Satrajit Roychoudhury

Robust exchangeability designs for early phase clinical trials with multiple strata.

Clinical trials with multiple strata are increasingly used in drug development. They may sometimes be the only option to study a new treatment, for example in small populations and rare diseases. In early phase trials, where data are often sparse, good statistical inference and subsequent decision-making can be challenging. Inferences from simple pooling or stratification are known to be inferior to hierarchical modeling methods, which build on exchangeable strata parameters and allow borrowing information across strata. However, the standard exchangeability (EX) assumption bears the risk of too much shrinkage and excessive borrowing for extreme strata. We propose the exchangeability-nonexchangeability (EXNEX) approach as a robust mixture extension of the standard EX approach. It allows each stratum-specific parameter to be exchangeable with other similar strata parameters or nonexchangeable with any of them. While EXNEX computations can be performed easily with standard Bayesian software, model specifications and prior distributions are more demanding and require a good understanding of the context. Two case studies from phases I and II (with three and four strata) show promising results for EXNEX. Data scenarios reveal tempered degrees of borrowing for extreme strata, and frequentist operating characteristics perform well for estimation (bias, mean-squared error) and testing (less type-I error inflation).

On the Use of Co-Data in Clinical Trials

ABSTRACT Historical data are important for the design of a clinical trial. Yet these data are rarely used in the analysis of the actual trial. While justifiable in certain situations, ignoring historical data can lead to less accurate inferences, and, therefore, suboptimal decisions. After a review of the main approaches to using historical data, the framework is extended to co-data, which comprise all relevant (historical and concurrent) trial-external data. These data can be used for the inference of the parameter in the actual trial via meta-analytic approaches. While the use of co-data in clinical trials is attractive, it is also ambitious. For example, avoiding undue weight of co-data (relative to actual trial data) is important, which can often be achieved by plausible assumptions about between-trial heterogeneity and allowance for nonexchangeability across trial parameters. Two applications with co-data will be discussed: phase III trials with interim decisions informed by co-data; and, a phase I combination trial in Oncology, which takes advantage of co-data from completed and ongoing phase I trials.

Designing and analysing clinical trials in mental health: an evidence synthesis approach

Objective When planning a clinical study, evidence on the treatment effect is often available from previous studies. However, this evidence is mostly ignored for the analysis of the new study. This is unfortunate, since using it could lead to a smaller study without compromising power. We describe a design that addresses this issue. Methods We use a Bayesian meta-analytic model to incorporate the available evidence in the analysis of the new study. The shrinkage estimate for the new study integrates the evidence from the other studies. At the planning phase of the study, it allows a statistically justified reduction of the sample size. Results The design is illustrated using data from an Food and Drug Administration (FDA) review of lurasidone for the treatment of schizophrenia. Three studies inform the meta-analysis before the new study is conducted. Results from an additional phase III study, which were not available at the time of the FDA review, are then used for the actual analysis. Conclusions In the presence of reliable and relevant evidence, the design offers a way to conduct a smaller study without compromising power. It therefore fills a gap between the assessment of evidence and its actual use in the design and analysis of studies.

Statistical analysis of data from dilution assays with censored correlated counts

Frequently, count data obtained from dilution assays are subject to an upper detection limit, and as such, data obtained from these assays are usually censored. Also, counts from the same subject at different dilution levels are correlated. Ignoring the censoring and the correlation may provide unreliable and misleading results. Therefore, any meaningful data modeling requires that the censoring and the correlation be simultaneously addressed. Such comprehensive approaches of modeling censoring and correlation are not widely used in the analysis of dilution assays data. Traditionally, these data are analyzed using a general linear model on a logarithmic-transformed average count per subject. However, this traditional approach ignores the between-subject variability and risks, providing inconsistent results and unreliable conclusions. In this paper, we propose the use of a censored negative binomial model with normal random effects to analyze such data. This model addresses, in addition to the censoring and the correlation, any overdispersion that may be present in count data. The model is shown to be widely accessible through the use of several modern statistical software.

Estimating Healthcare Demand for an Aging Population: A Flexible and Robust Bayesian Joint Model

In this paper, we analyse two frequently used measures of the demand for health care, namely hospital visits and out-of-pocket health care expenditure, which have been analysed separately in the existing literature. Given that these two measures of healthcare demand are highly likely to be closely correlated, we propose a framework to jointly model hospital visits and out-of-pocket medical expenditure. Furthermore, the joint framework allows for the presence of non-linear effects of covariates using splines to capture the effects of aging on healthcare demand. Sample heterogeneity is modelled robustly with the random effects following Dirichlet process priors with explicit cross-part correlation. The findings of our empirical analysis of the U.S. Health and Retirement Survey indicate that the demand for healthcare varies with age and gender and exhibits significant cross-part correlation that provides a rich understanding of how aging affects health care demand, which is of particular policy relevance in the context of an aging population.

Predictive Evidence Threshold Scaling: Does the Evidence Meet a Confirmatory Standard?

ABSTRACT Making better use of evidence is one of the tenets of modern drug development. This calls for an understanding of the evidential strength of nonconfirmatory evidence relative to a confirmatory standard. Such inferential comparisons can be done via predictive evidence threshold scaling (PETS). Under PETS, the evidence meets a confirmatory standard if the predictive probability of a positive effect reaches the predictive evidence threshold from hypothetical confirmatory data. These probabilities require plausible assumptions about between-trial heterogeneity and potential biases. Two examples are discussed. The first is breakthrough designation, illustrated by a recent Food and Drug Administration approval of crizotinib for the treatment of non-small-cell lung cancer based on phase I and II data. The second is childhood Guillain–Barré syndrome, with sparse children data enriched with adult data. The examples suggest that the evidential strength of nonconfirmatory data can meet a confirmatory standard. This is reassuring for modern drug development, which exploits various types of evidence to inform adaptive licensing decisions. Supplementary materials for this article are available online.