Search

In this paper, we aim to show the process of simulating survival data when the distribution of the overall population and one subgroup (called "positive subgroup") as well as the proportion of the subgroup is known, while the distribution of the other subgroup (called "negative subgroup") is unknown. We propose a combination method which generates survival data of the positive subgroup and negative subgroup, respectively, and survival data of the overall population are the combination of the two subgroups. The parameters of the overall population and the positive subgroup need to satisfy certain constraints, otherwise the parameters may lead to contradictions. From simulation, we show that our proposed combination method can reflect the correlation between the test statistics of overall population and positive subgroup, which makes the simulated data more realistic and the results of simulation more reliable. Moreover, for a multiplicity control in trial design, the combination method can help to determine the α splitting strategy between primary endpoints, and is helpful in designs of clinical trials as shown in three applications.

The delayed treatment effect, which manifests as a separation of survival curves after a change point, has often been observed in immunotherapy clinical trials. A late effect of this kind may violate the proportional hazards assumption, resulting in the non-negligible loss of statistical power of an ordinary log-rank test when comparing survival curves. The Fleming-Harrington (FH) test, a weighted log-rank test, is configured to mitigate the loss of power by incorporating a weight function with two parameters, one each for early and late treatment effects. The two parameters need to be appropriately determined, but no helpful guides have been fully established. Since the late effect is expected in immunotherapy trials, we focus on the late effect parameter in this study. We consider parameterizing the late effect in a readily interpretable fashion and determining the optimal late effect parameter in the FH test to maintain statistical power in reference to the asymptotic relative efficiency (ARE). The optimization is carried out under three lag models (i.e. linear, threshold, and generalized linear lag), where the optimal weights are proportional to the lag functions characterized by the change points. Extensive simulation studies showed that the FH test with the selected late parameter reliably provided sufficient power even when the change points in the lag models were misspecified. This finding suggests that the FH test with the ARE-guided late parameter may be a reasonable and practical choice for the primary analysis in immunotherapy clinical trials.

Inverse probability weighting (IPW) is frequently used to reduce or minimize the observed confounding in observational studies. IPW creates a pseudo-sample by weighting each individual by the inverse of the conditional probability of receiving the treatment level that he/she has actually received. In the pseudo-sample there is no variation among the multiple individuals generated by weighting the same individual in the original sample. This would reduce the variability of the data and therefore bias the variance estimate in the target population. Conventional variance estimation methods for IPW estimators generally ignore this underestimation and tend to produce biased estimates of variance. We here propose a more reasonable method that incorporates this source of variability by using parametric bootstrapping based on intra-stratum variability estimates. This approach firstly uses propensity score stratification and intra-stratum standard deviation to approximate the variability among multiple individuals generated based on a single individual whose propensity score falls within the corresponding stratum. The parametric bootstrapping is then used to incorporate the target variability by re-generating outcomes after adding a random error term to the original data. The performance of the proposed method is compared with three existing methods including the naïve model-based variance estimator, the nonparametric bootstrap variance estimator, and the robust variance estimator in the simulation section. An example of patients with sarcopenia is used to illustrate the implementation of the proposed approach. According to the results, the proposed approach has desirable statistical properties and can be easily implemented using the provided R code.

The failure rates of phase 3 trials are high. Incorrect sample size due to uncertainty of effect size could be a critical contributing factor. Adaptive sequential design (ASD), which may include one or more sample size re-estimations (SSR), has been a popular approach for dealing with such uncertainties. The operating characteristics (OCs) of ASD, including the unconditional power and mean sample size, can be substantially affected by many factors, including the planned sample size, the interim analysis schedule and choice of critical boundaries and rules for interim analysis. We propose a systematic, comprehensive strategy which uses iterative simulations to investigate the operating characteristics of adaptive designs and help achieve adequate unconditional power and cost-effective mean sample size if the effect size is in a pre-identified range.

