Informative Dropout Research Articles

Modeling informative dropout in longitudinal data: A joint model approach

In medical studies, a disease’s progress is often monitored through indicators that signify the improved or worsened condition of the patient. These are known as longitudinal biomarkers, which are observed along with an event of importance. Together, they form the framework of joint modeling, which has a longitudinal process and an inherently associated time-to-event process. In clinical studies, the change in biomarkers is often monitored in the form of a change in the patient’s plasma level after a drug is administered to the patient. Again, in such studies, patients also withdraw from the trials prematurely or at a later phase, thus giving rise to dropouts. In most cases, this dropout is not random (Missing Not at Random). A joint model has been considered to incorporate this informative dropout in longitudinal response. To demonstrate this approach, a one-compartmental pharmacokinetic (PK) nonlinear mixed-effects model consisting of time-dependent parameters has been used in this work. The dropout mechanism has been introduced using a proportional hazard model. A Bayesian model framework is adopted to study the model’s performance through detailed simulation. A PK study on the drug Divalproex subject to an informative dropout model has been discussed. Journal of Statistical Research 2024, Vol. 58, No. 1, pp. 97-110.

ℒ2 gain tracking control of linear completely unknown discrete‐time networked control systems with dropout

AbstractWe introduce an online, model‐free algorithm to address the gain optimal tracking problems in discrete linear networked control systems. The algorithm is specifically proposed to handle stochastic information dropout in the feedback loop. Our goal is to design a control law that achieves the system output reference tracking while attenuating the effect of the disturbance input. We first construct an augmented system consisting of the original system and the command generator system. Then, the performance index is expressed in a quadratic form taking into account packet loss. Next, we obtain the optimal solution by solving the dropout generalized algebraic riccati equation (GARE). Finally, a ‐learning algorithm is utilized to estimate the control and disturbance feedback gains of the system, using only measurement data with unknown system dynamics in the presence of dropout. Two algorithms are tested on a numerical example to demonstrate the validity and effectiveness of the proposed methodology.

Retracted: A Joint Model for Unbalanced Nested Repeated Measures with Informative Drop-Out Applied to Ambulatory Blood Pressure Monitoring Data.

[This retracts the article DOI: 10.1155/2022/4452158.].

Self-guided disentangled representation learning for single image dehazing

Image dehazing has received extensive research attention as images collected in hazy weather are limited by low visibility and information dropout. Recently, disentangled representation learning has made excellent progress in various vision tasks. However, existing networks for low-level vision tasks lack efficient feature interaction and delivery mechanisms in the disentanglement process or an evaluation mechanism for the degree of decoupling in the reconstruction process, rendering direct application to image dehazing challenging. We propose a self-guided disentangled representation learning (SGDRL) algorithm with a self-guided disentangled network to realize multi-level progressive feature decoupling through sharing and interaction. The self-guided disentangled (SGD) network extracts image features using the multi-layer backbone network, and attribute features are weighted using the self-guided attention mechanism for the backbone features. In addition, we introduce a disentanglement-guided (DG) module to evaluate the degree of feature decomposition and guide the feature fusion process in the reconstruction stage. Accordingly, we develop SGDRL-based unsupervised and semi-supervised single image dehazing networks. Extensive experiments demonstrate the superiority of the proposed method for real-world image dehazing. The source code is available at https://github.com/dehazing/SGDRL.

