Home Current Issue Previous Issues Published Ahead-of-Print Collections Podcasts Videos For Authors Journal Info
Skip Navigation LinksHome > September 2000 - Volume 11 - Issue 5 > Marginal Structural Models and Causal Inference in Epidemiol...
Epidemiology:
Editorial

Marginal Structural Models and Causal Inference in Epidemiology

Robins, James M.1 2; Hernán, Miguel Ángel1; Brumback, Babette2

Free Access
Article Outline
Collapse Box

Author Information

From the Departments of 1Epidemiology and 2Biostatistics, Harvard School of Public Health, Boston, MA.

This research was supported by NIH grant R01-A132475.

Submitted August 4, 1999; final version accepted February 28, 2000.

Address correspondence to: James M. Robins, Department of Epidemiology, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115.

Collapse Box

Abstract

In observational studies with exposures or treatments that vary over time, standard approaches for adjustment of confounding are biased when there exist time-dependent confounders that are also affected by previous treatment. This paper introduces marginal structural models, a new class of causal models that allow for improved adjustment of confounding in those situations. The parameters of a marginal structural model can be consistently estimated using a new class of estimators, the inverse-probability-of-treatment weighted estimators.

Marginal structural models (MSMs) are a new class of causal models for the estimation, from observational data, of the causal effect of a time-dependent exposure in the presence of time-dependent covariates that may be simultaneously confounders and intermediate variables. 1–3 The parameters of a MSM can be consistently estimated using a new class of estimators: the inverse-probability-of-treatment weighted (IPTW) estimators. MSMs are an alternative to structural nested models (SNMs), the parameters of which are estimated through the method of g-estimation. 4–6

The usual approach to the estimation of the effect of a time-varying exposure or treatment has been to model the probability of disease as a function of past exposure and past confounder history, using analytic methods such as stratified analysis and its parametric analogs (for example, logistic or proportional hazards regression). We will show in sections 4 and 7.1 that these standard approaches may be biased, whether or not one further adjusts for past confounder history in the analysis, when (1) there exists a time-dependent covariate that is a risk factor for, or predictor of, the event of interest and also predicts subsequent exposure, and (2) past exposure history predicts subsequent level of the covariate. We refer to covariates satisfying condition 1 as time-dependent confounders. Conditions 1 and 2 will be true in many observational studies, particularly those in which there is confounding by indication. For example, in a study of the effect of zidovudine (AZT) treatment on mortality among human immunodeficiency virus (HIV)-infected subjects, the time-dependent covariate CD4 lymphocyte count is both an independent predictor of survival and initiation of therapy with AZT and is itself influenced by prior AZT treatment. In a study of the effect of obesity on mortality, the development of clinical cardiac or respiratory disease is an independent predictor of both mortality and subsequent weight loss and is influenced by prior weight gain. Conditions 1 and 2 will always hold when there are time-dependent covariates that are simultaneously confounders and intermediate variables. A more detailed description of the bias of standard methods, as well as several additional epidemiologic examples of time-dependent confounding, has been presented elsewhere. 5,7

Back to Top | Article Outline

1. Time-Dependent Confounding

Consider a follow-up study of HIV-infected patients. Let Ak be the dose of the treatment or exposure of interest, say AZT, on the kth day since start of follow-up. Let Y be a dichotomous outcome of interest (for example, Y = 1 if HIV RNA is not detectable in the blood and is 0 otherwise) measured at end of follow-up on day K + 1. Our goal is to estimate the causal effect of the time-dependent treatment Ak on the outcome Y.

Figure 1 is a causal graph that represents our study with K = 1. A causal graph is a directed acyclic graph in which the vertices (nodes) of the graph represent variables and the directed edges (arrows) represent direct causal effects. 8 In Figure 1, Lk represents the value on day k of the vector of all measured risk factors for the outcome, such as age, CD4 lymphocyte count, white blood count (WBC), hematocrit, diagnosis of acquired immunodeficiency syndrome (AIDS), and the presence or absence of various symptoms or opportunistic infections such as oral candidiasis. Similarly, Uk represents the value on day k of all unmeasured causal risk factors for Y. Figure 1, b, differs from Figure 1, a, only in that the arrows from the unmeasured causal risk factors into the treatment variables have been removed. When, as in Figure 1, b, there is no arrow from unmeasured causal risk factors into treatment variables, we say that there are no unmeasured confounders given data on the measured confounders Lk. 9,10 Figure 1, c, differs from Figure 1, a and b, in that none of the risk factors for Y (measured or unmeasured) has arrows into any treatment variable. Note, however, that earlier treatment A0 can causally affect later treatment A1. When, as in Figure 1, c, there is no arrow from any (nontreatment) risk factor into any treatment variable, there is no confounding by either measured or unmeasured factors, in which case we say that treatment is unconfounded. 9,10

Figure 1
Figure 1
Image Tools

The distinctions drawn above apply equally to more familiar point-treatment studies in which the treatment is not time-dependent. As indicated in Figure 2, a point-treatment study is a special case of the general set-up in which K = 0. Figure 2, a–c, contains the analogs of Figure 1, a–c, for a point-treatment study.

Figure 2
Figure 2
Image Tools

As in any observational study, we cannot determine from the observed data on the Lk, Ak, and Y whether there is confounding by unmeasured risk factors. We can only hope that whatever residual confounding there may be due to the Uk is small. Under the untestable assumption that there is no unmeasured confounding given the Lk, we can, however, empirically test from the data whether treatment is unconfounded. Specifically, a sufficient condition for treatment to be unconfounded is that, at each time k, among subjects with the same past treatment history A0, …, Ak−1, the treatment Ak is unassociated with the past history of measured covariates L0, …, Lk. 9–11 For example, in our point-treatment study, treatment will be unconfounded if A0 is unassociated with L0.

Back to Top | Article Outline

2. Counterfactuals in Point-Treatment Studies

We begin with a review of how one would estimate the effect of A0 on Y in the point-treatment study of Figure 2. Suppose treatment A0 is dichotomous; suppose further that Figure 2, c, is the true causal graph, that is, that neither measured nor unmeasured covariates confound the relation between treatment and the outcome. Then the crude risk difference, risk ratio, and odds ratio each measure the causal effect of the treatment A0 on the outcome Y, although on different scales. The crude risk difference is cRD = pr[Y = 1|A0 = 1] − pr[Y = 1|A0 = 0], the crude risk ratio is cRR = pr[Y = 1|A0 = 1]/pr[Y = 1|A0 = 0], the crude odds ratio is cOR = pr[Y = 1|A0 = 1]pr[Y = 0|A0 = 0]/{pr[Y = 1|A0 = 0]pr[Y = 0|A0 = 1]}, and, for example, pr[Y = 1|A0 = 1] is the probability that Y = 1 among treated subjects (A0 = 1). We assume that the study subjects are a random sample from a large, possibly hypothetical, source population. Probabilities refer to proportions in the source population.

The causal contrasts that correspond to these associational parameters involve counterfactual variables. Specifically, the variable Ya0=1 denotes a subject’s outcome if treated and Ya0=0 denote a subject’s outcome if left untreated. For a given subject, the causal effect of treatment, measured on a difference scale, is Ya0=1 − Ya0=0. For no subject are both Ya0=0 and Ya0=1 observed. If a subject is treated (A0 = 1), the subject’s observed outcome Y equals Ya0=1, and Ya0=0 is unobserved. If A0 = 0, Y equals Ya0=0, and Ya0=1 is unobserved. Let pr(Ya0=1 = 1) and pr(Ya0=0 = 1), respectively, be the probability that Ya0=1 is equal to 1 and Ya0=0 is equal to 1. Then, if treatment A0 is unconfounded, the crude RD equals the causal risk difference pr[Ya0=1 = 1] − pr[Ya0=0 = 1] in the source population. The causal risk difference is the average of the individual causal risk differences Ya0=1 − Ya0=0. Similarly, the crude RR equals the causal RR, pr(Ya0=1 = 1)/pr(Ya0=0 = 1), and the crude OR equals the causal OR, pr(Ya0=1 = 1)pr(Ya0=0 = 0)/{pr(Ya0=1 = 0)pr(Ya0=0 = 1)}. Because of the possibility of effect modification, the population causal parameter need not equal the causal parameter within a stratum of the measured risk factors L0 even if treatment is unconfounded. Effect modification is considered in section 9.

Back to Top | Article Outline

3. Models for Point-Treatment Studies

The causal RD, RR, and OR can also be expressed in terms of the parameters of the following linear, log linear, and linear logistic models for the two counterfactual probabilities pr(Ya0=1 = 1) and pr(Ya0=0 = 1). MATH MATH MATH where Ya0 is Ya0=1 if a0 = 1 and Ya0 is Ya0=0 if a0 = 0. Specifically, the causal RD = ψ1, the causal RR = eθ1, and the causal OR = eβ1. Models 1–3 are saturated MSMs. They are marginal models, because they model the marginal distribution of the counterfactual random variables Ya0=1 and Ya0=0 rather than the joint distribution (that is, models 1–3 do not model the correlation of Ya0=1 and Ya0=0). They are structural models, because they model the probabilities of counterfactual variables and in the econometric and social science literature models for counterfactual variables are often referred to as structural. 8,12 Finally, they are saturated, because each has two unknown parameters and thus each model places no restriction on the possible values of the two unknown probabilities pr(Ya0=1 = 1) and pr(Ya0=0 = 1). Note that these models do not include covariates, because they are, by definition, models for causal effects on the entire source population; they are not models for observed associations.

Equation U1
Equation U1
Image Tools
Equation U2
Equation U2
Image Tools
Equation U3
Equation U3
Image Tools

The crude RD, RR, and OR can also be expressed in terms of the parameters of the following saturated linear, log linear, and linear logistic models for the observed outcome Y. MATH MATH MATH These are models for associations observed when comparing subpopulations (defined by levels of treatment) of the source population. The crude RD equals ψ1, the crude RR equals eθ′1, and the crude OR equals eβ′1. The parameters of the associational models 4–6 will differ from the parameters of the MSMs 1–3, except when treatment is unconfounded. Because models 4–6 are models for the observed data, (asymptotically) unbiased estimates of the model parameters can be obtained using standard statistical software (assuming no selection bias or measurement error). When treatment is unconfounded, these same estimates will also be unbiased for the corresponding causal parameters of models 1–3. For example, to fit models 4–6, one could use the general-purpose SAS program Proc Genmod, using the model statement Y = A0 with the outcome Y specified as a binomial variable. To estimate ψ1, one would specify the identity link; to estimate θ1, the log link; and for β1, the logit link. Programs analogous to Proc Genmod also exist in other packages, such as S-Plus, Gauss, and Stata. 13–16

Equation U4
Equation U4
Image Tools
Equation U5
Equation U5
Image Tools
Equation U6
Equation U6
Image Tools
Back to Top | Article Outline

4. No Unmeasured Confounders

Suppose now that treatment is confounded. Then the crude association parameter will not equal the corresponding causal parameter. Similarly, the parameters of the MSMs will fail to equal the parameters of the corresponding observed data models (for example, β0 ≠ β0 and β1 ≠ β1). Assuming we have no unmeasured confounders given data on measured confounders L0, unbiased estimates of the causal parameters ψ1, θ1, and β1 can, however, still be obtained using Proc Genmod by performing a weighted analysis. Specifically, using the weight statement (that is, option SCWGT) in Proc Genmod, each subject i is assigned a weight wi equal to the inverse of the conditional probability of receiving his or her own treatment. That is, wi = 1/pr[A0 = a0i|L0 = l0i], where, for example, l0i is the observed value of the variable L0 for subject i. The true weights wi are unknown but can be estimated from the data in a preliminary logistic regression of A0 on L0. For example, we might specify the logistic regression model MATH where A0 is AZT treatment, L0 is, for example, the column vector of covariates with components age, CD4 count, WBC count, hematocrit, and presence of symptoms, and α1 is a row vector of unknown parameters. We can then obtain estimates (α̂0, α̂1) of (α0, α1) using standard logistic regression software. Then, for a subject i with A0 = 0 and L0 = l0i, we would estimate wi = 1/pr[A0 = 0|L0 = l0i] by 1/{1/[1 + exp(α̂0 + α̂1l0i)]} = 1 + exp(α̂0 + α̂1l0i). For a subject with A0 = 1 and L0 = l0i, we would estimate wi = 1/pr[A0 = 1|L0 = l0i] by {1 + exp(α̂0 + α̂1l0i)}/exp(α̂0 + α̂1l0i) = 1 + exp(−α̂0 − α̂1l0i).

Equation U7
Equation U7
Image Tools

In summary, if there are no unmeasured confounders given data on L0, one can control confounding (due to L0) by modifying the crude analysis by weighting each subject i by wi. The denominator of wi is informally the probability that a subject had his or her own observed treatment. Thus, we refer to these weighted estimators as IPTW estimators.

Why does this approach work? The effect of weighting in Proc Genmod is to create a pseudopopulation consisting of wi copies of each subject i. That is, if, for a given subject, wi = 4, the subject contributes four copies of him- or herself to the pseudopopulation. This new pseudopopulation has the following two important properties. First, in the pseudopopulation, unlike the actual population, A0 is unconfounded by the measured covariates L0. Second, pr(Ya0=1 = 1) and pr(Ya0=0 = 1) in the pseudopopulation are the same as in the true study population so that the causal RD, RR, and OR are the same in both populations. Hence, it follows that we can unbiasedly estimate the causal RD, RR, and OR by a standard crude analysis in the pseudopopulation. But this is exactly what our IPTW estimator does, because the weights wi serve to create, as required, wi copies of each subject. In the Appendix, we present a detailed numerical example to clarify further our IPTW methodology, and we compare our methodology with the propensity score methodology of Rosenbaum and Rubin. 17

Back to Top | Article Outline

5. Unmeasured Confounding

In the presence of unmeasured confounding factors U0, one could still unbiasedly estimate the causal risk difference, risk ratio, and odds ratio as above if one used weights wi = 1/pr[A0 = a0i|L0 = l0i, U0 = u0i] in implementing the analysis in Proc Genmod. Nevertheless, because data on U0 are not observed, it is not possible to estimate these weights unbiasedly. Indeed, unbiased estimation by any method is impossible in the presence of unmeasured confounding factors without strong additional assumptions.

Back to Top | Article Outline

6. Multilevel Treatment and Unsaturated MSMs

Suppose again that treatment is unconfounded (as represented in Figure 2, c) but now A0 is an ordinal variable representing a subject’s daily dose in units of 100 mg of AZT. Possible values of A0 are 0, 1, …, 14, 15. In that case, the number of potential outcomes associated with each subject will be 16. Specifically, we let Ya0 be the value of Y that would have been observed had the subject received dose a0 rather than the observed dose. Thus, in principle, a subject has a separate counterfactual variable for each of the 16 possible AZT doses a0. The subject’s observed outcome Y is the outcome Ya0 corresponding to the dose a0 equal to the subject’s observed dose. For expositional convenience, we continue to refer to all of the Ya0’s as counterfactuals, even though for a0 equal to the observed dose, Ya0 is the factual variable Y. Because there are so many potential outcome variables Ya0, we can no longer conveniently perform a saturated analysis. Rather, we would usually assume a parsimonious dose-response relationship by specifying a linear logistic MSM such as MATH This model says that the probability of success had all subjects been treated with dose a0 is a linear logistic function of the dose with slope parameter β1 and intercept β0, so eβ1 is the causal OR associated with an increase in AZT dose of 100 mg.

Equation U8
Equation U8
Image Tools

We contrast the MSM model 8 with the following linear logistic association model for the observed data. MATH Assuming no selection bias or measurement error, we can unbiasedly estimate the associational parameters β0 and β1 by fitting the linear logistic model 9 using a standard logistic regression software package such as SAS Proc Logistic or Proc Genmod. If the treatment is unconfounded, then the parameters of models 8 and 9 are equal. As a consequence, our logistic regression estimate of β1 is also an unbiased estimate of our causal parameter β1.

Equation U9
Equation U9
Image Tools

If treatment is confounded by the measured variables L0, then β1 ≠ β1 and our standard logistic regression estimate of β1 is a biased estimate of the causal parameter β1 owing to confounding by L0. However, even when treatment is confounded, if there are no unmeasured confounders given L0, then one can obtain unbiased estimates of the causal parameter β1 of model 8 by fitting the logistic model 9 with Proc Genmod if one uses subject-specific weights wi = 1/pr(A0 = a0i|L0 = l0i). Again, in practice, wi is unknown and one must estimate it from the data by specifying a model.

For example, one might specify the following polytomous logistic model:MATH which can be fit in SAS using Proc Logistic or Genmod to obtain estimates of the parameters α01, α02, …, α015, and α1.

Equation U10
Equation U10
Image Tools
Back to Top | Article Outline
6.

1. Stabilized Weights

The probabilities pr[A0 = a0i|L0 = l0i] may vary greatly between subjects when components of L0 are strongly associated with A0. This variability can result in extremely large values of the weight wi for a few subjects. These few subjects will contribute a very large number of copies of themselves to the pseudopopulation and thus will dominate the weighted analysis, with the result that our IPTW estimator will have a large variance and will fail to be approximately normally distributed. If the MSM model is saturated (for example, models 1–3), this variability is unavoidable, because it reflects a lack of information in the data as a result of the confounders L0 being highly correlated with treatment A0. For unsaturated MSMs, such as model 8, however, this variability can, to a considerable extent, be mitigated by replacing the weight wi by the “stabilized weight” swi = pr[A0 = a0i]/pr[A0 = a0i|L0 = L0i]. To understand the stabilized weight, suppose A0 was unconfounded so that A0 and L0 are unassociated and pr[A0 = a0i] = pr[A0 = a0i|L0 = l0i]. Then swi = 1, and each subject contributes the same weight. When A0 is confounded, swi will not be constant but will vary around the number 1, depending on the subject’s value of L0. swi, however, will still tend to be much less variable than wi. Furthermore, Robins 1,2 shows that, when we use the weight swi rather than the weight wi in Proc Genmod, the estimates of the parameters β of an MSM remain unbiased and, in the case of an unsaturated MSM, will generally be less variable. For saturated MSMs, the variability of our estimate of β will be the same whether we use the stabilized or unstabilized weights.

Of course, pr[A0 = a0i] and pr[A0 = a0i|L0 = l0i] are unknown and must be estimated. pr[A0 = a0i|L0 = l0i] can be estimated as described above; pr[A0 = a0i] can be estimated as the proportion of subjects in the study sample with A0 equal to a0i. This estimate is equivalent to that obtained by fitting the polytomous model MATH where we place an asterisk on the parameter α*0a0 to indicate that this parameter will differ from the parameter α0a0 of model 10 when A0 is confounded.

Equation U11
Equation U11
Image Tools

In Ref 2, Robins introduces augmented IPTW estimators. These estimators are even more efficient than the IPTW estimator that uses stabilized weights but are more difficult to compute.

Back to Top | Article Outline
6.

2. Continuous Treatment

Suppose we were able to measure a subject’s daily intake of AZT to the nearest tenth of a milligram, so that now AZT is essentially a continuous treatment. Further, for expositional convenience, assume that no subject has AZT dose near 0, and that we can effectively model the distribution of A0 as normal. Now each individual has an extremely large number of counterfactual outcomes Ya0. One can still obtain unbiased estimates of the causal parameter β1 of model 8 by fitting the logistic model 9 with Proc Genmod if one uses the stabilized weights swi = f(a0i)/f(a0i|l0i), where f(a0|l0) is the conditional density of the continuous variable A0 given L0, and f(a0) is the marginal density of the continuous variable A0. To estimate f(a0|l0), one might specify that, given L0, A0 is normal with mean α0 + α1L0 and variance ς2. Then unbiased estimates (α̂0, α̂1, μ̂2) of (α0, α1, ς2) can be obtained by ordinary least-squares regression of A0 on L0 using, for example, Proc REG in SAS. Then f(a0i|l0i) would be estimated by the normal density (2πμ̂2)−1/2 exp{−[a0i − (α̂0 + α̂1l0i)]2/2μ̂2}. To estimate the numerator f(a0i) of the stabilized weight swi, one might specify that A0 is normal with mean α*0 and variance ς*2. f(a0i) could be estimated by the normal density (2πμ̂*2)−1/2 exp[−(a0i − α̂*0)2/2μ̂*2] where α̂*0 is the average of the observed A0s and μ̂*2 is their empirical variance. When A0 is continuous, estimates based on the unstabilized weights wi = 1/f(a0i|l0i) have infinite variance and thus cannot be used. 1,2

Back to Top | Article Outline
6.

