drtmle: Doubly-Robust Inference
in RIn many fields researchers are interested in assessing the population-level effect of a particular treatment on an outcome of interest. The treatment might correspond to a drug, a potentially harmful exposure exposure, or a policy intervention. Often, the treatment may not be randomized due to ethical or logistical reasons. This has sparked great interest in developing statistical methodology to estimate population-level effects from observational data. Estimation of these effects often requires accounting for putative confounding factors, that is, factors related to whether a participant receives treatment and to the participant’s outcome. In many settings, researchers measure many potential confounders, for example using administrative databases or genetic sequencing technology. Due to the curse-of-dimensionality, common statistical estimation techniques are not feasible for such high-dimensional data without restrictive modeling assumptions. When these assumptions are violated, estimates of the population-level effect of a treatment may have large bias.
This has sparked an interest in using adaptive estimation techniques from the machine learning literature to control for high-dimensional confounders when estimating treatment effects. However, facilitating proper inference (e.g., confidence intervals and hypothesis tests) about estimates of effects based on adaptive estimators is challenging. In particular, standard causal inference techniques, including both G-computation (GCOMP) estimators (Robins 1986) and inverse probability of treatment weighted (IPTW) estimators Robins, Hernan, and Brumback (2000), fail to yield valid inference for common estimators of effects. More complex proposals have been developed that are capable of utilizing adaptive methods to control for confounding while still yielding valid inference. These techniques include augmented inverse probability of treatment estimation (AIPTW) (Robins, Rotnitzky, and Zhao 1994) and targeted minimum loss-based estimation (TMLE) (van der Laan and Rubin 2006). Under assumptions, these estimators have limiting normal distributions with estimable variance. The asymptotic variance estimates then serve as the basis for constructing confidence intervals and hypothesis tests.
As an intermediate step, the AIPTW and TMLE estimators require estimation of two key nuisance parameters: the probability of treatment as a function of confounders (the propensity score, hereafter referred to as the PS) and the average outcome as a function of confounders and treatment (the outcome regression, OR). In addition to their desirable inferential properties, AIPTW and TMLE estimators also enjoy the property of double robustness. That is, the estimators are consistent for the population-level effect of interest if either of the PS or OR is consistently estimated. This property gives the AIPTW and TMLE estimators a natural appeal: inconsistency in the estimation of one nuisance parameter may be mitigated by the consistent estimation of the other. As such, these estimators have increased in popularity in recent years. However, recent work has shown that the double robustness property does not necessarily extend to inferential properties about these estimators when adaptive estimators of the PS and OR are used Benkeser et al. (2017). Instead, valid inference requires consistent estimation of both the PS and the OR. van der Laan (2014) proposed new TMLE estimators have an asymptotic normal sampling distribution with estimable variance under consistent estimation of either the PS or the OR, paving the way for inference that is doubly-robust. Benkeser et al. (2017) proposed modifications of these estimators with similar robustness properties and illustrated their practical performance.
The two proposed procedures that yield doubly-robust inference are
quite involved, notably involving iterative estimation of additional,
complex nuisance parameters. The drtmle package was
motivated by the need for a user-friendly tool to implement these
methods. The package implements the TMLE estimators of population-level
effects proposed in van der Laan (2014) and Benkeser et al. (2017), as
well as several extensions including cross-validated targeted minimum
loss-based estimators (CVTMLE). Additionally, drtmle
implements an IPTW estimator that overcomes shortcomings of existing
IPTW implementations with respect to statistical inference. In
particular, the estimator allows an adaptive estimator of the PS to be
used in construction of the IPTW estimator, while yielding valid
statistical inference about population-level effects. Both the TMLE and
IPTW estimators implemented in the drtmle package are
capable of estimating population-level effects for discrete-valued
treatments and the package provides convenient utilities for
constructing confidence intervals and hypothesis tests about contrasts
of the mean outcome under different levels of treatment. This document
explains some of the key concepts underlying doubly-robust inference and
illustrates usage of the drtmle package.
Suppose the observed data consist of n independent and identically distributed copies of the random variable O = (W, A, Y) ∼ P0. Here, W is a vector of baseline covariates, A ∈ {0, 1} is a binary (for the sake of simplicity) treatment, Y is an outcome, and P0 is the true data-generating distribution. The parameters of interest are erDef} \end{equation} where Q̄0(a, w) = E0(Y ∣ A = a, W = w) is the so-called outcome regression and QW(w) = Pr0(W ≤ w) is the distribution function of the covariates. The value ψ0(a) represents the marginal (i.e., population-level)treatment-specific average outcome, hereafter referred to as the marginal mean under treatment A = a. The marginal effect of the treatment on the outcome can be assessed by comparing marginal means under different treatment levels. For example, the average treatment effect is often defined as ψ0 := ψ0(1) − ψ0(0). Under additional assumptions, these parameters have richer interpretations as mean counterfactual outcomes under the different treatments and contrasts between these means define causal effects.
An estimator of ψ0(a) may be motivated directly by (1). Suppose we have available an estimate of the OR, Q̄n, and an estimate of the distribution function of baseline covariates, QW, n. An estimator of ψ0(a) may be obtained by plugging these into (1), ψn(a) = ∫Q̄0(a, w)dQW, n(w). Often the empirical distribution function of covariates is used so that this estimator can be equivalently written as This estimator is commonly referred to as the G-computation (GCOMP) estimator. Inference for GCOMP estimators based on parametric OR estimators may be facilitated through the nonparametric bootstrap. However, if adaptive approaches are used to estimate the OR, inference may not be available without further modification to the estimator.
An alternative estimator of the treatment-specific population-level mean may be obtained by noting that (1) can be equivalently written as We use g0(a ∣ W) := Pr0(A = a ∣ W) to denote the propensity score. Suppose we had available gn, an estimate of the PS. Equation (2) motivates the estimator which is referred to as the inverse probability of treatment (IPTW) estimator. Inference for IPTW estimators based on parametric PS estimators may be facilitated through the nonparametric bootstrap. If adaptive approaches are used to estimate the PS, inference may not be available without further modification of the propensity score estimator, as we discuss below.
The GCOMP and IPTW estimators suffer from two important shortcomings. The first is a lack of robustness. That is, validity of GCOMP estimators relies on consistent estimation of the OR, while validity of the IPTW relies on consistent estimation of the PS. In practice, we may not know which of the OR or PS is simpler to consistently estimate and thus may not know which of the two estimators to prefer. Furthermore, if adaptive estimators (e.g., based on machine learning) are used to estimate the OR or PS, then the GCOMP and IPTW estimators lack a basis for statistical inference.
This motivates a new class of estimators that combine both the GCOMP and IPTW estimators. One example is the augmented IPTW (AIPTW) estimator
which adds an augmentation term to the IPTW estimator. Equivalently, the estimator can be viewed as adding an augmentation to the GCOMP estimator The AIPTW estimator overcomes the two limitations of the GCOMP and IPTW estimators. The estimator exhibits double-robustness in that it is consistent for ψ0(a) if either Q̄n is consistent for the true OR or gn is consistent for the true PS. Furthermore, if both the OR and PS are consistently estimated and regularity conditions hold, one can establish a limiting normal distribution for the AIPTW estimator. Simple estimators of the variance of this limiting distribution can be derived that may be used to construct Wald-style confidence intervals and hypothesis tests.
A more recent proposal to improve robustness and inferential properties of the GCOMP and IPTW estimators utilizes targeted minimum loss-based estimation (TMLE). TMLE is a general methodology that may be used to modify an estimator of a regression function in such a way so as to ensure that the modified estimator satisfies user-selected equations. In the present problem, one can show that if an estimator of the OR Q̄n* satisfies then the GCOMP estimator based on Q̄n* will have the same asymptotic properties as the AIPTW estimator. That is, the estimator will be doubly-robust and, provided both the OR and PS are consistently estimated and regularity conditions are satisfied, will have a normal limiting distribution. TMLE provides a means of mapping an initial estimator of the OR into an estimator that satisfies (4). In addition to possessing the desirable asymptotic properties, TMLE estimators have been shown to outperform AIPTW estimators in finite samples (Porter et al. 2011). We refer readers to van der Laan and Rose (2011) for more about TMLE (van der Laan and Rose 2011)
van der Laan (2014) demonstrated that the double robustness property of AIPTW and TMLE estimators does not extend to their normal limiting distribution if adaptive estimators of the OR and PS are used to construct the estimators. However, in settings with high-dimensional and continuous-valued covariates, adaptive estimation is likely necessary in order to obtain a consistent estimate of either the OR or PS, and thus may be necessary to obtain minimally biased estimates of the marginal means of interest. van der Laan (2014) derived modified TMLE estimators that are doubly-robust not only with respect to consistency, but also with respect to asymptotic normality. Furthermore, he provided closed-form, doubly-robust variance estimators that may be used to construct doubly-robust confidence intervals and hypothesis tests. Benkeser et al. (2017) proposed simplifications of the van der Laan (2014) estimators and demonstrated the practical performance of estimators via simulation.