Platform trials offer a framework to study multiple interventions in one trial with the opportunity of opening and closing arms. The use of common controls can increase efficiency as compared to individual controls. The need for multiplicity adjustment because of common controls is currently a debate among researchers, pharmaceutical companies, and regulators. The impact of common controls on the type one error in a fixed platform trial, i.e. when all treatments start and end recruitment at the same time, has been discussed in the literature before. We complement these findings by investigating the impact of a common control on the type one error and power in a flexible platform trial, i.e. when one arm joins the platform later. We derived the correlation of test statistics to assess the impact of the overlap and compared the results to a trial with individual controls. Furthermore, we evaluate the power, and the impact of multiplicity adjustment on the power in fixed and flexible platform trials. These methodological considerations are complemented by a regulatory guideline review. With multiple arms, the FWER is inflated when no multiplicity adjustment is applied. However, the FWER inflation is smaller with common controls than with individual controls. Even after multiplicity adjustment, a trial with common controls is often beneficial in terms of sample size and power. However, in some cases, the trial with common controls loses the efficiency gain and it might be advisable to run a separate trial rather than joining a platform trial.

This paper discusses the problem of disease prevalence in clinical studies, focusing on multiple comparisons based on stratified partially validated series in the presence of a gold standard. Five test statistics, including two Wald-type test statistics, the inverse hyperbolic tangent transformation test statistic, likelihood ratio test statistic, and score test statistic, are proposed to conduct multiple comparisons. To control the overall type I error rate, several adjustment procedures are developed, namely the Bonferroni, Single-step adjusted MaxT, Single-step adjusted MinP, Holm's Step-down, and Hochberg's step-up procedures, based on these test statistics. The performance of the proposed methods is evaluated through simulation studies in terms of the empirical type I error rate and empirical power. Simulation results show that the Single-step adjusted MaxT procedure and Single-step adjusted MinP procedure generally outperform the other three procedures, and these two test procedures based on all test statistics have satisfactory performance. Notably, the Single-step adjusted MinP procedure tends to exhibit higher empirical power than the Single-step adjusted MaxT procedure. Furthermore, the Step-down and Step-up procedures show greater power compared to the Bonferroni method. The study also observes that as the validated ratio increases, the empirical type I errors of all test procedures approach the nominal level while maintaining higher power. Two real examples are presented to illustrate the proposed methods.

There are good reasons to perform a randomized controlled trial (RCT) even in early phases of clinical development. However, the low sample sizes in those settings lead to high variability of the treatment effect estimate. The variability could be reduced by adding external control data if available. For the common setting of suitable subject-level control group data only available from one external (clinical trial or real-world) data source, we evaluate different analysis options for estimating the treatment effect via hazard ratios. The impact of the external control data is usually guided by the level of similarity with the current RCT data. Such level of similarity can be determined via outcome and/or baseline covariate data comparisons. We provide an overview over existing methods, propose a novel option for a combined assessment of outcome and baseline data, and compare a selected set of approaches in a simulation study under varying assumptions regarding observable and unobservable confounder distributions using a time-to-event model. Our various simulation scenarios also reflect the differences between external clinical trial and real-world data. Data combinations via simple outcome-based borrowing or simple propensity score weighting with baseline covariate data are not recommended. Analysis options which conflate outcome and baseline covariate data perform best in our simulation study.

The Multiple Comparison Procedure and Modelling (MCPMod) approach has been shown to be a powerful statistical technique that can significantly improve the design and analysis of dose-finding studies under model uncertainty. Due to its frequentist nature, however, it is difficult to incorporate information into MCPMod from historical trials on the same drug. BMCPMod, a recently introduced Bayesian version of MCPMod, is designed to take into account historical information on the placebo dose group. We introduce a Bayesian hierarchical framework capable of incorporating historical information on an arbitrary number of dose groups, including both placebo and active ones, taking into account the relationship between responses of these dose groups. Our approach can also model both prognostic and predictive between-trial heterogeneity and is particularly useful in situations where the effect sizes of two trials are different. Our goal is to reduce the necessary sample size in the dose-finding trial while maintaining its target power.

