A web application for the design of multi-arm clinical trials

Statistical Considerations for the Design and Analysis of Pragmatic Trials in Aging Research

Pragmatic trials aim to assess intervention efficacy in usual patient care settings, contrasting with explanatory trials conducted under controlled conditions. In aging research, pragmatic trials are important designs for obtaining real-world evidence in elderly populations, which are often underrepresented in trials. In this review, we discuss statistical considerations from a frequentist approach for the design and analysis of pragmatic trials. When choosing the dependent variable, it is essential to use an outcome that is highly relevant to usual medical care while also providing sufficient statistical power. Besides traditionally used binary outcomes, ordinal outcomes can provide pragmatic answers with gains in statistical power. Cluster randomization requires careful consideration of sample size calculation and analysis methods, especially regarding missing data and outcome variables. Mixed effects models and generalized estimating equations (GEEs) are recommended for analysis to account for center effects, with tools available for sample size estimation. Multi-arm studies pose challenges in sample size calculation, requiring adjustment for design effects and consideration of multiple comparison correction methods. Secondary analyses are common but require caution due to the risk of reduced statistical power and false-discovery rates. Safety data collection methods should balance pragmatism and data quality. Overall, understanding statistical considerations is crucial for designing rigorous pragmatic trials that evaluate interventions in elderly populations under real-world conditions. In conclusion, this review focuses on various statistical topics of interest to those designing a pragmatic clinical trial, with consideration of aspects of relevance in the aging research field.

Multi-arm multi-stage (MAMS) randomised selection designs: impact of treatment selection rules on the operating characteristics

BACKGROUND

Multi-arm multi-stage (MAMS) randomised trial designs have been proposed to evaluate multiple research questions in the confirmatory setting. In designs with several interventions, such as the 8-arm 3-stage ROSSINI-2 trial for preventing surgical wound infection, there are likely to be strict limits on the number of individuals that can be recruited or the funds available to support the protocol. These limitations may mean that not all research treatments can continue to accrue the required sample size for the definitive analysis of the primary outcome measure at the final stage. In these cases, an additional treatment selection rule can be applied at the early stages of the trial to restrict the maximum number of research arms that can progress to the subsequent stage(s). This article provides guidelines on how to implement treatment selection within the MAMS framework. It explores the impact of treatment selection rules, interim lack-of-benefit stopping boundaries and the timing of treatment selection on the operating characteristics of the MAMS selection design.

METHODS

We outline the steps to design a MAMS selection trial. Extensive simulation studies are used to explore the maximum/expected sample sizes, familywise type I error rate (FWER), and overall power of the design under both binding and non-binding interim stopping boundaries for lack-of-benefit.

RESULTS

Pre-specification of a treatment selection rule reduces the maximum sample size by approximately 25% in our simulations. The familywise type I error rate of a MAMS selection design is smaller than that of the standard MAMS design with similar design specifications without the additional treatment selection rule. In designs with strict selection rules - for example, when only one research arm is selected from 7 arms - the final stage significance levels can be relaxed for the primary analyses to ensure that the overall type I error for the trial is not underspent. When conducting treatment selection from several treatment arms, it is important to select a large enough subset of research arms (that is, more than one research arm) at early stages to maintain the overall power at the pre-specified level.

CONCLUSIONS

Multi-arm multi-stage selection designs gain efficiency over the standard MAMS design by reducing the overall sample size. Diligent pre-specification of the treatment selection rule, final stage significance level and interim stopping boundaries for lack-of-benefit are key to controlling the operating characteristics of a MAMS selection design. We provide guidance on these design features to ensure control of the operating characteristics.

5-EPIFAT trial protocol: a multi-center, randomized, placebo-controlled trial of the efficacy of pharmacotherapy for fatigue using methylphenidate, bupropion, ginseng, and amantadine in advanced cancer patients on active treatment

BACKGROUND

Cancer-related fatigue (CRF) is still undertreated in most patients, as evidence for pharmacological treatments is limited and conflicting. Also, the efficacy of the pharmacological agents relative to each other is still unclear. Therefore, medications that may potentially contribute to improving CRF will be investigated in this head-to-head trial. Our main objective is to compare the efficacy of methylphenidate vs. bupropion vs. ginseng vs. amantadine vs. placebo in patients with advanced cancer.