Evaluating Edge Credibility in Evolving Noisy Social Networks

Despite the massive surge of evolving social network analysis in popularity, existing research usually represent the observed social interactions among individuals as completely credible edges. However, due to information inaccuracy, individual non-response and dropout, and sampling biases in observations, the evolving noisy social network that coexists true edges and spurious edges is pervasive in actual applications, where the ignoration of credibility otherness of observed edges could lead to the wrong estimates of social properties and misleading conclusions. To discover credible edge information to shape correct social interactions among individuals, we propose a universal and explainable multiple-neighbor evolutional filtering method ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MEFM</i> ) to evaluate how credible of observed edges to ‘truly’ exist in the evolving noisy social network. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MEFM</i> consists of an evolutional extractor and a filtering evaluator. To resist the noisy disturbance, the evolutional extractor exploits the evolutional states of edges from the perspective of evolution mechanisms within multiple-neighbor ranges, which applies different link prediction algorithms to fit the evolution mechanism in the formation of each edge. Further, the filtering evaluator reconstructs Kalman filter to predict and refine the evolutional states of edges based on their evolving local structures. As a result, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MEFM</i> combines the evolutional extractor and the filtering evaluator to analyze the evolutional fluctuations of the observed edges to evaluate their credibility. Extensive experiments on real-world datasets demonstrate that our proposed <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MEFM</i> can effectively and reasonably evaluate edge credibility in evolving noisy social networks.

An efficient doubly-robust imputation framework for longitudinal dropout, with an application to an Alzheimer’s clinical trial

We develop a novel doubly-robust (DR) imputation framework for longitudinal studies with monotone dropout, motivated by the informative dropout that is common in FDA-regulated trials for Alzheimer's disease. In this approach, the missing data are first imputed using a doubly-robust augmented inverse probability weighting (AIPW) estimator, then the imputed completed data are substituted into a full-data estimating equation, and the estimate is obtained using standard software. The imputed completed data may be inspected and compared to the observed data, and standard model diagnostics are available. The same imputed completed data can be used for several different estimands, such as subgroup analyses in a clinical trial, allowing for reduced computation and increased consistency across analyses. We present two specific DR imputation estimators, AIPW-I and AIPW-S, study their theoretical properties, and investigate their performance by simulation. AIPW-S has substantially reduced computational burden compared to many other DR estimators, at the cost of some loss of efficiency and the requirement of stronger assumptions. Simulation studies support the theoretical properties and good performance of the DR imputation framework. Importantly, we demonstrate their ability to address time-varying covariates, such as a time by treatment interaction. We illustrate using data from a large randomized Phase III trial investigating the effect of donepezil in Alzheimer's disease, from the Alzheimer's Disease Cooperative Study (ADCS) group.

A hidden Markov model for continuous longitudinal data with missing responses and dropout.

We propose a hidden Markov model for multivariate continuous longitudinal responses with covariates that accounts for three different types of missing pattern: (I) partially missing outcomes at a given time occasion, (II) completely missing outcomes at a given time occasion (intermittent pattern), and (III) dropout before the end of the period of observation (monotone pattern). The missing-at-random (MAR) assumption is formulated to deal with the first two types of missingness, while to account for the informative dropout, we rely on an extra absorbing state. Estimation of the model parameters is based on the maximum likelihood method that is implemented by an expectation-maximization (EM) algorithm relying on suitable recursions. The proposal is illustrated by a Monte Carlo simulation study and an application based on historical data on primary biliary cholangitis.

2b or not 2b? 2bRAD is an effective alternative to ddRAD for phylogenomics.

Restriction-site-associated DNA sequencing (RADseq) has become an accessible way to obtain genome-wide data in the form of single-nucleotide polymorphisms (SNPs) for phylogenetic inference. Nonetheless, how differences in RADseq methods influence phylogenetic estimation is poorly understood because most comparisons have largely relied on conceptual predictions rather than empirical tests. We examine how differences in ddRAD and 2bRAD data influence phylogenetic estimation in two non-model frog groups. We compare the impact of method choice on phylogenetic information, missing data, and allelic dropout, considering different sequencing depths. Given that researchers must balance input (funding, time) with output (amount and quality of data), we also provide comparisons of laboratory effort, computational time, monetary costs, and the repeatability of library preparation and sequencing. Both 2bRAD and ddRAD methods estimated well-supported trees, even at low sequencing depths, and had comparable amounts of missing data, patterns of allelic dropout, and phylogenetic signal. Compared to ddRAD, 2bRAD produced more repeatable datasets, had simpler laboratory protocols, and had an overall faster bioinformatics assembly. However, many fewer parsimony-informative sites per SNP were obtained from 2bRAD data when using native pipelines, highlighting a need for further investigation into the effects of each pipeline on resulting datasets. Our study underscores the importance of comparing RADseq methods, such as expected results and theoretical performance using empirical datasets, before undertaking costly experiments.