The key to the theory underlying these results is finding estimators of the OR and PS that simultaneously satisfy equation (4) in addition to two additional equations. These equations involve additional regression parameters that we refer to as the reduced outcome regression (R-OR) and the reduced propensity score (R-PS), defined respectively as In words, the R-OR is the regression of the residual from the outcome regression onto the estimated PS amongst observations with A = a, while the R-PS is the propensity for treatment A = a given the estimated OR and PS. We note that a key feature of these reduced-dimension regressions is that they are low-dimensional regardless of the dimension of W and they do not depend on the true OR nor PS. Thus, these regressions may be consistently estimated at reasonably fast rates irrespective of the dimension of W and irrespective of whether the OR or PS are consistently estimated. Based on these reduced-dimension regressions, van der Laan (2014) showed that if estimates of the OR Q̄n*, PS gn*, R-OR Q̄r, n*, and R-PS gr, n* satisfied then the GCOMP estimator based on Q̄n* would be doubly-robust not only with respect to consistency, but also with respect to its limiting distribution. The author additionally proposed a TMLE algorithm to map initial estimates of the OR, PS, R-OR, and R-PS into estimates that satisfy these key equations.
Benkeser et al. (2017) demonstrated alternative equations that can be satisfied to achieve this form of double-robustness. In particular, they formulated an alternative version of the R-PS,
which we refer to as R-PS1. They also introduced an additional regression of a weighted residual from the PS on the OR,
which we refer to as R-PS2. The key feature of R-PS1 and R-PS2 is that they are both univariate regressions, while R-PS is a bivariate regression. Benkeser and colleagues suggested that this may make R-PS1 and R-PS2 easier to estimate consistently than R-PS. Further, they showed that if estimators or the OR, PS, and R-OR satisfy equations (5)-(6), and if additionally, estimators of the OR, R-PS-1 gr, n, 1*, and R-PS-2 gr, n, 2* satisfy $$ \frac{1}{n} \sum\limits_{i=1}^n \frac{I(A_i = a)}{g_{r,n,1}^*(a \mid W_i)} g_{r,n,2}^*(a \mid W_i) \ \{Y_i - \bar{Q}_n^*(a,W_i)\} = 0 \ , $$ then the GCOMP estimator based on this Q̄n* also enjoys a doubly-robust limiting distribution. Benkeser et al. (2017) provided a TMLE algorithm that produced such estimates and provided closed-form, doubly-robust variance estimates.
In the previous section, we introduced AIPTW and TMLE as two methods for constructing doubly-robust (with respect to consistency) estimators of our parameter of interest. Theoretically, these two frameworks are closely related and practically the two estimators are generally seen as competitors for estimating marginal means. In equation (3), we were able to generate a doubly-robust estimator by adding a term to the GCOMP estimator. This term is exactly the left-hand-side of the key equation (4), the key equation that is solved by the standard TMLE. This is no coincidence; this term is crucial in studying the asymptotic properties of the GCOMP, AIPTW, and TMLE estimators (for discussion, see Benkeser et al. (2017) Section 2.2). Recall that a TMLE estimator achieves doubly-robust consistency by satisfying equation (4), while the AIPTW achieves doubly-robust consistency by combining the GCOMP estimator with the left-hand-side of this term. Now, we have argued that a TMLE estimator may achieve doubly-robust limiting behavior if it additionally satisfies equations (6) and (7). It is thus sensible to wonder whether an AIPTW-style estimator could be constructed by combining the left-hand-side of these two additional equations with the AIPTW estimator. Benkeser and colleagues showed that, surprisingly, this AIPTW-style estimator does not achieve the same doubly-robust properties as the TMLE estimator. Thus, while AIPTW and TMLE are generally competitors for producing estimators doubly-robust with respect to consistency, for doubly-robust inference, TMLE is preferred.
In recent work, we have proposed estimators of the average treatment effect to be used in settings where this effect is weakly identifiable, as evidenced by small estimated propensity scores. In these settings doubly robust estimators may exhibit non-robust behavior in small samples. The proposed solution was to target an alternative quantity in the propensity estimation step. If we are interested in estimating ψ0(a) for a = 0, 1, we can estimate Pr0(A = a ∣ Q̄0(1, W), Q̄0(0, W)), which describes the conditional probability of receiving treatment A = a given the value of the two outcome regressions. For this reason, we call this quantity an outcome-adaptive propensity score. We have shown that all of the usual asymptotics for TMLE carry through when an estimator of this quantity is substituted for the true propensity score. The resultant TMLE is super-efficient meaning that it will generally have greater asymptotic efficiency than a standard TMLE. Simulations indicate that this approach also delivers more stable behavior in settings where estimated propensities become very small. However, the estimator is not doubly robust – the asymptotics rely on the outcome regression being consistently estimated at n−1/4 rate. In this sense, this estimator trades off robustness in the name of achieving greater stability and efficiency. See our paper for further discussion.
The main function of the package is the eponymous drtmle
function. This function estimates the treatment-specific marginal mean
for user-specified levels of a discrete-valued treatment and computes a
doubly-robust covariance matrix for these estimates. The
drtmle function implements either the estimators of van der
Laan (2014) (if reduction = "bivariate") or the improved
estimators of Benkeser et al. (2017)
(reduction = "univariate"). The package includes utility
functions for combining these estimates into estimates of average
treatment effects, as well as other contrasts, as discussed below.
The main data arguments for drtmle are
W, A, Y, which, as above, represent the covariates, a
discrete-valued treatment, and an outcome, respectively. Missing data
are allowed in A and Y; this is discussed
further in a section below. Beyond the data arguments, it is necessary
to specify how to estimate each of the OR, PS, R-OR, and R-PS (or R-PS1,
R-PS2). The package provides flexibility in how these regression
parameters may be estimated. Once the user has specified the data
arguments and how to estimate each regression parameter, the function
proceeds as follows: 1. estimate the OR via the internal
estimateQ function; 2. estimate the PS via the internal
estimateG function; 3. estimate reduced-dimension
regression via estimateQrn and estimategrn
functions; 4. iteratively modify initial estimates of the OR and PS via
the fluctuateQ and fluctuateG functions until
equations (5)-(7) are approximately solved; 5. compute the TMLE
estimates and covariance estimates based on the final OR. We refer
interested readers to Benkeser et al. (2017) for a theoretical
derivation of how the iterative modification of initial OR and PS is
performed.
In addition to returning doubly-robust parameter and covariance
estimates, the drtmle function returns estimates of the
GCOMP, AIPTW, standard TMLE, and the modified AIPTW estimator discussed
in the remark above. While doubly-robust inference based on these
estimators is not available, the estimators are provided for comparison.
In the following subsections, we look at various options for making
calls to drtmle.
While the estimators computed by drtmle are consistent
and asymptotically normal under inconsistent estimation of either the OR
or PS, it is nevertheless advisable to strive for consistent estimation
of both the OR and PS. If both of the regression parameters are
consistently estimated then the estimators computed by
drtmle achieve the nonparametric efficiency bound,
resulting in narrower confidence intervals and more powerful hypothesis
tests. In order that the estimates of the OR and PS have the greatest
chance of consistency, the drtmle function facilitates
estimation using adaptive estimation techniques. In fact, there are
three ways to estimate each of the regression parameters: generalized
linear models, super learning, and a user-specified estimation
technique. We demonstrate each of these approaches in turn on a simple
simulated data set. The true parameter values of interest are ψ0(0)=
r round(EY0, 2) and ψ0(1)= 0.72.
set.seed(1234)
n <- 200
W <- data.frame(W1 = runif(n), W2 = rbinom(n, 1, 0.5))
A <- rbinom(n, 1, plogis(W$W1 + W$W2))
Y <- rbinom(n, 1, plogis(W$W1 + W$W2*A))glm to estimate regressionsGeneralized linear models are perhaps the most popular means of
estimating regression functions. Because of this legacy, these
estimators have been included as an option for estimating the regression
parameters. The user specifies the right-hand-side of a regression
formula as a character in the glm_Q, glm_g,
glm_Qr, and glm_gr options to estimate the OR,
PS, R-OR, and R-PS. These regressions are fit by calling the
glm function in base R. Thus, the model specification
should be able to be parsed as a valid formula, as required by
glm. Recall that the R-OR parameter involves a regression
onto the estimated propensity. Thus, the formula that is input via
glm_Qr should assume data with a single column named
"gn". Similarly, the R-PS1 and R-PS2 that are computed for
the estimators of Benkeser et al. (2017) (the default estimators,
computed whenever reduction = "univariate") involve
regressions onto the estimated outcome regression. Thus, the formula
glm_gr should assume data with a single column named
"Qn" if reduction = "univariate". If instead,
the estimators of van der Laan (2014) are preferred, we can set the
option reduction = "bivariate" and input a
glm_gr formula that assumes data with two columns named
"Qn" and "gn".
