Recommendations on multiple testing adjustment in multi-arm trials with a shared control group

The use of master protocols for efficient trial design to evaluate radiotherapy interventions: a systematic review

The aim of this review was to highlight why the use of master protocols trial design is particularly useful for radiotherapy intervention trials where complex setup pathways (including quality assurance, user training, and integrating multiple modalities of treatment) may hinder clinical advances. We carried out a systematic review according to Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines, reviewing the findings using a landscape analysis. Results were summarized descriptively, reporting on trial characteristics highlighting the benefits, limitations, and challenges of developing and implementing radiotherapy master protocols, with three case studies selected to explore these issues in more detail. Twelve studies were suitable for inclusion (4 platform trials, 3 umbrella trials, and 5 basket trials), evaluating a mix of solid tumor sites in both curative and palliative settings. The interventions were categorized into 1) novel agent and radiotherapy combinations; 2) radiotherapy dose personalization; and 3) device evaluation, with a case study provided for each intervention. Benefits of master protocol trials for radiotherapy intervention include protocol efficiency for implementation of novel radiotherapy techniques; accelerating the evaluation of novel agent drug and radiotherapy combinations; and more efficient translational research opportunities, leading to cost savings and research efficiency to improve patient outcomes. Master protocols offer an innovative platform under which multiple clinical questions can be addressed within a single trial. Due to the complexity of radiotherapy trial setup, cost and research efficiency savings may be more apparent than in systemic treatment trials. Use of this research approach may be the change needed to push forward oncological innovation within radiation oncology.

Adaptive platform trials: the impact of common controls on type one error and power

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.

How Much More Efficient Are Adaptive Platform Trials Than Multiple Stand-Alone Trials? A Comprehensive Simulation Study for Streamlining Drug Development During a Pandemic

With the coronavirus disease 2019 (COVID-19) pandemic, there is growing interest in utilizing adaptive platform clinical trials (APTs), in which multiple drugs are compared with a single common control group, such as a placebo or standard-of-care group. APTs evaluate several drugs for one disease and accept additions or exclusions of drugs as the trials progress; however, little is known about the efficiency of APTs over multiple stand-alone trials. In this study, we simulated the total development period, total sample size, and statistical operating characteristics of APTs and multiple stand-alone trials in drug development settings for hospitalized patients with COVID-19. Simulation studies using selected scenarios reconfirmed several findings regarding the efficiency of APTs. The APTs without staggered addition of drugs showed a shorter total development period than stand-alone trials, but the difference rapidly diminished if patient's enrollment was accelerated during the trials owing to the spread of infection. APTs with staggered addition of drugs still have the possibility of reducing the total development period compared with multiple stand-alone trials in some cases. Our study demonstrated that APTs could improve efficiency relative to multiple stand-alone trials regarding the total development period and total sample size without undermining statistical validity; however, this improvement varies depending on the speed of patient enrollment, sample size, presence/absence of family-wise error rate adjustment, allocation ratio between drug and placebo groups, and interval of staggered addition of drugs. Given the complexity of planning and implementing APT, the decision to implement APT during a pandemic must be made carefully.

Problem-solving therapy for pregnant women experiencing depressive symptoms and intimate partner violence: A randomised, controlled feasibility trial in rural Ethiopia

Evidence for the feasibility of brief psychological interventions for pregnant women experiencing intimate partner violence (IPV) in rural, low-income country settings is scarce. In rural Ethiopia, the prevalence of antenatal depressive symptoms and lifetime IPV are 29% and 61%, respectively. We aimed to assess the feasibility and related implementation outcomes of brief problem-solving therapy (PST) adapted for pregnant women experiencing IPV (PST-IPV) in rural Ethiopia, and of a randomised, controlled feasibility study design. We recruited 52 pregnant women experiencing depressive symptoms and past-year IPV from two antenatal care (ANC) services. Consenting women were randomised to PST-IPV (n = 25), 'standard' PST (not adapted for women experiencing IPV; n = 12) or enhanced usual care (information about sources of support; n = 15). Masked data collectors conducted outcome assessments nine weeks post-enrolment. Addis Ababa University (#032/19/CDT) and King's College London (#HR-18/19-9230) approved the study. Fidelity to randomisation was impeded by strong cultural norms about what constituted IPV. However, recruitment was feasible (recruitment rate: 1.5 per day; 37% of women screened were eligible). The intervention and trial were acceptable to women (4% declined initial screening, none declined to participate, and 76% attended all four sessions of either active intervention). PST-IPV was acceptable to ANC providers: none dropped out. Sessions lasting up to a mean 52 minutes raised questions about the appropriateness of the model to this context. Competence assessments recommended supplementary communication skills training. Fidelity assessments indicated high adherence, quality, and responsiveness but assessing risks and social networks, and discussing confidentiality needed improvement. Adjustments to optimise a future, fully powered, randomised controlled trial include staggering recruitment in line with therapist availability, more training on the types of IPV and how to discuss them, automating randomisation, a supervision cascade model, and conducting post-intervention outcome assessments immediately and three months postpartum. Registration: Pan African Clinical Trials Registry #PACTR202002513482084 (13/12/2019): https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=9601.

Improved efficiency for cross-arm comparisons via platform designs

Though platform trials have been touted for their flexibility and streamlined use of trial resources, their statistical efficiency is not well understood. We fill this gap by establishing their greater efficiency for comparing the relative efficacy of multiple interventions over using several separate, 2-arm trials, where the relative efficacy of an arbitrary pair of interventions is evaluated by contrasting their relative risks as compared to control. In theoretical and numerical studies, we demonstrate that the inference of such a contrast using data from a platform trial enjoys identical or better precision than using data from separate trials, even when the former enrolls substantially fewer participants. This benefit is attributed to the sharing of controls among interventions under contemporaneous randomization. We further provide a novel procedure for establishing the noninferiority of a given intervention relative to the most efficacious of the other interventions under evaluation, where this procedure is adaptive in the sense that it need not be a priori known which of these other interventions is most efficacious. Our numerical studies show that this testing procedure can attain substantially better power when the data arise from a platform trial rather than multiple separate trials. Our results are illustrated using data from two monoclonal antibody trials for the prevention of HIV.