Quantization-based tracking control for fuzzy singularly perturbed Markov jump systems with incomplete transition information and packet dropout

Quantization-based tracking control for fuzzy singularly perturbed Markov jump systems with incomplete transition information and packet dropout

When a joint model should be preferred over a linear mixed model for analysis of longitudinal health-related quality of life data in cancer clinical trials

BackgroundPatient-reported outcomes such as health-related quality of life (HRQoL) are increasingly used as endpoints in randomized cancer clinical trials. However, the patients often drop out so that observation of the HRQoL longitudinal outcome ends prematurely, leading to monotone missing data. The patients may drop out for various reasons including occurrence of toxicities, disease progression, or may die. In case of informative dropout, the usual linear mixed model analysis will produce biased estimates. Unbiased estimates cannot be obtained unless the dropout is jointly modeled with the longitudinal outcome, for instance by using a joint model composed of a linear mixed (sub)model linked to a survival (sub)model. Our objective was to investigate in a clinical trial context the consequences of using the most frequently used linear mixed model, the random intercept and slope model, rather than its corresponding joint model.MethodsWe first illustrate and compare the models on data of patients with metastatic pancreatic cancer. We then perform a more formal comparison through a simulation study.ResultsFrom the application, we derived hypotheses on the situations in which biases arise and on their nature. Through the simulation study, we confirmed and complemented these hypotheses and provided general explanations of the bias mechanisms.ConclusionsIn particular, this article reveals how the linear mixed model fails in the typical situation where poor HRQoL is associated with an increased risk of dropout and the experimental treatment improves survival. Unlike the joint model, in this situation the linear mixed model will overestimate the HRQoL in both arms, but not equally, misestimating the difference between the HRQoL trajectories of the two arms to the disadvantage of the experimental arm.

Multimorbid Patterns and Cognitive Performance in the Presence of Informative Dropout Among Community-Dwelling Taiwanese Older Adults.

Longitudinal studies among older adults often feature elevated dropout rates and multiple chronic conditions. How Taiwanese multimorbid patterns relate to different cognitive domains remains unclear. This study aims to identify sex-specific multimorbid patterns and associate them with cognitive performance while modeling the risk for dropout. A prospective cohort study (2011-19) in Taiwan recruited 449 Taiwanese older adults without dementia. Global and domain-specific cognition were assessed biennially. We used exploratory factor analysis to identify baseline sex-specific multimorbid patterns of 19 self-reported chronic conditions. We utilized a joint model incorporating longitudinal and time-to-dropout data to examine the association between multimorbid patterns and cognitive performance accounting for the informative dropout via the shared random effect. At the end of the study, 324 participants (72.1%) remained in the cohort, with an average annual attrition rate of 5.5%. We found that advanced age, low levels of physical activities, and poor cognition at baseline were associated with increased dropout risks. Besides, 6 multimorbid patterns were identified, labeled Mental, Renal-vascular, and Cancer-urinary patterns in men, and Mental, Cardiometabolic, and Cancer-endocrine patterns in women. For men, as the follow-up time increased, the Mental pattern was associated with poor global cognition and attention; the Renal-vascular pattern was associated with poor executive function. For women, the Mental pattern was associated with poor memory; as follow-up time increased, and Cardiometabolic patterns were related to poor memory. Sex-specific multimorbid patterns identified in the Taiwanese older population showed differences (notably Renal-vascular pattern in men) from patterns found in Western countries and were differentially associated with cognitive impairment over time. When informative dropout is suspected, appropriate statistical methods should be applied.