3. Confidence Intervals

As described above, we shall estimate the parameters of the MSM 8 by fitting the association model 9 in Proc Genmod using estimates of the stabilized weights swi. If we choose the option “repeated” and specify an independence working correlation matrix, the Proc Genmod program will also output a 95% “robust” Wald confidence interval for β1 given by β̂1 ± 1.96 MATH, where var (β̂1) is the so-called “robust”18 or “sandwich” estimator of the variance of β̂1. Robins 1,2 shows that the “robust” Wald interval will have coverage probability of at least 95%, although narrower valid intervals can be obtained with some additional programming. 1,2 The ordinary nonrobust model-based Wald confidence interval outputted by most weighted logistic regression programs will not be guaranteed to provide at least 95% coverage and thus should be avoided. Other software packages such as S-Plus, Gauss, and STATA also offer “robust” variance estimators and could be used in place of Proc Genmod. 13,19–21

Equation U12
Equation U12
Image Tools
Back to Top | Article Outline

7. Time-Dependent Treatments

We now return to the setting of section 1, in which Ak is the dose of treatment AZT on the kth day from start of follow-up and Y is a dichotomous outcome variable measured at end of follow-up on day K + 1. Similarly, Lk represents the value on day k of the vector of all measured risk factors for the outcome. Let Āk = (A0, A1, …, Ak) be the treatment or exposure history through day k and let Ā = ĀK. Define ¯Lk and ¯L similarly. Let Yā be the value of Y that would have been observed had all subjects received dose history ā = (a0, a1, …, aK) rather than their observed dose history Ā. Note that, even if ak is dichotomous on each day k (that is, on each day a subject is either on or off treatment), there will be 2K dose histories ā and thus 2K possible counterfactuals, only one of which is observed (that is, is factual). Thus, it may not be possible to estimate a saturated MSM. Therefore, we would generally assume some sort of parsimonious dose-response relationship by specifying a linear logistic MSM such as MATH where cum (ā) = ∑k=0K ak is the cumulative dose through end-of-follow-up associated with the dose history ā.

Equation U13
Equation U13
Image Tools

We contrast the MSM 12 with the following linear logistic association model for the observed data:MATH Assuming no loss to follow-up selection bias or measurement error, we can unbiasedly estimate the parameters β1 by fitting the linear logistic model 13 using a standard logistic regression software package. If the treatment is unconfounded, the parameters of models 12 and 13 are equal. As a consequence, our logistic regression estimate of β1 is also an unbiased estimate of our causal parameter β1. If the treatment is confounded, then β1 ≠ β1 and our standard logistic regression estimate of β1 is a biased estimate of the causal parameter β1 as a result of confounding by ¯Lk. When treatment is confounded, however, if there is no unmeasured confounder given the Lk, then one can still obtain unbiased estimates of the causal parameter β1 of model 12 by fitting the logistic model 13 with the stabilized weights MATH where ∏k=0K bk = b0 × b1 × b2 × … × bK and Ā−1 is defined to be 0. Note in the special case in which K = 0 (that is, a point-treatment study), models 12 and 13 reduce to our previous models 8 and 9 and 14 reduces to our previous swi. The denominator of swi is informally the conditional probability that a subject had his or her own observed treatment history through time K. With time-dependent treatments the variation in the unstabilized weights will often be enormous, with the result that the resulting estimator of β can be highly variable with a markedly non-normal sampling distribution. We therefore strongly recommend the use of stabilized weights.

Equation U14
Equation U14
Image Tools
Equation U15
Equation U15
Image Tools

We emphasize that when treatment is confounded, it is the parameter β1 of our MSM 12, as opposed to the parameter β1 of the association model 13 that is of policy importance. To see why, consider a new subject from the source or target population exchangeable with the N study subjects. We would like to administer to the subject the treatment ā that minimized probability that he or she has detectable HIV in the serum at the end of follow-up, that is, pr(Yā = 1). Thus, for example, if the parameter β1 of MSM 12 is positive (that is, the probability of having HIV in the blood increases with increasing duration of AZT treatment), we would withhold AZT treatment from our subject. In contrast, the parameter β1 of model 13 may be confounded by the association of covariates with treatment. For example, suppose physicians preferentially started AZT on subjects who, as indicated by their prognostic factor history (for example, CD4 count), were doing poorly. Further, suppose that AZT had no causal effect on Y (that is, β1 = 0). Then the parameter β1 (and thus our estimate from the unweighted logistic regression) will be positive but will have no causal interpretation as the effect of AZT on Y.

Back to Top | Article Outline
7.

1. Bias Induced by Controlling for a Variable Affected by Treatment

One might suppose that an alternative approach to controlling confounding by measured covariates is an unweighted logistic model that adjusts for confounder history ¯L ≡ ¯LK, such as MATH where, for simplicity, we here assume Lk consists of a single covariate CD4 count at time k. Nevertheless, even under our assumption of no unmeasured confounders, the parameter β1 differs from the parameter β1 of our MSM. What is worse is that the parameter β1 will generally not have a causal interpretation, even if the model for pr[Y = 1|Ā = ā, ¯L = ¯l] is correctly specified. This is because cum (Ā) depends on a subject’s entire treatment history, including A0, and A0 may affect the time-dependent covariates Lk and Lk−1. Fitting a logistic model that adjusts for a covariate that is both affected by treatment and is a risk factor for the outcome provides an unbiased estimate of the association parameter β1 but a biased estimate of the causal parameter β1. This is true even under the null hypothesis of no direct, indirect, or overall treatment effect (so that β1 of model 12 equals 0) when, as in Figure 1, a component of Lk (for example, red blood count) and the outcome Y have an unmeasured common cause U0 (for example, the baseline number of bone marrow stem cells). 5,7,22–26

Equation U16
Equation U16
Image Tools

To summarize, standard regression methods adjust for covariates by including them in the model as regressors. These standard methods may fail to adjust appropriately for confounding due to measured confounders Lk when treatment is time varying, because (1) Lk may be a confounder for later treatment and thus must be adjusted for, but (2) may also be affected by earlier treatment and thus should not be adjusted for by standard methods. A solution to this conundrum is to adjust for the time-dependent covariates Lk by using them to calculate the weights swi rather than by adding the Lk to the regression model as regressors.

Back to Top | Article Outline

8. Estimation of the Weights

We now describe how to estimate the weights swi. For simplicity, we again assume that the treatment Ak at each time k is dichotomous. Consider first the denominator of model 14. We begin by estimating the unknown probability pr[Ak = 1|Āk = 1|Āk−1 = āk−1, ¯Lk = ¯lk] using a pooled logistic model that treats each person-day as an observation. For example, we might fit the model MATH where, for example, lk is the vector of CD4 count, WBC, hematocrit, and an indicator for symptoms at time k, and the α4, α5, α6, and α7 are row vectors. This model says that the probability of being treated on day k depends in a linear logistic fashion on the day k, the previous 2 days’ treatment, the current and previous days’ covariates, an interaction between yesterday’s treatments and today’s covariates, and the baseline covariates.

Equation 17
Equation 17
Image Tools

One can fit model 15 using any standard logistic regression program. The numerator probabilities can be estimated similarly, except that, in fitting model 15, we remove the last four terms that are functions of the covariates. That is, we fit the model as follows:MATH

Equation U18
Equation U18
Image Tools

For each subject i, we then have our logistic program output the estimated predicted values p̂0i, …, p̂Ki from the fit of model 15, which are maximum likelihood estimates of pr[Ak = 1|Āk−1 = ā(k−1)i, ¯Lk = ¯lki]. Similarly, we have outputted the predicted values p̂*1i, …, p̂*Ki from model 16, which are estimates of the quantities pr[Ak = 1|Āk−1 = ā(k−1)i]. Then we estimate swi by MATH For example, 1 − p̂*ki is an estimate of the probability pr[Ak = akik−1 = ā(k−1)i] when aki = 0. The data analyst will need to write a small program to compute swi for each subject from the predicted values outputted from the fit of models such as 15 and 16.

Equation U19
Equation U19
Image Tools

Under our assumption of no unmeasured confounders, the resulting estimate of the causal parameter β1 will be unbiased, provided the model 15 for pr[Ak = 1|Āk−1 = āk−1, ¯Lk = ¯lk] is correctly specified. Furthermore, under these same conditions, the 95% robust Wald confidence interval will be guaranteed to cover β1 at least 95% of the time. The estimate of β1 will remain unbiased even if the model 16 for pr[Ak = 1|Āk−1 = āk−1] is misspecified. 1,2 Indeed, if model 15 is correct and treatment is confounded, model 16 is guaranteed to be somewhat misspecified because of the noncollapsibility of logistic models. 25

Back to Top | Article Outline

9. Effect Modification by Pretreatment Covariates

MSMs can be generalized to allow one to include pretreatment covariates. For example, model 12 could be generalized to MATH where V is a component of the vector of measured pretreatment covariates L0, and β3 denotes a treatment-covariate interaction. Note in model 18, β1 + β3v represents the effect of cumulative treatment on a linear logistic scale within level v of the baseline variable V. As our IPTW estimators already automatically adjust for any confounding due to V, the particular subset V of L0 that an investigator chooses to include in model 18 should only reflect the investigator’s substantive interest. For example, a variable V should be included in model 18 only if the investigator both believes that V may be an effect modifier and has greater substantive interest in the causal effect of treatment within levels of the covariate V than in the source population as a whole.

Equation U20
Equation U20
Image Tools

We obtain unbiased estimates of the parameters of model 18 under the assumption of no unmeasured confounders by fitting an association model such as MATH using Proc Genmod with the estimated weights swi of Eq 17, modified only in that p*k is now the estimated predicted value from the fit of a model such as MATH Elementary epidemiologic textbooks emphasize that effect modification is logically distinct from confounding. Nonetheless, many students have difficulty understanding the distinction, because the same statistical methods (stratification and regression adjustments) are used both for confounder control and detection of effect modification. Thus, there may be some advantage to teaching elementary epidemiologic methods using marginal structural models, because then methods for confounder control (inverse-probability-of-treatment weighting) are distinct from methods for detection of effect modification (adding treatment covariate interaction terms to an MSM).

Equation U21
Equation U21
Image Tools
Equation U22
Equation U22
Image Tools

Finally, an important caveat: MSMs cannot be used to model the interaction of treatment with a time-varying covariate. For this, structural nested models should be used. 4–6 Therefore, it is not valid to include in the covariate V in model 18 any components of the time-dependent covariate Lk measured at any time k > 0.

Back to Top | Article Outline

10. Censoring by Loss to Follow-Up

Heretofore, we have assumed that each study subject is observed until end of follow-up at time K + 1. In this section, we allow for censoring by loss to follow-up. Specifically, let Ck = 1 if a subject was lost to follow-up by day k and Ck = 0 otherwise. We assume that once a subject is lost to follow-up, the subject does not later re-enter follow-up. No new idea is required to account for censoring, provided we conceptualize censoring as just another time-varying treatment. From this point of view, to want to adjust for censoring is only to say that our interest is in the causal effect of the treatment Ā if, contrary to fact, all subjects had remained uncensored, rather than having followed their observed censoring history. Our goal remains to estimate the parameter β1 of the logistic MSM 12 except now Yā refers to a subject’s outcome if, possibly contrary to fact, the subject has followed treatment history ā and has never been censored. Again, we can do so if there are no unmeasured confounders for both treatment and censoring. To formalize this idea, one adds at each time k the variable Ck to the graph in Figure 1 just before Lk and after Ak−1. Then, the assumption of no unmeasured confounders for treatment and censoring is that no arrow arising from the unmeasured causal risk factors U goes directly into either Ck or Ak for any k. In that case, the measured covariates Lk are sufficient to adjust both for confounding and selection bias due to loss to follow-up.

Again, we can obtain unbiased estimates of the causal parameters β1 by fitting the linear logistic association model 13 with appropriate weights included. Because the outcome Y is unobserved unless the subject does not drop out, that is, ¯C = (C0, …, CK+1) = 0, our weighted logistic regression fit of model 13 is restricted to uncensored subjects. The required subject-specific weight is swi × swi, where MATH and, in addition, in defining and estimating swi, we now add to the right side of each conditioning event in models 14–16 the event ¯Ck = 0, because otherwise, Ak would not be observed. The unknown probabilities in swi can be estimated using a pooled logistic model that treats each person-day as an observation. Specifically, we fit analogs of models 15 and 16 for logit pr[Ck = 0|¯Ck−1 = 0, Āk−1 = āk−1, ¯Lk = ¯lk] and for logit pr[Ck = 0|¯Ck−1 = 0, Āk−1 = āk−1]. Note that the denominator of the product swi × swi is informally the conditional probability that an uncensored subject had his or her observed treatment and censoring history through time K + 1. Thus, we refer to our weighted logistic estimator as an inverse-probability-of-treatment-and-censoring weighted estimator. If we view (Ak, Ck) as a “joint treatment” at time k, then one can informally interpret this denominator as simply the probability that a subject follows his or her own treatment history, which is exactly the interpretation that we had previously in the absence of censoring.

Equation U23
Equation U23
Image Tools
Back to Top | Article Outline

11. Limitations of Marginal Structural Models

It is shown in Ref 2 and Appendix 2 that our IPTW estimators will be biased and thus MSMs should not be used in studies in which at each time k there is a covariate level lk such that all subjects with that level of the covariate are certain to receive the identical treatment ak. For example, this circumstance implies that MSMs should not be used in occupational cohort studies. To see why, consider an occupational cohort study in which Ak is the level of exposure to an industrial chemical at time k and Lk = 1 if a subject is off work at time k and Lk = 0 otherwise. Then all subjects with Lk = 1 have Ak = 0, because all subjects off work are unexposed. Similarly, in a study of the effect of screening on mortality from cervical cancer, women who have had their cervix operatively removed by time k (which we denote by Lk = 0) cannot receive exposure (that is, screening) at that time, so MSMs should not be used. Nevertheless, g-estimation of structural nested models can always be used to estimate exposure effects, even in studies in which MSMs cannot be used. In many studies, such as the analysis of the Multicenter AIDS Cohort Study data described in our companion paper, 27 we believe, based on substantive considerations, the above difficulty does not occur and MSMs are a practical method.

Back to Top | Article Outline

12. Conclusion

We have described how to use MSMs to estimate the causal effect of a time-varying exposure or treatment on a dichotomous outcome. In our companion paper, 27 we extend our results to survival time outcomes and compare and contrast methods based on MSMs to alternative, previously proposed methods, based on g-estimation of structural nested models and on estimation of the g-computation algorithm formula.

Back to Top | Article Outline

References

1. Robins JM. Marginal structural models. In: 1997 Proceedings of the Section on Bayesian Statistical Science, Alexandria, VA: American Statistical Association, 1998;1–10.

2. Robins JM. Marginal structural models versus structural nested models as tools for causal inference. In: Halloran E, Berry D, eds. Statistical Models in Epidemiology: The Environment and Clinical Trials. New York: Springer-Verlag, 1999; 95–134.

3. Robins JM. Correction for non-compliance in equivalence trials. Stat Med 1998; 17: 269–302.

4. Robins JM. Structural nested failure time models. In: Andersen PK, Keiding N, section eds. Survival Analysis. In: Armitage P, Colton T, eds. The Encyclopedia of Biostatistics. Chichester, UK: John Wiley and Sons, 1998;4372–4389.

5. Robins JM, Blevins D, Ritter G, Wulfsohn M. G-estimation of the effect of prophylaxis therapy for Pneumocystis carinii pneumonia on the survival of AIDS patients (erratum in Epidemiology 1993;3: 189). Epidemiology 1992; 3: 319–336.

6. Witteman JCM, D’Agostino RB, Stijnen T, Kannel WB, Cobb JC, de Ridder MAJ, Hofman A, Robins JM. G-estimation of causal effects: isolated systolic hypertension and cardiovascular death in the Framingham Heart Study. Am J Epidemiol 1998; 148: 390–401.

7. Robins JM. A graphical approach to the identification and estimation of causal parameters in mortality studies with sustained exposure periods. J Chron Dis 1987; 40 (suppl 2): 139S–161S.

8. Pearl J. Causal diagrams for empirical research. Biometrika 1995; 82: 669–688.

9. Pearl J, Robins JM. Probabilistic evaluation of sequential plans from causal models with hidden variables. In: Uncertainty in Artificial Intelligence: Proceedings of the Eleventh Conference on Artificial Intelligence. San Francisco: Morgan Kaufmann, 1995; 444–453.

10. Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology 1999; 10: 37–48.

11. Robins JM. A new approach to causal inference in mortality studies with sustained exposure periods: application to control of the healthy worker survivor effect (erratum appear in Math Modelling 1987;14:917–921). Mathematical Modelling 1987; 7: 1393–1512.

12. Robins JM. The analysis of randomized and non-randomized AIDS treatment trials using a new approach to causal inference in longitudinal studies. In: Sechrest L, Freeman H, Mulley A, eds. Health Services Research Methodology: A Focus on AIDS. Rockville, MD: National Center for Health Services Research, U.S. Public Health Service, 1989; 113–159.

13. SAS Institute Inc. SAS/STAT User’s Guide Version 8. Cary, NC: SAS Institute, 1999.

14. S-PLUS 4 Guide to Statistics. Seattle: Mathsoft, 1997.

15. GAUSS. Maple Valley, WA: Aptech Systems, 1996.

16. Stata. Stata Statistical Software: Release 6.0. College Station, TX: Stata Corporation, 1999.

17. Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika 1983; 70: 41–55.

18. Huber PJ. The behavior of maximum likelihood estimates under non-standard conditions. In: Proceedings of the Fifth Berkeley Symposium in Mathematical Statistics and Probability. Berkeley: University of California Press, 1967; 221–233.

19. Carey VJ. YAGS 1.5.1.5: GEE solver for S-plus. In http://biosun1.harvard.edu/carey/index.ssoft.html, 1998.

20. Källén A. Generalized estimating equations: marginal methods for the analysis of correlated data. In: Riemann Library. http://www.aptech.com, 1999.

21. Stata Corp. Obtaining robust variance estimates. In: Stata Statistical Software: Release 5.0 User’s Guide. College Station, TX: Stata Corporation, 1997;235–239.

22. Rosenbaum PR. The consequences of adjustment for a concomitant variable that has been affected by treatment. J R Stat Soc 1984; 147: 656–666.

23. Greenland S, Neutra RR. An analysis of detection bias and proposed corrections in the study of estrogens and endometrial cancer. J Chron Dis 1981; 34: 433–438.

24. Greenland S, Neutra RR. Control of confounding in the assessment of medical technology. Int J Epidemiol 1981; 9: 361–367.

25. Greenland S. Interpretation and choice of effect measures in epidemiologic analyses. Am J Epidemiol 1987; 125: 761–768.

26. Robins JM, Greenland S, Hu F-C. Estimation of the causal effect of a time-varying exposure on the marginal mean of a repeated binary outcome. J Am Stat Assoc 1999; 94: 687–700.

27. Hernán MA, Brumback B, Robins JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology 2000; 11: 561–570.

28. Horvitz DG, Thompson DJ. A generalization of sampling without replacement from a finite universe. J Am Stat Assoc 1952; 47: 663–685.

29. Rosenbaum PR. Model-based direct adjustment. J Am Stat Assoc 1987; 82: 387–394.

30. Robins JM, Rotnitzky A. Recovery of information and adjustment for dependent censoring using surrogate markers. In: Jewell N, Dietz K, Farewell V, eds. AIDS Epidemiology: Methodological Issues. Boston: Birkhäuser, 1992, pp 24–33.