Evolution of Phase II Oncology Trial Design: from Single Arm to Master Protocol

The recent development of novel anticancer treatments with diverse mechanisms of action has accelerated the detection of treatment candidates tremendously. The rapidly changing drug development landscapes and the high failure rates in Phase III trials both underscore the importance of more efficient and robust phase II designs. The goals of phase II oncology studies are to explore the preliminary efficacy and toxicity of the investigational product and to inform future drug development strategies such as go/no-go decisions for phase III development, or dose/indication selection. These complex purposes of phase II oncology designs call for efficient, flexible, and easy-to-implement clinical trial designs. Therefore, innovative adaptive study designs with the potential of improving the efficiency of the study, protecting patients, and improving the quality of information gained from trials have been commonly used in Phase II oncology studies. Although the value of adaptive clinical trial methods in early phase drug development is generally well accepted, there is no comprehensive review and guidance on adaptive design methods and their best practice for phase II oncology trials. In this paper, we review the recent development and evolution of phase II oncology design, including frequentist multistage design, Bayesian continuous monitoring, master protocol design, and innovative design methods for randomized phase II studies. The practical considerations and the implementation of these complex design methods are also discussed.

Online error rate control for platform trials

Platform trials evaluate multiple experimental treatments under a single master protocol, where new treatment arms are added to the trial over time. Given the multiple treatment comparisons, there is the potential for inflation of the overall type I error rate, which is complicated by the fact that the hypotheses are tested at different times and are not necessarily pre-specified. Online error rate control methodology provides a possible solution to the problem of multiplicity for platform trials where a relatively large number of hypotheses are expected to be tested over time. In the online multiple hypothesis testing framework, hypotheses are tested one-by-one over time, where at each time-step an analyst decides whether to reject the current null hypothesis without knowledge of future tests but based solely on past decisions. Methodology has recently been developed for online control of the false discovery rate as well as the familywise error rate (FWER). In this article, we describe how to apply online error rate control to the platform trial setting, present extensive simulation results, and give some recommendations for the use of this new methodology in practice. We show that the algorithms for online error rate control can have a substantially lower FWER than uncorrected testing, while still achieving noticeable gains in power when compared with the use of a Bonferroni correction. We also illustrate how online error rate control would have impacted a currently ongoing platform trial.

The West Africa Lassa fever Consortium pre-positioned protocol for a Phase II/III adaptive, randomised, controlled, platform trial to evaluate multiple Lassa fever therapeutics

Background: This is a standardized, pre-positioned protocol for the coordinated evaluation of Lassa fever therapeutics. The protocol is the product of discussions that took place in 2021 and 2022 among international investigators from a wide range of scientific and medical disciplines working together within the West Africa Lassa fever Consortium (WALC). Methods: This is a clinical Phase II/III multicentre randomised controlled platform trial using a superiority framework with an equal allocation ratio and a composite primary endpoint of all-cause mortality OR new onset of i) acute kidney failure (AKF), OR ii) acute respiratory failure (ARF), OR iii) shock assessed from enrolment (D0) to D28. Discussion: This pre-positioned protocol was developed by the WALC and made available for adaptation and implementation by the wider Lassa fever research community in order to generate efficient, reliable, and comparable evidence for Lassa fever therapeutics.

Alcohol-specific inhibition training in patients with alcohol use disorder: a multi-centre, double-blind randomized clinical trial examining drinking outcome and working mechanisms

AIMS

For the first time, to our knowledge, in a clinical sample with alcohol use disorder (AUD), this study compared the effects of two versions of alcohol-specific inhibition training (Alc-IT) on drinking outcomes and on experimental parameters assessing two possible working mechanisms: stimulus devaluation and inhibitory enhancement.

DESIGN

Multi-centre, double-blind, three-arm clinical RCT with 3-, 6- and 12-month follow-up comparing standard Alc-IT, improved Alc-IT and an active control condition.

SETTING

Three specialized AUD treatment centres in Switzerland.

PARTICIPANTS

A total of 242 detoxified, recently abstinent patients with severe AUD (18-60 years; 29.8% female).

INTERVENTION AND COMPARATOR

Both interventions [standard Alc-IT (n = 84) and improved Alc-IT (n = 79)] and the comparator [unspecific inhibition training (n = 79)] consisted of six sessions of a modified inhibitory task (Go/NoGo task) with alcohol-related and neutral stimuli. Both versions of Alc-IT required response inhibition in alcohol-related trials but differed in Go/NoGo ratios (standard: 50/50; improved: 75/25), with improved Alc-IT posing higher inhibitory demands. The control condition, an unspecific inhibition training, featured alcohol-related pictures in Go as well as NoGo trials.

MEASUREMENTS

The primary outcome, percentage of days abstinent, was assessed at 3-month follow-up with a time-line follow-back interview.

FINDINGS

The group receiving improved Alc-IT showed a significantly higher percentage of days abstinent at 3-month follow-up compared with the control group [γ_control  = 74.30; γ_improved  = 85.78; β = 11.48, 95% confidence interval (CI) = 2.57, 20.40, P = 0.012, adjusted r²  = 0.062], while for standard Alc-IT no effect significantly different from zero was detected (γ_standard  = 70.95; β = -3.35, 95% CI = -12.20, 5.50, P = 0.457, adjusted r²  = -0.04).

CONCLUSIONS

Alcohol-specific inhibition training with high inhibitory demands increased days abstinent at 3-month follow-up in patients with severe alcohol use disorder. Such an improved, inhibitory-demanding, alcohol-specific inhibition training outperformed the standard version of alcohol-specific inhibition training, suggesting an inhibitory working mechanism.

The West Africa Lassa fever Consortium pre-positioned protocol for a Phase II/III adaptive, randomised, controlled, platform trial to evaluate multiple Lassa fever therapeutics

Background: This is a standardized, pre-positioned protocol for the coordinated evaluation of Lassa fever therapeutics. The protocol is the product of discussions that took place in 2021 and 2022 among international investigators from a wide range of scientific and medical disciplines working together within the West Africa Lassa fever Consortium (WALC). Methods: This is a clinical Phase II/III multicentre randomised controlled platform trial using a superiority framework with an equal allocation ratio and a composite primary endpoint of all-cause mortality OR new onset of i) acute kidney failure (AKF), OR ii) acute respiratory failure (ARF), OR iii) shock assessed from enrolment (D0) to D28. Discussion: This pre-positioned protocol was developed by the WALC and made available for adaptation and implementation by the wider Lassa fever research community in order to generate efficient, reliable, and comparable evidence for Lassa fever therapeutics.