Inner crossover fusion network with pixel-wise sampling for dynamic scene deblurring

Despite dynamic scene deblurring methods based on multi-scale or scale-recursive or hierarchical multi-patch strategy have made great progress, they still have some shortcomings. In the multi-scale or scale-recursive networks, bicubic down-sampling operation may cause some important high frequency information loss, e.g., strong edges. In addition, these networks are limited to using the same receptive field information while neglecting the cross-scale and cross-receptive field relationships inside the subnetwork, which may lead to information drop-out. To alleviate the above problems, we propose an inner crossover fusion network with pixel-wise sampling, called ICFNet, for dynamic scene deblurring. More specifically, we first design a pixel-wise sampling operation to replace bicubic down-sampling in the multi-scale or scale-recursive networks, so as to avoid the loss of the strong edge information. In addition, different from multi-scale and multi-patch architecture, we design a novel inner crossover fusion architecture to make full use of blur information with cross-scale and cross-receptive field in the network. To maximize the advantages of the ICFNet architecture, we design three region-enhanced residual blocks with different receptive fields for learning multiple receptive field features. Finally, extensive experimental results on different datasets have demonstrated that the proposed network can obtain better deblurring effect.

Effects of systematic data reduction on trend estimation from German registration trials

Key messageVCU trials can provide unbiased estimates of post-breeding trends given that all data is used. Dropping data of genotypes tested for up to two years may result in biased post-breeding trend estimates.Increasing yield trends are seen on-farm in Germany. The increase is based on genetic trend in registered genotypes and changes in agronomic practices and climate. To estimate both genetic and non-genetic trends, historical wheat data from variety trials evaluating a varieties’ value for cultivation und use (VCU) were analyzed. VCU datasets include information on varieties as well as on genotypes that were submitted by breeders and tested in trials but could not make it to registration. Therefore, the population of registered varieties (post-registration population) is a subset of the population of genotypes tested in VCU trials (post-breeding population). To assess post-registration genetic trend, historical VCU trial datasets are often reduced, e.g. to registered varieties only. This kind of drop-out mechanism is statistically informative which affects variance component estimates and which can affect trend estimates. To investigate the effect of this informative drop-out on trend estimates, a simulation study was conducted mimicking the structure of German winter wheat VCU trials. Zero post-breeding trends were simulated. Results showed unbiased estimates of post-breeding trends when using all data. When restricting data to genotypes tested for at least three years, a positive genetic trend of 0.11 dt ha−1 year−1 and a negative non-genetic trend (− 0.11 dt ha−1 year−1) were observed. Bias increased with increasing genotype-by-year variance and disappeared with random selection. We simulated single-trait selection, whereas decisions in VCU trials consider multiple traits, so selection intensity per trait is considerably lower. Hence, our results provide an upper bound for the bias expected in practice.

Inverse weighting method with jackknife variance estimator for differential expression analysis of single-cell RNA sequencing data

Single-cell RNA sequencing (scRNA-seq) data exhibit an unusual abundance of zero counts with a considerable fraction due to the dropout events, which introduces challenges to differential expression analysis. To correct biases in differential expression due to the informative dropouts, an inverse non-dropout-probability weighting method is proposed given that the dropout rate is negatively dependent on the underlying gene expression magnitude in scRNA-seq data. The weights are estimated using the maximum likelihood method where dropout values are integrated out using the Gauss-Hermite quadrature. Linear, generalized linear and mixed regressions with the estimated weights are fitted on original or transformed scRNA-seq data. Variances of coefficient estimators from the weighted regressions are estimated using the jackknife method. Extensive simulation studies are carried out to compare the proposed method to five cutting-edge methods (Limma, edgeR, MAST, ZIAQ and scImpute), where the proposed method performs among the best under all scenarios in terms of AUC, sensitivity, specificity and FDR. Rate of detecting true positives is examined for the proposed method and five comparison methods using mouse embryonic stem cells and fibroblasts where differentially expressed (DE) genes detected in bulk RNA-seq data on the same set of genes under the same conditions from independent source serve as true positives. Specificity is compared for these methods on true negative data by random splitting of a real dataset. Furthermore, the proposed method is illustrated on a lineage study where cells in the same embryo are correlated and genes differentially expressed between cell division lineages are identified.