The code below illustrates how generalized linear models can be used
to estimate the regression parameters. In this call to
drtmle, we specify family = binomial() so that
logistic regression is used to estimate the OR. We specify
stratify = FALSE to indicate that we wish to fit a single
OR to estimate Q̄0(A, W),
as opposed to fitting two separate regressions to estimate Q̄0(1, W) and
Q̄0(0, W).
If the former fitting is specified, then the OR regression is fit using
a data frame that contains W and A. Thus, both
colnames(W) and "A" may appear in
glm_Q. However, if stratify = TRUE then the OR
model is fit separately in observations with A = 1 and A = 0. In this case, the
glm_Q option may not include "A". We
illustrate non-stratified regression for both the univariate and
bivariate reductions.
glm_fit_uni <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_g = "W1 + W2", glm_Q = "W1 + W2*A",
glm_gr = "Qn", glm_Qr = "gn", stratify = FALSE)
glm_fit_uni## $est
##
## 1 0.7111213
## 0 0.5383661
##
## $cov
## 1 0
## 1 1.741923e-03 -1.003719e-05
## 0 -1.003719e-05 4.081532e-03glm_fit_biv <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_g = "W1 + W2", glm_Q = "W1 + W2*A",
glm_gr = "Qn", glm_Qr = "gn", stratify = FALSE,
reduction = "bivariate")
glm_fit_biv## $est
##
## 1 0.7111231
## 0 0.5385794
##
## $cov
## 1 0
## 1 1.709464e-03 -1.523075e-05
## 0 -1.523075e-05 3.593222e-03SuperLearner to estimate regressionsSuper learning is a loss-based ensemble machine learning technique (van der Laan, Polley, and Hubbard 2007) that is a generalization of stacking and stacked regression (Wolpert 1992)breiman:1996:ml}. The user specifies many potential candidate regression estimators, referred to as a library of candidate estimators, and the super learner algorithm estimates the convex combination of regression estimators that minimizes V-fold cross-validated risk based on a user-selected loss function. Oracle inequalities show that the super learner estimate performs essentially as well as the unknown best convex combination of the candidate estimators (Vaart, Dudoit, and van der Laan 2006). Thus, the methodology may be seen as an asymptotically optimal way of performing estimator selection. Practically, the super learner provides the best chance for obtaining a consistent estimator for both the OR and PS, and thereby an efficient estimator of the marginal means of interest.
Super learning is implemented in the R package
SuperLearner (Polley et al.
2017) and drtmle provides the option to utilize
super learning for estimation of the OR, PS, and reduced-dimension
regressions via the SL_Q, SL_g,
SL_Qr, and SL_gr options. When calling
drtmle either the generalized linear model options for
regression modeling or the super learner options should be invoked. If
both are called, the function will default to the super learner option.
The inputs to the super learner options are passed to the
SL.library argument of the SuperLearner
function of the SuperLearner package. We refer readers to
the SuperLearner help file for information on how to
properly specify a super learner library. In the code below, we
illustrate a call to drtmle with a relatively simple
library. For the OR and PS, the library includes a main-terms
generalized linear model (via the function SL.glm from
SuperLearner) and an unadjusted generalized linear model
(SL.mean). For the reduced dimension regressions we use a
super learner library with a main terms generalized linear model
(SL.glm) and regression splines (SL.gam, which
utilizes the gam package (Hastie
2017)).
set.seed(1234)
sl_fit <- drtmle(W = W, A = A, Y = Y, family = binomial(),
SL_g = c("SL.glm","SL.mean"),
SL_Q = c("SL.glm","SL.mean"),
SL_gr = c("SL.glm", "SL.gam"),
SL_Qr = c("SL.glm", "SL.gam"),
stratify = FALSE)
sl_fit## $est
##
## 1 0.7109740
## 0 0.5381819
##
## $cov
## 1 0
## 1 1.737709e-03 -5.954220e-06
## 0 -5.954220e-06 4.142052e-03The drtmle package also allows users to pass in their
own functions for estimating regression parameters via the
SL_ options. These functions must be formatted properly for
compatibility with drtmle. The necessary formatting is
identical the formatting required for writing a
SuperLearner wrapper (see
SuperLearner::write.SL.template()). The drtmle
package includes SL.npreg, which is an example of how a
user can specify a function for estimating regression parameters. This
functions fits a kernel regression estimator using the np
package (Hayfield and Racine 2008). Below
we show the source code for this function.
## function (Y, X, newX, family = gaussian(), obsWeights = rep(1,
## length(Y)), rangeThresh = 1e-07, ...)
## {
## options(np.messages = FALSE)
## if (abs(diff(range(Y))) <= rangeThresh) {
## thisMod <- glm(Y ~ 1, data = X)
## }
## else {
## bw <- np::npregbw(stats::as.formula(paste("Y ~", paste(names(X),
## collapse = "+"))), data = X, ftol = 0.01, tol = 0.01,
## remin = FALSE)
## thisMod <- np::npreg(bw)
## }
## pred <- stats::predict(thisMod, newdata = newX)
## fit <- list(object = thisMod)
## class(fit) <- "SL.npreg"
## out <- list(pred = pred, fit = fit)
## return(out)
## }
## <bytecode: 0x563f1b6ade58>
## <environment: namespace:drtmle>In the above code, we see that a user-specified function should have
options Y, X, newX, family, obsWeights, ..., along with
other options specific to the function. The option Y
corresponds to the outcome of the regression and X to the
variables onto which the outcome is regressed. The option
newX should be of the same class and dimension as
X and is meant to contain new values of variables at which
to evaluate the regression fit. The family and
obsWeights options are not strictly necessary (indeed, they
are not used by SL.npreg), they are useful to include if
one wishes to utilize the function as part of a super learner library.
The SL.npreg function proceeds by checking whether there is
sufficient variation in Y to even both fitting the kernel
regression (as specified by option rangeThresh), and if so,
fits a kernel regression with cross-validated bandwidth selection via
the npregbw function. The output is then formatted in a
specific way so that drtmle is able to retrieve the
appropriate objects from the fitted regression. In particular the output
should be a list with named objects pred and
fit; the former including predictions made from the fitted
regression on newX, the latter including any information
that may be needed to predict new values based on the fitted object. A
class is assigned to the fit object, so that an S3
predict method may later be used to predict new values
based on the fitted regression object. The user must also specify this
predict method. Use drtmle:::predict.SL.npreg
to view the simple predict method for
SL.npreg.
In the code below, we demonstrate a call to drtmle with
SL.npreg used to estimate each nuisance parameter. Here, we
also illustrate the use of stratify = TRUE, which fits a
separate kernel regression of Y onto W in observations with A = 1 and A = 0.
npreg_fit <- drtmle(W = W, A = A, Y = Y, family = binomial(),
SL_g = "SL.npreg", SL_Q = "SL.npreg",
SL_gr = "SL.npreg", SL_Qr = "SL.npreg",
stratify = TRUE)
npreg_fit## $est
##
## 1 0.7159062
## 0 0.5343013
##
## $cov
## 1 0
## 1 1.674562e-03 -1.023615e-05
## 0 -1.023615e-05 3.963064e-03Some users may find it unsettling that when applying the super learner in real data analysis is that the results are dependent on the random seed that is set. To remove this dependence, one can instead average results over repeated super learner runs; more repeats will lead to less dependence on the random seed. There are generally two approaches that can be used: (i) averaging on the scale of the nuisance parameters and (ii) averaging on the scale of the point estimates.