Multi-arm trials are increasingly of interest because for many diseases; there are multiple experimental treatments available for testing efficacy. Several novel multi-arm multi-stage (MAMS) clinical trial designs have been proposed. However, a major hurdle to adopting the group sequential MAMS routinely is the computational effort of obtaining stopping boundaries. For example, the method of Jaki and Magirr for time-to-event endpoint, implemented in R package MAMS , requires complicated computational efforts to obtain stopping boundaries. In this study, we develop a group sequential MAMS survival trial design based on the sequential conditional probability ratio test. The proposed method is an improvement of the Jaki and Magirr's method in the following three directions. First, the proposed method provides explicit solutions for both futility and efficacy boundaries to an arbitrary number of stages and arms. Thus, it avoids complicated computational efforts for the trial design. Second, the proposed method provides an accurate number of events for the fixed sample and group sequential designs. Third, the proposed method uses a new procedure for interim analysis which preserves the study power.

We propose a new approach to select the regularization parameter using a new version of the generalized information criterion ( GIC ) in the subject of penalized regression. We prove the identifiability of bridge regression model as a prerequisite of statistical modeling. Then, we propose asymptotically efficient generalized information criterion ( AGIC ) and prove that it has asymptotic loss efficiency. Also, we verified the better performance of AGIC in comparison to the older versions of GIC . Furthermore, we propose MSE search paths to order the selected features by lasso regression based on numerical studies. The MSE search paths provide a way to cover the lack of feature ordering in lasso regression model. The performance of AGIC with other types of GIC is compared using MSE and model utility in simulation study. We exert AGIC and other criteria to analyze breast and prostate cancer and Parkinson disease datasets. The results confirm the superiority of AGIC in almost all situations.

In practice, the receiver operating characteristic (ROC) curve of a diagnostic test is widely used to show the performance of the test for discriminating two-class events. The area under the ROC curve (AUC) is proposed as an index for the assessment of the diagnostic accuracy of the test under consideration. Due to ethical and cost considerations associated with application of gold standard (GS) tests, only a subset of the patients initially tested have verified disease status. Statistical evaluation of the test performance based only on test results from subjects with verified disease status are typically biased. Various AUC estimation methods for tests with verification biased data have been developed over the last few decades. In this article, we develop new direct estimation methods for the volume under the ROC surface (VUS) by extending the AUC estimation methods for two-class diagnostic tests to three-class diagnostic tests in the presence of verification bias. The proposed methods will provide a comprehensive guide to deal with the verification bias in three-class diagnostic test accuracy studies and lead to a better choice of diagnostic tests.

Sequential, multiple assignment, randomized trial (SMART) designs are appropriate for comparing adaptive treatment interventions, in which intermediate outcomes (called tailoring variables) guide subsequent treatment decisions for individual patients. Within a SMART design, patients may be re-randomized to subsequent treatments following the outcomes of their intermediate assessments. In this paper, we provide an overview of statistical considerations necessary to design and implement a two-stage SMART design with a binary tailoring variable and a survival final endpoint. A chronic lymphocytic leukemia trial with a final endpoint of progression-free survival is used as an example for the simulations to assess how design parameters, including, choice of randomization ratios for each stage of randomization, and response rates of the tailoring variable affect the statistical power. We assess the choice of weights from restricted re-randomization on data analyses and appropriate hazard rate assumptions. Specifically, for a given first-stage therapy and prior to the tailoring variable assessment, we assume equal hazard rates for all patients randomized to a treatment arm. After the tailoring variable assessment, individual hazard rates are assumed for each intervention path. Simulation studies demonstrate that the response rate of the binary tailoring variable impacts power as it directly impacts the distribution of patients. We also confirm that when the first stage randomization is 1:1, it is not necessary to consider the first stage randomization ratio when applying the weights. We provide an R-shiny application for obtaining power for a given sample size for SMART designs.

In recent years, combined therapy shows expected treatment effect as they increase dose intensity, work on multiple targets and benefit more patients for antitumor treatment. However, dose -finding designs for combined therapy face a number of challenges. Therefore, under the framework of phase I-II, we propose a two-stage dose -finding design to identify the biologically optimal dose combination (BODC), defined as the one with the maximum posterior mean utility under acceptable safety. We model the probabilities of toxicity and efficacy by using linear logistic regression models and conduct Bayesian model selection (BMS) procedure to define the most likely pattern of dose-response surface. The BMS can adaptively select the most suitable model during the trial, making the results robust. We investigated the operating characteristics of the proposed design through simulation studies under various practical scenarios and showed that the proposed design is robust and performed well.

