A reanalysis of a longitudinal scleroderma clinical trial using non-ignorable missingness models

Experiments involving repeated observations of multivariate outcomes are common in biomedical and public health researches which lead to multivariate longitudinal data. These kinds of data have a unique property in the sense that they allow the researcher to study the joint evolution of the multiple outcomes over the time. Recently, there has been a considerable amount of interest on using Bayesian modelling of longitudinal data, data which commonly suffer from incomplete observations. Those Bayesian models for longitudinal data that rely on the ignorability assumption of the dropout mechanism might give misleading inferences. Hence, there is a need to further study the impact of departures from the ignorability assumption on the Bayesian estimates of the model parameters. Current methodology for Bayesian sensitivity analysis mostly involves single response variable in both cross-sectional and longitudinal studies. In this paper, we propose a multivariate extension of the Bayesian index of sensitivity to non-ignorability for the general case of multivariate longitudinal studies with the possibility of having mixed correlated outcomes and a vector of multiple non-ignorability parameters in the missing mechanism. To simultaneously model the mixed responses over the time, we use a random effect latent variable approach. We illustrate the method conducting some simulation studies and analyzing a real data set from a longitudinal study for the comparison of two oral treatments for toenail dermatophyte onychomycosis.

In the analyses of incomplete longitudinal clinical trial data, there has been a shift, away from simple ad hoc methods that are valid only if the data are missing completely at random (MCAR), to more principled (likelihood-based or Bayesian) ignorable analyses, which are valid under the less restrictive missing at random (MAR) assumption. The availability of the necessary standard statistical software allows for such analyses in practice. Although the possibility of data missing not at random (MNAR) cannot be ruled out, it is argued that analyses valid under MNAR are not well suited for the primary analysis in clinical trials. Therefore, rather than either forgetting about or blindly shifting to an MNAR framework, the optimal place for MNAR analyses is within a sensitivity analysis context. Such analyses can be used, for example, to assess how sensitive results from an ignorable analysis are to possible departures from MAR and how much results are affected by influential observations. In this article, we apply the local influence sensitivity tool (Verbeke et al., 2001) to a longitudinal depression trial, thereby applying it to continuous outcomes from clinical trials.

Valid inference can be drawn from a random-effects model for repeated measures that are incomplete if whether the data are missing or not, known as missingness, is independent of the missing data. Data that are missing completely at random or missing at random are two data types for which missingness is ignorable. Given ignorable missingness, statistical inference can proceed without addressing the source of the missing data in the model. If the missingness is not ignorable, however, recommendations are to fit multiple models that represent different plausible explanations of the missing data. A popular choice in methods for evaluating nonignorable missingness is a random-effects pattern-mixture model that extends a random-effects model to include one or more between-subjects variables that represent fixed patterns of missing data. Generally straightforward to implement, a fixed pattern-mixture model is one among several options for assessing nonignorable missingness, and when it is used as the sole model to address nonignorable missingness, understanding the impact of missingness is greatly limited. This paper considers alternatives to a fixed pattern-mixture model for nonignorable missingness that are generally straightforward to fit and encourage researchers to give greater attention to the possible impact of nonignorable missingness in longitudinal data analysis. Patterns of both monotonic and non-monotonic (intermittently) missing data are addressed. Empirical longitudinal psychiatric data are used to illustrate the models. A small Monte Carlo data simulation study is presented to help illustrate the utility of such methods.Supplementary InformationThe online version contains supplementary material available at 10.3758/s13428-023-02128-y.