31. Robins JM. Information recovery and bias adjustment in proportional hazards regression analysis of randomized trials using surrogate markers. In: Proceedings of the American Statistical Association, Biopharmaceutical Section, 1993;24–33.

Back to Top | Article Outline
Appendix 1
Example

We will analyze the data in Table A1 under the assumption of no unmeasured confounders given L0. For convenience, we shall ignore sampling variability and thus the distinction between parameters of the source population and their empirical estimates. Under the assumption of no unmeasured confounder, pr(Ya0=1 = 1) is a weighted average of the L0-stratum-specific risks among the treated with weights proportional to the distribution of L0 in the entire study population. That is, pr(Ya0=1 = 1) is given by MATH where the sum is over the possible values of L0. 17 We refer to Eq. A1 as the L0-standardized risk in the treated. Calculating from Table A1, we obtain that pr(Ya0=1 = 1) = 0.32. Similarly, pr(Ya0=0 = 1) is the L0-standardized risk in the untreated, MATH which, from Table A1, is 0.64. It follows that the causal risk difference, risk ratio, and odds ratio are −0.32, 0.50, and 0.26. Note that these differ from the crude parameters computed from Table A2. Thus, ψ1 = −0.32, θ1 = log 0.50, and β1 = log 0.26 in models 1–3 differ from the parameters ψ1 = −0.40, θ1 = log .044, and β1 = log 0.18 of models 4–6.

Equation A1
Equation A1
Image Tools
Equation U24
Equation U24
Image Tools
Table 1
Table 1
Image Tools
Table 2
Table 2
Image Tools

As is well known, the causal risk difference and causal risk ratio are also equal to weighted averages of the stratum-specific risk differences and risk ratios. For example, the causal RD equals the standardized risk difference (SRD) where MATH and RDl0 = pr[Y = 1|A0 = 1, L0 = l0] − pr[Y = 1|A0 = 0, L0 = l0] is the risk difference in stratum l0. Indeed, the usual way to estimate the causal RD is to calculate the SRD. Our IPTW method is an alternative approach to estimation of the causal RD that, in contrast to the approach based on calculating the SRD, appropriately generalizes to unsaturated MSMs in longitudinal studies with time-varying treatments, as discussed in section 7.

Equation U25
Equation U25
Image Tools

Table A3 displays the data from the study in a different format. In particular, it gives the number of subjects with each of the possible combinations of l0, a0, and y, as well as the weight w = 1/pr[A0 = a0|L0 = l0] associated with each. The final column of the table represents the number of subjects in the weighted pseudopopulation for each combination of (l0, a0, y). Note that the weights wi need not be whole numbers or sum to 1. As a consequence, the number of subjects in the pseudopopulation can be greater than the number in the actual population. Tables A4 and A5 display the data from the pseudopopulation in the same format as Tables A1 and A2. It can be seen that L0 and A0 are unassociated in the pseudopopulation, which implies that the treatment is unconfounded. Furthermore, the lack of association between L0 and A0 implies that in the pseudopopulation, the L0-standardized risk in the treated equals the crude risk pr(Y = 1|A0 = 1) = 0.32 and the L0-standardized risk in the untreated equals the crude risk pr(Y = 1|A0 = 0) = 0.64. Furthermore, the crude risk in the treated pseudopopulation equals the L0-standardized risk in the treated actual population and thus equals pr(Ya0=1 = 1). Similarly, the crude risk in the untreated pseudopopulation equals the L0-standardized risk in the untreated true population and thus equals pr(Ya0=0 = 1). It follows that, under the assumption of no unmeasured confounder given L0, the crude risk difference, risk ratio, and odds ratio in the pseudopopulation equal the causal risk difference, risk ratio, and odds ratio in the actual population. Finally, an IPTW analysis in Proc Genmod estimates a crude parameter of the pseudopopulation and thus a causal parameter of the actual population.

Table 3
Table 3
Image Tools
Table 4
Table 4
Image Tools
Table 5
Table 5
Image Tools
Back to Top | Article Outline
Relation to Propensity Score and Horvitz-Thompson Methods

Rosenbaum and Rubin 17 refer to the probability pi = pr[A0 = 1|L0 = l0i] that subject i would receive treatment as the propensity score. Note that IPTW weight wi is not simply the inverse of the propensity score. Specifically, although wi is the inverse of the propensity score for treated subjects, it is the inverse of 1 − pi for untreated subjects. Rosenbaum and Rubin 17 showed that, under the assumption of no unmeasured confounders, one can control for confounding due to measured covariates in a point-treatment study with a dichotomous treatment by regarding the propensity score as the sole confounder. Because the propensity score pi is a continuous covariate, however, they suggested that, in practice, one either approximately match treated with untreated subjects on the propensity score or stratify (that is, subclassify) on the basis of propensity score quintiles. Even when there are no unmeasured confounders and the propensity score is unbiasedly estimated, Rosenbaum and Rubin’s 17 approach, unlike our approach, suffers from the potential for substantial residual confounding due to the inability to obtain sufficiently close matches or to uncontrolled intrastratum confounding. More importantly, Rosenbaum and Rubin’s 17 propensity score methods, in contrast to our IPTW methods, do not generalize straightforwardly to studies with nondichotomous or time-dependent treatments or exposures.

In the special case of a dichotomous time-independent treatment, our IPTW estimator is essentially equivalent to estimating pr(Ya0=0 = 1) and pr(Ya0=1 = 1) separately among the untreated (a0 = 0) and treated (a0 = 1) using the Horvitz-Thompson estimator 28 from the sample survey literature. 29 Robins and Rotnitzky 30 and Robins 31 proposed generalizations of the Horvitz-Thompson estimator that could be used to estimate the parameters of a saturated MSM model with a time-varying treatment. Our IPTW estimators are further generalizations that allow the estimation of nonsaturated MSMs with both time-independent and time-dependent treatments.

Back to Top | Article Outline
Appendix 2
Bias of Inverse-Probability-of-Treatment Weighted Estimators in the Setting of Section 11

Consider a new study population for which pr(Ya0=1 = 1) and pr(Ya0=0 = 1), and therefore the causal risk difference, are the same as for the population in Tables 1–5. The observed data for the new population, however, given in Table A6, differs from the observed data for the population studied in Tables 1–5. Specifically, Table A6 differs from the observed data in Table A1 only in that no subject with L0 = 1 receives treatment A0 = 1, that is, MATH and thus represents the type of study discussed in section 11. We will show that when Eq A2 holds the IPTW estimator of the causal risk, difference is now biased.

Equation A2
Equation A2
Image Tools
Table 6
Table 6
Image Tools

In Table A6, pr(Ya0=0 = 1) is again 0.64, the L0-standardized risk in the untreated. Nevertheless, the L0-standardized risk in the treated pr(Y = 1|A0 = 1, L0 = 0)pr(L0 = 0) + pr(Y = 1|A0 = 1, L0 = 1)pr(L0 = 1) cannot be computed from the data in Table A6, because there is no subject with history (A0 = 1, L0 = 1), rendering pr(Y = 1|A0 = 1, L0 = 1) uncomputable. Similarly, the SRD is not computable, because the stratum-specific risk difference is undefined in the stratum L0 = 1. Thus, pr(Ya0=1 = 1) and the causal risk difference are not computable from the data in Table A6, although we know by assumption that they are still equal to the previous values 0.32 and −0.32. Table A7 displays the data in Table A6 in the format of Table A3. Tables A8 and A9 display the stratified and crude data for the pseudopopulation constructed from the last column of Table A7. Note that the SRD in Table A8 for the pseudopopulation is undefined. The pseudopopulation crude RD from Table A9 is −0.24, which differs from the true causal risk difference ψ1 = −0.32. As discussed previously, however, it is the crude RD in the pseudopopulation that our IPTW estimate of the parameter ψ1 in the MSM pr(Ya0 = 1) = ψ0 + ψ1a0 actually estimates. We conclude that our MSM estimate is biased for the causal risk difference ψ1.

Table 7
Table 7
Image Tools
Table 8
Table 8
Image Tools
Table 9
Table 9
Image Tools

Cited By:

This article has been cited 362 time(s).

Journal of the American Statistical Association
Marginal mean models for dynamic regimes
Murphy, SA; van der Laan, MJ; Robins, JM
Journal of the American Statistical Association, 96(): 1410-1423.

American Journal of Epidemiology
Estimating the effect of cardiovascular risk factors on all-cause mortality and incidence of coronary heart disease using G-estimation - The Atherosclerosis Risk in Communities Study
Tilling, K; Sterne, JAC; Szklo, M
American Journal of Epidemiology, 155(8): 710-718.

Psychologische Rundschau
Personal, proximal and distal factors in etiological research - In answer to Wietkunat
Amelang, M
Psychologische Rundschau, 54(3): 194-195.
10.1026//0033.54.3.194
CrossRef
British Journal of Haematology
Comparative analysis of clinical outcomes after allogeneic bone marrow transplantation versus peripheral blood stem cell transplantation from a related donor in Japanese patients
Tanimoto, TE; Yamaguchi, TS; Tanaka, Y; Saito, A; Tajima, K; Karasuno, T; Kasai, M; Kishi, K; Mori, T; Maseki, N; Morishima, S; Miyakoshi, S; Kasai, M; Ohno, Y; Kim, SW; Numata, A; Kami, M; Takaue, Y; Mori, S; Harada, M
British Journal of Haematology, 125(4): 480-493.
10.1111/j.1365-2141.2004.04943.x
CrossRef
Journal of Clinical Epidemiology
Commonalities in the classical, collapsibility and counterfactual concepts of confounding
Newman, SC
Journal of Clinical Epidemiology, 57(4): 325-329.
10.1016/j.jclinepi.2003.07.014
CrossRef
Environmental Health Perspectives
Lipid adjustment in the analysis of environmental contaminants and human health risks
Schisterman, EF; Whitcomb, BW; Louis, GMB; Louis, TA
Environmental Health Perspectives, 113(7): 853-857.
10.1289/ehp.7640
CrossRef
Sociological Methodology 2007, Vol 37
Identification and Estimation of Causal Effects With Time-Varying Treatments and Time-Varying Outcomes
Brand, JE; Xie, Y
Sociological Methodology 2007, Vol 37, 37(): 393-434.

Journal of Clinical Epidemiology
Methodological considerations, such as directed acyclic graphs, for studying "acute on chronic" disease epidemiology: Chronic obstructive pulmonary disease example
Tsai, CL; Camargo, CA
Journal of Clinical Epidemiology, 62(9): 982-990.
10.1016/j.jclinepi.2008.10.005
CrossRef
Biometrical Journal
Estimating Antihypertensive Effects of Combination Therapy in an Observational Study Using Marginal Structural Models
Sugihara, M; Kushiro, T; Saito, I; Matsushita, Y; Hiramatsu, K
Biometrical Journal, 51(5): 789-800.
10.1002/bimj.200900025
CrossRef
Journal of Clinical Epidemiology
A comparison of ad hoc methods to account for non-cancer AIDS and deaths as competing risks when estimating the effect of HAART on incident cancer AIDS among HIV-infected men
Shiels, MS; Cole, SR; Chmiel, JS; Margolick, J; Martinson, J; Zhang, ZF; Jacobson, LP
Journal of Clinical Epidemiology, 63(4): 459-467.
10.1016/j.jclinepi.2009.08.003
CrossRef
Educational Evaluation and Policy Analysis
Early-grade retention and children's reading and math learning in elementary years
Hong, G; Yu, B
Educational Evaluation and Policy Analysis, 29(4): 239-261.
10.3102/0162373707309073
CrossRef
Nephrology Dialysis Transplantation
Clinical research of kidney diseases III: Principles of regression and modelling
Ravani, P; Parfrey, P; Gadag, V; Malberti, F; Barrett, B
Nephrology Dialysis Transplantation, 22(): 3422-3430.
10.1093/ndt/gfm777
CrossRef
Clinical Infectious Diseases
Long-term effectiveness of highly active Antiretroviral therapy on the survival of children and adolescents with HIV infection: A 10-year follow-up study
Patel, K; Hernan, MA; Williams, PL; Seeger, JD; McIntosh, K; Van Dyke, RB; Seage, GR
Clinical Infectious Diseases, 46(4): 507-515.
10.1086/526524
CrossRef
AIDS Research and Human Retroviruses
Long-term efficacy and tolerance of efavirenz- and nevirapine-containing regimens in adult HIV type 1 Senegalese patients
de Beaudrap, P; Etard, JF; Gueye, FN; Gueye, M; Landman, R; Girard, PM; Sow, PS; Ndoye, I; Delaporte, E
AIDS Research and Human Retroviruses, 24(6): 753-760.
10.1089/aid.2007.0295
CrossRef
Psychiatric Annals
Balancing Treatment Comparisons in Longitudinal Studies
Marcus, SM; Siddique, J; Ten Have, TR; Gibbons, RD; Stuart, E; Normand, SLT
Psychiatric Annals, 38(): 805-811.

Social Science & Medicine
Job loss is bad for your health - Swedish evidence on cause-specific hospitalization following involuntary job loss
Eliason, M; Storrie, D
Social Science & Medicine, 68(8): 1396-1406.
10.1016/j.socscimed.2009.01.021
CrossRef
Revue D Epidemiologie Et De Sante Publique
Time-dependent covariates in the Cox proportional hazards model. Theory and practice
Desquilbet, L; Meyer, L
Revue D Epidemiologie Et De Sante Publique, 53(1): 51-68.

Clinical Trials
Estimation of treatment effect adjusting for dependent censoring using the IPCW method: an application to a large primary prevention study for coronary events (MEGA study)
Yoshida, M; Matsuyama, Y; Ohashi, Y
Clinical Trials, 4(4): 318-328.
10.1177/1740774507081224
CrossRef
Circulation
Long-Term Safety and Effectiveness of Unprotected Left Main Coronary Stenting With Drug-Eluting Stents Compared With Bare-Metal Stents
Kim, YH; Park, DW; Lee, SW; Yun, SC; Lee, CW; Hong, MK; Park, SW; Seung, KB; Gwon, HC; Jeong, MH; Jang, Y; Kim, HS; Seong, IW; Park, HS; Ahn, T; Chae, IH; Tahk, SJ; Chung, WS; Park, SJ
Circulation, 120(5): 400-407.
10.1161/CIRCULATIONAHA.108.800805
CrossRef
Tropical Medicine & International Health
Mortality after failure of antiretroviral therapy in sub-Saharan Africa
Keiser, O; Tweya, H; Braitstein, P; Dabis, F; MacPhail, P; Boulle, A; Nash, D; Wood, R; Luthi, R; Brinkhof, MWG; Schechter, M; Egger, M
Tropical Medicine & International Health, 15(2): 251-258.
10.1111/j.1365-3156.2009.02445.x
CrossRef
American Journal of Epidemiology
Secondhand Smoke, Vascular Disease, and Dementia Incidence: Findings From the Cardiovascular Health Cognition Study
Barnes, DE; Haight, TJ; Mehta, KM; Carlson, MC; Kuller, LH; Tager, IB
American Journal of Epidemiology, 171(3): 292-302.
10.1093/aje/kwp376
CrossRef
International Journal of Biostatistics
Special Issue on Causal Inference
Moodie, EEM; Stephens, DA
International Journal of Biostatistics, 6(2): -.
ARTN 1
CrossRef
AIDS and Behavior
Interventions Delivered in Clinical Settings are Effective in Reducing Risk of HIV Transmission Among People Living with HIV: Results from the Health Resources and Services Administration (HRSA)'s Special Projects of National Significance Initiative
Myers, JJ; Shade, SB; Rose, CD; Koester, K; Maiorana, A; Malitz, FE; Bie, J; Kang-Dufour, MS; Morin, SF
AIDS and Behavior, 14(3): 483-492.
10.1007/s10461-010-9679-y
CrossRef
Statistics in Medicine
Designing a sequentially randomized study with adherence enhancing interventions for diabetes patients
Buyze, J; Van Rompaye, B; Goetghebeur, E
Statistics in Medicine, 29(): 1114-1126.
10.1002/sim.3856
CrossRef
Kidney International
Vitamin D receptor activation and survival in chronic kidney disease
Kovesdy, CP; Kalantar-Zadeh, K
Kidney International, 73(): 1355-1363.
10.1038/ki.2008.35
CrossRef
Statistics in Medicine
Impact of mis-specification of the treatment model on estimates from a marginal structural model
Lefebvre, G; Delaney, JAC; Platt, RW
Statistics in Medicine, 27(): 3629-3642.
10.1002/sim.3200
CrossRef
Annals of Epidemiology
Time-Dependent Confounding in the Study of the Effects of Regular Physical Activity in Chronic Obstructive Pulmonary Disease: An Application of the Marginal Structural Model
Garcia-Aymerich, J; Lange, P; Serra, I; Schnohr, P; Anto, JM
Annals of Epidemiology, 18(): 775-783.
10.1016/j.annepidem.2008.05.003
CrossRef
American Journal of Kidney Diseases
Randomized and Observational Studies in Nephrology: How Strong Is the Evidence?
Greene, T
American Journal of Kidney Diseases, 53(3): 377-388.
10.1053/j.ajkd.2009.12.001
CrossRef
International Journal of Epidemiology
Role of breastfeeding cessation in mediating the relationship between maternal HIV disease stage and increased child mortality among HIV-exposed uninfected children
Fox, MP; Brooks, DR; Kuhn, L; Aldrovandi, G; Sinkala, M; Kankasa, C; Horsburgh, R; Thea, DM
International Journal of Epidemiology, 38(2): 569-576.
10.1093/ije/dyn249
CrossRef
Biostatistics
Estimating equation - based causality analysis with application to microarray time series data
Hu, JH; Hu, FF
Biostatistics, 10(3): 468-480.
10.1093/biostatistics/kxp005
CrossRef
Journal of Statistical Planning and Inference
Why prefer double robust estimators in causal inference?
Neugebauer, R; van der Laan, M
Journal of Statistical Planning and Inference, 129(): 405-426.
10.1016/j.jspi.2004.06.060
CrossRef
New England Journal of Medicine
A national evaluation of the effect of trauma-center care on mortality
MacKenzie, EJ; Rivara, FP; Jurkovich, GJ; Nathens, AB; Frey, KP; Egleston, BL; Salkever, DS; Scharfstein, DO
New England Journal of Medicine, 354(4): 366-378.

American Journal of Epidemiology
The birth weight "paradox" uncovered?
Hernandez-Diaz, S; Schisterman, EF; Hernan, MA
American Journal of Epidemiology, 164(): 1115-1120.
10.1093/aje/kwj275
CrossRef
Statistics in Medicine
Individualized treatment rules: Generating candidate clinical trials
Petersen, ML; Deeks, SG; van der Laan, MJ
Statistics in Medicine, 26(): 4578-4601.
10.1002/sim.2888
CrossRef
Journal of the American Society of Nephrology
Hemoglobin variability and mortality in ESRD
Yang, W; Israni, RK; Brunelli, SM; Joffe, MM; Fishbane, S; Feldman, HI
Journal of the American Society of Nephrology, 18(): 3164-3170.
10.1681/ASN.2007010058
CrossRef
Quality of Life Research
Association of clinical experiences with patient-reported outcomes among breast cancer surgery patients: breast cancer quality care study
Noh, DY; Nam, SJ; Ahn, SH; Park, BW; Lee, ES; Lee, MK; Kim, SH; Kim, YM; Park, SM; Yun, YH
Quality of Life Research, 17(2): 215-225.
10.1007/s11136-007-9299-9
CrossRef
Journal of Acquired Immune Deficiency Syndromes
The relationship of preventable opportunistic infections, HIV-1 RNA, and CD4 cell counts to chronic mortality
Seage, GR; Losina, E; Goldie, SJ; Paltiel, AD; Kimmel, AD; Freedberg, KA
Journal of Acquired Immune Deficiency Syndromes, 30(4): 421-428.
10.1097/01.QAI.0000021700.48448.69
CrossRef
British Medical Journal
Stable partnership and progression to AIDS or death in HIV infected patients receiving highly active antiretroviral therapy: Swiss HIV cohort study
Young, J; De Geest, S; Spirig, R; Flepp, M; Rickenbach, M; Furrer, H; Bernasconi, E; Hirschel, B; Telenti, A; Vernazza, P; Battegay, M; Bucher, HC
British Medical Journal, 328(): 15-18B.