Evaluating the effect of short-course rifapentine-based regimens with or without enhanced behaviour-targeted treatment support on adherence and completion of treatment for latent tuberculosis infection among adults in the UK (RID-TB: Treat): protocol for an open-label, multicentre, randomised controlled trial

INTRODUCTION

The successful scale-up of a latent tuberculosis (TB) infection testing and treatment programme is essential to achieve TB elimination. However, poor adherence compromises its therapeutic effectiveness. Novel rifapentine-based regimens and treatment support based on behavioural science theory may improve treatment adherence and completion.

METHODS AND ANALYSIS

A pragmatic multicentre, open-label, randomised controlled trial assessing the effect of novel short-course rifapentine-based regimens for TB prevention and additional theory-based treatment support on treatment adherence against standard-of-care. Participants aged between 16 and 65 who are eligible to start TB preventive therapy will be recruited in England. 920 participants will be randomised to one of six arms with allocation ratio of 5:5:6:6:6:6: daily isoniazid +rifampicin for 3 months (3HR), routine treatment support (control); 3HR, additional treatment support; weekly isoniazid +rifapentine for 3 months (3HP), routine treatment support; weekly 3HP, additional treatment support ; daily isoniazid +rifapentine for 1 month (1HP), routine treatment support; daily 1HP, additional treatment support. Additional treatment support comprises reminders using an electronic pillbox, a short animation, and leaflets based on the perceptions and practicalities approach. The primary outcome is adequate treatment adherence, defined as taking ≥90% of allocated doses within the pre-specified treatment period, measured by electronic pillboxes. Secondary outcomes include safety and TB incidence within 12 months. We will conduct process evaluation of the trial interventions and assess intervention acceptability and fidelity and mechanisms for effect and estimate the cost-effectiveness of novel regimens. The protocol was developed with patient and public involvement, which will continue throughout the trial.

ETHICS AND DISSEMINATION

Ethics approval has been obtained from The National Health Service Health Research Authority (20/LO/1097). All participants will be required to provide written informed consent. We will share the results in peer-reviewed journals.

TRIAL REGISTRATION NUMBER

EudraCT 2020-004444-29.

Which tendon plasty has the best outcome? A comparison of four tendon plasty techniques in a large cohort of patients with symptomatic trapeziometacarpal osteoarthritis

INTRODUCTION

Trapeziometacarpal osteoarthritis (TMC OA) is commonly treated with a trapeziectomy combined with a form of tendon plasty. The type of tendon plasty used is based on surgeons' preference. The purpose of this observational study is to compare the outcomes of four different tendon plasties combined with trapeziectomy used to treat TMC OA: the Weilby, Burton-Pellegrini, Zancolli and Anchovy plasty.

PATIENTS AND METHODS

Patients treated with a trapeziectomy followed by a tendon plasty completed patient-reported outcome measures at baseline and twelve months postoperatively. The primary outcome is the Michigan Hand outcomes Questionnaire (MHQ) pain subscale. Secondary outcomes are the minimal clinically important difference (MCID) of MHQ pain, MHQ hand function, satisfaction, and complication rate.

RESULTS

793 patients received a trapeziectomy with a tendon plasty between November 2013 and December 2018. There was no difference in pain score after twelve months between the four tendon plasty techniques. Patients undergoing an Anchovy plasty had a higher chance of reaching the MCID for MHQ pain compared to the other techniques (OR 2.3 (95% CI 1.2-4.6)). Overall, more than 80% of the patients were satisfied with the treatment outcome, independent of which technique was used. Complication rates of the different techniques were similar.

CONCLUSION

Surgical treatment of TMC OA reduced pain after twelve months, independent of which tendon plasty was used. Patients receiving an Anchovy plasty were more likely to experience a clinically relevant improvement in pain, whilst having a similar hand function, satisfaction, and complication rate. This suggests that Anchovy plasty is the preferred tendon plasty.

A study protocol for an ongoing multi-arm, randomized, double-blind, sham-controlled clinical trial with digital features, using portable transcranial electrical stimulation and internet-based behavioral therapy for major depression disorders: The PSYLECT study

BACKGROUND

Transcranial electrical stimulation (tES) is considered effective and safe for depression, albeit modestly, and prone to logistical burdens when performed in external facilities. Investigation of portable tES (ptES), and potentiation of ptES with remote psychological interventions have shown positive, but preliminary, results.

RESEARCH DESIGN

We report the rationale and design of an ongoing multi-arm, randomized, double-blind, sham-controlled clinical trial with digital features, using ptES and internet-based behavioral therapy (iBT) for major depressive disorder (MDD) (NCT04889976).

METHODS

We will evaluate the efficacy, safety, tolerability and usability of (1) active ptES + active iBT ('double-active'), (2) active ptES + sham iBT ('ptES-only'), and (3) sham ptES + sham iBT ('double-sham'), in adults with MDD, with a Hamilton Depression Rating Scale - 17 item version (HDRS-17) score ≥ 17 at baseline, during 6 weeks. Antidepressants are allowed in stable doses during the trial.

RESULTS

We primarily co-hypothesize changes in HDRS-17 will be greater in (1) 'double-active' compared to 'ptES-only,' (2) 'double-active' compared to 'double-sham,' and (3) 'ptES-only' compared to 'double-sham.' We aim to enroll 210 patients (70 per arm).

CONCLUSIONS

Our results should offer new insights regarding the efficacy and scalability of combined ptES and iBT for MDD, in digital mental health.

Decision rules for identifying combination therapies in open-entry, randomized controlled platform trials

Platform trials have become increasingly popular for drug development programs, attracting interest from statisticians, clinicians and regulatory agencies. Many statistical questions related to designing platform trials-such as the impact of decision rules, sharing of information across cohorts, and allocation ratios on operating characteristics and error rates-remain unanswered. In many platform trials, the definition of error rates is not straightforward as classical error rate concepts are not applicable. For an open-entry, exploratory platform trial design comparing combination therapies to the respective monotherapies and standard-of-care, we define a set of error rates and operating characteristics and then use these to compare a set of design parameters under a range of simulation assumptions. When setting up the simulations, we aimed for realistic trial trajectories, such that for example, a priori we do not know the exact number of treatments that will be included over time in a specific simulation run as this follows a stochastic mechanism. Our results indicate that the method of data sharing, exact specification of decision rules and a priori assumptions regarding the treatment efficacy all strongly contribute to the operating characteristics of the platform trial. Furthermore, different operating characteristics might be of importance to different stakeholders. Together with the potential flexibility and complexity of a platform trial, which also impact the achieved operating characteristics via, for example, the degree of efficiency of data sharing this implies that utmost care needs to be given to evaluation of different assumptions and design parameters at the design stage.

