Causal Mediation Analysis of Survival Outcome with Multiple Mediators.

Goodness-of-fit tests for Cox's proportional hazards model and Aalen's additive risk model, in which each model is compared on an equal footing with the best fitting fully nonparametric model, are developed. The goodness-of-fit statistics are based on differences each model, with a fully nonparametric estimator of d recently introduced by the authors. Here A( , z) denotes the conditional hazard function of the survival time of an individual with covariate vector z. Comparison of the results of the tests makes it possible to decide whether Cox's proportional hazards or Aalen's additive risk model gives a better fit to the data. In addition, a goodness-of-fit test for Cox's model within the family of all proportional hazards models A(t, z) = AO(t)r(z), where AO is a baseline hazard function and r is a general relative risk function, is developed. Additive hazards and proportional hazards regression models used in the analysis of censored survival data can give substantially different results. For instance, in connection with a study of cancer mortality among Japanese atomic bomb survivors, Muirhead & Darby (1987) have noted that the two models give substantially different estimates of the age-spe- cific probability that an individual will develop radiation induced cancer. Muirhead & Darby (see also Aranda-Ordaz, 1983) introduced a generalized parametric model which contains parametric additive hazards and proportional hazards models as special cases. The goodness- of-fit of each model is then obtained by comparing with the best fitting model within the generalized family, allowing the two special models to be treated on an equal footing. Beyond the parametric setting, much effort has been devoted to the development of goodness-of-fit tests for Cox's (1972) proportional hazards model

Causal mediation modeling has become a popular approach for studying the effect of an exposure on an outcome through mediators. Currently, the literature on mediation analyses with survival outcomes largely focused on settings with a single mediator and quantified the mediation effects on the hazard, log hazard and log survival time (Lange and Hansen 2011; VanderWeele 2011). In this article, we propose a multi-mediator model for survival data by employing a flexible semiparametric probit model. We characterize path-specific effects (PSEs) of the exposure on the outcome mediated through specific mediators. We derive closed form expressions for PSEs on a transformed survival time and the survival probabilities. Statistical inference on the PSEs is developed using a nonparametric maximum likelihood estimator under the semiparametric probit model and the functional Delta method. Results from simulation studies suggest that our proposed methods perform well in finite sample. We illustrate the utility of our method in a genomic study of glioblastoma multiforme survival.

Among several semiparametric models, the Cox proportional hazard model is widely used to assess the association between covariates and the time-to-event when the observed time-to-event is interval-censored. Often, covariates are measured with error. To handle this covariate uncertainty in the Cox proportional hazard model with the interval-censored data, flexible approaches have been proposed. To fill a gap and broaden the scope of statistical applications to analyze time-to-event data with different models, in this paper, a general approach is proposed for fitting the semiparametric linear transformation model to interval-censored data when a covariate is measured with error. The semiparametric linear transformation model is a broad class of models that includes the proportional hazard model and the proportional odds model as special cases. The proposed method relies on a set of estimating equations to estimate the regression parameters and the infinite-dimensional parameter. For handling interval censoring and covariate measurement error, a flexible imputation technique is used. Finite sample performance of the proposed method is judged via simulation studies. Finally, the suggested method is applied to analyze a real data set from an AIDS clinical trial.

In the disciplines of medicine, biology and social science, causal mediation analysis is widely used to examine the effect of an exposure on an outcome through different causal pathways. Recently, there is rapid development of mediation analysis with survival data. An early work of Lange and Hansen proposed to identify the direct and indirect causal effects under the additive hazards model with a single mediator. However, inaccurate measurements of mediators and confounders may lead to biased causal effect estimation. We propose a measurement error correction approach to tackle measurement error in the mediators and confounders under the additive hazards model where multiple mediators are present. We apply the proposed corrected method to the ACTG175 Study data set, and uncover interesting findings.