Human Reproduction
Risk of uterine leiomyomata in relation to tobacco, alcohol and caffeine consumption in the Black Women's Health Study
Wise, LA; Palmer, JR; Harlow, BL; Spiegelman, D; Stewart, EA; Adams-Campbell, LL; Rosenberg, L
Human Reproduction, 19(8): 1746-1754.
10.1093/humrep/deh309
CrossRef
Statistics in Medicine
Covariate adjustment in clinical trials with non-ignorable missing data and non-compliance
Levy, DE; O'Malley, AJ; Normand, SLT
Statistics in Medicine, 23(): 2319-2339.
10.1002/sim.1848
CrossRef
Journal of Lipid Research
Longitudinal impact of physical activity on lipid profiles in middle-aged adults: the Atherosclerosis Risk in Communities Study
Monda, KL; Ballantyne, CM; North, KE
Journal of Lipid Research, 50(8): 1685-1691.
10.1194/jlr.P900029-JLR200
CrossRef
Antiviral Therapy
Virological and immunological responses to efavirenz or boosted lopinavir as first-line therapy for patients with HIV
Young, J; Bucher, HC; Guenthard, HF; Rickenbach, M; Fux, CA; Hirschel, B; Cavassini, M; Vernazza, P; Bernasconi, E; Battegay, M
Antiviral Therapy, 14(6): 771-779.
10.3851/IMP1291
CrossRef
Kidney International
Shorter dialysis times are associated with higher mortality among incident hemodialysis patients
Brunelli, SM; Chertow, GM; Ankers, ED; Lowrie, EG; Thadhani, R
Kidney International, 77(7): 630-636.
10.1038/ki.2009.523
CrossRef
International Journal of Biostatistics
Accuracy of Conventional and Marginal Structural Cox Model Estimators: A Simulation Study
Xiao, YL; Abrahamowicz, M; Moodie, EEM
International Journal of Biostatistics, 6(2): -.
ARTN 13
CrossRef
International Journal of Epidemiology
Illustrating bias due to conditioning on a collider
Cole, SR; Platt, RW; Schisterman, EF; Chu, HT; Westreich, D; Richardson, D; Poole, C
International Journal of Epidemiology, 39(2): 417-420.
10.1093/ije/dyp334
CrossRef
American Sociological Review
Stained Red: A Study of Stigma by Association to Blacklisted Artists during the ''Red Scare'' in Hollywood, 1945 to 1960
Pontikes, E; Negro, G; Rao, H
American Sociological Review, 75(3): 456-478.
10.1177/0003122410368929
CrossRef
Biometrika
Population intervention models in causal inference
Hubbard, AE; Van der Laan, MJ
Biometrika, 95(1): 35-47.
10.1093/biomet/asm097
CrossRef
Statistics in Medicine
Estimation of treatment effect adjusting for treatment changes using the intensity score method: Application to a large primary prevention study for coronary events (MEGA study)
Tanaka, Y; Matsuyama, Y; Ohashi, Y
Statistics in Medicine, 27(): 1718-1733.
10.1002/sim.3065
CrossRef
Archives of Pediatrics & Adolescent Medicine
Propensity scores
Cummings, P
Archives of Pediatrics & Adolescent Medicine, 162(8): 734-737.

Journal of Educational and Behavioral Statistics
Causal inference for time-varying instructional treatments
Hong, GL; Raudenbush, SW
Journal of Educational and Behavioral Statistics, 33(3): 333-362.
10.3102/1076998607307355
CrossRef
Pharmacoepidemiology and Drug Safety
Traditional versus marginal structural models to estimate the effectiveness of beta-blocker use on mortality after myocardial infarction
Delaney, JAC; Daskalopoulou, SS; Suissa, S
Pharmacoepidemiology and Drug Safety, 18(1): 1-6.
10.1002/pds.1676
CrossRef
Biometrics
Related Causal Frameworks for Surrogate Outcomes
Joffe, MM; Greene, T
Biometrics, 65(2): 530-538.
10.1111/j.1541-0420.2008.01106.x
CrossRef
American Journal of Epidemiology
The effect on treatment comparisons of different measurement frequencies in human immunodeficiency virus observational databases
Griffin, JT; Fraser, C; Gras, L; de Wolf, F; Ghani, AC
American Journal of Epidemiology, 163(7): 676-683.
10.1093/aje/kwj083
CrossRef
American Journal of Epidemiology
Variable selection for propensity score models
Brookhart, MA; Schneeweiss, S; Rothman, KJ; Glynn, RJ; Avorn, J; Sturmer, T
American Journal of Epidemiology, 163(): 1149-1156.
10.1093/aje/kwj149
CrossRef
Journal of Managed Care Pharmacy
Observational study of the prevalence of febrile neutropenia in patients who received filgrastim or pegfilgrastim associated with 3-4 week chemotherapy regimens in community oncology practices
Morrison, VA; Wong, M; Hershman, D; Campos, LT; Ding, BY; Malin, J
Journal of Managed Care Pharmacy, 13(4): 337-348.

American Journal of Kidney Diseases
Protective effect of intravenous levocarnitine on subsequent-month hospitalization among prevalent hemodialysis patients, 1998 to 2003
Weinhandl, ED; Rao, M; Gilbertson, DT; Collins, AJ; Pereira, BJG
American Journal of Kidney Diseases, 50(5): 803-812.
10.1053/j.ajkd.2007.07.017
CrossRef
Clinical Gastroenterology and Hepatology
Effect of human immunodeficiency virus and antiretrovirals on outcomes of hepatitis C: A systematic review from an epidemiologic perspective
Kramer, JR; Giordano, TP; El-Serag, HB
Clinical Gastroenterology and Hepatology, 5(): 1321-1328.
10.1016/j.cgh.2007.08.006
CrossRef
Kidney International
The effect of epoetin dose on hematocrit
Cotter, D; Zhang, Y; Thamer, M; Kaufman, J; Hernan, MA
Kidney International, 73(3): 347-353.
10.1038/sj.ki.5002688
CrossRef
Journal of the American College of Cardiology
Long-Term Safety and Efficacy of Stenting Versus Coronary Artery Bypass Grafting for Unprotected Left Main Coronary Artery Disease 5-Year Results From the MAIN-COMPARE (Revascularization for Unprotected Left Main Coronary Artery Stenosis: Comparison of Percutaneous Coronary Angioplasty Versus Surgical Revascularization) Registry
Park, DW; Seung, KB; Kim, YH; Lee, JY; Kim, WJ; Kang, SJ; Lee, SW; Lee, CW; Park, SW; Yun, SC; Gwon, HC; Jeong, MH; Jang, YS; Kim, HS; Kim, PJ; Seong, IW; Park, HS; Ahn, T; Chae, IH; Tahk, SJ; Chung, WS; Park, SJ
Journal of the American College of Cardiology, 56(2): 117-124.
10.1016/j.jacc.2010.04.004
CrossRef
Journal of Epidemiology and Community Health
Effectiveness of highly active antiretroviral therapy among HIV-1 infected women
Gange, SJ; Barron, Y; Greenblatt, RM; Anastos, K; Minkoff, H; Young, M; Kovacs, A; Cohen, M; Meyer, WA
Journal of Epidemiology and Community Health, 56(2): 153-159.

Statistics in Medicine
Adjusting for differential proportions of second-line treatment in cancer clinical trials. Part I: Structural nested models and marginal structural models to test and estimate treatment arm effects
Yamaguchi, T; Ohashi, Y
Statistics in Medicine, 23(): 1991-2003.
10.1002/sim.1816
CrossRef
Statistica Sinica
Marginal structural quantile models for longitudinal observational studies with time-varying treatment
Hogan, JW; Lee, JY
Statistica Sinica, 14(3): 927-944.

Journal of Allergy and Clinical Immunology
Influences of earlier adherence and symptoms on current symptoms: A marginal structural models analysis
Kim, C; Feldman, HI; Joffe, M; Tenhave, T; Boston, R; Apter, AJ
Journal of Allergy and Clinical Immunology, 115(4): 810-814.
10.1016/j.jaci.2004.11.032
CrossRef
Journal of Epidemiology and Community Health
Estimating the potential impacts of intervention from observational data: methods for estimating causal attributable risk in a cross-sectional analysis of depressive symptoms in Latin America
Fleischer, NL; Fernald, LCH; Hubbard, AE
Journal of Epidemiology and Community Health, 64(1): 16-21.
10.1136/jech.2008.085985
CrossRef
Journal of Evaluation in Clinical Practice
Evaluating health management programmes over time: application of propensity score-based weighting to longitudinal data
Linden, A; Adams, JL
Journal of Evaluation in Clinical Practice, 16(1): 180-185.
10.1111/j.1365-2753.2009.01361.x
CrossRef
International Journal of Epidemiology
Causal thinking and complex system approaches in epidemiology
Galea, S; Riddle, M; Kaplan, GA
International Journal of Epidemiology, 39(1): 97-106.
10.1093/ije/dyp296
CrossRef
Technovation
Do royalties really foster university patenting activity? An answer from Italy
Baldini, N
Technovation, 30(2): 109-116.
10.1016/j.technovation.2009.09.007
CrossRef
Sociological Methods & Research
Direct and Indirect Effects for Neighborhood-Based Clustered and Longitudinal Data
VanderWeele, TJ
Sociological Methods & Research, 38(4): 515-544.
10.1177/0049124110366236
CrossRef
American Journal of Epidemiology
Generalizing Evidence From Randomized Clinical Trials to Target Populations
Cole, SR; Stuart, EA
American Journal of Epidemiology, 172(1): 107-115.
10.1093/aje/kwq084
CrossRef
Hiv Clinical Trials
The-incidence of HIV drug resistance and its impact on progression of HIV disease among antiretroviral-naive participants started on three different antiretroviral therapy strategies
Kozal, MJ; Hullsiek, KH; MacArthur, RD; van den Berg-Wolf, M; Peng, G; Xiang, Y; Baxter, JD; Uy, J; Telzak, EE; Novak, RM
Hiv Clinical Trials, 8(6): 357-370.
10.1310/hct0806-357
CrossRef
American Journal of Respiratory and Critical Care Medicine
The healthy worker effect in asthma - Work may cause asthma, but asthma may also influence work
Le Moual, N; Kauffmann, F; Eisen, EA; Kennedy, SM
American Journal of Respiratory and Critical Care Medicine, 177(1): 4-10.

Journal of Clinical Epidemiology
Fitting marginal structural models: estimating covariate-treatment associations in the reweighted data set can guide model fitting
Pullenayegum, EM; Lam, C; Manlhiot, C; Feldman, BM
Journal of Clinical Epidemiology, 61(9): 875-881.
10.1016/j.jclinepi.2007.10.024
CrossRef
Statistics in Medicine
Analyzing sequentially randomized trials based on causal effect models for realistic individualized treatment rules
Bembom, O; van der Laan, MJ
Statistics in Medicine, 27(): 3689-3716.
10.1002/sim.3268
CrossRef
Human Heredity
Improper Adjustment for Baseline in Genetic Association Studies of Change in Phenotype
McArdle, PF; Whitcomb, BW
Human Heredity, 67(3): 176-182.
10.1159/000181156
CrossRef
Statistics in Medicine
Inverse probability-of-censoring weights for the correction of time-varying noncompliance in the effect of randomized highly active antiretroviral therapy on incident AIDS or death
Cain, LE; Cole, SR
Statistics in Medicine, 28(): 1725-1738.
10.1002/sim.3585
CrossRef
Journal of the Royal Statistical Society Series B-Statistical Methodology
Causal inference with generalized structural mean models
Vansteelandt, S; Goetghebeur, E
Journal of the Royal Statistical Society Series B-Statistical Methodology, 65(): 817-835.

American Statistician
Model selection, confounder control, and marginal structural models: Review and new applications
Joffe, MM; Ten Have, TR; Feldman, HI; Kimmel, SE
American Statistician, 58(4): 272-279.

Statistical Methods in Medical Research
Estimating causal effects from multiple cycle data in studies of in vitro fertilization
Hogan, JW; Scharfstein, DO
Statistical Methods in Medical Research, 15(2): 195-209.
10.1191/0962280206sm439oa
CrossRef
Health Services Research
Hospice enrollment and evaluation of its causal effect on hospitalization of dying nursing home patients
Gozalo, PL; Miller, SC
Health Services Research, 42(2): 587-610.
10.1111/j.1475-6773.2006.00623.x
CrossRef
American Journal of Epidemiology
On the estimation and use of propensity scores in case-control and case-cohort studies
Mansson, R; Joffe, MM; Sun, WG; Hennessy, S
American Journal of Epidemiology, 166(3): 332-339.
10.1093/aje/kwm069
CrossRef
Medical Care
A simulation-based evaluation of methods to estimate the impact of an adverse event on hospital length of stay
Samore, MH; Shen, S; Greene, T; Stoddard, G; Sauer, B; Shinogle, J; Nebeker, J; Harbarth, S
Medical Care, 45(): S108-S115.

Journal of Evaluation in Clinical Practice
Using propensity score-based weighting in the evaluation of health management programme effectiveness
Linden, A; Adams, JL
Journal of Evaluation in Clinical Practice, 16(1): 175-179.
10.1111/j.1365-2753.2009.01219.x
CrossRef
Journal of Statistical Planning and Inference
Comparison of causal effect estimators under exposure misclassification
Babanezhad, M; Vansteelandt, S; Goetghebeur, E
Journal of Statistical Planning and Inference, 140(5): 1306-1319.
10.1016/j.jspi.2009.11.015
CrossRef
Journal of Clinical Epidemiology
Propensity score estimation: neural networks, support vector machines, decision trees (CART), and meta-classifiers as alternatives to logistic regression
Westreich, D; Lessler, J; Funk, MJ
Journal of Clinical Epidemiology, 63(8): 826-833.
10.1016/j.jclinepi.2009.11.020
CrossRef
Empirical Economics
Identification of the effects of dynamic treatments by sequential conditional independence assumptions
Lechner, M; Miquel, R
Empirical Economics, 39(1): 111-137.
10.1007/s00181-009-0297-3
CrossRef
International Journal of Obesity
Adjusting for reverse causality in the relationship between obesity and mortality
Flanders, WD; Augestad, LB
International Journal of Obesity, 32(): S42-S46.
10.1038/ijo.2008.84
CrossRef
Statistics in Medicine
Evaluating a surrogate endpoint at three levels, with application to vaccine development
Gilbert, PB; Qin, L; Self, SG
Statistics in Medicine, 27(): 4758-4778.
10.1002/sim.3122
CrossRef
American Journal of Epidemiology
Effect of Highly Active Antiretroviral Therapy on Incident AIDS Using Calendar Period as an Instrumental Variable
Cain, LE; Cole, SR; Greenland, S; Brown, TT; Chmiel, JS; Kingsley, L; Detels, R
American Journal of Epidemiology, 169(9): 1124-1132.
10.1093/aje/kwp002
CrossRef
Journal of the American Society of Nephrology
Resolved: Targeting a Higher Hemoglobin Is Associated with Greater Risk in Patients with CKD Anemia
Singh, AK; Himmelfarb, J; Szczecht, LA
Journal of the American Society of Nephrology, 20(7): 1436-1443.
10.1681/ASN.2009040444
CrossRef
Circulation
C-Reactive Protein and the Risk of Stent Thrombosis and Cardiovascular Events After Drug-Eluting Stent Implantation
Park, DW; Yun, SC; Lee, JY; Kim, WJ; Kang, SJ; Lee, SW; Kim, YH; Lee, CW; Kim, JJ; Park, SW; Park, SJ
Circulation, 120(): 1987-U29.
10.1161/CIRCULATIONAHA.109.876763
CrossRef
Journal of Quantitative Criminology
How Much Can We Trust Causal Interpretations of Fixed-Effects Estimators in the Context of Criminality?
Bjerk, D
Journal of Quantitative Criminology, 25(4): 391-417.
10.1007/s10940-009-9073-y
CrossRef
Antiviral Therapy
Lopinavir/ritonavir or efavirenz plus two nucleoside analogues as first-line antiretroviral therapy: a non-randomized comparison
De Luca, A; Cozzi-Lepri, A; Antinori, A; Zaccarelli, M; Bongiovanni, M; Di Giambenedetto, S; Marconi, P; Cicconi, P; Resta, F; Grisorio, B; Ciardi, M; Cauda, R; Monforte, AD
Antiviral Therapy, 11(5): 609-618.

Annals of Epidemiology
Secondhand smoke exposure, pulmonary function, and cardiovascular mortality
Eisner, MD; Wang, Y; Haight, TJ; Balmes, J; Hammond, SK; Tager, IB
Annals of Epidemiology, 17(5): 364-373.
10.1016/j.annepidem.2006.10.008
CrossRef
Psychological Medicine
Suicide and occupation: the impact of socio-economic, demographic and psychiatric differences
Ager, E; Gunnell, D; Bonde, JP; Mortensen, PB; Nordentoft, M
Psychological Medicine, 37(8): 1131-1140.
10.1017/S0033291707000487
CrossRef
Gaceta Sanitaria
Marginal structural models application to estimate the effects of antiretroviral therapy in 5 cohorts of HIV seroconverters
Perez-Hoyos, S; Ferreros, I; Hernan, MA
Gaceta Sanitaria, 21(1): 76-83.

Journal of Business & Economic Statistics
Sequential Causal Models for the Evaluation of Labor Market Programs
Lechner, M
Journal of Business & Economic Statistics, 27(1): 71-83.
10.1198/jbes.2009.0006
CrossRef
International Journal of Epidemiology
Opening the Black Box: a motivation for the assessment of mediation
Hafeman, DM; Schwartz, S
International Journal of Epidemiology, 38(3): 838-845.
10.1093/ije/dyn372
CrossRef
American Journal of Kidney Diseases
Evolving Statistical Methods to Facilitate Evaluation of the Causal Association Between Erythropoiesis-Stimulating Agent Dose and Mortality in Nonexperimental Research: Strengths and Limitations
Bradbury, BD; Brookhart, MA; Winkelmayer, WC; Critchlow, CW; Kilpatrick, RD; Joffe, MM; Feldman, HI; Acquavella, JF; Wang, OH; Rothman, KJ
American Journal of Kidney Diseases, 54(3): 554-560.
10.1053/j.ajkd.2009.05.010
CrossRef
Jama-Journal of the American Medical Association
Comparative Effectiveness of Minimally Invasive vs Open Radical Prostatectomy
Hu, JC; Gu, XM; Lipsitz, SR; Barry, MJ; D'Amico, AV; Weinberg, AC; Keating, NL
Jama-Journal of the American Medical Association, 302(): 1557-1564.

Journal of the American Society of Nephrology
Propensity-Matched Mortality Comparison of Incident Hemodialysis and Peritoneal Dialysis Patients
Weinhandl, ED; Foley, RN; Gilbertson, DT; Arneson, TJ; Snyder, JJ; Collins, AJ
Journal of the American Society of Nephrology, 21(3): 499-506.
10.1681/ASN.2009060635
CrossRef
Journal of Drug Issues
A Marginal Structural Modeling Approach to Assess the Cumulative Effect of Drug Treatment on Later Drug Use Abstinence
Li, LB; Evans, E; Hser, YI
Journal of Drug Issues, 40(1): 221-240.

Epidemiologia E Psichiatria Sociale-An International Journal for Epidemiology and Psychiatric Sciences
The methodological and conceptual aspects of mortality studies in psychiatry
Biggeri, A; Catelan, D
Epidemiologia E Psichiatria Sociale-An International Journal for Epidemiology and Psychiatric Sciences, 19(1): 16-20.