For (i), suppose that we have nSL estimated super learners for the OR, Q̄n, j, j = 1, …, nSL, and the PS gn, j, j = 1, …, nSL. We can take the average over these fits, $$ \bar{Q}_n = \frac{1}{n_{SL}} \sum_{j = 1}^{n_{SL}} \bar{Q}_{n,j} $$ as our estimate of the OR (similarly for the PS). These averaged estimates are then used to compute the AIPTW or TMLE, as described above.
For(ii), suppose for example that we have nSL TMLE (similarly, AIPTW) estimates ψn, j(1) of ψ0(1), each obtained based on one (or more) super learner of the OR and/or PS. We can take the average over these point estimates, $$ \psi_n(1) = \frac{1}{n_{SL}} \sum_{j=1}^{n_{SL}} \psi_{n,j}(1) $$ as our estimate of ψ0(1).
These feature is implemented via the n_SL and
avg_over options, where the former specifies the number of
super learners to repeat and the latter specifies the scale(s) to
average over. For example, if avg_over = "drtmle" and
n_SL = 2, then two point estimates are computed, each based
on a single super learner, and averaged to obtain a final estimate. If
instead avg_over = "SL" and n_SL = 2, then two
super learners are fit and averaged before a final point estimate is
obtained. Finally, if avg_over = c("drtmle", "SL") and
n_SL = 2 then averaging on both scales is
performed. That is, the final estimate is the average of two point
estimates, each obtained using OR and PS estimates that are averaged
over two super learners.
Not surprisingly, averaging over multiple super learners can add
considerable computational expense to calls to drtmle. The
package is currently parallelized at the level of individual calls to
SuperLearner. With parallelization over enough cores,
repeated super learning should not add too much to the computational
burden relative to a single call to SuperLearner.
If the outcome-adaptive PS is desired, the user may specify this via
the adapt_g option. In this case, the additional targeting
steps needed for doubly robust inference are “turned off” (i.e.,
guard = NULL), since, as discussed above, this estimator is
not doubly robust. Note that when the
adapt-g = TRUE the data used for the PS estimation will be
a data.frame with column names QaW where
a takes all the values specified in a_0. For
example if a_0 = c(0,1), then the PS estimation data frame
will have two columns called Q0W and Q1W. This
may be relevant for users writing their own SuperLearner
wrappers or those who make use of the glm_g setting, as
illustrated below.
# outcome adaptive propensity score fit using glms
adaptg_fit <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_Q = ".^2", glm_g = "Q1W + Q0W",
adapt_g = TRUE)
# outcome adaptive propensity score fit using super learner
adaptg_sl_fit <- drtmle(W = W, A = A, Y = Y, family = binomial(),
SL_Q = c("SL.glm", "SL.mean"),
SL_g = c("SL.glm", "SL.mean"),
adapt_g = TRUE)The drtmle package also allows users to pass in their
own estimated OR and PS via the Qn and gn
options, respectively. If Qn is specified by the user,
drtmle will ignore the nuisance parameter estimation
routines specified by SL_Q and glm_Q –
similarly for gn and SL_g, glm_g.
As noted in the package documentation, there is a specific format
necessary for inputting Qn and gn. For
Qn, the entries in the list should correspond to the OR
evaluated at A and the observed values of W, with order
determined by the input to a_0. For example, if
a_0 = c(0, 1) then Qn[[1]] should be OR at
A = 0 and Qn[[2]]
should be outcome regression at A = 1. The example below illustrates
this concept.
# fit a GLM for outcome regression outside of drtmle
out_glm_fit <- glm(Y ~ . , data = data.frame(A = A, W), family = binomial())
# generate Qn list
Qn10 <- list(
# first entry is predicted values setting A = 1
predict(out_glm_fit, newdata = data.frame(A = 1, W), type = "response"),
# second entry is predicted values setting A = 0
predict(out_glm_fit, newdata = data.frame(A = 0, W), type = "response")
)
# pass this list to drtmle to avoid internal estimation of
# outcome regression (note propensity score and reduced-dimension
# regressions are still estimated internally)
out_fit1 <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_g = ".", glm_Qr = "gn", glm_gr = "Qn", Qn = Qn10,
# crucial to set a_0 to match Qn's construction!
a_0 = c(1,0))
out_fit1## $est
##
## 1 0.7116610
## 0 0.5376892
##
## $cov
## 1 0
## 1 1.726990e-03 1.428573e-06
## 0 1.428573e-06 4.183006e-03# to be completely thorough let's re-order Qn10 and re-run
Qn01 <- list(Qn10[[2]], Qn10[[1]])
out_fit2 <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_g = ".", glm_Qr = "gn", glm_gr = "Qn",
# use re-ordered Qn list
Qn = Qn01,
# a_0 has to be reordered to match Qn01!
a_0 = c(0,1))
# should be the same as out_fit1, but re-ordered
out_fit2## $est
##
## 0 0.5376892
## 1 0.7116610
##
## $cov
## 0 1
## 0 4.183006e-03 1.428573e-06
## 1 1.428573e-06 1.726990e-03Similarly, for gn we need to pass in properly ordered
lists. For example, if a_0 = c(0, 1) then
gn[[1]] should be propensity for receiving A = 0 and gn[[2]]
should be the propensity for receiving A = 1. The example below illustrates
this concept.
# fit a GLM for propensity score outside of drtmle
out_glm_fit <- glm(A ~ . , data = data.frame(W), family = binomial())
# get P(A = 1 | W) by calling predict
gn1W <- predict(out_glm_fit, newdata = data.frame(W), type = "response")
# generate gn list
gn10 <- list(
# first entry is P(A = 1 | W)
gn1W,
# second entry is P(A = 0 | W) = 1 - P(A = 1 | W)
1 - gn1W
)
# pass this list to drtmle to avoid internal estimation of
# propensity score (note reduced-dimension regressions are
# still estimated internally)
out_fit3 <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_Qr = "gn", glm_gr = "Qn",
Qn = Qn10, gn = gn10,
# crucial to set a_0 to match Qn and gn's construction!
a_0 = c(1,0))
out_fit3## $est
##
## 1 0.7116610
## 0 0.5376892
##
## $cov
## 1 0
## 1 1.726990e-03 1.428573e-06
## 0 1.428573e-06 4.183006e-03# to be completely thorough let's re-order gn10 and re-run
gn01 <- list(gn10[[2]], gn10[[1]])
out_fit4 <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_Qr = "gn", glm_gr = "Qn",
# use re-ordered Qn/gn list
Qn = Qn01, gn = gn01,
# a_0 has to be reordered to match Qn01!
a_0 = c(0,1))
# should be the same as out_fit3, but re-ordered
out_fit4## $est
##
## 0 0.5376892
## 1 0.7116610
##
## $cov
## 0 1
## 0 4.183006e-03 1.428573e-06
## 1 1.428573e-06 1.726990e-03This flexibility allows users to use any regression technique to
externally obtain initial estimates of nuisance parameters. The key step
is making sure that the ordering of Qn and gn
is consistent with a_0.
This section contains some more advanced mathematical material and may be omitted by users interested only in implementation. For simplicity, we focus this discussion on the standard TMLE/AIPTW estimators. The ideas generalize immediately to the DRTMLE estimator, but the formulas are more complex.
Reported standard error estimates are based on an estimate the asymptotic variance of the estimators. This variance can be characterized by the estimators influence function. For example, let ψn(1) be an AIPTW estimate of ψ0(1). Under regularity conditions, $$ \psi_{n}(1) - \psi_0(1) = \frac{1}{n} \sum_{i=1}^n D^*(\bar{Q}_0, g_0, Q_{W,0})(O_i) + o_{p}(n^{-1/2}) \ , $$ where $$ D^*(\bar{Q}_0, g_0, Q_{W,0})(O_i) = \frac{I(A_i = 1)}{g_0(1 \mid W_i)} \{ Y_i - \bar{Q}_0(1, W_i)\} + \bar{Q}_0(1, W_i) - \int \bar{Q}_0(1, w) dQ_{W,0}(w) \ . $$ Thus, by the central limit theorem, n1/2ψn(1) converges in distribution to a Normal distribution with mean ψ0(1) and variance σ02 = E0([D*(Q̄0, g0, QW, 0)(O) − E0{D*(Q̄0, g0, QW, 0)(O)}]2) . An estimate of σ02 is naturally obtained by taking the empirical variance of Di = D*(Q̄n, gn, QW, n)(Oi), $$ \sigma^2_n = \frac{1}{n} \sum_{i=1}^n \left\{ D_i - \frac{1}{n} \sum_{j=1}^n D_j \right\}^2 \ . $$ Thus, a natural estimate of the standard error of ψn(1) is σn/n1/2. This idea generalizes naturally to estimating the asymptotic covariance matrix of a vector n1/2(ψn(a) : a).