Mastery versus invention learning: impacts on future learning of simulated procedural skills

BACKGROUND

Invention and mastery learning approaches differ in their foundational educational paradigms, proposed mechanisms of learning, and potential impacts on learning outcomes. They also differ in their resource requirements. We explored the relative effects of 'invent and problem-solve, followed by instruction' (PS-I) learning compared to mastery learning (i.e., standards-based training) on immediate post-test and Preparation for Future Learning (PFL) assessments. PFL assessments measure learners' capacity to use their existing knowledge and strategies to learn about and solve novel problems.

METHODS

In this non-inferiority trial, pre-clerkship medical students were randomized to either PS-I, Mastery Learning (ML), or instruction then practice (CON) during simulation-based training of infant lumbar puncture (LP). After a 2-week delay, participants returned to learn and complete a PFL assessment of simulated Knee Arthrocentesis. Two independent raters assessed performances with a 5-point global rating scale.

RESULTS

Based on our non-inferiority margin, analyses showed that for both the immediate post-test and the PFL assessment, the PS-I condition resulted in non-inferior outcomes relative to the ML condition. Results for the CON condition were mixed with respect to non-inferiority compared to either PS-I or ML.

CONCLUSIONS

We suggest cautiously that the PS-I approach was not inferior to the ML approach, based on skill acquisition and PFL assessment outcomes. With ML anecdotally and empirically requiring more time, greater faculty involvement, and higher costs, our findings question the preference ML has received relative to other instructional designs, especially in the healthcare simulation community. We encourage researchers to study the educational and resource impacts of instructional designs using non-inferiority designs.

Efficacy of Cognitive-Behavioral Therapy for the Prophylaxis of Migraine in Adults: A Three-Armed Randomized Controlled Trial

Background

Behavioral approaches are central to the preventive treatment of migraine but empirical evidence regarding efficacy and effectiveness is still sparse. This study aimed to evaluate the efficacy of a newly developed migraine-specific, integrative cognitive-behavioral therapy program (miCBT) combining several approaches (trigger and stress management, coping with fear of attacks, relaxation training) by comparing it with a single behavioral approach (relaxation training, RLX) as an active control group and a waiting-list control group (WLC).

Methods

In a three-armed open-label randomized controlled trial, 121 adults with migraine were assigned to either miCBT, RLX or WLC. The outpatient group therapy (miCBT or RLX) consisted of seven sessions each 90 min. Participants who completed the WLC were subsequently randomized to one of the two treatment groups. Primary outcomes were headache days, headache-related disability, emotional distress, and self-efficacy. The baseline was compared to post-treatment, and followed by assessments 4- and 12-months post-treatment to compare miCBT and RLX.

Results

Mixed-model analyses (intention-to-treat sample, 106 participants) showed significantly stronger pre-post improvements in self-efficacy (assessed by the Headache Management Self-Efficacy Scale, HMSE-G-SF) in both treatment groups compared to the WLC (mean difference at post; miCBT: 4.67 [0.55–8.78], p = 0.027; RLX: 4.42 [0.38 to 8.46], p = 0.032), whereas no other significant between-group differences were observed. The follow-up analyses revealed significant within-group improvements from baseline to 12-month follow-up in all four primary outcomes for both treatments. However, between-group effects (miCBT vs. RLX) were not significant at follow-up.

Conclusion

The miCBT has no better treatment effects compared to RLX in migraine-prophylaxis. Both treatments effectively increase patients' self-efficacy.

Trial Registration

German Clinical Trials Register (www.drks.de; DRKS-ID: DRKS00011111).

Multiplicity adjustments in parallel-group multi-arm trials sharing a control group: Clear guidance is needed

Multi-arm, parallel-group clinical trials are an efficient way of testing several new treatments, treatment regimens or doses. However, guidance on the requirement for statistical adjustment to control for multiple comparisons (type I error) using a shared control group is unclear. We argue, based on current evidence, that adjustment is not always necessary in such situations. We propose that adjustment should not be a requirement in multi-arm, parallel-group trials testing distinct treatments and sharing a control group, and we call for clearer guidance from stakeholders, such as regulators and scientific journals, on the appropriate settings for adjustment of multiplicity.

Statistical considerations of phase 3 umbrella trials allowing adding one treatment arm mid-trial

Master protocols, in particular umbrella trials and platform trials, when evaluating multiple experimental treatments with a common control, could save patient resource, increase trial efficiency, and reduce drug development cost. Compared to the phase 3 platform trials that allow unlimited number of experimental arms to be added, it is more practical for individual companies to evaluate two experimental arms with a common control in an umbrella trial and allow the second experimental arm to be added at a later time. There have been limited research done in this type of trials in terms of statistical properties and guidance. In this article, we present statistical considerations of a phase 3 three-arm umbrella design including Type I error control and power, as well as the optimal allocation ratio. We intend to not only complement the existing literature, but more importantly to provide practical guidance to pave the way for its implementation by individual companies.

Recommendations for designing and analysing multi-arm non-inferiority trials: a review of methodology and current practice

Background and purpose

Multi-arm non-inferiority (MANI) trials, here defined as non-inferiority trials with multiple experimental treatment arms, can be useful in situations where several viable treatments exist for a disease area or for testing different dose schedules. To maintain the statistical integrity of such trials, issues regarding both design and analysis must be considered, from both the multi-arm and the non-inferiority perspectives. Little guidance currently exists on exactly how these aspects should be addressed and it is the aim of this paper to provide recommendations to aid the design of future MANI trials.

Methods

A comprehensive literature review covering four databases was conducted to identify publications associated with MANI trials. Literature was split into methodological and trial publications in order to investigate the required design and analysis considerations for MANI trials and whether they were being addressed in practice.

Results

A number of issues were identified that if not properly addressed, could lead to issues with the FWER, power or bias. These ranged from the structuring of trial hypotheses at the design stage to the consideration of potential heterogeneous treatment variances at the analysis stage. One key issue of interest was adjustment for multiple testing at the analysis stage. There was little consensus concerning whether more powerful p value adjustment methods were preferred to approximate adjusted CIs when presenting and interpreting the results of MANI trials.