Plos One
Is Antiretroviral Therapy Causing Long-Term Liver Damage? A Comparative Analysis of HIV-Mono-Infected and HIV/Hepatitis C Co-Infected Cohorts
Moodie, EEM; Pai, NP; Klein, MB
Plos One, 4(2): -.
ARTN e4517
CrossRef
Journal of the National Cancer Institute
Intermediacy and Gene-Environment Interaction: The Example of CHRNA5-A3 Region, Smoking, Nicotine Dependence, and Lung Cancer
Wacholder, S; Chatterjee, N; Caporaso, N
Journal of the National Cancer Institute, 100(): -.
10.1093/jnci/djn380
CrossRef
Statistics & Probability Letters
A note on using the estimated versus the known propensity score to estimate the average treatment effect
Brumback, BA
Statistics & Probability Letters, 79(4): 537-542.
10.1016/j.spl.2008.09.032
CrossRef
European Journal of Epidemiology
The impact of unmeasured baseline effect modification on estimates from an inverse probability of treatment weighted logistic model
Delaney, JAC; Platt, RW; Suissa, S
European Journal of Epidemiology, 24(7): 343-349.
10.1007/s10654-009-9341-z
CrossRef
Journal of Asthma
Ethnic Disparities in Asthma Morbidity in Chicago
Evans, AT; Sadowski, LS; VanderWeele, TJ; Curtis, LM; Sharp, LK; Kee, RA; Grammer, LC; Lyttle, CS; Weiss, KB; Shannon, JJ
Journal of Asthma, 46(5): 448-454.
10.1080/02770900802492061
CrossRef
American Journal of Epidemiology
Time-modified Confounding
Platt, RW; Schisterman, EF; Cole, SR
American Journal of Epidemiology, 170(6): 687-694.
10.1093/aje/kwp175
CrossRef
International Journal of Biostatistics
Comparing Mortality in Renal Patients on Hemodialysis versus Peritoneal Dialysis Using a Marginal Structural Model
van der Wal, WM; Noordzij, M; Dekker, FW; Boeschoten, EW; Krediet, RT; Korevaar, JC; Geskus, RB
International Journal of Biostatistics, 6(1): -.
ARTN 2
CrossRef
Journal of the National Cancer Institute
Highly active antiretroviral therapy and cervical squamous intraepithelial lesions in human immunodeficiency virus-positive women
Ahdieh-Grant, L; Li, R; Levine, AM; Massad, LS; Strickler, HD; Minkoff, H; Moxley, M; Palefsky, J; Sacks, H; Burk, RD; Gange, SJ
Journal of the National Cancer Institute, 96(): 1070-1076.
10.1093/jnci/djh192
CrossRef
American Journal of Epidemiology
Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect
Kurth, T; Walker, AM; Glynn, RJ; Chan, KA; Gaziano, JM; Berger, K; Robins, JM
American Journal of Epidemiology, 163(3): 262-270.
10.1093/aje/kwj047
CrossRef
Pharmacoepidemiology and Drug Safety
Insights into different results from different causal contrasts in the presence of effect-measure modification
Sturmer, T; Rothman, KJ; Glynn, RJ
Pharmacoepidemiology and Drug Safety, 15(): 698-709.
10.1002/pds.1231
CrossRef
American Journal of Epidemiology
History-adjusted marginal structural models for estimating time-varying effect modification
Petersen, ML; Deeks, SG; Martin, JN; van der Laan, MJ
American Journal of Epidemiology, 166(9): 985-993.
10.1093/aie/kwm232
CrossRef
Proceedings of the National Academy of Sciences of the United States of America
Durable effects of concentrated disadvantage on verbal ability among African-American children
Sampson, RJ; Sharkey, P; Raudenbush, SW
Proceedings of the National Academy of Sciences of the United States of America, 105(3): 845-852.
10.1073/pnas.0710189104
CrossRef
Psychiatria Danubina
Acute phase results from storm, a multicountry observational study of bipolar disorder treatment and outcomes
Treuer, T; Oruc, L; Loza, N; El Saidi, MA; Kovacs, Z; Akkay, C; Guelseren, S; Saylan, M; Colman, SA; Harrison, GA
Psychiatria Danubina, 19(4): 282-295.

Annals of the Rheumatic Diseases
Early treatment with, and time receiving, first disease-modifying antirheumatic drug predicts long-term function in patients with inflammatory polyarthritis
Farragher, TM; Lunt, M; Fu, B; Bunn, D; Symmons, DPM
Annals of the Rheumatic Diseases, 69(4): 689-695.
10.1136/ard.2009.108639
CrossRef
American Journal of Epidemiology
Use of a marginal structural model to determine the effect of aspirin on cardiovascular mortality in the physicians' health study
Cook, NR; Cole, SR; Hennekens, CH
American Journal of Epidemiology, 155(): 1045-1053.

Economic Journal
The impact of active labour market programmes on the duration of unemployment in Switzerland
Lalive, R; van Ours, JC; Zweimuller, J
Economic Journal, 118(): 235-257.

American Journal of Epidemiology
Constructing inverse probability weights for marginal structural models
Cole, SR; Hernan, MA
American Journal of Epidemiology, 168(6): 656-664.
10.1093/aje/kwn164
CrossRef
Value in Health
Good Research Practices for Comparative Effectiveness Research: Analytic Methods to Improve Causal Inference from Nonrandomized Studies of Treatment Effects Using Secondary Data Sources: The ISPOR Good Research Practices for Retrospective Database Analysis Task Force Report-Part III
Johnson, ML; Crown, W; Martin, BC; Dormuth, CR; Siebert, U
Value in Health, 12(8): 1062-1073.
10.1111/j.1524-4733.2009.00602.x
CrossRef
Biometrika
A note on the variance of doubly-robust G-estimators
Moodie, EEM
Biometrika, 96(4): 998-1004.
10.1093/biomet/asp043
CrossRef
Lifetime Data Analysis
Adjusting for time-varying confounding in the subdistribution analysis of a competing risk
Bekaert, M; Vansteelandt, S; Mertens, K
Lifetime Data Analysis, 16(1): 45-70.
10.1007/s10985-009-9130-8
CrossRef
Anesthesia and Analgesia
The Effect of Bispectral Index Monitoring on Long-Term Survival in the B-Aware Trial
Leslie, K; Myles, PS; Forbes, A; Chan, MTV
Anesthesia and Analgesia, 110(3): 816-822.
10.1213/ANE.0b013e3181c3bfb2
CrossRef
Social Psychiatry and Psychiatric Epidemiology
Calculating control variables with age at onset data to adjust for conditions prior to exposure
Hofler, M; Brueck, T; Lieb, R; Wittchen, HU
Social Psychiatry and Psychiatric Epidemiology, 40(9): 731-736.
10.1007/s00127-005-0944-8
CrossRef
Statistical Methods in Medical Research
Analysis of repeated pregnancy outcomes
Louis, GB; Dukic, V; Heagerty, PJ; Louis, TA; Lynch, CD; Ryan, LM; Schisterman, EF; Trumble, A
Statistical Methods in Medical Research, 15(2): 103-126.
10.1191/0962270206sm434oa
CrossRef
Criminology
Does marriage reduce crime? A counterfactual approach to within-individual causal effects
Sampson, RJ; Laub, JH; Wimer, C
Criminology, 44(3): 465-508.

Journal of Statistical Planning and Inference
Nonparametric causal effects based on marginal structural models
Neugebauer, R; van der Laan, M
Journal of Statistical Planning and Inference, 137(2): 419-434.
10.1016/j.jspi.2005.12.008
CrossRef
Environmental Health Perspectives
The epidemiology of lead toxicity in adults: Measuring dose and consideration of other methodologic issues
Hu, H; Shih, R; Rothenberg, S; Schwartz, BS
Environmental Health Perspectives, 115(3): 455-462.
10.1289/ehp.9783
CrossRef
Journal of Evaluation in Clinical Practice
Estimating causal parameters without target populations
Shahar, E
Journal of Evaluation in Clinical Practice, 13(5): 814-816.
10.1111/j.1365-2753.2007.00796.x
CrossRef
Circulation-Cardiovascular Quality and Outcomes
Vision and Creation of the American Heart Association Pharmaceutical Roundtable Outcomes Research Centers
Peterson, ED; Spertus, JA; Cohen, DJ; Hlatky, MA; Go, AS; Vickrey, BG; Saver, JL; Hinton, PC
Circulation-Cardiovascular Quality and Outcomes, 2(6): 663-670.
10.1161/CIRCOUTCOMES.109.868612
CrossRef
Social Science & Medicine
The (mis)estimation of neighborhood effects: causal inference for a practicable social epidemiology
Oakes, JM
Social Science & Medicine, 58(): 1929-1952.
10.1016/j.socscimed.2003.08.004
CrossRef
Biometrics
Estimating mean response as a function of treatment duration in an observational study, where duration may be informatively censored
Johnson, BA; Tsiatis, AA
Biometrics, 60(2): 315-323.

Journal of Clinical Epidemiology
Hematocrit was not validated as a surrogate end point for survival among epoetin-treated hemodialysis patients
Cotter, DJ; Stefanik, K; Zhang, Y; Thamer, M; Scharfstein, D; Kaufman, J
Journal of Clinical Epidemiology, 57(): 1086-1095.
10.1016/j.jclinepi.2004.05.002
CrossRef
Journal of Adolescent Health
Virginity Pledges Among the Willing: Delays in First Intercourse and Consistency of Condom Use
Martino, SC; Elliott, MN; Collins, RL; Kanouse, DE; Berry, SH
Journal of Adolescent Health, 43(4): 341-348.
10.1016/j.jadohealth.2008.02.018
CrossRef
Journal of the American College of Surgeons
Survival Advantage in Trauma Centers: Expeditious Intervention or Experience?
Haas, B; Jurkovich, GJ; Wang, J; Rivara, FP; MacKenzie, EJ; Nathens, AB
Journal of the American College of Surgeons, 208(1): 28-36.
10.1016/j.jamcollsurg.2008.09.004
CrossRef
American Journal of Epidemiology
Different Methods of Balancing Covariates Leading to Different Effect Estimates in the Presence of Effect Modification
Lunt, M; Solomon, D; Rothman, K; Glynn, R; Hyrich, K; Symmons, DPM; Sturmer, T
American Journal of Epidemiology, 169(7): 909-917.
10.1093/aje/kwn391
CrossRef
Clinical Infectious Diseases
Tuberculosis Treatment and Risk of Stavudine Substitution in First-Line Antiretroviral Therapy
Westreich, DJ; Sanne, I; Maskew, M; Malope-Kgokong, B; Conradie, F; Majuba, P; Funk, MJ; Kaufman, JS; Van Rie, A; MacPhail, P
Clinical Infectious Diseases, 48(): 1617-1623.
10.1086/598977
CrossRef
Journal of Industrial Economics
The Impact of Academic Patenting on the Rate, Quality and Direction of (Public) Research Output
Azoulay, P; Ding, W; Stuart, T
Journal of Industrial Economics, 57(4): 637-676.

Biomarkers
Physical activity and lung cancer among non-smokers: a pilot molecular epidemiological study within EPIC
Rundle, A; Richie, J; Steindorf, K; Peluso, M; Overvad, K; Raaschou-Nielsen, O; Clavel-Chapelon, F; Linseisen, JP; Boeing, H; Trichopoulou, A; Palli, D; Krogh, V; Tumino, R; Panico, S; Bueno-De-Mesquita, HB; Peeters, PH; Lund, E; Gonzalez, CA; Martinez, C; Dorronsoro, M; Barricarte, A; Tormo, MJ; Quiros, JR; Agudo, A; Berglund, G; Jarvholm, B; Bingham, S; Key, TJ; Gormally, E; Saracci, R; Kaaks, R; Riboli, E; Vineis, P
Biomarkers, 15(1): 20-30.
10.3109/13547500903186452
CrossRef
Clinical Infectious Diseases
Long-term effects of highly active antiretroviral therapy on CD4(+) cell evolution among children and adolescents infected with HIV: 5 years and counting
Patel, K; Hernan, MA; Williams, PL; Seeger, JD; McIntosh, K; Van Dyke, RB; Seage, GR
Clinical Infectious Diseases, 46(): 1751-1760.
10.1086/587900
CrossRef
American Journal of Epidemiology
A longitudinal study of vaginal douching and bacterial vaginosis - A marginal structural modeling analysis
Brotman, RM; Klebanoff, MA; Nansel, TR; Andrews, WW; Schwebke, JR; Zhang, J; Yu, KF; Zenilman, JM; Scharfstein, DO
American Journal of Epidemiology, 168(2): 188-196.
10.1093/aje/kwn103
CrossRef
American Heart Journal
alpha-Adducin polymorphism associated with increased risk of adverse cardiovascular outcomes: Results from GENEtic Substudy of the INternational VErapamil SR-trandolapril STudy (INVEST-GENES)
Gerhard, T; Gong, Y; Beitelshees, AL; Mao, X; Lobmeyer, MT; Cooper-DeHoff, RM; Langaee, TY; Schork, NJ; Shriver, MD; Pepine, CJ; Johnson, JA
American Heart Journal, 156(2): 397-404.
10.1016/j.ahj.2008.03.007
CrossRef
American Journal of Sociology
Moving to inequality: Neighborhood effects and experiments meet social structure
Sampson, RJ
American Journal of Sociology, 114(1): 189-231.

Journal of Epidemiology and Community Health
Glossary for econometrics and epidemiology
Gunasekara, FI; Carter, K; Blakely, T
Journal of Epidemiology and Community Health, 62(): 858-861.
10.1136/jech.2008.077461
CrossRef
Physical Therapy
Benchmarking physical therapy clinic performance: Statistical methods to enhance internal validity when using observational data
Resnik, L; Liu, DW; Hart, DL; Mor, V
Physical Therapy, 88(9): 1078-1087.
10.2522/ptj.20070327
CrossRef
Clinical Trials
Estimation of the average causal effect among subgroups defined by post-treatment variables
Matsuyama, Y; Morita, S
Clinical Trials, 3(1): 1-9.
10.1191/1740774506cn135oa
CrossRef
Prevention Science
Assessing the total effect of time-varying predictors in prevention research
Bray, BC; Almirall, D; Zimmerman, RS; Lynam, D; Murphy, SA
Prevention Science, 7(1): 1-17.
10.1007/s11121-005-0023-0
CrossRef
Scandinavian Journal of Statistics
What can statistics contribute to a causal understanding?
Aalen, OO; Frigessi, A
Scandinavian Journal of Statistics, 34(1): 155-168.
10.1111/j.1467-9469.2006.00549.x
CrossRef
Scandinavian Journal of Statistics
On confounding, prediction and efficiency in the analysis of longitudinal and cross-sectional clustered data
Vansteelandt, S
Scandinavian Journal of Statistics, 34(3): 478-498.
10.1111/j.1467-9469.2006.00555.x
CrossRef
Bmc Psychiatry
Evaluating dose response from flexible dose clinical trials
Lipkovich, I; Adams, DH; Mallinckrodt, C; Faries, D; Baron, D; Houston, JP
Bmc Psychiatry, 8(): -.
ARTN 3
CrossRef
American Journal of Epidemiology
Invited commentary: Variable selection versus shrinkage in the control of multiple confounders
Greenland, S
American Journal of Epidemiology, 167(5): 523-529.
10.1093/aje/kwm355
CrossRef
International Journal of Epidemiology
Pre-hospital coronary care and coronary fatality in the Belfast and Glasgow MONICA populations
Moore, W; Kee, F; Evans, AE; McCrum-Gardner, EE; Morrison, C; Tunstall-Pedoc, H
International Journal of Epidemiology, 34(2): 422-430.
10.1093/ije/dyh377
CrossRef
Health Services Research
Developing a quality measure for clinical inertia in diabetes care
Berlowitz, DR; Ash, AS; Glickman, M; Friedman, RH; Pogach, LM; Nelson, AL; Wong, AT
Health Services Research, 40(6): 1836-1853.
10.1111/j.1475-6773.2005.00436.x
CrossRef
Journal of Clinical Epidemiology
A comparison of four prenatal care indices in birth outcome models: Comparable results for predicting small-for-gestational-age outcome but different results for preterm birth or infant mortality
VanderWeele, TJ; Lantos, JD; Siddique, J; Lauderdale, DS
Journal of Clinical Epidemiology, 62(4): 438-445.
10.1016/j.jclinepi.2008.08.001
CrossRef
Journal of Human Resources
Does Job Loss Shorten Life?
Eliason, M; Storrie, D
Journal of Human Resources, 44(2): 277-302.

Nicotine & Tobacco Research
Smoking status and body mass index: A longitudinal study
Munafo, MR; Tilling, K; Ben-Shlomo, Y
Nicotine & Tobacco Research, 11(6): 765-771.
10.1093/ntr/ntp062
CrossRef
Review of Economics and Statistics
Estimating Treatment Effects From Contaminated Multiperiod Education Experiments: the Dynamic Impacts of Class Size Reductions
Ding, WL; Lehrer, SF
Review of Economics and Statistics, 92(1): 31-42.