METHODS

The 5-EPIFAT study is a 5-arm, randomized, multi-blind, placebo-controlled, multicenter trial that will use a parallel-group design with an equal allocation ratio comparing the efficacy and safety of four medications (Methylphenidate vs. Bupropion vs. Ginseng vs. Amantadine) versus placebo for management of CRF. We will recruit 255 adult patients with advanced cancer who experience fatigue intensity ≥ 4 based on a 0-10 scale. The study period includes a 4-week intervention and a 4-week follow-up with repeated measurements over time. The primary outcome is the cancer-related fatigue level over time, which will be measured by the functional assessment of chronic illness therapy-fatigue (FACIT-F) scale. To evaluate safety, the secondary outcome is the symptomatic adverse events, which will be assessed using the Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events in cancer clinical trials (PRO-CTCAE). Also, a subgroup analysis based on a decision tree-based machine learning algorithm will be employed for the clinical prediction of different agents in homogeneous subgroups.

DISCUSSION

The findings of the 5-EPIFAT trial could be helpful to guide clinical decision-making, personalization treatment approach, design of future trials, as well as the development of CRF management guidelines.

TRIAL REGISTRATION

IRCT.ir IRCT20150302021307N6. Registered on 13 May 2023.

CITIES: Clinical trials with intercurrent events simulator

Although clinical trials are often designed with randomization and well-controlled protocols, complications will inevitably arise in the presence of intercurrent events (ICEs) such as treatment discontinuation. These can lead to missing outcome data and possibly confounding causal inference when the missingness is a function of a latent stratification of patients defined by intermediate outcomes. The pharmaceutical industry has been focused on developing new methods that can yield pertinent causal inferences in trials with ICEs. However, it is difficult to compare the properties of different methods developed in this endeavor as real-life clinical trial data cannot be easily shared to provide benchmark data sets. Furthermore, different methods consider distinct assumptions for the underlying data-generating mechanisms, and simulation studies often are customized to specific situations or methods. We develop a novel, general simulation model and corresponding Shiny application in R for clinical trials with ICEs, aptly named the Clinical Trials with Intercurrent Events Simulator (CITIES). It is formulated under the Rubin Causal Model where the considered treatment effects account for ICEs in clinical trials with repeated measures. CITIES facilitates the effective generation of data that resemble real-life clinical trials with respect to their reported summary statistics, without requiring the use of the original trial data. We illustrate the utility of CITIES via two case studies involving real-life clinical trials that demonstrate how CITIES provides a comprehensive tool for practitioners in the pharmaceutical industry to compare methods for the analysis of clinical trials with ICEs on identical, benchmark settings that resemble real-life trials.

Mapping Oral health and Local Area Resources (MOLAR): protocol for a randomised controlled trial connecting emergency department patients with social and dental resources

INTRODUCTION

There are substantial inequities in oral health access and outcomes in the USA, including by income and racial and ethnic identity. People with adverse social determinants of health (aSDoH), such as housing or food insecurity, are also more likely to have unmet dental needs. Many patients with dental problems present to the emergency department (ED), where minimal dental care or referral is usually available. Nonetheless, the ED represents an important point of contact to facilitate screening and referral for unmet oral health needs and aSDoH, particularly for patients who may not otherwise have access to care.

METHODS AND ANALYSIS

Mapping Oral health and Local Area Resources is a randomised controlled trial enrolling 2049 adult and paediatric ED patients with unmet oral health needs into one of three trial arms: (a) a standard handout of nearby dental and aSDoH resources; (b) a geographically matched listing of aSDoH resources and a search link for identification of geographically matched dental resources; or (c) geographically matched resources along with personalised care navigation. Follow-up at 3, 6, 9 and 12 months will evaluate oral health-related quality of life, linkage to resources and dental treatment, ED visits for dental problems and the association between linkage and neighbourhood resource density.

ETHICS AND DISSEMINATION

All sites share a single human subjects review board protocol which has been fully approved by the Mass General Brigham Human Subjects Review Board. Informed consent will be obtained from all adults and adult caregivers, and assent will be obtained from age-appropriate child participants. Results will demonstrate the impact of addressing aSDoH on oral health access and the efficacy of various forms of resource navigation compared with enhanced standard care. Our findings will facilitate sustainable, scalable interventions to identify and address aSDoH in the ED to improve oral health and reduce oral health inequities.

TRIAL REGISTRATION NUMBER

NCT05688982.