Longitudinal designs for biomarker traits are very appealing for a number of reasons. First, the collection of repeated biomarker measurements over time allows investigators to characterize both the overall and individual trajectories in these measurements. This helps investigators to understand overall effects as well as how individuals vary around these average effects. Distinguishing individual heterogeneity from average trend is impossible without longitudinal data. This distinction is important for understanding individual between-subject in its own right, as well as in developing predictors of disease outcome from longitudinal marker data. There are many statistical issues that are unique to longitudinal data analysis ranging from correctly specifying realistic models which characterize average changes over time and individual variation, accounting for observations that are taken at unequally spaced observation times which may differ between individuals, to accounting for missing data due to occasionally missed measurements or loss to follow-up. When the reason for missing data is not related to the underlying observation process, most longitudinal data analysis methods can be routinely applied. However, complex modelling is necessary when the reason for missing an observation or dropout is directly related to the underlying outcome process [1–2]. Statistical techniques that appropriately analyze continuous and discrete longitudinal data have been discussed by many authors. Diggle, et al. [3] provides a good summary of these techniques for the practitioner, while Fitzmaurice et al. [4] discusses recent advances from a more technical perspective. The three papers in this section present innovative approaches for solving important methodological problems in biomarker studies. Most standard longitudinal data analysis techniques assume that the measurement process is independent of the outcome process. However, when the outcome is related to the underlying observation process, failure to properly account for the observation process can result in bias [5]. Schildcrout et al.[6] proposed an approach for regression modelling for longitudinal binary data where the outcome is measured with probabilities determined by an auxiliary variable that is related, but not identical to the binary outcome. Modelling multivariate longitudinal biomarker data is an increasingly common issue with the advent of an increasing number of important biomarker techniques (e.g., multiple cytokines in a multiplex assay or multiple reproductive hormones that describe complex biological phenomenon). Although various authors have proposed techniques for analyzing multivariate longitudinal data [7–8 and others], little work has focused on estimating multivariate linear mixed models under known constraints in the relationship between the multivariate outcomes. Roy et al [9] proposes such an approach for analyzing multivariate reproductive hormones over the menstrual cycle, where established relationships have solid biological justification. Finally, the collection of longitudinal biomarker data as disease predictors is becoming an increasingly important design in diagnostic medicine. Pfeiffer et al. [10] proposes flexible non-parametric methodology for combining biomarkers that exploits the longitudinal structure in formulating a score that is used for predicting a binary outcome.

Multivariate binary longitudinal data sets are comprised of repeated observations of a vector outcome and a set of multi-dimensional covariates under each of many independent clusters. The multivariate longitudinal data of this type exhibit two-way correlations:first, at a given point of time, the multivariate responses under a cluster are correlated; second, these correlated responses are repeatedly observed under a cluster over a period of time and the repeated responses are also correlated. One objective is to describe the marginal expectation of the outcome variable as a function of the covariates while accounting for the correlations among the repeated multi-dimensional responses for a given cluster. Based on the assumption that each pair of the multi-dimensional responses in a cluster follow the bivariate beta correlated binary distribution, this paper proposes a general class of estimating equations approach for inference on response probabilities and the scale parameter of the beta correlated binary distribution. The estimating equations have solutions which are consistent and asymptotically Gaussian. We illustrate the use of the estimating equations approach by analyzing a multivariate longitudinal data set on health care utilization by families in the city of St. John's, Canada.

Asian Americans, Native Hawaiians, and Pacific Islanders (AA/NH/PI) contribute significantly to the state’s ethnic diversity. Despite their substantial presence, these groups face considerable health disparities, particularly in access to and participation in clinical trials. Furthermore, detailed accrual data based on ethnicity is missing as national datasets often aggregate AA/NH/PI populations into a single racial category. The aim of this study is to investigate change in accrual rates to interventional oncology clinical trials over time and determine the patient demographic factors that caused disparities in the trial participation. Through the Clinical Trial Management System at the University of Hawai'i Cancer Center, we collected patient enrollment data for NCI-funded interventional clinical trials (i.e., National Cancer Trials Network, NCTN and NCI Community Oncology Research Program, NCORP trials) conducted within the Hawai'i Cancer Consortium (HCC) between 2005 and 2020. The HCC is constituted by seven entities, including 2 major healthcare organizations in Hawai'i. The dataset comprised of demographic details such as biological sex, age, and race. Among 2, 341 study participants who were identified by screening, 870 were enrolled in non-therapeutic studies, while 1, 471 were involved in therapeutic trials. Nearly 50% identified as Asian American, 25% as White, and 12% as Native Hawaiian, and 3% as Pacific Islander. The remaining 10% were categorized as Black, American Indian, Multiracial, Other, or Unknown. The number of Asian American participants showed a declining trend until 2014, after which it began to recover. The number of Native Hawaiian and Pacific Islander participants has been stable over this time period. Female participation consistently exceeded male participation over the study period. The annual enrollment of all participants remained low until 2014 but steadily increased, reaching its peak in 2019. This study revealed that local AA/NH/PI accrual has a different trend from the national accrual over time. While the findings suggest that local, continuous community outreach efforts have been effective, male patient accrual remains inequitable. To address this disparity, our center is currently implementing an educational program aimed at Native Hawaiian men to help eliminate mistrust in clinical trial participation. Citation Format: Aiesha Pradhan, Jangsoon Lee, Naoto T. Ueno, Toshiaki Iwase. The change in Asian American/Native American/Pacific Islander demographics in interventional clinical trials in Hawai'i [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7092.