International Journal of Biostatistics
Targeted Maximum Likelihood Based Causal Inference: Part I
van der Laan, MJ
International Journal of Biostatistics, 6(2): -.
ARTN 2
CrossRef
International Journal of Biostatistics
Attributable Fractions for Sufficient Cause Interactions
VanderWeele, TJ
International Journal of Biostatistics, 6(2): -.
ARTN 5
CrossRef
Biostatistics
Marginal structural models for partial exposure regimes
Vansteelandt, S; Mertens, K; Suetens, C; Goetghebeur, E
Biostatistics, 10(1): 46-59.
10.1093/biostatistics/kxn012
CrossRef
Journal of Clinical Epidemiology
Quantitative assessment of unobserved confounding is mandatory in nonrandomized intervention studies
Groenwold, RHH; Hak, E; Hoes, AW
Journal of Clinical Epidemiology, 62(1): 22-28.
10.1016/j.jclinepi.2008.02.011
CrossRef
Circulation
Disparity in Outcomes of Surgical Revascularization for Limb Salvage Race and Gender Are Synergistic Determinants of Vein Graft Failure and Limb Loss
Nguyen, LL; Hevelone, N; Rogers, SO; Bandyk, DF; Clowes, AW; Moneta, GL; Lipsitz, S; Conte, MS
Circulation, 119(1): 123-130.
10.1161/CIRCULATIONAHA.108.810341
CrossRef
Journal of the American Society of Nephrology
Relationship between Dialysis Modality and Mortality
McDonald, SP; Marshall, MR; Johnson, DW; Polkinghorne, KR
Journal of the American Society of Nephrology, 20(1): 155-163.
10.1681/ASN.2007111188
CrossRef
Ieee Transactions on Engineering Management
The Unequal Benefits of Academic Patenting for Science and Engineering Research
Calderini, M; Franzoni, C; Vezzulli, A
Ieee Transactions on Engineering Management, 56(1): 16-30.
10.1109/TEM.2008.2009889
CrossRef
International Journal of Epidemiology
Epidemiological methods to tackle causal questions
Rutter, M
International Journal of Epidemiology, 38(1): 3-6.
10.1093/ije/dyn253
CrossRef
Annals of Epidemiology
How Have Changes in Air Bag Designs Affected Frontal Crash Mortality?
Braver, ER; Shardell, M; Teoh, ER
Annals of Epidemiology, 20(7): 499-510.
10.1016/j.annepidem.2010.04.005
CrossRef
Seminars in Dialysis
Reverse epidemiology in patients with chronic kidney disease who are not yet on dialysis
Kovesdy, CP; Anderson, JE
Seminars in Dialysis, 20(6): 566-569.
10.1111/j.1525-139X.2007.00335.x
CrossRef
Clinical Journal of the American Society of Nephrology
History-adjusted marginal structural analysis of the association between hemoglobin variability and mortality among chronic hemodialysis patients
Brunelli, SM; Joffe, MM; Israni, RK; Yang, W; Fishbane, S; Berns, JS; Feldman, HI
Clinical Journal of the American Society of Nephrology, 3(3): 777-782.
10.2215/CJN.04281007
CrossRef
Journal of Clinical Epidemiology
Marginal structural models might overcome confounding when analyzing multiple treatment effects in observational studies
Suarez, D; Haro, JM; Novick, D; Ochoa, S
Journal of Clinical Epidemiology, 61(6): 525-530.
10.1016/j.jclinepi.2007.11.007
CrossRef
Neuroepidemiology
Evaluation of the Impact of Memantine Treatment Initiation on Psychotropics Use: A Study from the French National Health Care Database
Vidal, JS; Lacombe, JM; Dartigues, JF; Pasquier, F; Robert, P; Tzourio, C; Alperovitch, A
Neuroepidemiology, 31(3): 193-200.
10.1159/000158226
CrossRef
Biostatistics
Case series analysis for censored, perturbed, or curtailed post-event exposures
Farrington, CP; Whitaker, HJ; Hocine, MN
Biostatistics, 10(1): 3-16.
10.1093/biostatistics/kxn013
CrossRef
Journal of Clinical Epidemiology
Survivor treatment bias, treatment selection bias, and propensity scores in observational research
Austin, PC; Platt, RW
Journal of Clinical Epidemiology, 63(2): 136-138.
10.1016/j.jclinepi.2009.05.009
CrossRef
Journal of Clinical Epidemiology
Propensity score analysis with a time-dependent intervention is an acceptable although not an optimal analytical approach when treatment selection bias and survivor bias coexist
Tleyjeh, IM; Ghomrawi, HMK; Steckelberg, JA; Montori, VM; Hoskin, TL; Enders, F; Huskins, WC; Mookadam, F; Wilson, WR; Zimmerman, V; Baddour, LM
Journal of Clinical Epidemiology, 63(2): 139-140.
10.1016/j.jclinepi.2009.07.017
CrossRef
American Heart Journal
Triple antiplatelet therapy reduces ischemic events after drug-eluting stent implantation: Drug-Eluting stenting followed by Cilostazol treatment REduces Adverse Serious cardiac Events (DECREASE registry)
Lee, SW; Park, SW; Yun, SC; Kim, YH; Park, DW; Kim, WJ; Lee, JY; Lee, CW; Hong, MK; Kim, JJ; Park, SJ
American Heart Journal, 159(2): -.
10.1016/j.ahj.2009.11.014
CrossRef
Value in Health
Use of Stabilized Inverse Propensity Scores as Weights to Directly Estimate Relative Risk and Its Confidence Intervals
Xu, S; Ross, C; Raebel, MA; Shetterly, S; Blanchette, C; Smith, D
Value in Health, 13(2): 273-277.
10.1111/j.1524-4733.2009.00671.x
CrossRef
Bmc Health Services Research
Association of shared decision-making with type of breast cancer surgery: a cross-sectional study
Lee, MK; Noh, DY; Nam, SJ; Ahn, SH; Park, BW; Lee, ES; Yun, YH
Bmc Health Services Research, 10(): -.
ARTN 48
CrossRef
Supportive Care in Cancer
Impact of caregivers' unmet needs for supportive care on quality of terminal cancer care delivered and caregiver's workforce performance
Park, SM; Kim, YJ; Kim, S; Choi, JS; Lim, HY; Choi, YS; Hong, YS; Kim, SY; Heo, DS; Kang, KM; Jeong, HS; Lee, CG; Moon, DH; Choi, JY; Kong, IS; Yun, YH
Supportive Care in Cancer, 18(6): 699-706.
10.1007/s00520-009-0668-5
CrossRef
Psychological Methods
Average Causal Effects From Nonrandomized Studies: A Practical Guide and Simulated Example
Schafer, JL; Kang, J
Psychological Methods, 13(4): 279-313.
10.1037/a0014268
CrossRef
Educational Evaluation and Policy Analysis
Reading Instruction Time and Homogeneous Grouping in Kindergarten: An Application of Marginal Mean Weighting Through Stratification
Hong, GL; Hong, YH
Educational Evaluation and Policy Analysis, 31(1): 54-81.
10.3102/0162373708328259
CrossRef
Health Services Research
Anesthesia Provider Model, Hospital Resources, and Maternal Outcomes
Needleman, J; Minnick, AF
Health Services Research, 44(2): 464-482.
10.1111/j.1475-6773.2008.00919.x
CrossRef
Research Policy
Is commercialization good or bad for science? Individual-level evidence from the Max Planck Society
Buenstorf, G
Research Policy, 38(2): 281-292.
10.1016/j.respol.2008.11.006
CrossRef
Biometrical Journal
Risk Factor Adjustment in Marginal Structural Model Estimation of Optimal Treatment Regimes
Moodie, EEM
Biometrical Journal, 51(5): 774-788.
10.1002/bimj.200800182
CrossRef
Annals of Internal Medicine
Medicare Spending for Previously Uninsured Adults
McWilliams, JM; Meara, E; Zaslavsky, AM; Ayanian, JZ
Annals of Internal Medicine, 151(): 757-W250.

Journal of Infectious Diseases
Initial treatment of HIV infection: Randomized trials with clinical end points are still needed
Hughes, MD
Journal of Infectious Diseases, 194(5): 542-544.

American Journal of Kidney Diseases
Exploring relative mortality and epoetin alfa dose among Hemodialysis patients
Bradbury, BD; Wang, O; Critchlow, CW; Rothman, KJ; Heagerty, P; Keen, M; Acquavella, JF
American Journal of Kidney Diseases, 51(1): 62-70.
10.1053/j.ajkd.2007.09.015
CrossRef
Rheumatology
Effects of switching between anti-TNF therapies on HAQ response in patients who do not respond to their first anti-TNF drug
Hyrich, KL; Lunt, M; Dixon, WG; Watson, KD; Symmons, DPM
Rheumatology, 47(7): 1000-1005.
10.1093/rheumatology/ken127
CrossRef
Environmental Health Perspectives
Ambient ozone concentrations cause increased hospitalizations for asthma in children: An 18-year study in Southern California
Moore, K; Neugebauer, R; Lurmann, F; Hall, J; Brajer, V; Alcorn, S; Tager, I
Environmental Health Perspectives, 116(8): 1063-1070.
10.1289/ehp.10497
CrossRef
Preventive Veterinary Medicine
Linking causal concepts, study design, analysis and inference in support of one epidemiology for population health
Martin, W
Preventive Veterinary Medicine, 86(): 270-288.
10.1016/j.prevetmed.2008.02.013
CrossRef
Pharmaceutical Statistics
Estimation of the marginal survival time in the presence of dependent competing risks using inverse probability of censoring weighted (IPCW) methods
Matsuyama, Y; Yamaguchi, T
Pharmaceutical Statistics, 7(3): 202-214.
10.1002/pst.290
CrossRef
Biometrics
Estimating causal treatment effects from longitudinal HIV natural history studies using marginal structural models
Ko, HJ; Hogan, JW; Mayer, KH
Biometrics, 59(1): 152-162.

Rheumatic Disease Clinics of North America
Data collection, maintenance, and analysis for rheumatic disease research
Wolfe, F; Michaud, K
Rheumatic Disease Clinics of North America, 30(4): 753-+.
10.1016/j.rdc.2004.07.007
CrossRef
Journal of Epidemiology and Community Health
Effect of psychiatric illness and labour market status on suicide: a healthy worker effect?
Agerbo, E
Journal of Epidemiology and Community Health, 59(7): 598-602.
10.1136/jech.2004.025288
CrossRef
Sociological Methodology, Vol 34
Adjusting for time-varying confounding in survival analysis
Barber, JS; Murphy, SA; Verbitsky, N
Sociological Methodology, Vol 34, 34(): 163-192.

American Journal of Epidemiology
Effects of body composition and leisure-time physical activity on transitions in physical functioning in the elderly
Haight, T; Tager, I; Sternfeld, B; Satariano, W; van der Laan, M
American Journal of Epidemiology, 162(7): 607-617.

Drug Development Research
Propensity score for the analysis of observational data: An introduction and an illustrative example
Cavuto, S; Bravi, F; Grassi, MC; Apolone, G
Drug Development Research, 67(3): 208-216.
10.1002/ddr.20079
CrossRef
Canadian Journal of Statistics-Revue Canadienne De Statistique
Proportional hazards models based on biased samples and estimated selection probabilities
Pan, Q; Schaubel, DE
Canadian Journal of Statistics-Revue Canadienne De Statistique, 36(1): 111-127.

Journal of the American Society of Nephrology
Statin Use Is Associated with Prolonged Survival of Renal Transplant Recipients
Wiesbauer, F; Heinze, G; Mitterbauer, C; Harnoncourt, F; Horl, WH; Oberbauer, R
Journal of the American Society of Nephrology, 19(): 2211-2218.

American Journal of Health-System Pharmacy
Bias: Considerations for research practice
Gerhard, T
American Journal of Health-System Pharmacy, 65(): 2159-2168.
10.2146/ajhp070369
CrossRef
Bmc Medical Research Methodology
Reducing bias through directed acyclic graphs
Shrier, I; Platt, RW
Bmc Medical Research Methodology, 8(): -.
ARTN 70
CrossRef
Medicina Clinica
What is a confounding variable?
de Irala, J; Martinez-Gonzalez, MA; Grima, FG
Medicina Clinica, 117(): 377-385.

Community Dentistry and Oral Epidemiology
Directed acyclic graphs (DAGs): an aid to assess confounding in dental research
Merchant, AT; Pitiphat, W
Community Dentistry and Oral Epidemiology, 30(6): 399-404.

Statistical Methods in Medical Research
Instrumental variables and inverse probability weighting for causal inference from longitudinal observational studies
Hogan, JW; Lancaster, T
Statistical Methods in Medical Research, 13(1): 17-48.
10.1191/0962280204sm351ra
CrossRef
Pharmacoepidemiology and Drug Safety
An application of propensity score matching using claims data
Seeger, JD; Williams, PL; Walker, AM
Pharmacoepidemiology and Drug Safety, 14(7): 465-476.
10.1002/pds.1062
CrossRef
Computational Statistics & Data Analysis
G-computation estimation for causal inference with complex longitudinal data
Neugebauer, R; van der Laan, MJ
Computational Statistics & Data Analysis, 51(3): 1676-1697.
10.1016/j.csda.2006.06.016
CrossRef
Seminars in Perinatology
Epidemiologic approaches for studying recurrent pregnancy outcomes: Challenges and implications for research
Ananth, CV
Seminars in Perinatology, 31(3): 196-201.
10.1053/j.semperi.2007.03.008
CrossRef
AIDS
Incident and prevalent herpes simplex virus type 2 infection increases risk of HIV acquisition among women in Uganda and Zimbabwe
Brown, JM; Wald, A; Hubbard, A; Rungruengthanakit, K; Chipato, T; Rugpao, S; Mmiro, F; Celentano, DD; Salata, RS; Morrison, CS; Richardson, BA; Padian, NS
AIDS, 21(): 1515-1523.

Journal of the Royal Statistical Society Series B-Statistical Methodology
Maximum likelihood estimation in semiparametric regression models with censored data - Discussion on the paper by Zeng and Lin
Henderson, R
Journal of the Royal Statistical Society Series B-Statistical Methodology, 69(): 536-564.

International Journal of Epidemiology
Fallibility in estimating direct effects
Cole, SR; Hernan, MA
International Journal of Epidemiology, 31(1): 163-165.

Epidemiology and Infection
Associations between water-treatment methods and diarrhoea in HIV-positive individuals
Eisenberg, JNS; Wade, TJ; Hubbard, A; Abrams, DI; Leiser, RJ; Charles, S; Vu, M; Saha, S; Wright, CC; Levy, DA; Jensen, P; Colford, JM
Epidemiology and Infection, 129(2): 315-323.
10.1017/S0950268802007422
CrossRef
Statistics in Medicine
Adjusting for differential proportions of second-line treatment in cancer clinical trials. Part II: An application in a clinical trial of unresectable non-small-cell lung cancer
Yamaguchi, T; Ohashi, Y
Statistics in Medicine, 23(): 2005-2022.
10.1002/sim.1817
CrossRef
American Journal of Epidemiology
An application of model-fitting procedures for marginal structural models
Mortimer, KM; Neugebauer, R; van der Laan, M; Tager, IB
American Journal of Epidemiology, 162(4): 382-388.
10.1093/aje/kwi208
CrossRef
American Journal of Epidemiology
Commentary: Considerations for use of racial/ethnic classification in etiologic research
Kaufman, JS; Cooper, RS
American Journal of Epidemiology, 154(4): 291-298.

American Journal of Epidemiology
Mammographic Screening and Risk Factors for Breast Cancer
Cook, NR; Rosner, BA; Hankinson, SE; Colditz, GA
American Journal of Epidemiology, 170(): 1422-1432.
10.1093/aje/kwp304
CrossRef
International Journal of Epidemiology
Commentary: Reducing HIV-associated tuberculosis in children
Boulle, A; Eley, B
International Journal of Epidemiology, 38(6): 1621-1623.
10.1093/ije/dyp301
CrossRef
International Journal of Clinical Practice
Adherence to statin therapy: the key to survival?
Andersohn, F
International Journal of Clinical Practice, 64(7): 843-847.
10.1111/j.1742-1241.2009.02302.x
CrossRef
Controlled Clinical Trials
Design and analysis of two-period studies of potentially disease-modifying treatments
McDermott, MP; Hall, WJ; Oakes, D; Eberly, S
Controlled Clinical Trials, 23(6): 635-649.
PII S0197-2456(02)00238-6
CrossRef
Journal of the Royal Statistical Society Series A-Statistics in Society
Another look into the effect of premarital cohabitation on duration of marriage: an approach based on matching
Mazzuco, S
Journal of the Royal Statistical Society Series A-Statistics in Society, 172(): 255-273.
10.1111/j.1467-985X.2008.00568.x
CrossRef
Nephrology Dialysis Transplantation
The survival advantage for haemodialysis patients taking vitamin D is questioned: findings from the Dialysis Outcomes and Practice Patterns Study
Tentori, F; Albert, JM; Young, EW; Blayney, MJ; Robinson, BM; Pisoni, RL; Akiba, T; Greenwood, RN; Kimata, N; Levin, NW; Piera, LM; Saran, R; Wolfe, RA; Port, FK
Nephrology Dialysis Transplantation, 24(3): 963-972.
10.1093/ndt/gfn592
CrossRef
Clinical Journal of the American Society of Nephrology
Estimated Effect of Epoetin Dosage on Survival among Elderly Hemodialysis Patients in the United States
Zhang, Y; Thamer, M; Cotter, D; Kaufman, J; Hernan, MA
Clinical Journal of the American Society of Nephrology, 4(3): 638-644.
10.2215/CJN.05071008
CrossRef
Circulation
Longitudinal Tracking of Left Ventricular Mass Over the Adult Life Course Clinical Correlates of Short- and Long-Term Change in the Framingham Offspring Study
Lieb, W; Xanthakis, V; Sullivan, LM; Aragam, J; Pencina, MJ; Larson, MG; Benjamin, EJ; Vasan, RS
Circulation, 119(): 3085-U73.
10.1161/CIRCULATIONAHA.108.824243
CrossRef
Information Systems Research
From Association to Causation via a Potential Outcomes Approach
Mithas, S; Krishnan, MS
Information Systems Research, 20(2): 295-313.
10.1287/isre.1080.0184
CrossRef
Journal of Affective Disorders
Early discontinuation of antidepressant treatment and suicide risk among persons aged 50 and over: A population-based register study
Erlangsen, A; Agerbo, E; Hawton, K; Conwell, Y
Journal of Affective Disorders, 119(): 194-199.
10.1016/j.jad.2009.03.011
CrossRef
Contemporary Clinical Trials
Medical cost analysis: Application to colorectal cancer data from the SEER Medicare database
Bang, H
Contemporary Clinical Trials, 26(5): 586-597.
10.1016/j.cct.2005.05.004
CrossRef
Bioinformatics
A causal inference approach for constructing transcriptional regulatory networks
Xing, B; van der Laan, MJ
Bioinformatics, 21(): 4007-4013.
10.1093/bioinformatics/bti648
CrossRef
Thorax
Regular physical activity reduces hospital admission and mortality in chronic obstructive pulmonary disease: a population based cohort study
Garcia-Aymerich, J; Lange, P; Benet, M; Schnohr, P; Anto, JM
Thorax, 61(9): 772-778.
10.1136/thx.2006.060145
CrossRef
Addiction
Cessation of injection drug use and change in injection frequency: the Chicago Needle Exchange Evaluation Study
Huo, DZ; Bailey, SL; Ouellet, LJ
Addiction, 101(): 1606-1613.
10.1111/j.1360-0443.2006.01577.x
CrossRef
Neurology
Cumulative effect of lead on cognition - Is bone more revealing than blood?
Weisskopf, MG; Myers, G
Neurology, 67(9): 1536-1537.

Journal of the American Geriatrics Society
The relationship between exercise and risk of venous thrombosis in elderly people
van Stralen, KJ; Doggen, CJM; Lumley, T; Cushman, M; Folsom, AR; Psaty, BM; Siscovick, D; Rosendaal, FR; Heckbert, SR
Journal of the American Geriatrics Society, 56(3): 517-522.
10.1111/j.1532-5415.2007.01588.x
CrossRef
American Journal of Epidemiology
Statistical issues in life course epidemiology
De Stavola, BL; Nitsch, D; Silva, ID; McCormack, V; Hardy, R; Mann, V; Cole, TJ; Morton, S; Leon, DA
American Journal of Epidemiology, 163(1): 84-96.
10.1093/aje/kwj003
CrossRef
Basic & Clinical Pharmacology & Toxicology
Indications for propensity scores and review of their use in pharmacoepidemiology
Glynn, RJ; Schneeweiss, S; Sturmer, T
Basic & Clinical Pharmacology & Toxicology, 98(3): 253-259.

Statistical Methods in Medical Research
Methodologic issues in the design and analysis of epidemiologic studies of pregnancy outcome
Savitz, DA; Dole, N; Herring, AH
Statistical Methods in Medical Research, 15(2): 93-102.
10.1191/0962280206sm433oa
CrossRef
Revue D Epidemiologie Et De Sante Publique
Neighborhoods and health: where are we and were do we go from here?
Roux, AVD
Revue D Epidemiologie Et De Sante Publique, 55(1): 13-21.
10.1016/j.respe.2006.12.003
CrossRef
AIDS
Virologic efficacy of boosted double versus boosted single protease inhibitor therapy
Petersen, ML; Wang, Y; van der Laan, MJ; Rhee, SY; Shafer, RW; Fessel, WJ
AIDS, 21(): 1547-1554.

Biometrics
Presidential Address: XXI International Biometric Conference, Freiburg, Germany, July 2002 - Are statistical contributions to medicine undervalued?
Breslow, NE
Biometrics, 59(1): 1-8.

Journal of the American Society of Nephrology
Administration of parenteral iron and mortality among hemodialysis patients
Feldman, HI; Joffe, M; Robinson, B; Knauss, J; Cizman, B; Guo, WS; Franklin-Becker, E; Faich, G
Journal of the American Society of Nephrology, 15(6): 1623-1632.
10.1097/01.ASN.0000128009.69594.BE
CrossRef
Computer Methods and Programs in Biomedicine
Adjusted survival curves with inverse probability weights
Cole, SR; Hernan, MA
Computer Methods and Programs in Biomedicine, 75(1): 45-49.
10.1016/j.cmpb.2003.10.004
CrossRef
Epidemiologic Reviews
The study of group-level factors in epidemiology: Rethinking variables, study designs, and analytical approaches
Roux, AVD
Epidemiologic Reviews, 26(): 104-111.
10.1093/epirev/mxh006
CrossRef
Statistics in Medicine
Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study
Lunceford, JK; Davidian, M
Statistics in Medicine, 23(): 2937-2960.
10.1002/SIM.1903
CrossRef
American Journal of Epidemiology
Analytic strategies to adjust confounding using exposure propensity scores and disease risk scores: Nonsteroidal antiinflammatory drugs and short-term mortality in the elderly
Sturmer, T; Schneeweiss, S; Brookhart, MA; Rothman, KJ; Avorn, J; Glynn, RJ
American Journal of Epidemiology, 161(9): 891-898.
10.1093/aje/kwi106
CrossRef
Kidney International
Anemia and mortality in hemodialysis patients: Accounting for morbidity and treatment variables updated over time
Robinson, BM; Joffe, MM; Berns, JS; Pisoni, RL; Port, FK; Feldman, HI
Kidney International, 68(5): 2323-2330.

