptions, the LATE can be estimated via two-stage least squares (TSLS) using cluster-level summaries of outcomes and treatment received. Implementation needs to account for this, as well as the possible heteroscedasticity, to obtain valid inferences. We use simulations to assess the performance of TSLS of cluster-level summaries under cluster-level or individual-level non-adherence, with and without weighting and robust standard errors. We also explore the impact of adjusting for cluster-level covariates and of appropriate degrees of freedom correction for inference. We find that TSLS estimation using cluster-level summaries provides estimates with small to negligible bias and coverage close to nominal level, provided small sample degrees of freedom correction is used for inference, with appropriate use of robust standard errors. We illustrate the methods by re-analysing a CRT in UK primary health settings., Comment: 21 pages, 6 Figures

We consider the estimation of the average treatment effect in the treated as a function of baseline covariates, where there is a valid (conditional) instrument. We describe two doubly robust (DR) estimators: a locally efficient g-estimator, and a targeted minimum loss-based estimator (TMLE). These two DR estimators can be viewed as generalisations of the two-stage least squares (TSLS) method to semi-parametric models that make weaker assumptions. We exploit recent theoretical results that extend to the g-estimator the use of data-adaptive fits for the nuisance parameters. A simulation study is used to compare standard TSLS with the two DR estimators' finite-sample performance, (1) when fitted using parametric nuisance models, and (2) using data-adaptive nuisance fits, obtained from the Super Learner, an ensemble machine learning method. Data-adaptive DR estimators have lower bias and improved coverage, when compared to incorrectly specified parametric DR estimators and TSLS. When the parametric model for the treatment effect curve is correctly specified, the g-estimator outperforms all others, but when this model is misspecified, TMLE performs best, while TSLS can result in large biases and zero coverage. Finally, we illustrate the methods by reanalysing the COPERS (COping with persistent Pain, Effectiveness Research in Self-management) trial to make inference about the causal effect of treatment actually received, and the extent to which this is modified by depression at baseline., Comment: 24 pages, 3 figues

In Randomised Controlled Trials (RCT) with treatment non-compliance, instrumental variable approaches are used to estimate complier average causal effects. We extend these approaches to cost-effectiveness analyses, where methods need to recognise the correlation between cost and health outcomes. We propose a Bayesian full likelihood (BFL) approach, which jointly models the effects of random assignment on treatment received and the outcomes, and a three-stage least squares (3sls) method, which acknowledges the correlation between the endpoints, and the endogeneity of the treatment received. This investigation is motivated by the REFLUX study, which exemplifies the setting where compliance differs between the RCT and routine practice. A simulation is used to compare the methods performance. We find that failure to model the correlation between the outcomes and treatment received correctly can result in poor CI coverage and biased estimates. By contrast, BFL and 3sls methods provide unbiased estimates with good coverage., Comment: 40 pages, 2 figures, includes R and Stata code to fit the models