A Bayesian network interpretation of the Cox's proportional hazard model

The effect of socioeconomic status on adolescent substance abuse may be mediated by family socialization practices. However, traditional mediation analysis using a product or difference method is susceptible to bias when assumptions are not addressed. We aimed to use a potential outcomes framework to assess assumptions of exposure-mediator interaction and of no confounding of the results. We revisited a traditional mediation analysis with a multiple mediator causal mediation approach using data from 17,761 Norwegian young people (13-18 years), 51% female. Data were collected through a print questionnaire. Socioeconomic status was operationalized as parental education and employment status (employed or receiving welfare); drinking behavior as the frequency of alcohol consumption and frequency of intoxication in the past year; and socialization practices as general parenting measures, alcohol-related parental permissiveness, and parent drinking behavior. There was no consistent evidence of exposure-mediator interaction. Formal sensitivity analysis of mediator-outcome confounding was not possible in the multiple mediator model, and this analysis supported the hypothesis that socioeconomic status effects on adolescent substance abuse are fully mediated by family socialization practices, with apparently stronger effects in younger age groups observed in plots. We found that the effect of socioeconomic status on adolescent substance abuse was fully mediated by family socialization practices. While our analysis provides more rigorous support for causal inferences than past work, we could not completely rule out the possibility of unmeasured confounding.

Longitudinal association between sleep duration and fall risk in middle-aged and older adults: A multiple mediation pathways analysis.

To determine and compare the clinical and immunologic outcomes for HIV-infected women initiated on antiretroviral therapy (ART), with and without previous exposure to single-dose nevirapine in the MTCT-Plus programme - Kampala, Uganda, from 2003 to 2011. Retrospective comparison of prospectively collected programmatic data of clinical and immunologic treatment outcomes among HIV-infected Ugandan women, with and without prior exposure to sdNVP, who received NNRTI-based ART for a median follow-up of 6years. Of the 408 women in the programme, 289 (70.8%) were started on ART, of whom 205 (70.9%) had prior exposure to sdNVP. Clinical, immunologic and combined (clinical and or immunologic) treatment failure occurred in 29 (10.0%), 132 (45.7%) and 142 (49.1%) women, respectively. There was no significant difference in the distribution of time to immunologic failure for women by exposure to sdNVP (log-rank P=0.98). In Cox proportional hazard modelling, exposure to sdNVP was not associated with immunologic failure [adjusted hazard ratio (HR)=0.89, 95% confidence interval (CI): 0.61-1.30]. CD4 count >100cells/mm(3) at initiation was associated with reduced incidence of immunologic failure in adjusted analyses (HR=0.32, 95% CI: 0.22-0.48). HIV-infected Ugandan women initiated on an NVP-based ART regimen had similar immunologic treatment outcomes irrespective of previous NVP exposure. CD4 cell count prior to initiating HAART was a key prognostic factor for successful long-term immunologic treatment outcomes. In poor settings, regular follow-up of patients on HAART with adequate counselling to promote adherence and safe disclosure may promote low clinical failure rates.

Questions about the mode of action (MoA) of a drug have interest to scientific communities as well as regulatory authorities. In the absence of an already established MoA such questions may be enlightened by causal mediation analysis in clinical trials. In this paper, we present a general framework that facilitates causal mediation analysis in a setting of a clinical trial where both the outcome of interest, one or more mediators and additional potential post-baseline confounders are measured repeatedly at planned visits. The targeted estimands are path-specific effects implicitly defined from a causal diagram that includes all the longitudinally measured variables. The framework including estimation method is developed with inspiration from a similar approach for time-to-event outcomes. The novelty of our approach compared to many currently used methods is that it directly establishes MoA, takes the full amount of longitudinal data into account, and properly adjusts for sources of confounding. We motivate and illustrate the approach with examples from the STEP 2 clinical trial. Modular SAS code is provided that can be used in a general setting.