The last-observation-carried-forward imputation method is commonly used for imputing data missing due to dropouts in longitudinal clinical trials. The method assumes that outcome remains constant at the last observed value after dropout which is unlikely in many clinical trials. Recently, random-effects regression models have become popular for analysis of longitudinal clinical trial data with dropouts. However, inference obtained from random-effects regression models is valid when the missing-at-random dropout process is present. The random-effects pattern-mixture model, on the other hand provides an approach that is valid under more general missingness mechanisms. In this article we describe the use of random-effects pattern-mixture models under different patterns for dropouts. First, subjects are divided into groups depending on their missing-data patterns and then model parameters are estimated for each pattern. Finally, overall estimates are obtained by averaging over the missing-data patterns and corresponding standard errors are obtained using the delta method. A typical longitudinal clinical trial data set is used to illustrate and compare the above methods of data analyses in the presence of missing data due to dropouts.

During the early postnatal period, the human brain undergoes rapid and dynamic development. Over the past decades, there has been increased attention in studying the cognitive and cortical development of infants. However, accurate prediction of the infant cognitive and cortical development at an individual-level is a significant challenge, due to the huge complexities in highly irregular and incomplete longitudinal data that is commonly seen in current studies. Besides, joint prediction of cognitive scores and cortical morphology is barely investigated, despite some studies revealing the tight relationship between cognitive ability and cortical morphology and suggesting their potential mutual benefits. To tackle this challenge, we develop a flexible multi-task framework for joint prediction of cognitive scores and cortical morphological maps, namely, disentangled intensive triplet spherical adversarial autoencoder (DITSAA). First, we extract the mixed representative latent vector through a triplet spherical ResNet and further disentangles latent vector into identity-related and age-related features with an attention-based module. The identity recognition and age estimation tasks are introduced as supervision for a reliable disentanglement of the two components. Then we formulate the individualized cortical profile at a specific age by combining disentangled identity-related information and corresponding age-related information. Finally, an adversarial learning strategy is integrated to achieve a vivid and realistic prediction of cortical morphology, while a cognitive module is employed to predict cognitive scores. Extensive experiments are conducted on a public dataset, and the results affirm our method's ability to predict cognitive scores and cortical morphology jointly and flexibly using incomplete longitudinal data.

Missing data are frequently encountered in longitudinal clinical trials. To better monitor and understand the progress over time, one must handle the missing data appropriately and examine whether the missing data mechanism is ignorable or nonignorable. In this article, we develop a new probit model for longitudinal binary response data. It resolves a challenging issue for estimating the variance of the random effects, and substantially improves the convergence and mixing of the Gibbs sampling algorithm. We show that when improper uniform priors are specified for the regression coefficients of the joint multinomial model via a sequence of one-dimensional conditional distributions for the missing data indicators under nonignorable missingness, the joint posterior distribution is improper. A variation of Jeffreys prior is thus established as a remedy for the improper posterior distribution. In addition, an efficient Gibbs sampling algorithm is developed using a collapsing technique. Two model assessment criteria, the deviance information criterion (DIC) and the logarithm of the pseudomarginal likelihood (LPML), are used to guide the choices of prior specifications and to compare the models under different missing data mechanisms. We report on extensive simulations conducted to investigate the empirical performance of the proposed methods. The proposed methodology is further illustrated using data from an HIV prevention clinical trial.

Can selenium supplementation prevent colorectal cancer?