The benefits of covariate adjustment for adaptive multi-arm designs

Covariate adjustment via a regression approach is known to increase the precision of statistical inference when fixed trial designs are employed in randomized controlled studies. When an adaptive multi-arm design is employed with the ability to select treatments, it is unclear how covariate adjustment affects various aspects of the study. Consider the design framework that relies on pre-specified treatment selection rule(s) and a combination test approach for hypothesis testing. It is our primary goal to evaluate the impact of covariate adjustment on adaptive multi-arm designs with treatment selection. Our secondary goal is to show how the Uniformly Minimum Variance Conditionally Unbiased Estimator can be extended to account for covariate adjustment analytically. We find that adjustment with different sets of covariates can lead to different treatment selection outcomes and hence probabilities of rejecting hypotheses. Nevertheless, we do not see any negative impact on the control of the familywise error rate when covariates are included in the analysis model. When adjusting for covariates that are moderately or highly correlated with the outcome, we see various benefits to the analysis of the design. Conversely, there is negligible impact when including covariates that are uncorrelated with the outcome. Overall, pre-specification of covariate adjustment is recommended for the analysis of adaptive multi-arm design with treatment selection. Having the statistical analysis plan in place prior to the interim and final analyses is crucial, especially when a non-collapsible measure of treatment effect is considered in the trial.

Vitamin D Supplementation Does Not Influence SARS-CoV-2 Vaccine Efficacy or Immunogenicity: Sub-Studies Nested within the CORONAVIT Randomised Controlled Trial

Vitamin D deficiency has been reported to associate with the impaired development of antigen-specific responses following vaccination. We aimed to determine whether vitamin D supplements might boost the immunogenicity and efficacy of SARS-CoV-2 vaccination by conducting three sub-studies nested within the CORONAVIT randomised controlled trial, which investigated the effects of offering vitamin D supplements at a dose of 800 IU/day or 3200 IU/day vs. no offer on risk of acute respiratory infections in UK adults with circulating 25-hydroxyvitamin D concentrations <75 nmol/L. Sub-study 1 (n = 2808) investigated the effects of vitamin D supplementation on the risk of breakthrough SARS-CoV-2 infection following two doses of SARS-CoV-2 vaccine. Sub-study 2 (n = 1853) investigated the effects of vitamin D supplementation on titres of combined IgG, IgA and IgM (IgGAM) anti-Spike antibodies in eluates of dried blood spots collected after SARS-CoV-2 vaccination. Sub-study 3 (n = 100) investigated the effects of vitamin D supplementation on neutralising antibody and cellular responses in venous blood samples collected after SARS-CoV-2 vaccination. In total, 1945/2808 (69.3%) sub-study 1 participants received two doses of ChAdOx1 nCoV-19 (Oxford−AstraZeneca); the remainder received two doses of BNT162b2 (Pfizer). Mean follow-up 25(OH)D concentrations were significantly elevated in the 800 IU/day vs. no-offer group (82.5 vs. 53.6 nmol/L; mean difference 28.8 nmol/L, 95% CI 22.8−34.8) and in the 3200 IU/day vs. no offer group (105.4 vs. 53.6 nmol/L; mean difference 51.7 nmol/L, 45.1−58.4). Vitamin D supplementation did not influence the risk of breakthrough SARS-CoV-2 infection in vaccinated participants (800 IU/day vs. no offer: adjusted hazard ratio 1.28, 95% CI 0.89 to 1.84; 3200 IU/day vs. no offer: 1.17, 0.81 to 1.70). Neither did it influence IgGAM anti-Spike titres, neutralising antibody titres or IFN-γ concentrations in the supernatants of S peptide-stimulated whole blood. In conclusion, vitamin D replacement at a dose of 800 or 3200 IU/day effectively elevated 25(OH)D concentrations, but it did not influence the protective efficacy or immunogenicity of SARS-CoV-2 vaccination when given to adults who had a sub-optimal vitamin D status at baseline.

Effect of a test-and-treat approach to vitamin D supplementation on risk of all cause acute respiratory tract infection and covid-19: phase 3 randomised controlled trial (CORONAVIT)

OBJECTIVE

To determine the effect of population level implementation of a test-and-treat approach to correction of suboptimal vitamin D status (25-hydroxyvitamin D (25(OH)D) <75 nmol/L) on risk of all cause acute respiratory tract infection and covid 19.

DESIGN

Phase 3 open label randomised controlled trial.

