Estimation of Controlled Direct Effects in Longitudinal Mediation Analyses with Latent Variables in Randomized Studies

A Causal Mediation Approach to Account for Interaction of Treatment and Intercurrent Events: Using Hypothetical Strategy

Hypothetical strategy is a common strategy for handling intercurrent events (IEs). No current guideline or study considers treatment-IE interaction to target the estimand in any one IE-handling strategy. Based on the hypothetical strategy, we aimed to (1) assess the performance of three estimators with different considerations for the treatment-IE interaction in a simulation and (2) compare the estimation of these estimators in a real trial. Simulation data were generalized based on realistic clinical trials of Alzheimer's disease. The estimand of interest was the effect of treatment with no IE occurring under the hypothetical strategy. Three estimators, namely, G-estimation with and without interaction and IE-ignored estimation, were compared in scenarios where the treatment-IE interaction effect was set as -50% to 50% of the main effect. Bias was the key performance measure. The real case was derived from a randomized trial of methadone maintenance treatment. Only G-estimation with interaction exhibited unbiased estimations regardless of the existence, direction or magnitude of the treatment-IE interaction in those scenarios. Neglecting the interaction and ignoring the IE would introduce a bias as large as 0.093 and 0.241 (true value, -1.561) if the interaction effect existed. In the real case, compared with G-estimation with interaction, G-estimation without interaction and IE-ignored estimation increased the estimand of interest by 33.55% and 34.36%, respectively. This study highlights the importance of considering treatment-IE interaction in the estimand framework. In practice, it would be better to include the interaction in the estimator by default.

Adjusting for Baseline Measurements of the Mediators and Outcome as a First Step Toward Eliminating Confounding Biases in Mediation Analysis

Mediation analysis prevails for researchers probing the etiological mechanisms through which treatment affects an outcome. A central challenge of mediation analysis is justifying sufficient baseline covariates that meet the causal assumption of no unmeasured confounding. But current practices routinely overlook this assumption. In this article, we suggest a relatively easy way to mitigate the risks of incorrect inferences resulting from unmeasured confounding: include pretreatment measurements of the mediator(s) and the outcome as baseline covariates. We explain why adjusting for pretreatment baseline measurements is a necessary first step toward eliminating confounding biases. We hope that such a practice can encourage explication, justification, and reflection of the causal assumptions underpinning mediation analysis toward improving the validity of causal inferences in psychology research.

A Causal Approach to Functional Mediation Analysis with Application to a Smoking Cessation Intervention

The increase in the use of mobile and wearable devices now allows dense assessment of mediating processes over time. For example, a pharmacological intervention may have an effect on smoking cessation via reductions in momentary withdrawal symptoms. We define and identify the causal direct and indirect effects in terms of potential outcomes on the mean difference and odds ratio scales, and present a method for estimating and testing the indirect effect of a randomized treatment on a distal binary variable as mediated by the nonparametric trajectory of an intensively measured longitudinal variable (e.g., from ecological momentary assessment). Coverage of a bootstrap test for the indirect effect is demonstrated via simulation. An empirical example is presented based on estimating later smoking abstinence from patterns of craving during smoking cessation treatment. We provide an R package, funmediation, available on CRAN at https://cran.r-project.org/web/packages/funmediation/index.html, to conveniently apply this technique. We conclude by discussing possible extensions to multiple mediators and directions for future research.

Evaluating interpersonal discrimination and depressive symptoms as partial mediators of the effects of education on cognition: Evidence from the Study of Healthy Aging in African Americans (STAR)

INTRODUCTION

Education is correlated with positive health outcomes, but associations are sometimes weaker among African Americans. The extent to which exposure to discrimination and depressive symptoms attenuates the education-cognition link has not been investigated.

METHODS

Study of Healthy Aging in African Americans (STAR) participants (n = 764; average age 69 years) completed the Spanish and English Neuropsychological Assessment Scales. We assessed everyday and major lifetime discrimination and depressive symptoms as mediators of education effects on cognition using G-estimation with measurement error corrections.

RESULTS

Education was correlated with greater major lifetime and everyday discrimination but lower depressive symptoms. Accounting for discrimination and depressive symptoms slightly reduced the estimated effect of education on cognition. The estimated total effect of graduate education (vs 

DISCUSSION

Education has robust effects on later-life cognition after controlling multiple mediating pathways and offsetting mechanisms.