For TMLE-based estimators, it is standard practice to construct Di above using
the targeted version of the OR (denoted by Q̄n*
above). Recall that such estimates are obtained by modifying an initial
OR estimate in such a way as to ensure a particular (set of) equation(s)
is (are) solved. However, simulations have shown that better standard
error estimates for TMLE may be achieved by instead using the
initial OR estimate in the computation of standard errors, as
described in the previous subsection. The drtmle function
includes the targeted_se option to this end.
glm_fit_uni_nontargeted_se <-
drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_g = "W1 + W2", glm_Q = "W1 + W2*A",
glm_gr = "Qn", glm_Qr = "gn", stratify = FALSE,
targeted_se = FALSE)
# compare to original call
list(
glm_fit_uni$drtmle$cov, # based on targeted OR
glm_fit_uni_nontargeted_se$drtmle$cov # untargeted OR
)## [[1]]
## [,1] [,2]
## [1,] 1.741923e-03 -1.003719e-05
## [2,] -1.003719e-05 4.081532e-03
##
## [[2]]
## [,1] [,2]
## [1,] 1.745301e-03 -9.407578e-06
## [2,] -9.407578e-06 4.053834e-03In simulations, we often find that when employing machine learning to estimate the OR and/or PS, the estimated standard errors are too small in small to moderate-sized samples, leading to under-coverage of confidence intervals and possibly too-large type I errors of hypothesis tests that are based on these standard error estimates. This should not be too surprising when studying the form of the standard error estimates. In particular, they will include a term that looks like a (weighted) mean squared-error (MSE) of the OR. Because machine learning algorithms often set out to minimize empirical MSE as part of their training, we often find that the empirical MSE is biased too small relative to the true MSE. Thus, in order to more accurately estimate MSE of a machine learning algorithm, we typically rely on cross-validation. The same trick can be used here to obtain a more conservative standard error estimate.
Suppose we use V-fold cross-validation to obtained V training-sample-specific estimates of the OR, Q̄n, v, v = 1, …, V and the PS, gn, v, v = 1, …, V. For i = 1, …, n, let Q̄n, icv(1) denote the estimate of Q̄0(a, Wi) obtained when observation i was in the validation fold. That is, Q̄n, icv is an out-of-sample prediction of Yi. We define gn, icv similarly. Using these estimates, we can define $$ D_i^{cv} = \frac{I(A_i = 1)}{g_{n,i}^{cv}} \{ Y_i - \bar{Q}_{n,i}^{cv}(1)\} + \bar{Q}_{n,i}^{cv}(1) - \frac{1}{n} \sum_{i=1}^n \bar{Q}_{n,i}^{cv}(1) $$ and compute an estimate of σ02 using $$ \sigma^2_{n,cv} = \frac{1}{n} \sum_{i=1}^n \left\{ D_i^{cv} - \frac{1}{n} \sum_{j=1}^n D_j^{cv} \right\}^2 \ . $$
Cross-validated standard errors can be obtained by setting
se_cv = "full" and se_cvFolds to a number
greater than 1.
glm_fit_uni_cv <- drtmle(W = W, A = A, Y = Y, family = binomial(),
glm_g = "W1 + W2", glm_Q = "W1 + W2*A",
glm_gr = "Qn", glm_Qr = "gn",
se_cv = "full", se_cvFolds = 2, stratify = FALSE,
targeted_se = FALSE) # needed for this to work
# compare variance to previously obtained
list(glm_fit_uni$drtmle$cov, glm_fit_uni_cv$drtmle$cov)## [[1]]
## [,1] [,2]
## [1,] 1.741923e-03 -1.003719e-05
## [2,] -1.003719e-05 4.081532e-03
##
## [[2]]
## [,1] [,2]
## [1,] 0.0020102441 0.0002597992
## [2,] 0.0002597992 0.0052698821Obtaining fully cross-validated standard errors can be fairly time
intensive when super learner is used to estimate the OR and PS. In
drtmle we give an option to compute standard error
estimates that are partially cross-validated and do not require
any additional fitting. Here we describe briefly how these estimates are
obtained and show calls to drtmle to obtain these
estimates.
Suppose we have K candidate regressions in our super learner library for both the OR and PS (in practice, the number could be different for the OR as for the PS). Recall that a super learner estimate of for example Q̄0(a, w) is obtained as a weighted combination of estimates from each candidate regression. Let Q̄n, k denote the estimate of Q̄0 obtained by training algorithm k using the full data set. The super learner estimate of Q̄0(a, w) is $$ \bar{Q}_{n,SL}(a, w) = \sum_{k = 1}^K \alpha_{n,k} \bar{Q}_{n,k}(a, w) \ , $$ where αn = (αn, 1, …, αn, K) is a weight vector that is selected by minimizing a cross-validated risk criteria. That is, each of the K algorithms is trained according to a V-fold cross-validation scheme, resulting in V + 1 fits of each algorithm (one in each of V training folds plus one obtained by fitting the algorithm using the full data). Let Q̄n, k, v denote the k-th algorithm trained using the v-th training fold. For i = 1, …, n, let Q̄n, k, icv(1) denote the estimate of Q̄0(1, Wi) implied by the fitted algorithm Q̄n, k, v, which was obtained when observation i was in the validation sample. That is, for observations with Ai = 1, Q̄n, k, icv(1) is an out of sample prediction of Yi based on the k-th algorithm. Now, we can construct a partially cross-validated super learner prediction as follows. For i = 1, …, n let $$ \bar{Q}_{n,SL,i}^{cv}(1) = \sum_{k = 1}^K \alpha_{n,k} \bar{Q}_{n,k,i}^{cv}(1) $$ be the estimate obtained by ensembling the training-set-specific algorithm fits based on αn. We can use these estimates to produce our partially cross-validated standard error estimate as follows. We define $$ D_i^{pcv} = \frac{I(A_i = 1)}{g_{n,SL,i}^{cv}} \{ Y_i - \bar{Q}_{n,SL,i}^{cv}(1)\} + \bar{Q}_{n,SL,i}^{cv}(1) - \frac{1}{n} \sum_{i=1}^n \bar{Q}_{n,SL,i}^{cv}(1) $$ and compute an estimate of σ02 using $$ \sigma^2_{n,pcv} = \frac{1}{n} \sum_{i=1}^n \left\{ D_i^{pcv} - \frac{1}{n} \sum_{j=1}^n D_j^{pcv} \right\}^2 \ . $$
Note that is not a proper cross-validated estimate, since observation i is used to compute the weight vector αn. However, such estimates can be obtained based on the output of a single round of super learning, thereby providing improved standard error estimates at minimal additional computational cost.
Here, we demonstrate how to obtain such estimates using
drtmle.
set.seed(1234)
sl_fit_pcv <- drtmle(W = W, A = A, Y = Y, family = binomial(),
SL_g = c("SL.glm","SL.mean"),
SL_Q = c("SL.glm","SL.mean"),
SL_gr = c("SL.glm", "SL.gam"),
SL_Qr = c("SL.glm", "SL.gam"),
stratify = FALSE,
se_cv = "partial")
# compare estimates
# pt estimates should be same
# variance should be larger for pcv
list(
sl_fit, # no cv
sl_fit_pcv # partial cv
)## [[1]]
## $est
##
## 1 0.7109740
## 0 0.5381819
##
## $cov
## 1 0
## 1 1.737709e-03 -5.954220e-06
## 0 -5.954220e-06 4.142052e-03
##
##
## [[2]]
## $est
##
## 1 0.7109740
## 0 0.5381819
##
## $cov
## 1 0
## 1 1.818141e-03 -1.846086e-05
## 0 -1.846086e-05 4.332436e-03The drtmle package contains methods that compute
doubly-robust confidence intervals. The ci method computes
confidence intervals about the estimated marginal means in a fitted
drtmle object, in addition to contrasts between those
means. By default the method gives confidence intervals about each
mean.
## $drtmle
## est cil ciu
## 1 0.716 0.636 0.796
## 0 0.534 0.411 0.658Alternatively the contrast option allows users to
specify a linear combination of marginal means, which can be used to
estimate the average treatment effect as we show below.