We found 65 examples of previous MANI trials, of which 31 adjusted for multiple testing out of the 39 that were adjudged to require it. Trials generally preferred to utilise simple, well-known methods for study design and analysis and while some awareness was shown concerning FWER inflation and choice of power, many trials seemed not to consider the issues and did not provide sufficient definition of their chosen design and analysis approaches.

Conclusions

While MANI trials to date have shown some awareness of the issues raised within this paper, very few have satisfied the criteria of the outlined recommendations. Going forward, trials should consider the recommendations in this paper and ensure they clearly define and reason their choices of trial design and analysis techniques.

Practical Considerations and Recommendations for Master Protocol Framework: Basket, Umbrella and Platform Trials

Master protocol, categorized as basket trial, umbrella trial or platform trial, is an innovative clinical trial framework that aims to expedite clinical drug development, enhance trial efficiency, and eventually bring medicines to patients faster. Despite a clear uptake on the advantages in the concepts and designs, master protocols are still yet to be widely used. Part of that may be due to the fact that the master protocol framework comes with the need for new statistical designs and considerations for analyses and operational challenges. In this article, we provide an overview of the master protocol framework, unify the definitions with some examples, review the statistical methods for the designs and analyses, and focus our discussions on some practical considerations and recommendations of master protocols to help practitioners better design and implement such studies.

A platform trial in practice: adding a new experimental research arm to the ongoing confirmatory FLAIR trial in chronic lymphocytic leukaemia

BACKGROUND

The FLAIR trial in chronic lymphocytic leukaemia has a randomised, controlled, open-label, confirmatory, platform design. FLAIR was successfully amended to include an emerging promising experimental therapy to expedite its assessment, greatly reducing the time to reach the primary outcome compared to running a separate trial and without compromising the validity of the research or the ability to recruit to the trial and report the outcomes. The methodological and practical issues are presented, describing how they were addressed to ensure the amendment was a success.

METHODS

FLAIR was designed as a two-arm trial requiring 754 patients. In stage 2, two new arms were added: a new experimental arm and a second control arm to protect the trial in case of a change in practice. In stage 3, the original experimental arm was closed as its planned recruitment target was reached. In total, 1516 participants will be randomised to the trial.

RESULTS

The changes to the protocol and randomisation to add and stop arms were made seamlessly without pausing recruitment. The statistical considerations to ensure the results for the original and new hypotheses are unbiased were approved following peer review by oversight committees, Cancer Research UK, ethical and regulatory committees and pharmaceutical partners. These included the use of concurrent comparators in case of any stage effect, appropriate control of the type I error rate and consideration of analysis methods across trial stages. The operational aspects of successfully implementing the amendments are described, including gaining approvals and additional funding, data management requirements and implementation at centres.

CONCLUSIONS

FLAIR is an exemplar of how an emerging experimental therapy can be assessed within an existing trial structure without compromising the conduct, reporting or validity of the trial. This strategy offered considerable resource savings and allowed the new experimental therapy to be assessed within a confirmatory trial in the UK years earlier than would have otherwise been possible. Despite the clear efficiencies, treatment arms are rarely added to ongoing trials in practice. This paper demonstrates how this strategy is acceptable, feasible and beneficial to patients and the wider research community.

TRIAL REGISTRATION

ISRCTN Registry ISRCTN01844152 . Registered on August 08, 2014.

Controlling type I error rates in multi-arm clinical trials: A case for the false discovery rate

Multi-arm trials are an efficient way of simultaneously testing several experimental treatments against a shared control group. As well as reducing the sample size required compared to running each trial separately, they have important administrative and logistical advantages. There has been debate over whether multi-arm trials should correct for the fact that multiple null hypotheses are tested within the same experiment. Previous opinions have ranged from no correction is required, to a stringent correction (controlling the probability of making at least one type I error) being needed, with regulators arguing the latter for confirmatory settings. In this article, we propose that controlling the false-discovery rate (FDR) is a suitable compromise, with an appealing interpretation in multi-arm clinical trials. We investigate the properties of the different correction methods in terms of the positive and negative predictive value (respectively how confident we are that a recommended treatment is effective and that a non-recommended treatment is ineffective). The number of arms and proportion of treatments that are truly effective is varied. Controlling the FDR provides good properties. It retains the high positive predictive value of FWER correction in situations where a low proportion of treatments is effective. It also has a good negative predictive value in situations where a high proportion of treatments is effective. In a multi-arm trial testing distinct treatment arms, we recommend that sponsors and trialists consider use of the FDR.

Non-adjustment for multiple testing in multi-arm trials of distinct treatments: Rationale and justification

There is currently a lack of consensus and uncertainty about whether one should adjust for multiple testing in multi-arm trials of distinct treatments. A detailed rationale is presented to justify non-adjustment in this situation. We argue that non-adjustment should be the default starting position in simple multi-arm trials of distinct treatments.

Design, analysis and reporting of multi-arm trials and strategies to address multiple testing

BACKGROUND

It is unclear how multiple treatment comparisons are managed in the analysis of multi-arm trials, particularly related to reducing type I (false positive) and type II errors (false negative).

METHODS

We conducted a cohort study of clinical-trial protocols that were approved by research ethics committees in the UK, Switzerland, Germany and Canada in 2012. We examined the use of multiple-testing procedures to control the overall type I error rate. We created a decision tool to determine the need for multiple-testing procedures. We compared the result of the decision tool to the analysis plan in the protocol. We also compared the pre-specified analysis plans in trial protocols to their publications.

RESULTS

Sixty-four protocols for multi-arm trials were identified, of which 50 involved multiple testing. Nine of 50 trials (18%) used a single-step multiple-testing procedures such as a Bonferroni correction and 17 (38%) used an ordered sequence of primary comparisons to control the overall type I error. Based on our decision tool, 45 of 50 protocols (90%) required use of a multiple-testing procedure but only 28 of the 45 (62%) accounted for multiplicity in their analysis or provided a rationale if no multiple-testing procedure was used. We identified 32 protocol-publication pairs, of which 8 planned a global-comparison test and 20 planned a multiple-testing procedure in their trial protocol. However, four of these eight trials (50%) did not use the global-comparison test. Likewise, 3 of the 20 trials (15%) did not perform the multiple-testing procedure in the publication. The sample size of our study was small and we did not have access to statistical-analysis plans for the included trials in our study.