Collapse

Key Words

• African Americans
• cognition
• depression
• discrimination
• education
• psychosocial stress

Collapse

MESH Headings

• Aged
• Humans
• Black or African American
• Cognition
• Depression/psychology
• Educational Status
• Healthy Aging
• Racism/ethnology
• Racism/psychology

Collapse

Grants

• R01 AG066132 NIA NIH HHS
• R01 AG050782 NIA NIH HHS
• RF1 AG050782 NIA NIH HHS
• P30 AG072972 NIA NIH HHS
• R00 AG073457 NIA NIH HHS

Collapse

Affiliation(s)

Dakota W. Cintron Center for Health and Community, University of California, San Francisco, San Francisco, CA Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA
Camilla Calmasini Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA
Lisa L. Barnes Rush Alzheimer’s Disease Center, Rush University Medical Center, Chicago, IL, USA Department of Neurology, Rush University Medical Center, Chicago, IL, USA
Dan M. Mungas Department of Neurology, University of California Davis Health, Sacramento, CA, USA
Rachel A. Whitmer Department of Public Health Sciences, University of California Davis, Davis, CA, USA Division of Research, Kaiser Permanente Northern California, Oakland, CA, USA
Chloe W. Eng Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA
Paola Gilsanz Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA Division of Research, Kaiser Permanente Northern California, Oakland, CA, USA
Kristen M. George Department of Neurology, University of California Davis Health, Sacramento, CA, USA Department of Public Health Sciences, University of California Davis, Davis, CA, USA
Rachel Peterson Department of Neurology, University of California Davis Health, Sacramento, CA, USA Department of Public Health Sciences, University of California Davis, Davis, CA, USA
M. Maria Glymour Center for Health and Community, University of California, San Francisco, San Francisco, CA Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA

Collapse

Integrating Technology Adoption Models Into Implementation Science Methodologies: A Mixed-Methods Preimplementation Study

BACKGROUND AND OBJECTIVES

Sustainable implementation of patient-oriented technologies in health care settings is challenging. Preimplementation studies guided by the Consolidated Framework for Implementation Research (CFIR) can provide opportunities to address barriers and leverage facilitators that can maximize the likelihood of successful implementation. When looking to implement patient-oriented technologies, preimplementation studies may also benefit from guidance from a conceptual framework specific to technology adoption such as the Unified Theory of Acceptance and Use of Technology. This study was, therefore, aimed at identifying determinants for the successful implementation of a patient-oriented technology (i.e., automated pain behavior monitoring [APBM] system) within a health care setting (i.e., long-term care [LTC] facility).

RESEARCH DESIGN AND METHODS

Using a mixed-methods study design, 164 LTC nurses completed a set of questionnaires and 68 LTC staff participated in individual interviews involving their perceptions of an APBM system in LTC environments. Quantitative data were analyzed using a series of mediation analyses and narrative responses were examined using directed content analysis.

RESULTS

Performance expectancy and effort expectancy partially and fully mediated the influence of implementation, readiness for organizational change, and technology readiness constructs on behavioral intentions to use the APBM system in LTC environments. Findings from the qualitative portion of this study provide guidance for the development of an intervention that is grounded in the CFIR.

DISCUSSION AND IMPLICATIONS

Based on our results, we offer recommendations for the implementation of patient-oriented technologies in health care settings.

Quantifying direct and indirect effect for longitudinal mediator and survival outcome using joint modeling approach

Longitudinal biomarkers are widely used in biomedical and translational researches to monitor the progressions of diseases. Methods have been proposed to jointly model longitudinal data and survival data, but its causal mechanism is yet to be investigated rigorously. Understanding how much of the total treatment effect is through the biomarker is important in understanding the treatment mechanism and evaluating the biomarker. In this work, we propose a causal mediation analysis method to compute the direct and indirect effects, when a joint modeling approach is used to take the longitudinal biomarker as the mediator and the survival endpoint as the outcome. Such a joint modeling approach allows us to relax the commonly used "sequential ignorability" assumption. We demonstrate how to evaluate longitudinally measured biomarkers using our method with two case studies, an AIDS study and a liver cirrhosis study.

Estimators for handling COVID-19-related intercurrent events with a hypothetical strategy

The COVID-19 pandemic has affected clinical trials across disease areas, raising the questions how interpretable results can be obtained from impacted studies. Applying the estimands framework, analyses may seek to estimate the treatment effect in the hypothetical absence of such impact. However, no established estimators exist. This simulation study, based on an ongoing clinical trial in patients with Tourette syndrome, compares the performance of candidate estimators for estimands including either a continuous or binary variable and applying a hypothetical strategy for COVID-19-related intercurrent events (IE). The performance is investigated in a wide range of scenarios, under the null and the alternative hypotheses, including different modeling assumptions for the effect of the IE and proportions of affected patients ranging from 10% to 80%. Bias and type I error inflation were minimal or absent for most estimators under most scenarios, with only multiple imputation- and weighting-based methods displaying a type I error inflation in some scenarios. Of more concern, all methods that discarded post-IE data displayed a sharp decrease of power proportional to the proportion of affected patients, corresponding to both a reduced precision of estimation and larger confidence intervals. The simulation study shows that de-mediation via g-estimation is a promising approach. Besides showing the best performance in our simulation study, these approaches allow to estimate the effect of the IE on the outcome and cross-compare between different studies affected by similar IEs. Importantly, the results can be extrapolated to IEs not related to COVID-19 that follow a similar causal structure.