## $drtmle
## est cil ciu
## E[Y(1)]-E[Y(0)] 0.182 0.034 0.329The contrast option can also be input as a list of
functions in order to compute a confidence interval for a user-specified
contrast. For example, we might be interested in the risk ratio ψ0(1)/ψ0(0).
When constructing Wald style confidence intervals for ratios, it is
common to compute the interval on the log scale and back-transform. That
is, the confidence interval is of the form $$
\mbox{exp}\biggl[\mbox{log}\biggl\{ \frac{\psi_0(1)}{\psi_0(0)} \biggr\}
\pm
z_{1-\alpha/2} \sigma^{\text{log}}_n \biggr] \ ,
$$ where σnlog
is the estimated standard error on the log-scale. By the delta method,
an estimate of this standard error is given by σnlog = ∇g(ψn)TΣn∇g(ψn),
where Σn
is the doubly-robust covariance matrix estimate output from
drtmle and ∇g(ψ) is the gradient of
log{ψ(1)/ψ(0)}. To
obtain this confidence interval, we may use the following code.
riskRatio <- list(f = function(eff){ log(eff) },
f_inv = function(eff){ exp(eff) },
h = function(est){ est[1]/est[2] },
fh_grad = function(est){ c(1/est[1],-1/est[2]) })
ci(npreg_fit, contrast = riskRatio)## $drtmle
## est cil ciu
## user contrast 1.34 1.036 1.733More generally this method of inputting the contrast
option to the ci method can be used to compute confidence
intervals of the form f−1{f(h(ψn)) ± z1 − α/2∇f(h(ψn))TΣn∇f(h(ψn))} ,
where f
(contrast$f) is the transformation of the confidence
interval, f−1
(contrast$f_inv) is the back-transformation of the
interval, h
(contrast$h) is the contrast of marginal means, and ∇f(h(ψn))
(contrast$fh_grad) defines the gradient of the transformed
contrast at the estimated marginal means.
We note that this method of computing confidence intervals can also be used to obtain a transformed confidence interval for a particular marginal mean. For example, in our running example Y is binary. Thus, we might wish to enforce that our confidence interval be computed on a scale which ensures that the confidence limits are bounded between zero and one. To that end, we can consider an interval of the form construct an interval on the logit scale and back-transform, $$ \mbox{expit}\biggl[ \mbox{log}\biggl\{ \frac{\psi_n(1)}{1 - \psi_n(1)} \biggr\} \pm z_{1-\alpha/2} \sigma^{\text{logit}}_n \biggr] \ . $$ This confidence interval may be implemented as follows.
logitMean <- list(f = function(eff){ qlogis(eff) },
f_inv = function(eff){ plogis(eff) },
h = function(est){ est[1] },
fh_grad = function(est){ c(1/(est[1] - est[1]^2), 0) })
ci(npreg_fit, contrast = logitMean)## $drtmle
## est cil ciu
## user contrast 0.716 0.629 0.789Doubly-robust Wald-style hypothesis tests may be implemented using
the wald_test method. As with confidence intervals, there
are three options for testing. First, we may test the null hypothesis
that the marginal means equal a fixed value (or values). Below, we test
the null hypothesis that ψ0(1) = 0.5 and that
ψ0(0) = 0.6 by
inputting a vector to the null option.
## $drtmle
## zstat pval
## H0: E[Y(1)]=0.5 5.276 0.000
## H0: E[Y(0)]=0.6 -1.044 0.297As with the ci method, the wald_test method
allows users to test linear combinations of marginal means by inputting
a vector of ones and negative ones in the contrast option.
We can use this to test the null hypothesis of no average treatment
effect or to test the null hypothesis that the average treatment effect
equals 0.05 (by specifying the null option)
## $drtmle
## zstat pval
## H0:E[Y(1)]-E[Y(0)]=0 2.414 0.016## $drtmle
## zstat pval
## H0:E[Y(1)]-E[Y(0)]=0.05 1.75 0.08The wald_test method also allows for user-specified
contrasts to be tested using syntax similar to the ci
method. The only difference syntactically is that it is not strictly
necessary to specify f_inv, as the hypothesis test is
performed on the transformed scale. That is, we can generally test the
null hypothesis that f(h(ψ0))
equals f0 (the
function f applied to the
value passed to null) using the Wald statistic $$
Z_n := \frac{\{f(h(\psi_n)) - f_0\}}{\nabla f(h(\psi_n))^T \Sigma_n
\nabla
f(h(\psi_n))} \ .
$$ We can use the riskRatio contrast defined
previously to test the null hypothesis that ψ0(1)/ψ0(0) = 1.
## $drtmle
## zstat pval
## H0: user contrast = 1 2.231 0.026The drtmle function supports missing data in
A and Y. Such missing data are common in
practice and, if ignored, can bias estimates of the marginal means of
interest. The estimators previously discussed can be modified to allow
this missingness to depend on covariates. Let ΔA and ΔY be indicators
that A and Y are observed, respectively. We can
redefine the OR to be Q̄n(a, w) := E(ΔYY ∣ ΔAA = a, ΔA = 1, ΔY = 1, W = w)
and similarly redefine the PS to be We introduce missing values to
A and Y in our running example.
DeltaA <- rbinom(n, 1, plogis(2 + W$W1))
DeltaY <- rbinom(n, 1, plogis(2 + W$W2 + A))
A[DeltaA == 0] <- NA
Y[DeltaY == 0] <- NAThe only modification of the call to drtmle is in how
the PS estimation is performed. The options glm_g and
SL_g are still used to fit generalized linear models or a
super learner, respectively. However, the code allows for separate
regression fits for each of the three components of the PS in (8). To
perform separate generalized linear model fits for each of the three
components, glm_g should be a named list with entries
"DeltaA", "A", and "DeltaY".
Respectively these should provide a regression formula for the
regression of ΔA on W, A on W in observations with ΔA = 1, and a
regression of ΔY on W and A in observations with ΔA = ΔY = 1
(if stratify = FALSE) or a regression of ΔY on W to be fit in observations with
ΔA = ΔY = 1
and A = a for each
a specified in option
a_0 (if stratify = TRUE). If only a single
formula is passed to glm_g, it is used for each of the
three regressions. We illustrate a call to drtmle with
missing data and generalized linear model fits below.
glm_fit_wNAs <- drtmle(W = W, A = A, Y = Y, stratify = FALSE,
glm_g = list(DeltaA = "W1 + W2", A = "W1 + W2",
DeltaY = "W1 + W2 + A"),
glm_Q = "W1 + W2*A", glm_Qr = "gn",
glm_gr = "Qn", family = binomial())
glm_fit_wNAs## $est
##
## 1 0.7353583
## 0 0.5213146
##
## $cov
## 1 0
## 1 1.836839e-03 -2.380141e-05
## 0 -2.380141e-05 4.917920e-03It is possible to mix generalized linear models and super learning
when fitting each of the components of the PS. In the below example we
use the wrapper function SL.glm, which fits a main terms
logistic regression, to fit the regression of ΔA on W, use SL.npreg to fit
the regression of A on W, and use a super learner with
library SL.glm, SL.mean, and
SL.gam to fit the regression of DeltaY
on W and A.
mixed_fit_wNAs <- drtmle(W = W, A = A, Y = Y, stratify = FALSE,
SL_g = list(DeltaA = "SL.glm", A = "SL.npreg",
DeltaY = c("SL.glm", "SL.mean", "SL.gam")),
glm_Q = "W1 + W2*A", glm_Qr = "gn",
glm_gr = "Qn", family = binomial())
mixed_fit_wNAs## $est
##
## 1 0.7386462
## 0 0.5187046
##
## $cov
## 1 0
## 1 0.0017938723 -2.77481e-05
## 0 -0.0000277481 4.88416e-03We can also fit the PS regressions externally and pass in via
gn, though this process is slightly more complicated than
when no missing data are present. The basic idea is to fit one
regression for (i) the indicator of missing A, (ii) A itself, and (iii) the indicator of
missing the outcome. The three PS’s are then multiplied together and an
appropriately ordered list is constructed as before. We illustrate with
the following simple example.