CONCLUSIONS

Strategies to reduce type I and type II errors are inconsistently employed in multi-arm trials. Important analytical differences exist between planned analyses in clinical-trial protocols and subsequent publications, which may suggest selective reporting of analyses.

Problem solving therapy (PST) tailored for intimate partner violence (IPV) versus standard PST and enhanced usual care for pregnant women experiencing IPV in rural Ethiopia: protocol for a randomised controlled feasibility trial

BACKGROUND

In rural Ethiopia, 72% of women are exposed to lifetime intimate partner violence (IPV); IPV is most prevalent during pregnancy. As well as adversely affecting women's physical and mental health, IPV also increases the risk of child morbidity and mortality associated with maternal depression, thus making antenatal care an important opportunity for intervention. Adapting generic, task-shared, brief psychological interventions for perinatal depression and anxiety to address the needs and experiences of women affected by IPV may improve acceptability to women and feasibility for health workers. This randomised controlled feasibility trial will compare brief problem solving therapy (PST) specifically adapted for pregnant women experiencing IPV (PST-IPV) with standard PST and enhanced usual care to determine the feasibility of a future fully powered randomised controlled trial.

METHODS

Seventy-five pregnant women scoring five or more on the Patient Health Questionnaire, endorsing a tenth question about functional impact and reporting past-year IPV, will be recruited from antenatal care clinics in predominantly rural districts in Ethiopia. Consenting participants will be randomised to either four sessions of PST-IPV, four sessions of standard PST or information about sources of support (enhanced usual care) in a three-arm design. The interventions will be delivered by trained, supervised antenatal care staff using a task-sharing model. Assessments will be made at baseline and after 9 weeks by masked outcome assessors and will include measures of depression symptoms (primary outcome), post-traumatic stress, anxiety symptoms, functional impact, past-month IPV and hypothesised mediators (secondary outcomes). A mixed-method process evaluation will determine the feasibility of a future randomised controlled trial, assess the feasibility, acceptability, fidelity and quality of implementation of PST-IPV, generate testable hypotheses about causal mechanisms, and identify potential contextual factors influencing outcomes.

DISCUSSION

Despite mental health being a critical concern for women experiencing IPV, there is limited evidence for brief, task-shared psychological interventions adapted for their needs in low- and middle-income countries. Contextually tailored interventions for pregnant women experiencing IPV in low- and middle-income countries require development and process evaluation. This randomised controlled feasibility trial will yield results on the feasibility of conducting a fully powered trial, relevant to researchers, primary and antenatal care clinicians in resource-limited settings.

TRIAL REGISTRATION

Pan-African clinical trials registry: PACTR202002513482084. Prospectively registered on 13 December 2019.

Hypofractionated breast radiotherapy for 1 week versus 3 weeks (FAST-Forward): 5-year efficacy and late normal tissue effects results from a multicentre, non-inferiority, randomised, phase 3 trial

BACKGROUND

We aimed to identify a five-fraction schedule of adjuvant radiotherapy (radiation therapy) delivered in 1 week that is non-inferior in terms of local cancer control and is as safe as an international standard 15-fraction regimen after primary surgery for early breast cancer. Here, we present 5-year results of the FAST-Forward trial.

METHODS

FAST-Forward is a multicentre, phase 3, randomised, non-inferiority trial done at 97 hospitals (47 radiotherapy centres and 50 referring hospitals) in the UK. Patients aged at least 18 years with invasive carcinoma of the breast (pT1-3, pN0-1, M0) after breast conservation surgery or mastectomy were eligible. We randomly allocated patients to either 40 Gy in 15 fractions (over 3 weeks), 27 Gy in five fractions (over 1 week), or 26 Gy in five fractions (over 1 week) to the whole breast or chest wall. Allocation was not masked because of the nature of the intervention. The primary endpoint was ipsilateral breast tumour relapse; assuming a 2% 5-year incidence for 40 Gy, non-inferiority was predefined as ≤1·6% excess for five-fraction schedules (critical hazard ratio [HR] of 1·81). Normal tissue effects were assessed by clinicians, patients, and from photographs. This trial is registered at isrctn.com, ISRCTN19906132.

FINDINGS

Between Nov 24, 2011, and June 19, 2014, we recruited and obtained consent from 4096 patients from 97 UK centres, of whom 1361 were assigned to the 40 Gy schedule, 1367 to the 27 Gy schedule, and 1368 to the 26 Gy schedule. At a median follow-up of 71·5 months (IQR 71·3 to 71·7), the primary endpoint event occurred in 79 patients (31 in the 40 Gy group, 27 in the 27 Gy group, and 21 in the 26 Gy group); HRs versus 40 Gy in 15 fractions were 0·86 (95% CI 0·51 to 1·44) for 27 Gy in five fractions and 0·67 (0·38 to 1·16) for 26 Gy in five fractions. 5-year incidence of ipsilateral breast tumour relapse after 40 Gy was 2·1% (1·4 to 3·1); estimated absolute differences versus 40 Gy in 15 fractions were -0·3% (-1·0 to 0·9) for 27 Gy in five fractions (probability of incorrectly accepting an inferior five-fraction schedule: p=0·0022 vs 40 Gy in 15 fractions) and -0·7% (-1·3 to 0·3) for 26 Gy in five fractions (p=0·00019 vs 40 Gy in 15 fractions). At 5 years, any moderate or marked clinician-assessed normal tissue effects in the breast or chest wall was reported for 98 of 986 (9·9%) 40 Gy patients, 155 (15·4%) of 1005 27 Gy patients, and 121 of 1020 (11·9%) 26 Gy patients. Across all clinician assessments from 1-5 years, odds ratios versus 40 Gy in 15 fractions were 1·55 (95% CI 1·32 to 1·83, p<0·0001) for 27 Gy in five fractions and 1·12 (0·94 to 1·34, p=0·20) for 26 Gy in five fractions. Patient and photographic assessments showed higher normal tissue effect risk for 27 Gy versus 40 Gy but not for 26 Gy versus 40 Gy.

INTERPRETATION

26 Gy in five fractions over 1 week is non-inferior to the standard of 40 Gy in 15 fractions over 3 weeks for local tumour control, and is as safe in terms of normal tissue effects up to 5 years for patients prescribed adjuvant local radiotherapy after primary surgery for early-stage breast cancer.