SETTING

United Kingdom.

PARTICIPANTS

6200 people aged ≥16 years who were not taking vitamin D supplements at baseline.

INTERVENTIONS

Offer of a postal finger prick test of blood 25(OH)D concentration with provision of a six month supply of lower dose vitamin D (800 IU/day, n=1550) or higher dose vitamin D (3200 IU/day, n=1550) to those with blood 25(OH)D concentration <75 nmol/L, compared with no offer of testing or supplementation (n=3100). Follow-up was for six months.

MAIN OUTCOME MEASURES

The primary outcome was the proportion of participants with at least one swab test or doctor confirmed acute respiratory tract infection of any cause. A secondary outcome was the proportion of participants with swab test confirmed covid-19. Logistic regression was used to calculate odds ratios and associated 95% confidence intervals. The primary analysis was conducted by intention to treat.

RESULTS

Of 3100 participants offered a vitamin D test, 2958 (95.4%) accepted and 2674 (86.3%) had 25(OH)D concentrations <75 nmol/L and received vitamin D supplements (n=1328 lower dose, n=1346 higher dose). Compared with 136/2949 (4.6%) participants in the no offer group, at least one acute respiratory tract infection of any cause occurred in 87/1515 (5.7%) in the lower dose group (odds ratio 1.26, 95% confidence interval 0.96 to 1.66) and 76/1515 (5.0%) in the higher dose group (1.09, 0.82 to 1.46). Compared with 78/2949 (2.6%) participants in the no offer group, 55/1515 (3.6%) developed covid-19 in the lower dose group (1.39, 0.98 to 1.97) and 45/1515 (3.0%) in the higher dose group (1.13, 0.78 to 1.63).

CONCLUSIONS

Among people aged 16 years and older with a high baseline prevalence of suboptimal vitamin D status, implementation of a population level test-and-treat approach to vitamin D supplementation was not associated with a reduction in risk of all cause acute respiratory tract infection or covid-19.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04579640.

Adaptive Designs: Benefits and Cautions for Neurosurgery Trials

BACKGROUND

It is well accepted that randomized controlled trials provide the greatest quality of evidence about effectiveness and safety of new interventions. In neurosurgery, randomized controlled trials face challenges, with their use remaining relatively low compared with other clinical areas. Adaptive designs have emerged as a method for improving the efficiency and patient benefit of trials. They allow modifications to the trial design to be made as patient outcome data are collected. The benefit they provide is highly variable, predominantly governed by the time taken to observe the primary endpoint compared with the planned recruitment rate. They also face challenges in design, conduct, and reporting.

METHODS

We provide an overview of the benefits and challenges of adaptive designs, with a focus on neurosurgery applications. To investigate how often an adaptive design may be advantageous in neurosurgery, we extracted data on recruitment rates and endpoint lengths for ongoing neurosurgery trials registered in ClinicalTrials.gov.

RESULTS

We found that a majority of neurosurgery trials had a relatively short endpoint length compared with the planned recruitment period and therefore may benefit from an adaptive trial. However, we did not identify any ongoing ClinicalTrials.gov registered neurosurgery trials that mentioned using an adaptive design.

CONCLUSIONS

Adaptive designs may provide benefits to neurosurgery trials and should be considered for use more widely. Use of some types of adaptive design, such as multiarm multistage, may further increase the number of interventions that can be tested with limited patient and financial resources.

Systematic review of available software for multi-arm multi-stage and platform clinical trial design

BACKGROUND

In recent years, the popularity of multi-arm multi-stage, seamless adaptive, and platform trials has increased. However, many design-related questions and questions regarding which operating characteristics should be evaluated to determine the potential performance of a specific trial design remain and are often further complicated by the complexity of such trial designs.

METHODS

A systematic search was conducted to review existing software for the design of platform trials, whereby multi-arm multi-stage trials were also included. The results of this search are reported both on the literature level and the software level, highlighting the software judged to be particularly useful.

RESULTS

In recent years, many highly specialized software packages targeting single design elements on platform studies have been released. Only a few of the developed software packages provide extensive design flexibility, at the cost of limited access due to being commercial or not being usable as out-of-the-box solutions.

CONCLUSIONS

We believe that both an open-source modular software similar to OCTOPUS and a collaborative effort will be necessary to create software that takes advantage of and investigates the impact of all the flexibility that platform trials potentially provide.