In recent years, adaptive randomization methods have gained significant popularity in clinical research and trial design due to their ability to provide both efficiency and flexibility in adjusting the statistical procedures of ongoing clinical trials. For a study to compare multiple treatments, a multi-arm two-stage design could be utilized to select the best treatment from the first stage and further compare that treatment with control in the second stage. The traditional design used equal randomization in both stages. To better utilize the interim results from the first stage, we propose to develop response adaptive randomization two-stage designs for a multi-arm clinical trial with binary outcome. Two allocation methods are considered: (1) an optimal allocation based on a sequential design; (2) the play-the-winner rule. Optimal multi-arm two-stage designs are obtained under three criteria: minimizing the expected number of failures, minimizing the average expected sample size, and minimizing the expected sample size under the null hypothesis. Simulation studies show that the proposed adaptive design based on the play-the-winner rule has good performance. A phase II trial for patients with pancreas adenocarcinoma and a germline BRCA / PALB2 mutation was used to illustrate the application of the proposed response adaptive randomization designs.

This manuscript consists of two topics. Firstly, we explore the utility of internal pilot study (IPS) approach for reestimating sample size at an interim stage when a reliable estimate of the nuisance shape parameter of the Weibull distribution for modeling survival data is unavailable during the planning phase of a study. Although IPS approach can help rescue the study power, it is noted that the adjusted sample size can be as much as twice the initially planned sample size, which may put substantial practical constraints to continue the study. Secondly, we discuss Bayesian predictive probability for conducting interim analyses to obtain preliminary evidence of efficacy or futility of an experimental treatment warranting early termination of a clinical trial. In the context of single-arm clinical trials with time-to-event endpoints following Weibull distribution, we present the calculation of the Bayesian predictive probability when the shape parameter of the Weibull distribution is unknown. Based on the data accumulated at the interim, we propose two approaches which rely on the posterior mode or the entire posterior distribution of the shape parameter. To account for uncertainty in the shape parameter, it is recommended to incorporate its entire posterior distribution in our calculation.

Clinical trials powered to detect subgroup effects provide the most reliable data on heterogeneity of treatment effect among different subpopulations. However, pre-specified subgroup analysis is not always practical and post hoc analysis results should be examined cautiously. Bayesian hierarchical modelling provides grounds for defining a controlled post hoc analysis plan that is developed after seeing outcome data for the population but before unblinding the outcome by subgroup. Using simulation based on the results from a tobacco cessation clinical trial conducted among the general population, we defined an analysis plan to assess treatment effect among American Indians and Alaska Natives (AI/AN) enrolled in the study. Patients were randomized into two arms using Bayesian adaptive design. For the opt-in arm, clinicians offered a cessation treatment plan after verifying that a patient was ready to quit. For the opt-out arm, clinicians provided all participants with free cessation medications and referred them to a Quitline. The study was powered to test a hypothesis of significantly higher quit rates for the opt-out arm at one-month post randomization. Overall, one-month abstinence rates were 15.9% and 21.5% (opt-in and opt-out arm, respectively). For AI/AN, one-month abstinence rates were 10.2% and 22.0% (opt-in and opt-out arm, respectively). The posterior probability that the abstinence rate in the treatment arm is higher is 0.96, indicating that AI/AN demonstrate response to treatment at almost the same probability as the whole population.

Selecting a safe and clinically beneficial dose can be difficult in drug development. Dose justification often relies on dose-response modeling where parametric assumptions are made in advance which may not adequately fit the data. This is especially problematic in longitudinal dose-response models, where additional parametric assumptions must be made. This paper proposes a class of longitudinal dose-response models to be used in the Bayesian model averaging paradigm which improve trial operating characteristics while maintaining flexibility a priori . A new longitudinal model for non-monotonic longitudinal profiles is proposed. The benefits and trade-offs of the proposed approach are demonstrated through a case study and simulation.