Joint models for the longitudinal analysis of measurement scales in the presence of informative dropout

In health cohort studies, repeated measures of markers are often used to describe the natural history of a disease. Joint models allow to study their evolution by taking into account the possible informative dropout usually due to clinical events. However, joint modeling developments mostly focused on continuous Gaussian markers while, in an increasing number of studies, the actual quantity of interest is non-directly measurable; it constitutes a latent variable evaluated by a set of observed indicators from questionnaires or measurement scales. Classical examples include anxiety, fatigue, cognition. In this work, we explain how joint models can be extended to the framework of a latent quantity measured over time by indicators of different nature (e.g. continuous, binary, ordinal). The longitudinal submodel describes the evolution over time of the quantity of interest defined as a latent process in a structural mixed model, and links the latent process to each observation of the indicators through appropriate measurement models. Simultaneously, the risk of multi-cause event is modelled via a proportional cause-specific hazard model that includes a function of the mixed model elements as linear predictor to take into account the association between the latent process and the risk of event. Estimation, carried out in the maximum likelihood framework and implemented in the R-package JLPM, has been validated by simulations. The methodology is illustrated in the French cohort on Multiple-System Atrophy (MSA), a rare and fatal neurodegenerative disease, with the study of dysphagia progression over time stopped by the occurrence of death.

Accounting for drop-out using inverse probability censoring weights in longitudinal clustered data with informative cluster size

Periodontal disease is a serious gum infection impacting half of the U.S. adult population that may lead to loss of teeth. Using standard marginal models to study the association between patient-level predictors and tooth-level outcomes can lead to biased estimates because the independence assumption between the outcome (periodontal disease) and cluster size (number of teeth per patient) is violated. Specifically, the baseline number of teeth of a patient is informative. In this setting a cluster-weighted generalized estimating equations (CWGEE) approach can be used to obtain unbiased marginal inference from data with informative cluster size (ICS). However, in many longitudinal studies of dental health, including the Veterans Affairs Dental Longitudinal Study, the rate of tooth-loss or tooth drop-out over time is also informative, creating a missing at random data mechanism. Here, we propose a novel modeling approach that incorporates the technique of inverse probability censoring weights into CWGEE with binary outcomes to account for ICS and informative drop-out over time. In an extensive simulation study we demonstrate that results obtained from our proposed method yield lower bias and excellent coverage probability, compared to those obtained from traditional methods which do not account for ICS or drop-out.

A Joint Model for Unbalanced Nested Repeated Measures with Informative Drop-Out Applied to Ambulatory Blood Pressure Monitoring Data.

This study proposes a Bayesian joint model with extended random effects structure that incorporates nested repeated measures and provides simultaneous inference on treatment effects over time and drop-out patterns. The proposed model includes flexible splines to characterize the circadian variation inherent in blood pressure sequences, and we assess the effectiveness of an intervention to resolve pediatric obstructive sleep apnea. We demonstrate that the proposed model and its conventional two-stage counterpart provide similar estimates of nighttime blood pressure but estimates on the mean evolution of daytime blood pressure are discrepant. Our simulation studies tailored to the motivating data suggest reasonable estimation and coverage probabilities for both fixed and random effects. Computational challenges of model implementation are discussed.

Joint modelling with competing risks of dropout for longitudinal analysis of health-related quality of life in cancer clinical trials.