Jama-Journal of the American Medical Association
Adjuvant Chemotherapy Use and Adverse Events Among Older Patients With Stage III Colon Cancer
Kahn, KL; Adams, JL; Weeks, JC; Chrischilles, EA; Schrag, D; Ayanian, JZ; Kiefe, CI; Ganz, PA; Bhoopalam, N; Potosky, AL; Harrington, DP; Fletcher, RH
Jama-Journal of the American Medical Association, 303(): 1037-1045.

International Journal of Biostatistics
Comparing Approaches to Causal Inference for Longitudinal Data: Inverse Probability Weighting versus Propensity Scores
Ertefaie, A; Stephens, DA
International Journal of Biostatistics, 6(2): -.
ARTN 14
CrossRef
Journal of Quantitative Criminology
The Impact of Imprisonment on Marriage and Divorce: A Risk Set Matching Approach
Apel, R; Blokland, AAJ; Nieuwbeerta, P; van Schellen, M
Journal of Quantitative Criminology, 26(2): 269-300.
10.1007/s10940-009-9087-5
CrossRef
Bmc Genetics
Multiple imputation methods for longitudinal blood pressure measurements from the Framingham Heart Study
Kang, T; Kraft, P; Gauderman, WJ; Thomas, D
Bmc Genetics, 4(): -.
ARTN S43
CrossRef
Supportive Care in Cancer
Self-reported experience and outcomes of care among stomach cancer patients at a median follow-up time of 27 months from diagnosis
Kim, S; Bae, JM; Kim, YW; Ryu, KW; Lee, JH; Noh, JH; Sohn, TS; Hong, SK; Lee, MK; Park, SM; Yun, YH
Supportive Care in Cancer, 16(7): 831-839.
10.1007/s00520-007-0340-x
CrossRef
American Journal of Kidney Diseases
Disordered mineral metabolism in hemodialysis patients: An analysis of cumulative effects in the Hemodialysis (HEMO) Study
Wald, R; Sarnak, MJ; Tighiouart, H; Cheung, AK; Levey, AS; Eknoyan, G; Miskulin, DC
American Journal of Kidney Diseases, 52(3): 531-540.
10.1053/j.ajkd.2008.05.020
CrossRef
Physical Therapy
Predictors of physical therapy clinic performance in the treatment of patients with low back pain syndromes
Resnik, L; Liu, DW; Mor, V; Hart, DL
Physical Therapy, 88(9): 989-1004.
10.2522/ptj.20070110
CrossRef
Clinical Infectious Diseases
Drug toxicity, HIV progression, or comorbidity of aging: Does tipranavir use increase the risk of intracranial hemorrhage?
Justice, AC; Zingmond, DS; Gordon, KS; Fultz, SL; Goulet, JL; King, JT; Bravata, DM; Valdez, H; Kraft, M; Mattocks, KM
Clinical Infectious Diseases, 47(9): 1226-1230.
10.1086/592302
CrossRef
American Journal of Epidemiology
Marginal Structural Models for Sufficient Cause Interactions
VanderWeele, TJ; Vansteelandt, S; Robins, JM
American Journal of Epidemiology, 171(4): 506-514.
10.1093/aje/kwp396
CrossRef
Annals of Internal Medicine
Coronary Heart Disease in Postmenopausal Recipients of Estrogen Plus Progestin Therapy: Does the Increased Risk Ever Disappear? A Randomized Trial
Toh, S; Hernandez-Diaz, S; Logan, R; Rossouw, JE; Hernan, MA
Annals of Internal Medicine, 152(4): 211-U30.

Contemporary Clinical Trials
Changes in beliefs identify unblinding in randomized controlled trials: a method to meet CONSORT guidelines
Rees, JR; Wade, TJ; Levy, DA; Colford, JM; Hilton, JF
Contemporary Clinical Trials, 26(1): 25-37.
10.1016/j.cct.2004.11.020
CrossRef
Lancet
Long-term effectiveness of potent antiretroviral therapy in preventing AIDS and death: a prospective cohort study
Sterne, JAC; Hernan, MA; Ledergerber, B; Tilling, K; Weber, R; Sendi, P; Rickenbach, M; Robins, JM; Egger, M
Lancet, 366(): 378-384.

Journal of Statistical Planning and Inference
Double robust estimation in longitudinal marginal structural models
Yu, Z; van der Laan, M
Journal of Statistical Planning and Inference, 136(3): 1061-1089.
10.1016/j.jspi.2004.08.011
CrossRef
American Journal of Epidemiology
Effect of tuberculosis on the survival of women infected with human immunodeficiency virus
Lopez-Gatell, H; Cole, SR; Hessol, NA; French, AL; Greenblatt, RM; Landesman, S; Preston-Martin, S; Anastos, K
American Journal of Epidemiology, 165(): 1134-1142.
10.1093/aje/kwm116
CrossRef
American Journal of Epidemiology
Determining the effect of highly active antiretroviral therapy on changes in human immunodeficiency virus type 1 RNA viral load using a marginal structural left-censored mean model
Cole, SR; Hernan, MA; Anastos, K; Jamieson, BD; Robins, JM
American Journal of Epidemiology, 166(2): 219-227.
10.1093/aje/kwm047
CrossRef
Clinical Infectious Diseases
Pillbox organizers are associated with improved adherence to HIV antiretroviral therapy and viral suppression: A marginal structural model analysis
Petersen, ML; Wang, Y; van der Laan, MJ; Guzman, D; Riley, E; Bangsberg, DR
Clinical Infectious Diseases, 45(7): 908-915.
10.1086/521250
CrossRef
Kidney International
Methodological issues in studying the epidemiology of mild to moderate chronic renal insufficiency
Hsu, CY; Chertow, GM; Curhan, GC
Kidney International, 61(5): 1567-1576.

Neuroendocrinology Letters
The darkness at the end of the tunnel: Summary and evaluation of an International Symposium on Light, Endocrine Systems and Cancer
Poole, C
Neuroendocrinology Letters, 23(): 71-78.
PII NEL230802E02
CrossRef
American Journal of Kidney Diseases
Is the evidence for high hematocrit targets valid?
Cotter, DJ
American Journal of Kidney Diseases, 41(2): 520.

Biostatistics
Analysis of longitudinal marginal structural models
Bryan, J; Yu, Z; Van Der Laan, MJ
Biostatistics, 5(3): 361-380.
10.1093/biostatistics/kxg041
CrossRef
Statistics in Medicine
Sense and sensitivity when correcting for observed exposures in randomized clinical trials
Vansteelandt, S; Goetghebeur, E
Statistics in Medicine, 24(2): 191-210.
10.1002/sim.1829
CrossRef
American Journal of Epidemiology
Time Scale and Adjusted Survival Curves for Marginal Structural Cox Models
Westreich, D; Cole, SR; Tien, PC; Chmiel, JS; Kingsley, L; Funk, MJ; Anastos, K; Jacobson, LP
American Journal of Epidemiology, 171(6): 691-700.
10.1093/aje/kwp418
CrossRef
Pharmaceutical Statistics
Survival analysis using inverse probability of treatment weighted methods based on the generalized propensity score
Sugihara, M
Pharmaceutical Statistics, 9(1): 21-34.
10.1002/pst.365
CrossRef
Lancet
Daily co-trimoxazole prophylaxis in severely immunosuppressed HIV-infected adults in Africa started on combination antiretroviral therapy: an observational analysis of the DART cohort
Walker, AS; Ford, D; Gilks, CF; Munderi, P; Ssali, F; Reid, A; Katabira, E; Grosskurth, H; Mugyenyi, P; Hakim, J; Darbyshire, JH; Gibb, DM; Babiker, AG
Lancet, 375(): 1278-1286.
10.1016/S0140-6736(10)60057-8
CrossRef
Journal of Drug Issues
Differences in Mortality Among Heroin, Cocaine, and Methamphetamine Users: A Hierarchical Bayesian Approach
Liang, LJ; Huang, D; Brecht, ML; Hser, YI
Journal of Drug Issues, 40(1): 121-140.

American Journal of Kidney Diseases
Vascular Access Practice in Hemodialysis: Instrumental in Determining Patient Mortality
Polkinghorne, KR; Epi, MC
American Journal of Kidney Diseases, 53(3): 359-362.
10.1053/j.ajkd.2009.01.010
CrossRef
Kdd-2007 Proceedings of the Thirteenth Acm Sigkdd International Conference on Knowledge Discovery and Data Mining
Making Generative Classifiers Robust to Selection Bias
Smith, A; Elkan, C
Kdd-2007 Proceedings of the Thirteenth Acm Sigkdd International Conference on Knowledge Discovery and Data Mining, (): 657-666.

American Journal of Epidemiology
The Author Replies
Suissa, S
American Journal of Epidemiology, 170(5): 668-669.
10.1093/aje/kwp240
CrossRef
Journal of Evaluation in Clinical Practice
Motivational interviewing-based health coaching as a chronic care intervention
Linden, A; Butterworth, SW; Prochaska, JO
Journal of Evaluation in Clinical Practice, 16(1): 166-174.
10.1111/j.1365-2753.2009.01300.x
CrossRef
Clinical Journal of the American Society of Nephrology
Relationship between Epoetin Alfa Dose and Mortality: Findings from a Marginal Structural Model
Wang, OH; Kilpatrick, RD; Critchlow, CW; Ling, X; Bradbury, BD; Gilbertson, DT; Collins, AJ; Rothman, KJ; Acquavella, JF
Clinical Journal of the American Society of Nephrology, 5(2): 182-188.
10.2215/CJN.03040509
CrossRef
American Journal of Epidemiology
The Association Between Persistent Fetal Occiput Posterior Position and Perinatal Outcomes: An Example of Propensity Score and Covariate Distance Matching
Cheng, YW; Hubbard, A; Caughey, AB; Tager, IB
American Journal of Epidemiology, 171(6): 656-663.
10.1093/aje/kwp437
CrossRef
Journal of Biopharmaceutical Statistics
Analysis of treatment effectiveness in longitudinal observational data
Faries, D; Ascher-Svanum, H; Belger, M
Journal of Biopharmaceutical Statistics, 17(5): 809-826.
10.1080/10543400701513967
CrossRef
Occupational and Environmental Medicine
Reducing healthy worker survivor bias by restricting date of hire in a cohort study of Vermont granite workers
Applebaum, KM; Malloy, EJ; Eisen, EA
Occupational and Environmental Medicine, 64(): 681-687.
10.1136/oem.2006.031369
CrossRef
Medical Care
Using propensity scores subclassification to estimate effects of longitudinal treatments - An example using a new diabetes medication
Segal, JB; Griswold, M; Achy-Brou, A; Herbert, R; Bass, EB; Dy, SM; Millman, AE; Wu, AW; Frangakis, CE
Medical Care, 45(): S149-S157.

Health Services Research
How do proxy responses and proxy-assisted responses differ from what Medicare beneficiaries might have reported about their health care?
Elliott, MN; Beckett, MK; Chong, K; Hambarsoomians, K; Hays, RD
Health Services Research, 43(3): 833-848.
10.1111/j.1475-6773.2007.00820.x
CrossRef
American Journal of Epidemiology
Adolescent cannabis problems and young adult depression: Male-female stratified propensity score analyses
Harder, VS; Stuart, EA; Anthony, JC
American Journal of Epidemiology, 168(6): 592-601.
10.1093/aje/kwn184
CrossRef
Infection Control and Hospital Epidemiology
Drug toxicity, HIV progression, or comorbidity of aging: Does tipranavir use increase the risk of intracranial hemorrhage?
Justice, AC; Zingmond, DS; Gordon, KS; Fultz, SL; Goulet, JL; King, JT; Bravata, DM; Valdez, H; Kraft, M; Mattocks, KM
Infection Control and Hospital Epidemiology, 29(): 1226-1230.
10.1086/592302
CrossRef
European Journal of Epidemiology
Predictors of follow-up and assessment of selection bias from dropouts using inverse probability weighting in a cohort of university graduates
Alonso, A; Segui-Gomez, M; de Irala, J; Sanchez-Villegas, A; Beunza, JJ; Martinez-Gonzalez, MA
European Journal of Epidemiology, 21(5): 351-358.
10.1007/s10654-006-9008-y
CrossRef
Jaids-Journal of Acquired Immune Deficiency Syndromes
Assessing the effectiveness of antiretroviral adherence interventions - Using marginal structural models to replicate the findings of randomized controlled trials
Petersen, ML; Wang, Y; van der Laan, MJ; Bangsberg, DR
Jaids-Journal of Acquired Immune Deficiency Syndromes, 43(): S96-S103.

Computational Statistics & Data Analysis
Causal effects in longitudinal studies: Definition and maximum likelihood estimation
Neugebauer, R; van der Laan, MJ
Computational Statistics & Data Analysis, 51(3): 1664-1675.
10.1016/j.csda.2006.06.013
CrossRef
Journal of Thoracic and Cardiovascular Surgery
Suture bicuspidization of the tricuspid valve versus ring annuloplasty for repair of functional tricuspid regurgitation: Midterm results of 237 consecutive patients
Ghanta, RK; Chen, R; Narayanasamy, N; McGurk, S; Lipsitz, S; Chen, FY; Cohn, LH
Journal of Thoracic and Cardiovascular Surgery, 133(1): 117-126.
10.1016/j.jctvs.2006.08.068
CrossRef
Environmental Health Perspectives
Environmental determinants of infectious disease: A framework for tracking causal links and guiding public health research
Eisenberg, JNS; Desai, MA; Levy, K; Bates, SJ; Liang, S; Naumoff, K; Scott, JC
Environmental Health Perspectives, 115(8): 1216-1223.
10.1289/ehp.9806
CrossRef
American Journal of Kidney Diseases
EPO adjustments in patients with elevated hemoglobin levels: Provider practice patterns compared with recommended practice guidelines
Collins, AJ; Ebben, JP; Gilbertson, DT
American Journal of Kidney Diseases, 49(1): 135-142.
10.1053/j.ajkd.2006.09.020
CrossRef
Archives of Internal Medicine
Does the evidence for anemia treatment support a survival benefit?
Cotter, D
Archives of Internal Medicine, 163(): 2395-2396.

Social Science & Medicine
Estimating neighborhood health effects: the challenges of causal inference in a complex world - Commentary
Roux, AVD
Social Science & Medicine, 58(): 1953-1960.
10.1016/S0277-9536(03)00414-3
CrossRef
Statistical Methods in Medical Research
Structural mean models for compliance analysis in randomized clinical trials and the impact of errors on measures of exposure
Goetghebeur, E; Vansteelandt, S
Statistical Methods in Medical Research, 14(4): 397-415.
10.1191/0962280205sm407oa
CrossRef
Computational Statistics & Data Analysis
A semiparametric model selection criterion with applications to the marginal structural model
Brookhart, MA; van der Laan, MJ
Computational Statistics & Data Analysis, 50(2): 475-498.
10.1016/j.csda.2004.08.013
CrossRef
American Journal of Epidemiology
Invited commentary: Beyond frequencies and coefficients-toward meaningful descriptions for life course epidemiology
Wang, C
American Journal of Epidemiology, 164(2): 122-125.
10.1093/aje/kwj194
CrossRef
American Journal of Epidemiology
Invited Commentary: The Search for Preventable Causes of Cardiovascular Disease-Whither Work?
Cullen, MR
American Journal of Epidemiology, 169(): 1422-1425.
10.1093/aje/kwp078
CrossRef
Biometrika
Improving efficiency and robustness of the doubly robust estimator for a population mean with incomplete data
Cao, WH; Tsiatis, AA; Davidian, M
Biometrika, 96(3): 723-734.
10.1093/biomet/asp033
CrossRef
American Journal of Epidemiology
Neighborhood Poverty and Injection Cessation in a Sample of Injection Drug Users
Nandi, A; Glass, TA; Cole, SR; Chu, HT; Galea, S; Celentano, DD; Kirk, GD; Vlahov, D; Latimer, WW; Mehta, SH
American Journal of Epidemiology, 171(4): 391-398.
10.1093/aje/kwp416
CrossRef
Journal of Medical Virology
Prediction of the Response to Peg-Interferon-Alfa in Patients With HBeAg Positive Chronic Hepatitis B Using Decline of HBV DNA During Treatment
Hansen, BE; Buster, EHCJ; Steyerberg, EW; Lesaffre, E; Janssen, HLA
Journal of Medical Virology, 82(7): 1135-1142.
10.1002/jmv.21778
CrossRef
Comptes Rendus Biologies
Observation versus intervention in the evaluation of drugs: the story of hormone replacement therapy
Costaghola, D
Comptes Rendus Biologies, 330(4): 347-355.
10.1016/j.crvi.2007.02.011
CrossRef
Cancer
Quality of life and sexual problems in disease-free survivors of cervical cancer compared with the general population
Park, SY; Bae, DS; Nam, JH; Park, CT; Cho, CH; Lee, JM; Lee, MK; Kim, SH; Park, SM; Yun, YH
Cancer, 110(): 2716-2725.
10.1002/cncr.23094
CrossRef
Ieee Transactions on Knowledge and Data Engineering
Explaining classifications for individual instances
Robnik-Sikonja, M; Kononenko, I
Ieee Transactions on Knowledge and Data Engineering, 20(5): 589-600.
10.1109/TKDE.2007.190734
CrossRef
Journal of Epidemiology and Community Health
Using directed acyclic graphs to guide analyses of neighbourhood health effects: an introduction
Fleischer, NL; Roux, AVD
Journal of Epidemiology and Community Health, 62(9): 842-846.
10.1136/jech.2007.067371
CrossRef
Journal of Rheumatology
Prednisone, Lupus Activity, and Permanent Organ Damage
Thamer, M; Hernan, MA; Zhang, Y; Cotter, D; Petri, M
Journal of Rheumatology, 36(3): 560-564.
10.3899/jrheum.080828
CrossRef
Annals of Epidemiology
Marginal Structural Models for Estimating Effect Modification
Chiba, Y; Azuma, K; Okumura, J
Annals of Epidemiology, 19(5): 298-303.
10.1016/j.annepidem.2009.01.025
CrossRef
Biometrics
The intensity-score approach to adjusting for confounding
Brumback, B; Greenland, S; Redman, M; Kiviat, N; Diehr, P
Biometrics, 59(2): 274-285.

Perspectives on Psychological Science
Proceeding From Observed Correlation to Causal Inference The Use of Natural Experiments
Rutter, M
Perspectives on Psychological Science, 2(4): 377-395.
10.1111/j.1745-6916.2007.00050.x
CrossRef
American Journal of Epidemiology
Marginal structural models for analyzing causal effects of time-dependent treatments: An application in perinatal epidemiology
Bodnar, LM; Davidian, M; Siega-Riz, AM; Tsiatis, AA
American Journal of Epidemiology, 159(): 926-934.
10.1093/aje/kwh131
CrossRef
European Journal of Epidemiology
Epidemiologic methods: Beyond clinical medicine, beyond epidemiology
Bolumar, F; Porta, M
European Journal of Epidemiology, 19(8): 733-735.

