Methods for handling missing binary data in substance use disorder trials

Missing data are a ubiquitous problem in longitudinal substance use disorder (SUD) clinical trials. In particular, the rates of missingness are often high and study participants often intermittently skip their scheduled outcome assessments, leading to so-called "non-monotone" missing data patterns. Moreover, when the primary outcome is a measure of substance use, study investigators often have strong prior beliefs based on their clinical experience that those participants with missing data are more likely to be using substances at those occasions, i.e., data are missing not at random (MNAR). Although approaches for handling missing data are well-developed when the missing data patterns are monotone, arising primarily from study participants withdrawing from the trial prematurely, fewer methods are available for non-monotone missingness. In this paper we review some conventional, as well as more novel, methods for handling non-monotone missingness in SUD trials when the repeatedly measured outcome variable is binary (e.g., denoting presence/absence of substance use). We compare and contrast the different approaches using data from a longitudinal clinical trial of four psychosocial treatments from the Collaborative Cocaine Treatment Study. We conclude by making some recommendations to the SUD research community concerning how more principled methods for handling missing data can be incorporated in the analysis and reporting of trial results.

Statistical Modeling of Longitudinal Data with Non-ignorable Non-monotone Missingness with Semiparametric Bayesian and Machine Learning Components

In longitudinal studies, outcomes are measured repeatedly over time and it is common that not all the patients will be measured throughout the study. For example patients can be lost to follow-up (monotone missingness) or miss one or more visits (non-monotone missingness); hence there are missing outcomes. In the longitudinal setting, we often assume the missingness is related to the unobserved data, which is non-ignorable. Pattern-mixture models (PMM) analyze the joint distribution of outcome and patterns of missingness in longitudinal data with non-ignorable nonmonotone missingness. Existing methods employ PMM and impute the unobserved outcomes using the distribution of observed outcomes, conditioned on missing patterns. We extend the existing methods using latent class analysis (LCA) and a shared-parameter PMM. The LCA groups patterns of missingness with similar features and the shared-parameter PMM allows a subset of parameters to be different between latent classes when fitting a model. We also propose a method for imputation using distribution of observed data conditioning on latent class. Our model improves existing methods by accommodating data with small sample size. In a simulation study our estimator had smaller mean squared error than existing methods. Our methodology is applied to data from a phase II clinical trial that studies quality of life of patients with prostate cancer receiving radiation therapy.

Evaluation of treatment effect modification by biomarkers measured pre- and post-randomization in the presence of non-monotone missingness

In vaccine studies, an important research question is to study effect modification of clinical treatment efficacy by intermediate biomarker-based principal strata. In settings where participants entering a trial may have prior exposure and therefore variable baseline biomarker values, clinical treatment efficacy may further depend jointly on a biomarker measured at baseline and measured at a fixed time after vaccination. This makes it important to conduct a bivariate effect modification analysis by both the intermediate biomarker-based principal strata and the baseline biomarker values. Existing research allows this assessment if the sampling of baseline and intermediate biomarkers follows a monotone pattern, i.e., if participants who have the biomarker measured post-randomization would also have the biomarker measured at baseline. However, additional complications in study design could happen in practice. For example, in a dengue correlates study, baseline biomarker values were only available from a fraction of participants who have biomarkers measured post-randomization. How to conduct the bivariate effect modification analysis in these studies remains an open research question. In this article, we propose approaches for bivariate effect modification analysis in the complicated sampling design based on an estimated likelihood framework. We demonstrate advantages of the proposed method over existing methods through numerical studies and illustrate our method with data sets from two phase 3 dengue vaccine efficacy trials.

Simulation-based study comparing multiple imputation methods for non-monotone missing ordinal data in longitudinal settings

The application of multiple imputation (MI) techniques as a preliminary step to handle missing values in data analysis is well established. The MI method can be classified into two broad classes, the joint modeling and the fully conditional specification approaches. Their relative performance for the longitudinal ordinal data setting under the missing at random (MAR) assumption is not well documented. This article intends to fill this gap by conducting a large simulation study on the estimation of the parameters of a longitudinal proportional odds model. The two MI methods are also illustrated in quality of life data from a cancer clinical trial.