Health-related quality of life (HRQoL) is an important endpoint in cancer clinical trials. Analysis of HRQoL longitudinal data is plagued by missing data, notably due to dropout. Joint models are increasingly receiving attention for modelling longitudinal outcomes and the time-to-dropout. However, dropout can be informative or non-informative depending on the cause. We propose using a joint model that includes a competing risks sub-model for the cause-specific time-to-dropout. We compared a competing risks joint model (CR JM) that distinguishes between two causes of dropout with a standard joint model (SJM) that treats all the dropouts equally. First, we applied the CR JM and SJM to data from 267 patients with advanced oesophageal cancer from the randomized clinical trial PRODIGE 5/ACCORD 17 to analyse HRQoL data in the presence of dropouts unrelated and related to a clinical event. Then, we compared the models using a simulation study. We showed that the CR JM performed as well as the SJM in situations where the risk of dropout was the same whatever the cause. In the presence of both informative and non-informative dropouts, only the SJM estimations were biased, impacting the HRQoL estimated parameters. The systematic collection of the reasons for dropout in clinical trials would facilitate the use of CR JMs, which could be a satisfactory approach to analysing HRQoL data in presence of both informative and non-informative dropout. This study is registered with ClinicalTrials.gov, number NCT00861094.

Phylogenomics and loci dropout patterns of deeply diverged Zodarion ant-eating spiders suggest a high potential of RAD-seq for genus-level spider phylogenetics.

RAD sequencing yields large amounts of genome-wide data at a relatively low cost and without requiring previous taxon-specific information, making it ideal for evolutionary studies of highly diversified and neglected organisms. However, concerns about information decay with phylogenetic distance have discouraged its use for assessing supraspecific relationships. Here, using Double Digest Restriction Associated DNA (ddRAD) data, we perform the first deep-level approach to the phylogeny of Zodarion, a highly diversified spider genus. We explore the impact of loci and taxon filtering across concatenated and multispecies coalescent reconstruction methods and investigate the patterns of information dropout in reference to both the time of divergence and the mitochondrial divergence between taxa. We found that relaxed loci-filtering and nested taxon-filtering strategies maximized the amount of molecular information and improved phylogenetic inference. As expected, there was a clear pattern of allele dropout towards deeper time and mitochondrial divergences, but the phylogenetic signal remained strong throughout the phylogeny. Therefore, we inferred topologies that were almost fully resolved, highly supported, and noticeably congruent between setups and inference methods, which highlights overall inconsistency in the taxonomy of Zodarion. Because Zodarion appears to be among the oldest and most mitochondrially diversified spider genera, our results suggest that ddRAD data show high potential for inferring intra-generic relationships across spiders and probably also in other taxonomic groups.

Trajectories of Structural Disease Progression in Knee Osteoarthritis.

Knee osteoarthritis (OA) is a heterogeneous disease, with most patients experiencing slow disease progression and some with rapid deterioration. We aimed to identify groups of patients with symptomatic knee OA experiencing rapid structural progression. We selected participants from the Osteoarthritis Initiative with baseline Kellgren/Lawrence (K/L) grades 1-3 and knee pain, and with joint space width (JSW) on fixed-flexion knee radiographs assessed at baseline and with ≥1 follow-up over 8 years. We used latent class growth analysis to identify subgroups of JSW progression, jointly modeling time to knee replacement (KR) to account for potential informative dropouts. After identifying trajectories, we used logistic regression to assess the association between baseline characteristics and the JSW trajectory group. We used data from 1,578 participants. Baseline radiographic severity was K/L grade 1 in 17%, K/L grade 2 in 50%, and K/L grade 3 in 33%. We identified 3 distinct JSW trajectories: 86% stable, 6% with stable JSW followed by late progression, and 8% with early progression. Incorporating information about KR resulted in 47% of KRs initially classified as stable being reclassified to 1 of the progressing trajectories. Prior knee surgery was associated with being in the late-progressing versus the stable trajectory, while obesity was associated with being in the early-progressing versus stable trajectory. In addition to a subgroup of individuals experiencing early structural progression, 8-year longitudinal data allowed the identification of a late-progressing trajectory. Incorporating information about KR was important to properly identify longitudinal structural trajectories in knee OA.