Cancer medicine continues to evolve into a more molecularly based discipline with many new treatments being developed aimed at specific cancer-related pathways and genetic alterations. In order to meet the challenges of the evolving science, new clinical trial designs and approaches for treatment trials are needed to evaluate the promise of these potential therapies. The National Cancer Institute (NCI) with the national NCI Clinical Trials Network (NCTN) Groups will launch two clinical research initiatives in lung cancer in 2014 that demonstrate innovative approaches to the design and conduct of clinical trials. The first initiative is the ALCHEMIST trial (Adjuvant Lung Cancer Enrichment Marker Identification and Sequencing Trial) which is aimed at patients with lung adenocarcinoma that has been entirely removed by surgery. In this study, samples of tumor tissue from thousands of patients will be tested for two specific genetic alterations, one in the EGFR gene, and the other in the ALK gene. These alterations are not common; about 10% of the lung cancer patient population in the US will have tumors with alterations in the EGFR gene and 5% will have alterations in the ALK gene. Rather than conduct a single trial for each alteration, the screening component of the ALCHEMIST trial will screen patients for both alterations at the same time. The ALCHEMIST is connected to 2 specific NCI-supported NCTN Group clinical treatment trials that are each testing a targeted agent specific for one of the alterations in conjunction with adjuvant therapy. Patients found to have tumors containing one of the alterations will be offered enrollment onto the appropriate clinical treatment trial. Every patient in the ALCHEMIST trial, including those without an EGFR or ALK alteration, will also be studied for cancer risk characteristics and their tumor tissue will be analyzed in a research genomics initiative conducted by the NCI Center for Cancer Genomics (CCG). This effort will employ next-generation sequencing technology to capitalize on the foundation of the CCG's earlier effort, the Cancer Genome Atlas. The second initiative, the Advanced Squamous Cell Carcinoma Lung Master Protocol, is a public-private effort made possible through collaboration with stakeholders including the NCI, NCTN Groups, FDA, NIH Foundation, the Friends of Cancer patient advocacy group, and the pharmaceutical industry. By pooling resources to screen tumor samples from about a thousand patients per year with advanced squamous cell carcinoma for multiple genetic alterations, this master protocol will enable several, randomized, phase II clinical trials with promising new agents to take place in parallel within a single study management framework. Agents will be evaluated rapidly and either be further evaluated in a phase III clinical trial or replaced by other new agents for phase II evaluation, without having to develop individual trials de novo each time. Compared with the usual process of developing clinical trials, these two new design approaches will hopefully allow more patients to be enrolled in trials with agents targeted for important molecular characteristics specific for their tumors and the trials will yield clinically useful results more rapidly. Citation Format: Margaret M. Mooney, Jack Welch, Jeffrey S. Abrams. Clinical trial design and master protocols in NCI clinical treatment trials. [abstract]. In: Proceedings of the AACR-IASLC Joint Conference on Molecular Origins of Lung Cancer; 2014 Jan 6-9; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2014;20(2Suppl):Abstract nr IA08.

Accurate prediction of patient outcomes in clinical trials is crucial for the timely assessment of treatment efficacy. This study proposes a novel approach to predict patient response using longitudinal clinical data. We construct temporal trajectories from longitudinal data and extrapolate these trajectories to forecast individual patient outcomes. Additionally, we assess when new patients align with established response patterns. The approach is evaluated using data from the MGTX trial involving patients with myasthenia gravis. Our analysis demonstrates the predictability of patient trajectories and enables automatic clustering of patients based on treatment success. The clustering reveals potential associations with age and smoking status. These findings highlight the potential of trajectory-based methods for early prediction of treatment response in clinical trials. We also discuss possible confounding factors that may influence the observed associations and predictive performance.

The main challenge in biomedical and clinical studies which involve collection of longitudinal data is the premature withdrawal of the subjects from the study resulting in incomplete data. Standard statistical analysis approaches usually give biased estimates of the model parameters if the mechanisms that led to dropout are ignored. In this discussion, we consider nonlinear mixed-effects models for multivariate longitudinal data in the presence of subject dropout. We present techniques for estimation of model parameters. These procedures are applied to estimate the parameters in the HIV dynamic system using routine observational data from an HIV clinic.

Web-delivered trials are an important component in eHealth services. These trials, mostly behavior-based, generate big heterogeneous data that are longitudinal, high dimensional with missing values. Unsupervised learning methods have been widely applied in this area, however, validating the optimal number of clusters has been challenging. Built upon our multiple imputation (MI) based fuzzy clustering, MIfuzzy, we proposed a new multiple imputation based validation (MIV) framework and corresponding MIV algorithms for clustering big longitudinal eHealth data with missing values, more generally for fuzzy-logic based clustering methods. Specifically, we detect the optimal number of clusters by auto-searching and -synthesizing a suite of MI-based validation methods and indices, including conventional (bootstrap or cross-validation based) and emerging (modularity-based) validation indices for general clustering methods as well as the specific one (Xie and Beni) for fuzzy clustering. The MIV performance was demonstrated on a big longitudinal dataset from a real web-delivered trial and using simulation. The results indicate MI-based Xie and Beni index for fuzzy-clustering are more appropriate for detecting the optimal number of clusters for such complex data. The MIV concept and algorithms could be easily adapted to different types of clustering that could process big incomplete longitudinal trial data in eHealth services.

The analysis of longitudinal data with non-ignorable missingness remains an active area in biostatistics research. This article discusses various random effects and latent process models which have been proposed for analyzing longitudinal binary data subject to both non-ignorable intermittent missing data and dropout. These models account for non-ignorable missingness by introducing random effects or a latent process which is shared between the response model and the model for the missing-data mechanism. We discuss various random effects and latent processes approaches and compare these approaches with analyses from an opiate clinical trial data set, which had high proportion of intermittent missingness and dropout. We also compare these random effect and latent process approaches with other methods for accounting for non-ignorable missingness using this data set.