Bayesian predictive probabilities have become a ubiquitous tool for design and monitoring of clinical trials. The typical procedure is to average predictive probabilities over the prior or posterior distributions. In this paper, we highlight the limitations of relying solely on averaging, and propose the reporting of intervals or quantiles for the predictive probabilities. These intervals formalize the intuition that uncertainty decreases with more information. We present four different applications (Phase 1 dose escalation, early stopping for futility, sample size re-estimation, and assurance/probability of success) to demonstrate the practicality and generality of the proposed approach.

Statistical methods have been well developed for comparing the predictive values of two binary diagnostic tests under a paired design. However, existing methods do not make allowance for incomplete data. Although maximum likelihood based method can be used to deal with incomplete data, it requires iterative algorithm for implementation. A simple and easily implemented statistical method is therefore needed. Simple methods exist for comparing two sensitivities or specificities with incomplete data but such simple methods are not available for comparing two predictive values with incomplete data. In this paper, we propose two simple methods for comparing two predictive values with incomplete data. The test statistics derived by these two methods are simple to compute, only involving some minor modification of the existing weighted generalized score statistics with complete data. Simulation results demonstrate that the proposed methods are more efficient than the ad-hoc method that only uses the subjects wit complete data. As an illustration, the proposed methods are applied to an observational study comparing two non-invasive methods in detecting endometriosis.

We propose a simple approach to assess whether a nonlinear parametric model is appropriate to depict the dose-response relationships and whether two parametric models can be applied to fit a dataset via nonparametric regression. The proposed approach can compensate for the ANOVA, which is sometimes conservative, and is very easy to implement. We illustrate the performance by analyzing experimental examples and a small simulation study.

Conducting a well-powered and adequately controlled clinical trial in children is often challenging. Bayesian approaches are an attractive option for addressing such challenges as they provide a quantitatively rigorous and integrated framework that makes use of current control data to check and borrow information from historical control data. However various practical concerns and related statistical issues emerge when implementing such Bayesian borrowing approaches. In this manuscript we use a motivating case study to discuss a rigorous stepwise approach on how to address those issues within the Bayesian framework. Specifically, a comprehensive quantitative framework is proposed to assess the extent, synergy, and impact of borrowing. Steps on computing the measures to interpret borrowing are illustrated. Those measures can further help to determine whether additional discounting of external information is necessary.

The use of Bayesian methodology to design and analyze pediatric efficacy trials is one of the possible options to reduce their sample size. This reduction of the sample size results from the use of an informative prior for the parameters of interest. In most of the applications, the principle of 'information borrowing' from adults' trials is applied, which means that the informative prior is constructed using efficacy results in adult of the drug under investigation. This implicitly assumes similarity in efficacy between the selected pediatric dose and the efficacious dose in adults. The goal of this article is to propose a method to construct prior distribution for the parameter of interest, not directly constructed from the efficacy results of the efficacious dose in adult patients but using pharmacodynamic modeling of a bridging biomarker using early phase pediatric data. When combined with a model bridging the biomarker with the clinical endpoints, the prior is constructed using a variational method after simulation of the parameters of interest. A use case application illustrates how the method can be used to construct a realistic informative prior.

The main purpose of this paper is to survey the statistical inference for covariate-specific time-dependent receiver operating characteristic (ROC) curves with nonignorable missing continuous biomarker values. To construct time-dependent ROC curves, we consider a joint model which assumes that the failure time depends on the continuous biomarker and the covariates through a Cox proportional hazards model and that the continuous biomarker depends on the covariates through a semiparametric location model. Assuming a purely parametric model on the propensity score, we utilize instrumental variables to deal with the identifiable issue and estimate the unknown parameters of the propensity score by a simple and efficient method. In addition, when the propensity score is estimated, we develop HT and AIPW approaches to estimate our interested quantities. In the presence of nonignorable missing biomarker, our AIPW estimators of the interested quantities are still doubly robust when the true propensity score is a special parametric logistic model. At last, simulation studies are conducted to assess the performance of our proposed approaches, and a real data analysis is also carried out to illustrate its application.