Journal of the American Statistical Association
Structural equation models: A review with applications to environmental epidemiology
Sanchez, BN; Budtz-Jorgensen, E; Ryan, LM; Hu, H
Journal of the American Statistical Association, 100(): 1443-1455.
10.1198/0162145050000001005
CrossRef
International Journal of Cancer
Smoking and survival after breast cancer diagnosis
Holmes, MD; Morin, S; Chen, WY; Kroenke, CH; Spiegelman, D; Colditz, GA
International Journal of Cancer, 120(): 2672-2677.
10.1002/ijc.22575
CrossRef
Journal of Clinical Oncology
Aspirin Intake and Survival After Breast Cancer
Holmes, MD; Chen, WY; Li, L; Hertzmark, E; Spiegelman, D; Hankinson, SE
Journal of Clinical Oncology, 28(9): 1467-1472.
10.1200/JCO.2009.22.7918
CrossRef
International Journal of Public Health
Effects of time-varying exposures adjusting for time-varying confounders: the case of alcohol consumption and risk of incident human immunodeficiency virus infection
Howe, CJ; Sander, PM; Plankey, MW; Cole, SR
International Journal of Public Health, 55(3): 227-228.
10.1007/s00038-010-0120-0
CrossRef
Heart
Mitral valve replacement with or without a concomitant Maze procedure in patients with atrial fibrillation
Kim, JB; Ju, MH; Yun, SC; Jung, SH; Chung, CH; Choo, SJ; Lee, TY; Song, H; Lee, JW
Heart, 96(): 1126-1131.
10.1136/hrt.2010.192435
CrossRef
Circulation-Cardiovascular Interventions
Increased Risk of Bleeding in Patients on Clopidogrel Therapy After Drug-Eluting Stents Implantation Insights From the HMO Research Network-Stent Registry (HMORN-Stent)
Tsai, TT; Ho, PM; Xu, S; Powers, JD; Carroll, NM; Shetterly, SM; Maddox, TM; Rumsfeld, JS; Margolis, K; Go, AS; Magid, DJ
Circulation-Cardiovascular Interventions, 3(3): 230-235.
10.1161/CIRCINTERVENTIONS.109.919001
CrossRef
International Journal of Epidemiology
Commentary: When brilliant insights lead astray
Hertz-Picciotto, I
International Journal of Epidemiology, 30(6): 1243-1244.

Statistics in Medicine
Sensitivity analysis for the estimation of rates of change with non-ignorable drop-out: an application to a randomized clinical trial of the vitamin D-3
Matsuyama, Y
Statistics in Medicine, 22(5): 811-827.
10.1002/sim.1367
CrossRef
Lifetime Data Analysis
Sensitivity analysis for unmeasured confounding in a marginal structural Cox proportional hazards model
Klungsoyr, O; Sexton, J; Sandanger, I; Nygard, JF
Lifetime Data Analysis, 15(2): 278-294.
10.1007/s10985-008-9109-x
CrossRef
International Journal of Epidemiology
Anti-retroviral therapy reduces incident tuberculosis in HIV-infected children
Edmonds, A; Lusiama, J; Napravnik, S; Kitetele, F; Van Rie, A; Behets, F
International Journal of Epidemiology, 38(6): 1612-1621.
10.1093/ije/dyp208
CrossRef
American Journal of Public Health
Mass Media as an HIV-Prevention Strategy: Using Culturally Sensitive Messages to Reduce HIV-Associated Sexual Behavior of At-Risk African American Youth
Romer, D; Sznitman, S; DiClemente, R; Salazar, LF; Vanable, PA; Carey, MP; Hennessy, M; Brown, LK; Valois, RF; Stanton, BF; Fortune, T; Juzang, I
American Journal of Public Health, 99(): 2150-2159.
10.2105/AJPH.2008.155036
CrossRef
American Journal of Epidemiology
Using Marginal Structural Measurement-Error Models to Estimate the Long-term Effect of Antiretroviral Therapy on Incident AIDS or Death
Cole, SR; Jacobson, LP; Tien, PC; Kingsley, L; Chmiel, JS; Anastos, K
American Journal of Epidemiology, 171(1): 113-122.
10.1093/aje/kwp329
CrossRef
Statistics in Medicine
Missing data in the exposure of interest and marginal structural models: A simulation study based on the Framingham Heart Study
Shortreed, SM; Forbes, AB
Statistics in Medicine, 29(4): 431-443.
10.1002/sim.3801
CrossRef
Journal of General Internal Medicine
Diabetes and Drug-Associated Hyperkalemia: Effect of Potassium Monitoring
Raebel, MA; Ross, C; Xu, S; Roblin, DW; Cheetham, C; Blanchette, CM; Saylor, G; Smith, DH
Journal of General Internal Medicine, 25(4): 326-333.
10.1007/s11606-009-1228-x
CrossRef
Annals of the Rheumatic Diseases
Evidence for differential acquired drug resistance to anti-tumour necrosis factor agents in rheumatoid arthritis
Finckh, A; Simard, JF; Gabay, C; Guerne, PA
Annals of the Rheumatic Diseases, 65(6): 746-752.
10.1136/ard.2005.045062
CrossRef
International Journal of Methods in Psychiatric Research
The use of propensity score methods in psychiatric research
VanderWeele, T
International Journal of Methods in Psychiatric Research, 15(2): 95-103.
10.1002/mpr.183
CrossRef
Medical Decision Making
Estimation of mortality rates for disease simulation models using Bayesian evidence synthesis
McMahon, PM; Zaslavsky, AM; Weinstein, MC; Kuntz, KM; Weeks, JC; Gazelle, GS
Medical Decision Making, 26(5): 497-511.
10.1177/0272989X06291326
CrossRef
Jama-Journal of the American Medical Association
Health of previously uninsured adults after acquiring medicare coverage
McWilliams, JM; Meara, E; Zaslavsky, AM; Ayanian, JZ
Jama-Journal of the American Medical Association, 298(): 2886-2894.

Health Services Research
Hospice effect on government expenditures among nursing home residents
Gozalo, PL; Miller, SC; Intrator, O; Barber, JP; Mor, V
Health Services Research, 43(1): 134-153.
10.1111/j.1475-6773.2007.00746.x
CrossRef
Educational Evaluation and Policy Analysis
Does NBPTS certification affect the number of colleagues a teacher helps with instructional matters?
Frank, KA; Sykes, G; Anagnostopoulos, D; Cannata, M; Chard, L; Krause, A; McCrory, R
Educational Evaluation and Policy Analysis, 30(1): 3-30.
10.3102/0162373707313781
CrossRef
AIDS
Effect of pulmonary tuberculosis on mortality in patients receiving HAART
Rubel, D; Boulmé, R; Eron, J; Sanne, I; Westreich, D; MacPhail, P; Van Rie, A; Malope-Kgokong, B; Ive, P
AIDS, 23(6): 707-715.
10.1097/QAD.0b013e328325d115
PDF (574) | CrossRef
AIDS
Breastfeeding and maternal HIV-1 disease progression and mortality
Sedgh, G; Spiegelman, D; Larsen, U; Msamanga, G; Fawzi, WW
AIDS, 18(7): 1043-1049.

PDF (84)
AIDS
Effect on mortality and virological response of delaying antiretroviral therapy initiation in children receiving tuberculosis treatment
Yotebieng, M; Van Rie, A; Moultrie, H; Cole, SR; Adimora, A; Behets, F; Meyers, T
AIDS, 24(9): 1341-1349.
10.1097/QAD.0b013e328339e576
PDF (295) | CrossRef
AIDS
The effect of HIV infection on overdose mortality
Galea, S; Wang, C; Vlahov, D; Galai, N; Cole, SR; Bareta, J; Pollini, R; Mehta, SH; Nelson, KE
AIDS, 19(9): 935-942.

PDF (107)
AIDS
Hormonal contraception and HIV acquisition: reanalysis using marginal structural modeling
Morrison, CS; Chen, P; Kwok, C; Richardson, BA; Chipato, T; Mugerwa, R; Byamugisha, J; Padian, N; Celentano, DD; Salata, RA
AIDS, 24(11): 1778-1781.
10.1097/QAD.0b013e32833a2537
PDF (219) | CrossRef
AIDS
Effect of tuberculosis on the survival of HIV-infected men in a country with low tuberculosis incidence
López-Gatell, H; Cole, SR; Margolick, JB; Witt, MD; Martinson, J; Phair, JP; Jacobson, LP
AIDS, 22(14): 1869-1873.
10.1097/QAD.0b013e32830e010c
PDF (216) | CrossRef
AIDS
CD38+CD8+ T-cells negatively correlate with CD4 central memory cells in virally suppressed HIV-1-infected individuals
Kolber, MA
AIDS, 22(15): 1937-1941.
10.1097/QAD.0b013e32830f97e2
PDF (105) | CrossRef
Anesthesiology
Perioperative Use of Dobutamine in Cardiac Surgery and Adverse Cardiac Outcome: Propensity-adjusted Analyses
Fellahi, J; Parienti, J; Hanouz, J; Plaud, B; Riou, B; Ouattara, A
Anesthesiology, 108(6): 979-987.
10.1097/ALN.0b013e318173026f
PDF (382) | CrossRef
Annals of Surgery
The Impact of an Intensivist-Model ICU on Trauma-Related Mortality
Nathens, AB; Rivara, FP; MacKenzie, EJ; Maier, RV; Wang, J; Egleston, B; Scharfstein, DO; Jurkovich, GJ
Annals of Surgery, 244(4): 545-554.
10.1097/01.sla.0000239005.26353.49
PDF (353) | CrossRef
Critical Care Medicine
The impact of development of acute lung injury on hospital mortality in critically ill trauma patients
Shah, CV; Localio, AR; Lanken, PN; Kahn, JM; Bellamy, S; Gallop, R; Finkel, B; Gracias, VH; Fuchs, BD; Christie, JD
Critical Care Medicine, 36(8): 2309-2315.
10.1097/CCM.0b013e318180dc74
PDF (294) | CrossRef
Critical Care Medicine
Impact of admission hyperglycemia on hospital mortality in various intensive care unit populations*
Whitcomb, BW; Pradhan, EK; Pittas, AG; Roghmann, M; Perencevich, EN
Critical Care Medicine, 33(12): 2772-2777.
10.1097/01.CCM.0000189741.44071.25
PDF (234) | CrossRef
European Journal of Cardiovascular Prevention & Rehabilitation
Beneficial effects of statins after percutaneous coronary intervention
Zhang, Z; Marroquin, OC; Weissfeld, JL; Stone, RA; Mulukutla, SR; Williams, DO; Selzer, F; Kip, KE
European Journal of Cardiovascular Prevention & Rehabilitation, 16(4): 445-450.
10.1097/HJR.0b013e32832a4e3b
PDF (174) | CrossRef
Epidemiology
The Consistency Statement in Causal Inference: A Definition or an Assumption?
Cole, SR; Frangakis, CE
Epidemiology, 20(1): 3-5.
10.1097/EDE.0b013e31818ef366
PDF (122) | CrossRef
Epidemiology
Marginal Structural Models for the Estimation of Direct and Indirect Effects
VanderWeele, TJ
Epidemiology, 20(1): 18-26.
10.1097/EDE.0b013e31818f69ce
PDF (345) | CrossRef
Epidemiology
Estimation of Direct Causal Effects
Petersen, ML; Sinisi, SE; van der Laan, MJ
Epidemiology, 17(3): 276-284.
10.1097/01.ede.0000208475.99429.2d
PDF (964) | CrossRef
Epidemiology
A Structural Approach to the Familial Coaggregation of Disorders
Hudson, JI; Javaras, KN; Laird, NM; VanderWeele, TJ; Pope, HG; Hernán, MA
Epidemiology, 19(3): 431-439.
10.1097/EDE.0b013e31816a9de7
PDF (362) | CrossRef
Epidemiology
Leisure-time Physical Activity and All-cause Mortality in an Elderly Cohort
Bembom, O; van der Laan, M; Haight, T; Tager, I
Epidemiology, 20(3): 424-430.
10.1097/EDE.0b013e31819e3f28
PDF (273) | CrossRef
Epidemiology
Bias due to Secondary Transmission in Estimation of Attributable Risk From Intervention Trials
Eisenberg, JN; Lewis, BL; Porco, TC; Hubbard, AH; Colford, JM
Epidemiology, 14(4): 442-450.
10.1097/01.ede.0000071411.19255.4c
PDF (363) | CrossRef
Epidemiology
Causal Analysis of Individual Change Using the Difference Score
Clarke, PS
Epidemiology, 15(4): 414-421.
10.1097/01.ede.0000120882.86399.e8
PDF (317) | CrossRef
Epidemiology
Effects of Physical Activity and Body Composition on Functional Limitation in the Elderly: Application of the Marginal Structural Model
Tager, IB; Haight, T; Sternfeld, B; Yu, Z; van Der Laan, M
Epidemiology, 15(4): 479-493.
10.1097/01.ede.0000128401.55545.c6
PDF (546) | CrossRef
Epidemiology
Properties of 2 Counterfactual Effect Definitions of a Point Exposure
Flanders, WD; Klein, M
Epidemiology, 18(4): 453-460.
10.1097/01.ede.0000261472.07150.4f
PDF (309) | CrossRef
Epidemiology
On the Relative Nature of Overadjustment and Unnecessary Adjustment
VanderWeele, TJ
Epidemiology, 20(4): 496-499.
10.1097/EDE.0b013e3181a82f12
PDF (234) | CrossRef
Epidemiology
The Relationship Between Neighborhood Poverty and Alcohol Use: Estimation by Marginal Structural Models
Cerdá, M; Diez-Roux, AV; Tchetgen Tchetgen, E; Gordon-Larsen, P; Kiefe, C
Epidemiology, 21(4): 482-489.
10.1097/EDE.0b013e3181e13539
PDF (299) | CrossRef
Epidemiology
Marginalia: Comparing Adjusted Effect Measures
Kaufman, JS
Epidemiology, 21(4): 490-493.
10.1097/EDE.0b013e3181e00730
PDF (157) | CrossRef
Epidemiology
Estimating Absolute Risks in the Presence of Nonadherence: An Application to a Follow-up Study With Baseline Randomization
Toh, S; Hernández-Díaz, S; Logan, R; Robins, JM; Hernán, MA
Epidemiology, 21(4): 528-539.
10.1097/EDE.0b013e3181df1b69
PDF (628) | CrossRef
Epidemiology
Marginal Structural Models to Estimate the Causal Effect of Zidovudine on the Survival of HIV-Positive Men
Hernán, MÁ; Brumback, B; Robins, JM
Epidemiology, 11(5): 561-570.

PDF (175)
Epidemiology
A Structural Approach to Selection Bias
Hernán, MA; Hernández-Díaz, S; Robins, JM
Epidemiology, 15(5): 615-625.
10.1097/01.ede.0000135174.63482.43
PDF (617) | CrossRef
Epidemiology
Potential Confounding by Exposure History and Prior Outcomes: An Example From Perinatal Epidemiology
Howards, PP; Schisterman, EF; Heagerty, PJ
Epidemiology, 18(5): 544-551.
10.1097/EDE.0b013e31812001e6
PDF (321) | CrossRef
Epidemiology
Marginal Structural Models as a Tool for Standardization
Sato, T; Matsuyama, Y
Epidemiology, 14(6): 680-686.
10.1097/01.EDE.0000081989.82616.7d
PDF (269) | CrossRef
Epidemiology
Estimating Direct Effects in Cohort and Case–Control Studies
Vansteelandt, S
Epidemiology, 20(6): 851-860.
10.1097/EDE.0b013e3181b6f4c9
PDF (480) | CrossRef
Epidemiology
On the Distinction Between Interaction and Effect Modification
VanderWeele, TJ
Epidemiology, 20(6): 863-871.
10.1097/EDE.0b013e3181ba333c
PDF (192) | CrossRef
JAIDS Journal of Acquired Immune Deficiency Syndromes
HIV Infection and Human Herpesvirus-8 Oral Shedding Among Men Who Have Sex with Men
Morrow, RA; Corey, L; Kiviat, N; Wald, A; Casper, C; Redman, M; Huang, M; Pauk, J; Lampinen, TM; Hawes, SE; Critchlow, CW
JAIDS Journal of Acquired Immune Deficiency Syndromes, 35(3): 233-238.

PDF (317)
JAIDS Journal of Acquired Immune Deficiency Syndromes
The Combined Effect of Modern Highly Active Antiretroviral Therapy Regimens and Adherence on Mortality Over Time
Lima, VD; Harrigan, R; Bangsberg, DR; Hogg, RS; Gross, R; Yip, B; Montaner, JS
JAIDS Journal of Acquired Immune Deficiency Syndromes, 50(5): 529-536.
10.1097/QAI.0b013e31819675e9
PDF (189) | CrossRef
JAIDS Journal of Acquired Immune Deficiency Syndromes
Effect of HAART on Incident Cancer and Noncancer AIDS Events Among Male HIV Seroconverters
Shiels, MS; Cole, SR; Wegner, S; Armenian, H; Chmiel, JS; Ganesan, A; Marconi, V; Martinez-Maza, O; Martinson, J; Weintrob, A; Jacobson, L; Crum-Cianflone, N
JAIDS Journal of Acquired Immune Deficiency Syndromes, 48(4): 485-490.
10.1097/QAI.0b013e31817dc42b
PDF (246) | CrossRef
JAIDS Journal of Acquired Immune Deficiency Syndromes
The Pediatric AIDS Severity Score (PASS): A Multidimensional AIDS-Severity Adjustment for Pediatric HIV Infection
for the Pediatric AIDS Clinical Trials Group 219 Study Team, ; Seage, GR; Buchacz, K; Weinberg, GA; Patel, K; McIntosh, K; Dankner, WM
JAIDS Journal of Acquired Immune Deficiency Syndromes, 43(5): 603-610.
10.1097/01.qai.0000242453.20521.4f
PDF (236) | CrossRef
Pharmacogenetics and Genomics
A systems biology network model for genetic association studies of nicotine addiction and treatment
Benowitz, N; Tyndale, RF; Bergen, AW; Conti, DV; for the Pharmacogenetics of Nicotine Addiction and Treatment Consortium, ; Thomas, PD; Mi, H; Swan, GE; Lerman, C
Pharmacogenetics and Genomics, 19(7): 538-551.
10.1097/FPC.0b013e32832e2ced
PDF (353) | CrossRef
Journal of Trauma and Acute Care Surgery
The National Study on Costs and Outcomes of Trauma
MacKenzie, EJ; Rivara, FP; Jurkovich, GJ; Nathens, AB; Frey, KP; Egleston, BL; Salkever, DS; Weir, S; Scharfstein, DO
Journal of Trauma and Acute Care Surgery, 63(6): S54-S67.
10.1097/TA.0b013e31815acb09
PDF (527) | CrossRef
Medical Care
Is Survival Better at Hospitals With Higher “End-of-Life” Treatment Intensity?
Barnato, AE; Chang, CH; Farrell, MH; Lave, JR; Roberts, MS; Angus, DC
Medical Care, 48(2): 125-132.
10.1097/MLR.0b013e3181c161e4
PDF (1133) | CrossRef
Medical Care
Confounding Control in Healthcare Database Research: Challenges and Potential Approaches
Brookhart, MA; Stürmer, T; Glynn, RJ; Rassen, J; Schneeweiss, S
Medical Care, 48(6): S114-S120.
10.1097/MLR.0b013e3181dbebe3
PDF (503) | CrossRef
Transplantation
Glucose Control is Associated With Patient Survival in Diabetic Patients After Renal Transplantation
Wiesbauer, F; Heinze, G; Regele, H; Hörl, WH; Schernthaner, GH; Schwarz, C; Kainz, A; Kramar, R; Oberbauer, R
Transplantation, 89(5): 612-619.
10.1097/TP.0b013e3181c6ffa4
PDF (950) | CrossRef
Back to Top | Article Outline
Keywords:

causality; counterfactuals; epidemiologic methods; longitudinal data; structural models; confounding; intermediate variables

© 2000 Lippincott Williams & Wilkins, Inc.

Twitter  Facebook

Login

Article Tools

Images

Share

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.