FUNDING

National Institute for Health Research Health Technology Assessment Programme.

Designs for adding a treatment arm to an ongoing clinical trial

Background

For many disease areas, there are often treatments in different stages of the development process. We consider the design of a two-arm parallel group trial where it is planned to add a new experimental treatment arm during the trial. This could potentially save money, patients, time and resources; however, the addition of a treatment arm creates a multiple comparison problem. Current practice in trials when a new treatment arm has been added is to compare the new treatment only to controls randomised concurrently, and this is the setting we consider here. Furthermore, for standard multi-arm trials, optimal allocation randomises a larger number of patients to the control arm than to each experimental treatment arm.

Methods

In this paper we propose an adaptive design, the aim of which is to adapt the sample size of the trial when the new treatment arm is added to control the family-wise error rate (FWER) in the strong sense, whilst maintaining the marginal power of each treatment-to-control comparison at the level of the original study. We explore optimal allocation for designs where a treatment arm is added with the aim of increasing the overall power of the study, where we define the overall power to be the probability of detecting all treatments that are better than the control.

Results and conclusions

An increase in sample size is required to maintain the marginal power for each pairwise comparison when adding a treatment arm if control of the FWER is required at the level of the type I error in the original study. When control of the FWER is required in a single trial which adds an additional experimental treatment arm, but control of the FWER is not required in separate trials, depending on the design characteristics, it may be better to run a separate trial for each experimental treatment, in terms of the number of patients required. An increase in overall power can be achieved when optimal allocation is used once a treatment arm has been added to the trial, rather than continuing with equal allocation to all treatment arms.

Adding new experimental arms to randomised clinical trials: Impact on error rates

BACKGROUND

Experimental treatments pass through various stages of development. If a treatment passes through early-phase experiments, the investigators may want to assess it in a late-phase randomised controlled trial. An efficient way to do this is adding it as a new research arm to an ongoing trial while the existing research arms continue, a so-called multi-arm platform trial. The familywise type I error rate is often a key quantity of interest in any multi-arm platform trial. We set out to clarify how it should be calculated when new arms are added to a trial some time after it has started.

METHODS

We show how the familywise type I error rate, any-pair and all-pairs powers can be calculated when a new arm is added to a platform trial. We extend the Dunnett probability and derive analytical formulae for the correlation between the test statistics of the existing pairwise comparison and that of the newly added arm. We also verify our analytical derivation via simulations.

RESULTS

Our results indicate that the familywise type I error rate depends on the shared control arm information (i.e. individuals in continuous and binary outcomes and primary outcome events in time-to-event outcomes) from the common control arm patients and the allocation ratio. The familywise type I error rate is driven more by the number of pairwise comparisons and the corresponding (pairwise) type I error rates than by the timing of the addition of the new arms. The familywise type I error rate can be estimated using Šidák's correction if the correlation between the test statistics of pairwise comparisons is less than 0.30.

CONCLUSIONS

The findings we present in this article can be used to design trials with pre-planned deferred arms or to add new pairwise comparisons within an ongoing platform trial where control of the pairwise error rate or familywise type I error rate (for a subset of pairwise comparisons) is required.

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

BACKGROUND

Multi-arm designs provide an effective means of evaluating several treatments within the same clinical trial. Given the large number of treatments now available for testing in many disease areas, it has been argued that their utilisation should increase. However, for any given clinical trial there are numerous possible multi-arm designs that could be used, and choosing between them can be a difficult task. This task is complicated further by a lack of available easy-to-use software for designing multi-arm trials.

RESULTS

To aid the wider implementation of multi-arm clinical trial designs, we have developed a web application for sample size calculation when using a variety of popular multiple comparison corrections. Furthermore, the application supports sample size calculation to control several varieties of power, as well as the determination of optimised arm-wise allocation ratios. It is built using the Shiny package in the R programming language, is free to access on any device with an internet browser, and requires no programming knowledge to use. It incorporates a variety of features to make it easier to use, including help boxes and warning messages. Using design parameters motivated by a recently completed phase II oncology trial, we demonstrate that the application can effectively determine and evaluate complex multi-arm trial designs.

CONCLUSIONS

The application provides the core information required by statisticians and clinicians to review the operating characteristics of a chosen multi-arm clinical trial design. The range of designs supported by the application is broader than other currently available software solutions. Its primary limitation, particularly from a regulatory agency point of view, is its lack of validation. However, we present an approach to efficiently confirming its results via simulation.

Testing monotherapy and combination therapy in one trial with biomarker consideration

It is a common scenario that an experimental oncology therapy, as a monotherapy, may be more effective than standard of care (SOC) in a biomarker positive population but less so or even inferior to SOC in biomarker negative population. At the same time, due to synergistic or additive effect, the combination of the two may be more effective than SOC alone in the all-comer population. The conventional development paradigm is to conduct two separate Phase III trials, one with the monotherapy versus SOC in the biomarker positive population, and the other with the combination therapy versus SOC in the all-comer population. In this manuscript, we propose a one-trial design that stratifies by biomarker status and randomizes biomarker positive patients into three arms (combination therapy, monotherapy, and SOC) and biomarker negative patients into two arms (combination therapy and SOC). There are two hypotheses in the proposed design and each addresses a different question. The family-wise type-I error rate (FWER) is smaller, due to shared control, than that of two separate trials. Therefore, no FWER adjustment is necessary in the proposed design and each hypothesis can be tested at the conventional 2.5% (one-sided) alpha level. The population for comparison between the combination therapy and SOC is skewed in the proposed design. A two-step log-rank statistic is proposed to account for the skewness. Power and sample size of the proposed design are evaluated in comparison with the two-trial paradigm. The proposed design is more efficient.

Reporting of Multi-Arm Parallel-Group Randomized Trials: Extension of the CONSORT 2010 Statement

IMPORTANCE

The quality of reporting of randomized clinical trials is suboptimal. In an era in which the need for greater research transparency is paramount, inadequate reporting hinders assessment of the reliability and validity of trial findings. The Consolidated Standards of Reporting Trials (CONSORT) 2010 Statement was developed to improve the reporting of randomized clinical trials, but the primary focus was on parallel-group trials with 2 groups. Multi-arm trials that use a parallel-group design (comparing treatments by concurrently randomizing participants to one of the treatment groups, usually with equal probability) but have 3 or more groups are relatively common. The quality of reporting of multi-arm trials varies substantially, making judgments and interpretation difficult. While the majority of the elements of the CONSORT 2010 Statement apply equally to multi-arm trials, some elements need adaptation, and, in some cases, additional issues need to be clarified.