SUMMARY When there are covariate effects to be considered, multi-state survival analysis is dominated either by parametric Markov regression models or by semiparametric Markov regression models using Cox's (1972) proportional hazards models for transition intensities between the states. The purpose of this research work is to study alternatives to Cox's model in a general finite-state Markov process setting. We shall look at two alternative models, Aalen's (1989) nonparametric additive hazards model and Lin & Ying's (1994) semiparametric additive hazards model. The former allows the effects of covariates to vary freely over time, while the latter assumes that the regression coefficients are constant over time. With the basic tools of the product integral and the functional delta-method, we present an estimator of the transition probability matrix and develop the large-sample theory for the estimator under each of these two models. Data on 1459 HLA identical sibling transplants for acute leukaemia from the International Bone Marrow Transplant Registry serve as illustration.

BackgroundMapping expression quantitative trait loci (eQTLs) has provided insight into gene regulation. Compared to cis-eQTLs, the regulatory mechanisms of trans-eQTLs are less known. Previous studies suggest that trans-eQTLs may regulate expression of remote genes by altering the expression of nearby genes. Trans-association has been studied in the mediation analysis with a single mediator. However, prior applications with one mediator are prone to model misspecification due to correlations between genes. Motivated from the observation that trans-eQTLs are more likely to associate with more than one cis-gene than randomly selected SNPs in the GTEx dataset, we developed a computational method to identify trans-eQTLs that are mediated by multiple mediators.ResultsWe proposed two hypothesis tests for testing the total mediation effect (TME) and the component-wise mediation effects (CME), respectively. We demonstrated in simulation studies that the type I error rates were controlled in both tests despite model misspecification. The TME test was more powerful than the CME test when the two mediation effects are in the same direction, while the CME test was more powerful than the TME test when the two mediation effects are in opposite direction. Multiple mediator analysis had increased power to detect mediated trans-eQTLs, especially in large samples. In the HapMap3 data, we identified 11 mediated trans-eQTLs that were not detected by the single mediator analysis in the combined samples of African populations. Moreover, the mediated trans-eQTLs in the HapMap3 samples are more likely to be trait-associated SNPs. In terms of computation, although there is no limit in the number of mediators in our model, analysis takes more time when adding additional mediators. In the analysis of the HapMap3 samples, we included at most 5 cis-gene mediators. Majority of the trios we considered have one or two mediators.ConclusionsTrans-eQTLs are more likely to associate with multiple cis-genes than randomly selected SNPs. Mediation analysis with multiple mediators improves power of identification of mediated trans-eQTLs, especially in large samples.

BackgroundMediation analysis methodology underwent many advancements throughout the years, with the most recent and important advancement being the development of causal mediation analysis based on the counterfactual framework. However, a previous review showed that for experimental studies the uptake of causal mediation analysis remains low. The aim of this paper is to review the methodological characteristics of mediation analyses performed in observational epidemiologic studies published between 2015 and 2019 and to provide recommendations for the application of mediation analysis in future studies.MethodsWe searched the MEDLINE and EMBASE databases for observational epidemiologic studies published between 2015 and 2019 in which mediation analysis was applied as one of the primary analysis methods. Information was extracted on the characteristics of the mediation model and the applied mediation analysis method.ResultsWe included 174 studies, most of which applied traditional mediation analysis methods (n = 123, 70.7%). Causal mediation analysis was not often used to analyze more complicated mediation models, such as multiple mediator models. Most studies adjusted their analyses for measured confounders, but did not perform sensitivity analyses for unmeasured confounders and did not assess the presence of an exposure-mediator interaction.ConclusionsTo ensure a causal interpretation of the effect estimates in the mediation model, we recommend that researchers use causal mediation analysis and assess the plausibility of the causal assumptions. The uptake of causal mediation analysis can be enhanced through tutorial papers that demonstrate the application of causal mediation analysis, and through the development of software packages that facilitate the causal mediation analysis of relatively complicated mediation models.

