A5016431348- Statistical Methods in Clinical Trials
- Health Systems, Economic Evaluations, Quality of Life
- Advanced Causal Inference Techniques
- Cancer Genomics and Diagnostics
- Meta-analysis and systematic reviews
- Optimal Experimental Design Methods
- Statistical Methods and Inference
- Economic and Financial Impacts of Cancer
- SARS-CoV-2 and COVID-19 Research
- Biosimilars and Bioanalytical Methods
- Viral gastroenteritis research and epidemiology
- Advanced Statistical Process Monitoring
- Pharmaceutical Economics and Policy
- Animal Virus Infections Studies
- Statistical Methods and Bayesian Inference
- Rheumatoid Arthritis Research and Therapies
- Pharmaceutical studies and practices
- Radiomics and Machine Learning in Medical Imaging
- Computational Drug Discovery Methods
- Colorectal Cancer Treatments and Studies
- Lymphoma Diagnosis and Treatment
- Genetic Associations and Epidemiology
- Bayesian Methods and Mixture Models
- Cancer Immunotherapy and Biomarkers
- Biomedical Ethics and Regulation
Pfizer (United States)
2018-2025
Statistical Research (United States)
2024
Johns Hopkins University
2020-2021
SUNY Upstate Medical University
2020-2021
Fundação Oswaldo Cruz
2021
Novartis (United States)
2012-2016
Novartis (Switzerland)
2015
Novartis (France)
2014
Tris Pharma (United States)
2014
St Thomas' Hospital
2004
BackgroundSevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and the resulting disease 2019 (Covid-19) have afflicted tens of millions people in a worldwide pandemic. Safe effective vaccines are needed urgently.MethodsIn an ongoing multinational, placebo-controlled, observer-blinded, pivotal efficacy trial, we randomly assigned persons 16 years age or older 1:1 ratio to receive two doses, 21 days apart, either placebo BNT162b2 vaccine candidate (30 μg per dose). is lipid...
BNT162b2 is a lipid nanoparticle-formulated, nucleoside-modified RNA vaccine encoding prefusion-stabilized, membrane-anchored severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) full-length spike protein. highly efficacious against disease 2019 (Covid-19) and currently approved, conditionally or authorized for emergency use worldwide. At the time of initial authorization, data beyond months after vaccination were unavailable.In an ongoing, placebo-controlled, observer-blinded,...
Historical information is always relevant for clinical trial design. Additionally, if incorporated in the analysis of a new trial, historical data allow to reduce number subjects. This decreases costs and duration, facilitates recruitment, may be more ethical. Yet, under prior-data conflict, too optimistic use inappropriate. We address this challenge by deriving Bayesian meta-analytic-predictive prior from data, which then combined with data. prospective approach equivalent...
ABSTRACT Background BNT162b2 is a lipid nanoparticle-formulated, nucleoside-modified RNA vaccine encoding prefusion-stabilized, membrane-anchored SARS-CoV-2 full-length spike protein. highly efficacious against COVID-19 and currently authorized for emergency use or conditional approval worldwide. At the time of authorization, data beyond 2 months post-vaccination were unavailable. Methods In an ongoing, placebo-controlled, observer-blinded, multinational, pivotal efficacy study, 44,165...
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 small populations and rare diseases. In early phase trials, where data often sparse, good statistical inference subsequent decision-making can challenging. Inferences from simple pooling or stratification known inferior hierarchical modeling methods, which build on exchangeable parameters allow borrowing information across strata....
Abstract A randomized trial allows estimation of the causal effect an intervention compared to a control in overall population and subpopulations defined by baseline characteristics. Often, however, clinical questions also arise regarding treatment patients, which would experience or disease related events post‐randomization. Events that occur after initiation potentially affect interpretation existence measurements are called intercurrent ICH E9(R1) guideline. If event is consequence...
Loss of power and clear description treatment differences are key issues in designing analyzing a clinical trial where nonproportional hazard (NPH) is possibility. A log-rank test may be inefficient interpretation the ratio estimated using Cox regression potentially problematic. In this case, current ICH E9 (R1) addendum would suggest with clinically relevant estimand, for example, expected life gain. This approach considers appropriate analysis methods supporting chosen estimand. However,...
Historical data are important for the design of a clinical trial. Yet these rarely used in analysis actual While justifiable certain situations, ignoring historical can lead to less accurate inferences, and, therefore, suboptimal decisions. After review main approaches using data, framework is extended co-data, which comprise all relevant (historical and concurrent) trial-external data. These be inference parameter trial via meta-analytic approaches. use co-data trials attractive, it also...
Abstract Introduction As patient-reported outcomes (PROs) are increasingly used in the evaluation of medical treatments, it is important that PROs carefully analyzed and interpreted. This may be challenging due to substantial missing values. The missingness often closely related patients’ disease status. In case, using observed information about intercurrent events (ICEs) such as progression death will improve handling PRO data. Therefore, aim this study was develop imputation models for...
ABSTRACT G‐computation has become a widely used robust method for estimating unconditional (marginal) treatment effects with covariate adjustment in the analysis of randomized clinical trials. Statistical inference this context typically relies on Wald test or interval, which can be easily implemented using consistent variance estimator. However, existing literature suggests that when sample sizes are small parameters interest near boundary values, Wald‐based methods may less reliable due to...
The log-rank test is most powerful under proportional hazards (PH). In practice, non-PH patterns are often observed in clinical trials, such as immuno-oncology; therefore, alternative methods needed to restore the efficiency of statistical testing. Three categories testing were evaluated, including weighted tests, Kaplan-Meier curve-based tests (including and Restricted Mean Survival Time, RMST), combination Breslow test, Lee's combo MaxCombo test). Nine scenarios representing PH various...
Objective Post hoc analysis of pooled data from nine randomised controlled trials to assess the effect tofacitinib (oral Janus kinase inhibitor for treatment rheumatoid arthritis (RA) and psoriatic (PsA)) on residual pain in patients with RA or PsA abrogated inflammation. Methods Patients who received ≥1 dose 5 mg twice daily, adalimumab placebo with/without background conventional synthetic disease-modifying antirheumatic drugs had inflammation (swollen joint count (SJC)=0 C reactive...
Abstract Background Patient-reported outcomes (PROs) play an increasing role in the evaluation of oncology treatments. At same time, single-arm trials are commonly included regulatory approval submissions. Because high risk biases, results from require careful interpretation. This benefits a clearly defined estimand , or target estimation. In this case study, we demonstrated how ICH E9 (R1) framework can be implemented SATs with PRO endpoints. Methods For global quality life outcome real...
Background Well-designed phase II trials must have acceptable error rates relative to a pre-specified success criterion, usually statistically significant p-value. Such standard designs may not always suffice from clinical perspective because relevance call for more. For example, proof-of-concept in often requires only statistical significance but also sufficiently large effect estimate. Purpose We propose dual-criterion complement with relevance, discuss their methodology, and illustrate...
Summary We analyse two frequently used measures of the demand for health—hospital visits and out‐of‐pocket health care expenditure—which have been analysed separately in existing literature. Given that these are highly likely to be closely correlated, we propose a framework jointly model hospital medical expenditure, which allows presence nonlinear effects covariates using splines capture aging on demand. The findings from our empirical analysis US Health Retirement Survey indicate varies...