Clinical studies sometimes provide clustered data with censored failure times. A crucial factor of the randomized design that lessens selection bias is the random allocation rule. Given this, the weighted rank tests' p-values for stratified survival clustered sampling based on the random allocation rule are approximated using the double saddle-point approximation technique. For tests of significance and confidence intervals for the treatment effect, this approximation can be utilized. Through simulation experiments, the accuracy of the saddle-point approximation is examined by comparing saddle-point and normal approximations to the exact underlying permutation distribution.

Estimands play an important role for aligning study objectives, study design and analyses through a precise definition of the quantity of interest. For COVID-19 studies, apart from intercurrent events, high volume of missing data has been observed. We explore their impact on several estimands through a synthetic COVID-19 data generated from a discrete-time multi-state model. We compare estimators of these estimands based on their ability to closely match the true response rates and retain assumed power. The final choice of the estimand then needs to be aligned with clinically meaningful quantities of interest to patients, clinicians, regulators and payers.

Use of historical data has become a hot topic recently, considered to provide a way to reduce patient burden, lower drug development cost, and make innovative therapies available to patients earlier. In a single-arm study designed to examine the benefit of an experimental treatment, there is often a desire to compare the outcomes of patients receiving the new intervention with those receiving a control treatment, which can be extracted from sources such as historical trials or electronic medical records. Since the treatment is not randomly assigned, there is a need to adjust for the potential imbalance in key patient characteristics between the current study and historical controls. If the outcome of interest is measured longitudinally and subject to random missing, the required adjustment becomes more complicated. In this paper, we propose a doubly robust adjustment procedure specifically designed for longitudinal data analysis with missing data. The proposed method yields valid analysis results, if either the propensity score model or the mixed effects model for repeated measures (MMRM) regression model is correctly specified. An extensive numerical study is conducted to examine the performance of the proposed method. Data from a real clinical trial comparing with historical data are analyzed as an example applying the proposed procedure.

For ordered categorical data from randomized clinical trials, the relative effect, the probability that observations in one group tend to be larger, has been considered appropriate for a measure of an effect size. Although the Wilcoxon-Mann-Whitney test is widely used to compare two groups, the null hypothesis is not just the relative effect of 50%, but the identical distribution between groups. The null hypothesis of the Brunner-Munzel test, another rank-based method used for arbitrary types of data, is just the relative effect of 50%. In this study, we compared actual type I error rates (or 1 - coverage probability) of the profile-likelihood-based confidence intervals for the relative effect and other rank-based methods in simulation studies at the relative effect of 50%. The profile-likelihood method, as with the Brunner- Munzel test, does not require any assumptions on distributions. Actual type I error rates of the profile-likelihood method and the Brunner-Munzel test were close to the nominal level in large or medium samples, even under unequal distributions. Those of the Wilcoxon-Mann-Whitney test largely differed from the nominal level under unequal distributions, especially under unequal sample sizes. In small samples, the actual type I error rates of Brunner-Munzel test were slightly larger than the nominal level and those of the profile-likelihood method were even larger. We provide a paradoxical numerical example: only the Wilcoxon-Mann-Whitney test was significant under equal sample sizes, but by changing only the allocation ratio, it was not significant but the profile-likelihood method and the Brunner-Munzel test were significant. This phenomenon might reflect the nature of the Wilcoxon-Mann-Whitney test in the simulation study, that is, the actual type I error rates become over and under the nominal level depending on the allocation ratio.

Generalized pairwise comparisons and win statistics (i.e., win ratio, win odds and net benefit) are advantageous in analyzing and interpreting a composite of multiple outcomes in clinical trials. An important limitation of these statistics is their inability to adjust for covariates other than by stratified analysis. Because the win ratio does not account for ties, the win odds, a modification that includes ties, has attracted attention. We review and combine information on the win odds to articulate the statistical inferences for the win odds. We also show alternative variance estimators based on the exact permutation and bootstrap as well as statistical inference via the probabilistic index. Finally, we extend multiple-covariate regression probabilistic index models to the win odds with a univariate outcome. As an illustration we apply the regression models to the data in the CHARM trial.