# first regress indicator of missing A on W
fit_DeltaA <- glm(DeltaA ~ . , data = W, family = binomial())
# get estimated propensity for observing A
ps_DeltaA <- predict(fit_DeltaA, type = "response")
# now regress A on W | DeltaA = 1
fit_A <- glm(A[DeltaA == 1] ~ . , data = W[DeltaA == 1, , drop = FALSE],
family = binomial())
# get estimated propensity for receiving A = 1
ps_A1 <- predict(fit_A, newdata = W, type = "response")
# propensity for receiving A = 0
ps_A0 <- 1 - ps_A1
# now regress DeltaY on A + W | DeltaA = 1. Here we are pooling over
# values of A (i.e., this is what drtmle does if stratify = FALSE).
# We could also fit two regressions, one of DeltaY ~ W | DeltaA = 1, A = 1
# and one of DeltaY ~ W | DeltaA = 1, A = 0 (this is what drtmle does if
# stratify = TRUE).
fit_DeltaY <- glm(DeltaY[DeltaA == 1] ~ . ,
data = data.frame(A = A, W)[DeltaA == 1, ],
family = binomial())
# get estimated propensity for observing outcome if A = 1
ps_DeltaY_A1 <- predict(fit_DeltaY, newdata = data.frame(A = 1, W),
type = "response")
# get estimated propensity for observing outcome if A = 0
ps_DeltaY_A0 <- predict(fit_DeltaY, newdata = data.frame(A = 0, W),
type = "response")
# now combine all results into a single propensity score
gn <- list(
# propensity for DeltaA = 1, A = 1, DeltaY = 1
ps_DeltaA * ps_A1 * ps_DeltaY_A1,
# propensity for DeltaA = 1, A = 0, DeltaY = 1
ps_DeltaA * ps_A0 * ps_DeltaY_A0
)
# pass in this gn to drtmle
out_fit_ps <- drtmle(W = W, A = A, Y = Y, stratify = FALSE,
# make sure a_0/gn are ordered properly!
gn = gn, a_0 = c(1, 0),
glm_Q = "W1 + W2*A", glm_Qr = "gn",
glm_gr = "Qn", family = binomial())
out_fit_ps## $est
##
## 1 0.7353583
## 0 0.5213146
##
## $cov
## 1 0
## 1 1.836839e-03 -2.380141e-05
## 0 -2.380141e-05 4.917920e-03So far in our examples, we have only considered binary treatments.
However, drtmle is equipped to handle treatments with an
arbitrary number of discrete values. Suppose that A assumes values in a set 𝒜, the PS regression estimation is modified
to ensure that $$
\sum\limits_{a \in \mathcal{A}} g_n(a \mid w) = 1 \ \mbox{for all $w$}
\ .
$$ If this condition is true, we say that the estimates of the PS
are compatible. Many regression methodologies that have been developed
work well with binary outcomes, but have not been extended to
multi-level outcomes. To ensure that users of drtmle have
access to the large number of binary regression techniques when
estimating the PS, we have taken an alternative approach to estimation
of the propensity for each level of the treatment. Specifically, rather
than fitting a single multinomial-type regression, we fit a series of
binomial regressions. As a concrete example, consider the case of no
missing data and where 𝒜 = {0, 1, 2}.
We can ensure compatible estimates of the PS by generating an estimate
gn(0 ∣ w)
of Pr0(A = 0 ∣ W = w)
and g̃n(1 ∣ w)
of Pr0(A = 1 ∣ A > 0, W = w).
The latter estimate may be generated by regression I(A = 1) on W in observations with A > 0. Because the outcome is
binary, this regression may be estimated using any technique suitable
for binary outcomes. By simple rules of conditional probability, we know
that g0(1 ∣ w) = Pr0(A = 1 ∣ A > 0, W = w){1 − Pr0(A = 0 ∣ W)}
and thus an estimate of g0(1 ∣ w) can be
computed as gn(1 ∣ w) = g̃n(1 ∣ w){1 − gn(0 ∣ w)}.
Finally, we let gn(2 ∣ w) = 1 − gn(0 ∣ w) − gn(1 ∣ w),
which ensures compatibility of the estimates for each level of
covariates. This approach generalizes to an arbitrary number of
treatment levels. Below, we provide an illustration using simulated data
with a treatment that has three levels. The true parameter values in
this simulation are ψ0(0)= 0.62, ψ0(1)= 0.72, and ψ0(2)= 0.77.
set.seed(1234)
n <- 200
W <- data.frame(W1 = runif(n), W2 = rbinom(n, 1, 0.5))
A <- rbinom(n, 2, plogis(W$W1 + W$W2))
Y <- rbinom(n, 1, plogis(W$W1 + W$W2*A))The multi-level PS can still be estimated using any of the three techniques discussed previously. Here, we illustrate with simple generalized linear models.
glm_fit_multiA <- drtmle(W = W, A = A, Y = Y, stratify = FALSE,
glm_g = "W1 + W2", glm_Q = "W1 + W2*A",
glm_gr = "Qn", glm_Qr = "gn",
family = binomial(), a_0 = c(0,1,2))
glm_fit_multiA## $est
##
## 0 0.4563335
## 1 0.7357243
## 2 0.7960642
##
## $cov
## 0 1 2
## 0 1.483804e-02 -3.223757e-05 -6.660173e-05
## 1 -3.223757e-05 2.178154e-03 6.376441e-05
## 2 -6.660173e-05 6.376441e-05 2.531909e-03Above, we used the option a_0 to specify the levels of
the treatment at which we were interested in estimating marginal means.
There is no need for a_0 to contain all unique values of
A. For example, certain levels of the treatment may not be
of scientific interest, but nevertheless we would like to use these data
to help estimate the OR and PS (by setting
stratify = FALSE).
The confidence interval and testing procedures extend to multi-level
treatments with minor modifications. We can still obtain a confidence
interval and hypothesis test for each of the marginal means by not
specifying a contrast.
## $drtmle
## est cil ciu
## 0 0.456 0.218 0.695
## 1 0.736 0.644 0.827
## 2 0.796 0.697 0.895## $drtmle
## zstat pval
## H0: E[Y(0)]=0.4 0.462 0.644
## H0: E[Y(1)]=0.5 5.051 0.000
## H0: E[Y(2)]=0.6 3.896 0.000Alternatively, we can specify a contrast to estimate
confidence intervals about the average treatment effect of A = 1 vs. A = 0. Calls to
wald_test can also be made.
## $drtmle
## est cil ciu
## E[Y(1)]-E[Y(0)] 0.279 0.023 0.536Similarly, we can compute confidence intervals comparing the average treatment effect of A = 2 vs. A = 0.
## $drtmle
## est cil ciu
## E[Y(2)]-E[Y(0)] 0.34 0.08 0.599The user-specified contrasts are also available. For example, we can
modify the riskRatio object we used previously for a
two-level treatment to compute the risk ratio comparing A = 1 to A = 0 and comparing A = 2 to A = 0.
riskRatio_1v0 <- list(f = function(eff){ log(eff) },
f_inv = function(eff){ exp(eff) },
h = function(est){ est[2]/est[1] },
fh_grad = function(est){ c(1/est[2], -1/est[1], 0) })
ci(glm_fit_multiA, contrast = riskRatio_1v0)## $drtmle
## est cil ciu
## user contrast 1.612 1.1 2.363Similarly, we can compute confidence intervals for the risk ratio comparing the marginal mean for A = 2 to A = 1.
riskRatio_2v0 <- list(f = function(eff){ log(eff) },
f_inv = function(eff){ exp(eff) },
h = function(est){ est[3]/est[1] },
fh_grad = function(est){ c(0, -1/est[1], 1/est[3]) })
ci(glm_fit_multiA, contrast = riskRatio_2v0)## $drtmle
## est cil ciu
## user contrast 1.744 1.382 2.202When considering adaptive estimation techniques, one runs the risk of
overfitting the initial regressions. While the cross validation used by
super learning attempts to guard against overfitting, there are no
guarantees in finite-samples. To definitively avoid such overfitting,
one can use cross-validated estimates of the regression parameters.
Theoretically, the use of cross-validated regression weakens the
assumptions necessary to achieve asymptotic normality (Zheng and van der Laan 2010). This is true of
the regular TMLE and AIPTW, as well as of the TMLE estimator with
doubly-robust inference. The drtmle function implements
cross-validated estimation of the regression parameters through the
cvFolds option, as shown in the code below, where we
continue with the multi-level treatment example.
cv_sl_fit <- drtmle(W = W, A = A, Y = Y, family = binomial(),
SL_g = c("SL.glm", "SL.glm.interaction", "SL.gam"),
SL_Q = c("SL.glm", "SL.glm.interaction", "SL.gam"),
SL_gr = c("SL.glm", "SL.mean"),
SL_Qr = c("SL.glm", "SL.mean"),
stratify = FALSE, cvFolds = 2, a_0 = c(0, 1, 2))
cv_sl_fit## $est
##
## 0 0.5018057
## 1 0.7366243
## 2 0.7882366
##
## $cov
## 0 1 2
## 0 1.193714e-02 -7.941656e-06 1.535563e-05
## 1 -7.941656e-06 2.235493e-03 6.137198e-05
## 2 1.535563e-05 6.137198e-05 2.758834e-03Cross-validation may significantly increase the computation time
necessary for running drtmle. Parallelized regression
fitting is available by setting the option
use_future = TRUE. Since, internally, all computations are
carried out using futures (with the API provided by the “future” R
package), setting this option amounts only to parallelizing with futures
(Bengtsson 2017a) via
future.apply::lapply. Selecting this option results in
parallelization of the estimation of the OR, PS, and reduced-dimension
regressions.