OBJECTIVE

To present an extension to the CONSORT 2010 Statement for reporting multi-arm trials to facilitate the reporting of such trials.

DESIGN

A guideline writing group, which included all authors, formed following the CONSORT group meeting in 2014. The authors met in person and by teleconference bimonthly between 2014 and 2018 to develop and revise the checklist and the accompanying text, with additional discussions by email. A draft manuscript was circulated to the wider CONSORT group of 36 individuals, plus 5 other selected individuals known for their specialist knowledge in clinical trials, for review. Extensive feedback was received from 14 individuals and, after detailed consideration of their comments, a final revised version of the extension was prepared.

FINDINGS

This CONSORT extension for multi-arm trials expands on 10 items of the CONSORT 2010 checklist and provides examples of good reporting and a rationale for the importance of each extension item. Key recommendations are that multi-arm trials should be identified as such and require clear objectives and hypotheses referring to all of the treatment groups. Primary treatment comparisons should be identified and authors should report the planned and unplanned comparisons resulting from multiple groups completely and transparently. If statistical adjustments for multiplicity are applied, the rationale and method used should be described.

CONCLUSIONS AND RELEVANCE

This extension of the CONSORT 2010 Statement provides specific guidance for the reporting of multi-arm parallel-group randomized clinical trials and should help provide greater transparency and accuracy in the reporting of such trials.

Does self-monitoring and self-management of blood pressure after stroke or transient ischemic attack improve control? TEST-BP, a randomized controlled trial

The therapeutic benefit of self-monitoring blood pressure in stroke patients is uncertain. We investigated the effect of self-monitoring with or without guided antihypertensive management compared with usual care in patients with a recent cerebrovascular event. No between-group differences in blood pressure at outcome were found, but blood pressure self-monitoring and management was well tolerated.

Dressings and securements for the prevention of peripheral intravenous catheter failure in adults (SAVE): a pragmatic, randomised controlled, superiority trial

BACKGROUND

Two billion peripheral intravenous catheters (PIVCs) are used globally each year, but optimal dressing and securement methods are not well established. We aimed to compare the efficacy and costs of three alternative approaches to standard non-bordered polyurethane dressings.

METHODS

We did a pragmatic, randomised controlled, parallel-group superiority trial at two hospitals in Queensland, Australia. Eligible patients were aged 18 years or older and required PIVC insertion for clinical treatment, which was expected to be required for longer than 24 h. Patients were randomly assigned (1:1:1:1) via a centralised web-based randomisation service using random block sizes, stratified by hospital, to receive tissue adhesive with polyurethane dressing, bordered polyurethane dressing, a securement device with polyurethane dressing, or polyurethane dressing (control). Randomisation was concealed before allocation. Patients, clinicians, and research staff were not masked because of the nature of the intervention, but infections were adjudicated by a physician who was masked to treatment allocation. The primary outcome was all-cause PIVC failure (as a composite of complete dislodgement, occlusion, phlebitis, and infection [primary bloodstream infection or local infection]). Analysis was by modified intention to treat. This trial is registered with the Australian New Zealand Clinical Trials Registry, number ACTRN12611000769987.

FINDINGS

Between March 18, 2013, and Sept 9, 2014, we randomly assigned 1807 patients to receive tissue adhesive with polyurethane (n=446), bordered polyurethane (n=454), securement device with polyurethane (n=453), or polyurethane (n=454); 1697 patients comprised the modified intention-to-treat population. 163 (38%) of 427 patients in the tissue adhesive with polyurethane group (absolute risk difference -4·5% [95% CI -11·1 to 2·1%], p=0·19), 169 (40%) of 423 of patients in the bordered polyurethane group (-2·7% [-9·3 to 3·9%] p=0·44), 176 (41%) of 425 patients in the securement device with poplyurethane group (-1·2% [-7·9% to 5·4%], p=0·73), and 180 (43%) of 422 patients in the polyurethane group had PIVC failure. 17 patients in the tissue adhesive with polyurethane group, two patients in the bordered polyurethane group, eight patients in the securement device with polyurethane group, and seven patients in the polyurethane group had skin adverse events. Total costs of the trial interventions did not differ significantly between groups.

INTERPRETATION

Current dressing and securement methods are commonly associated with PIVC failure and poor durability, with simultaneous use of multiple products commonly required. Cost is currently the main factor that determines product choice. Innovations to achieve effective, durable dressings and securements, and randomised controlled trials assessing their effectiveness are urgently needed.

FUNDING

Australian National Health and Medical Research Council.

Evaluation of biases present in the cohort multiple randomised controlled trial design: a simulation study

Background

The cohort multiple randomised controlled trial (cmRCT) design provides an opportunity to incorporate the benefits of randomisation within clinical practice; thus reducing costs, integrating electronic healthcare records, and improving external validity. This study aims to address a key concern of the cmRCT design: refusal to treatment is only present in the intervention arm, and this may lead to bias and reduce statistical power.

Methods

We used simulation studies to assess the effect of this refusal, both random and related to event risk, on bias of the effect estimator and statistical power. A series of simulations were undertaken that represent a cmRCT trial with time-to-event endpoint. Intention-to-treat (ITT), per protocol (PP), and instrumental variable (IV) analysis methods, two stage predictor substitution and two stage residual inclusion, were compared for various refusal scenarios.

Results

We found the IV methods provide a less biased estimator for the causal effect when refusal is present in the intervention arm, with the two stage residual inclusion method performing best with regards to minimum bias and sufficient power. We demonstrate that sample sizes should be adapted based on expected and actual refusal rates in order to be sufficiently powered for IV analysis.

Conclusion

We recommend running both an IV and ITT analyses in an individually randomised cmRCT as it is expected that the effect size of interest, or the effect we would observe in clinical practice, would lie somewhere between that estimated with ITT and IV analyses. The optimum (in terms of bias and power) instrumental variable method was the two stage residual inclusion method. We recommend using adaptive power calculations, updating them as refusal rates are collected in the trial recruitment phase in order to be sufficiently powered for IV analysis.