Introduction: The Hypertension Control Program in Argentina (HCPIA), a cluster-randomized trial of a community health worker-led multicomponent intervention on blood pressure (BP) control in low-income patients in Argentina, previously reported a significant reduction in systolic BP in the intervention compared to the control group over 18 months (6.6 mm Hg net difference; p<0.001). Understanding which elements of a multicomponent intervention are important contributors to its success is critical for future BP control efforts. We conducted a mediation analysis to determine the contributions of seven measured intervention targets to the systolic BP intervention effect. Methods: HCPIA randomized 18 primary healthcare clinics treating uninsured patients to either a multicomponent intervention, consisting of health coaching, home blood pressure monitoring, blood pressure audit and feedback, physician education, and a mobile health intervention, or usual care. A total of 1,357 hypertensive participants completed the trial and are included in the mediation analyses. Changes in seven potential mediators over the 18-month intervention were assessed: medication adherence score, medication adjustments (addition of new medication or dose titration), waist circumference, physical activity, fruit and vegetable consumption, sodium consumption, and alcohol consumption. Generalized estimating equations were used to compare changes in potential mediators between intervention groups. Path analytic models were used to test single-mediator and multiple mediator indirect intervention effects using the Lavaan package in R statistical software. Mediation was considered significant at p<0.10. Results: All potential mediators were significantly improved in the intervention group compared to the control group. Fruit and vegetable consumption, medication adherence, physical activity, and salt consumption are each significant mediators accounting for 4.3, 3.8, 3.3, and 3.2%, respectively, of the intervention effect in single-mediator models. The multiple mediation model explained 15.5% (1.3 mm Hg) of the total intervention effect. Changes in fruit and vegetable consumption and medication adherence remained significant mediators in multiple mediation analysis, responsible for 3.8% and 4.3% of the net reduction in systolic BP, respectively. Other potential mediators did not account for a significant proportion of the intervention effect. Conclusions: Increases in medication adherence and fruit and vegetable consumption were responsible for the largest proportions of the explained intervention effect.

To investigate the factors related to viral load becoming undetectable among patients from Southern Alberta who started a protease inhibitor for the first time, and to determine the factors related to subsequent re-emergence of detectable viral load amongst those patients whose viral load initially became undetectable. A total of 243 patients from the Southern Alberta Clinic had started a protease inhibitor for the first time and had been followed up for a median time of 32 weeks. Standard survival techniques including Kaplan-Meier techniques and Cox proportional hazards models were used to determine which factors were related to viral load becoming undetectable. At 24 weeks after first exposure to a protease inhibitor, 52.8% of the patients [95% confidence interval (Cl), 45.2-56.6] had achieved an undetectable viral load. In a multivariate analysis, those with a higher initial viral load were less likely to become undetectable [relative hazard (RH), 0.50; 95% Cl, 0.35-0.70; P < 0.0001], whereas those starting more new drugs (RH per new drug, 1.54; 95% Cl, 1.01-2.11; P = 0.048) were significantly more likely to achieve an undetectable viral load. Amongst 111 patients whose viral load became undetectable, Kaplan-Meier analysis indicated that 15.5% of patients experienced re-emergence of detectable viral load at 24 weeks after the first undetectable viral load. A higher CD4 cell count was associated with a lower risk of viral load becoming detectable (RH, 0.73; 95% Cl, 0.53-1.00; P = 0.049), as was treatment with indinavir (versus any other protease inhibitor RH, 0.17; 95% Cl, 0.03-0.86; P = 0.033). A significant proportion of patients in a routine clinic setting achieved an undetectable viral load measurement after first starting a protease inhibitor; viral load in patients with a higher CD4 cell count was more likely to become and stay undetectable. There was no evidence that patients who were drug-naive experienced significantly worse virological effects than drug-experienced patients, as long as the same number of new drugs was started at the date of first exposure to a protease inhibitor. Further follow-up of these patients is warranted to study the longer term effects of treatment with protease inhibitors.

Optimal Duration of Imatinib Treatment / Deep Molecular Response for Treatment-Free Remission after Imatinib Discontinuation from a Canadian Tyrosine Kinase Inhibitor Discontinuation Trial