If no adjustments are made prior to invoking the drtmle
function, futures are computed sequentially. It is up to the user to
specify appropriate system-specific future back-ends that may better
suit their problem (e.g., using the future.batchtools
package (Bengtsson 2017b)). Below we
demonstrate registration using an ad-hoc cluster on the local system.
The choice of back-end is flexible, and submission of the job to a wide
variety of schedulers (e.g., SLURM, TORQUE, etc.) is also permitted.
Interested readers are invited to consult the documentation of the
future.batchtools package for further information.
## Commented out to avoid build errors
# library(future.batchtools)
# library(parallel)
# cl <- makeCluster(2, type = "SOCK")
# plan(cluster, workers = cl)
# clusterEvalQ(cl, library("SuperLearner"))
# pcv_sl_fit <- drtmle(W = W, A = A, Y = Y, family = binomial(),
# SL_g = c("SL.glm", "SL.glm.interaction","SL.gam"),
# SL_Q = c("SL.glm", "SL.glm.interaction","SL.gam"),
# SL_gr = c("SL.glm", "SL.gam", "SL.mean"),
# SL_Qr = c("SL.glm", "SL.gam", "SL.mean"),
# stratify = FALSE, a_0 = c(0,1,2),
# cvFolds = 2, use_future = TRUE)Doubly-robust estimators of marginal means are appealing in their robustness and efficiency properties. However, researchers may prefer the IPTW estimator for its intuitive derivation and implementation. Heretofore, inference for IPTW estimators precluded the use of adaptive estimators of the PS. However, van der Laan (2014) proposed modified IPTW estimators that allow for adaptive PS estimation. Similarly as above, this estimator is based on an additional regression parameter, Q̃r, 0n(a, w) := E0{Y ∣ A = a, gn(W) = gn(w)}, that is the regression of the outcome Y on the estimated propensity amongst observations with A = a. The author showed that if gn*(a ∣ w), an estimator of the propensity for treatment A = a, satisfied that $$ \frac{1}{n} \sum\limits_{i=1}^n \frac{\tilde{Q}_{r,0}(a,W_i)}{g_n(a \mid W = W_i)} \ \{ I(A_i = a) - g_n(a \mid W = W_i)\} = 0 \ ,$$ then the IPTW estimator based on gn* has an asymptotic normal distribution with an estimable variance. In fact, van der Laan (2014) proposed a TMLE algorithm that mapped an initial PS estimate into an estimate that satisfies this key equation and variance estimators that can be used to generate Wald-style confidence intervals and hypothesis tests.
An estimator with equivalent asymptotic properties based on a PS
estimate, gn(a ∣ w),
can be computed as $$
\psi_{n,IPTW}(a) = \psi_{n,IPTW}(a) - \frac{1}{n} \sum\limits_{i=1}^n
\frac{\tilde{Q}_{r,0}(a,W_i)}{g_n(a \mid W = W_i)} \ \{ I(A_i = a) -
g_n(a
\mid W = W_i)\} \ .
$$ Both this estimator and the TMLE-based IPTW estimator above
are implemented in drtmle.
The IPTW estimators based on adaptive estimation of the PS are
implemented in the adaptive_iptw function. The
adaptive_iptw function shares many options with
drtmle. However, as the IPTW estimator does not rely on an
estimate of the OR, these options are omitted. We continue with the
multilevel treatment example.
npreg_iptw_multiA <- adaptive_iptw(W = W, A = A, Y = Y, stratify = FALSE,
SL_g = "SL.npreg", SL_Qr = "SL.npreg",
family = binomial(), a_0 = c(0, 1, 2))
npreg_iptw_multiA## $est
##
## 0 0.4646874
## 1 0.7052869
## 2 0.8030194
##
## $cov
## 0 1 2
## 0 1.414756e-02 -3.045258e-05 -4.361100e-05
## 1 -3.045258e-05 1.852371e-03 2.263188e-05
## 2 -4.361100e-05 2.263188e-05 2.371211e-03The "adaptive_iptw" class contains identical methods for
computing confidence intervals and Wald tests as the
"drtmle" class and we refer readers back to confidence
intervals and hypothesis tests section for reminders on the various
options for these tests. By default, these methods are implemented for
the iptw_tmle estimator of van der Laan (2014). We expect
that this estimator will have improved finite-sample behavior relative
to iptw_os (the alternative estimator described in the
previous subsection), though further study is needed.
As with drtmle, the adaptive_iptw function
allows missing data in A and Y. Similarly,
adaptive_iptw also allows for parallelized, cross-validated
estimation of the regression parameters via cvFolds and
parallel options. As with the estimators implemented in
drtmle, cross-validation allows for valid inference under
weaker assumptions.
The drtmle package was designed to provide a
user-friendly implementation of the TMLE algorithm that facilitates
doubly-robust inference about marginal means and average treatment
effects. While the theory underlying doubly-robust inference is complex,
the implementation of the methods only requires users to provide the
data and specify how to estimate regressions. While estimation of the OR
and PS is familiar to those familiar with causal inference-related
estimation, the reduced-dimension regressions may appear strange.
Indeed, it is difficult to provide intuition for these parameters and
more difficult to determine what scientific background knowledge might
guide users in how to estimate these regression parameters. Therefore,
we recommended estimating these quantities adaptively, e.g., with the
super learner. The super learner library should contain several
relatively smooth estimators such as SL.mean,
SL.glm, and SL.gam. In particular, including
SL.mean (an unadjusted generalized linear model) may be
important due to the fact that, when the OR and PS are consistently
estimated, the reduced-dimension regressions are identically equal to
zero. Thus, this unadjusted regression estimator may do a good job
estimating the reduced-dimension regressions in situations where the OR
and PS are estimated well. In addition to relatively smooth estimators,
we recommend additionally including several adaptive estimators such as
SL.npreg or SL.gam.
Planned extensions of the drtmle package include
extensions of longitudinal settings involving treatments at multiple
time points and inclusion of the collaborative targeted maximum
likelihood estimator (CTMLE) of van der Laan (2014) that also
facilitates collaborative doubly-robust inference.
## R version 4.4.1 (2024-06-14)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.1 LTS
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so; LAPACK version 3.12.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=C
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## time zone: Etc/UTC
## tzcode source: system (glibc)
##
## attached base packages:
## [1] splines stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] quadprog_1.5-8 nloptr_2.1.1 SuperLearner_2.0-29
## [4] gam_1.22-5 foreach_1.5.2 nnls_1.6
## [7] drtmle_1.1.2 rmarkdown_2.28
##
## loaded via a namespace (and not attached):
## [1] Matrix_1.7-1 np_0.60-17 future.apply_1.11.3
## [4] jsonlite_1.8.9 compiler_4.4.1 Rcpp_1.0.13-1
## [7] MatrixModels_0.5-3 parallel_4.4.1 jquerylib_0.1.4
## [10] globals_0.16.3 boot_1.3-31 yaml_2.3.10
## [13] fastmap_1.2.0 lattice_0.22-6 R6_2.5.1
## [16] knitr_1.48 MASS_7.3-61 iterators_1.0.14
## [19] future_1.34.0 maketools_1.3.1 bslib_0.8.0
## [22] rlang_1.1.4 cachem_1.1.0 xfun_0.49
## [25] sass_0.4.9 sys_3.4.3 cli_3.6.3
## [28] digest_0.6.37 grid_4.4.1 cubature_2.1.1
## [31] quantreg_5.99 lifecycle_1.0.4 SparseM_1.84-2
## [34] evaluate_1.0.1 listenv_0.9.1 codetools_0.2-20
## [37] buildtools_1.0.0 survival_3.7-0 parallelly_1.38.0
## [40] tools_4.4.1 htmltools_0.5.8.1