A Short History of Bernard Fisher’s Contributions to Randomized Clinical Trials

Dr. Bernard Fisher (1918-2019) was an early proponent of evidence-based medicine using the mechanism of prospective, multicenter, randomized clinical trials to test biological and clinical hypotheses. In this article, I trace how his early scientific work in striving to understand the nature of cancer metastasis through animal experiments led to a new, testable, clinical hypothesis: that surgery to remove only the tumor and a small amount of tissue around it was as effective as the more disfiguring operations that were then the standard treatment. Fisher’s work with the National Surgical Adjuvant Breast and Bowel Project (NSABP) using large, randomized clinical trials to demonstrate the veracity of this hypothesis led to a new paradigm in which the emphasis was placed on how systemic therapies used at an early stage of disease could effectively eradicate breast cancer for many patients. This new therapeutic approach led to the successful development of new treatments, many of which are widely used today. Ultimately, the new paradigm led to successfully preventing breast cancer in women who were at high risk for the disease but who had not yet been diagnosed with the disease. Throughout his entire career, Fisher championed the use of large prospective, randomized clinical trials despite criticism from many in the medical community who strongly criticized his use of randomization as a mechanism for testing clinical hypotheses. The approach he and the NSABP employed is still considered to be the highest standard of evidence in conducting clinical studies.

The representation of Indigenous peoples in chronic disease clinical trials in Australia, Canada, New Zealand, and the United States

Background Indigenous peoples are overrepresented with chronic health conditions and experience suboptimal outcomes compared with non-Indigenous peoples. Genetic variations influence therapeutic responses, thus there are potential risks and harm when extrapolating evidence from the general population to Indigenous peoples. Indigenous population–specific clinical studies, and inclusion of Indigenous peoples in general population clinical trials, are perceived to be rare. Our study (1) identified and characterized Indigenous population–specific chronic disease trials and (2) identified the representation of Indigenous peoples in general population chronic disease trials conducted in Australia, Canada, New Zealand, and the United States. Methods For Objective 1, publicly available clinical trial registries were searched from May 2010 to May 2020 using Indigenous population–specific terms and included for data extraction if in pre-specified chronic disease. For identified trials, we extracted Indigenous population group identity and characteristics, type of intervention, and funding type. For Objective 2, a random selection of 10% of registered clinical trials was performed and the proportion of Indigenous population participants enrolled extracted. Results In total, 170 Indigenous population–specific chronic disease trials were identified. The clinical trials were predominantly behavioral interventions (n = 95). Among general population studies, 830 studies were randomly selected. When race was reported in studies (n = 526), Indigenous individuals were enrolled in 172 studies and constituted 5.6% of the total population enrolled in those studies. Conclusion Clinical trials addressing chronic disease conditions in Indigenous populations are limited. It is crucial to ensure adequate representation of Indigenous peoples in clinical trials to ensure trial data are applicable to their clinical care.

Marginal modeling in community randomized trials with rare events: Utilization of the negative binomial regression model

Background/aims This work is motivated by the HEALing Communities Study, which is a post-test only cluster randomized trial in which communities are randomized to two different trial arms. The primary interest is in reducing opioid overdose fatalities, which will be collected as a count outcome at the community level. Communities range in size from thousands to over one million residents, and fatalities are expected to be rare. Traditional marginal modeling approaches in the cluster randomized trial literature include the use of generalized estimating equations with an exchangeable correlation structure when utilizing subject-level data, or analogously quasi-likelihood based on an over-dispersed binomial variance when utilizing community-level data. These approaches account for and estimate the intra-cluster correlation coefficient, which should be provided in the results from a cluster randomized trial. Alternatively, the coefficient of variation or R coefficient could be reported. In this article, we show that negative binomial regression can also be utilized when communities are large and events are rare. The objectives of this article are (1) to show that the negative binomial regression approach targets the same marginal regression parameter(s) as an over-dispersed binomial model and to explain why the estimates may differ; (2) to derive formulas relating the negative binomial overdispersion parameter k with the intra-cluster correlation coefficient, coefficient of variation, and R coefficient; and (3) analyze pre-intervention data from the HEALing Communities Study to demonstrate and contrast models and to show how to report the intra-cluster correlation coefficient, coefficient of variation, and R coefficient when utilizing negative binomial regression. Methods Negative binomial and over-dispersed binomial regression modeling are contrasted in terms of model setup, regression parameter estimation, and formulation of the overdispersion parameter. Three specific models are used to illustrate concepts and address the third objective. Results The negative binomial regression approach targets the same marginal regression parameter(s) as an over-dispersed binomial model, although estimates may differ. Practical differences arise in regard to how overdispersion, and hence the intra-cluster correlation coefficient is modeled. The negative binomial overdispersion parameter is approximately equal to the ratio of the intra-cluster correlation coefficient and marginal probability, the square of the coefficient of variation, and the R coefficient minus 1. As a result, estimates corresponding to all four of these different types of overdispersion parameterizations can be reported when utilizing negative binomial regression. Conclusion Negative binomial regression provides a valid, practical, alternative approach to the analysis of count data, and corresponding reporting of overdispersion parameters, from community randomized trials in which communities are large and events are rare.

Time trends with response-adaptive randomization: The inevitability of inefficiency

Response-adaptive randomization, which changes the randomization ratio as a randomized clinical trial progresses, is inefficient as compared to a fixed 1:1 randomization ratio in terms of increased required sample size. It is also known that response-adaptive randomization leads to biased treatment effects if there are time trends in the accruing outcome data, for example, due to changes in the patient population being accrued, evaluation methods, or concomitant treatments. Response-adaptive-randomization analysis methods that account for potential time trends, such as time-block stratification or re-randomization, can eliminate this bias. However, as shown in this Commentary, these analysis methods cause a large additional inefficiency of response-adaptive randomization, regardless of whether a time trend actually exists.

Influential methods reports for group-randomized trials and related designs

Background. This article identifies the most influential methods reports for group-randomized trials and related designs published through 2020. Many interventions are delivered to participants in real or virtual groups or in groups defined by a shared interventionist so that there is an expectation for positive correlation among observations taken on participants in the same group. These interventions are typically evaluated using a group- or cluster-randomized trial, an individually randomized group treatment trial, or a stepped wedge group- or cluster-randomized trial. These trials face methodological issues beyond those encountered in the more familiar individually randomized controlled trial. Methods. PubMed was searched to identify candidate methods reports; that search was supplemented by reports known to the author. Candidate reports were reviewed by the author to include only those focused on the designs of interest. Citation counts and the relative citation ratio, a new bibliometric tool developed at the National Institutes of Health, were used to identify influential reports. The relative citation ratio measures influence at the article level by comparing the citation rate of the reference article to the citation rates of the articles cited by other articles that also cite the reference article. Results. In total, 1043 reports were identified that were published through 2020. However, 55 were deemed to be the most influential based on their relative citation ratio or their citation count using criteria specific to each of the three designs, with 32 group-randomized trial reports, 7 individually randomized group treatment trial reports, and 16 stepped wedge group-randomized trial reports. Many of the influential reports were early publications that drew attention to the issues that distinguish these designs from the more familiar individually randomized controlled trial. Others were textbooks that covered a wide range of issues for these designs. Others were “first reports” on analytic methods appropriate for a specific type of data (e.g. binary data, ordinal data), for features commonly encountered in these studies (e.g. unequal cluster size, attrition), or for important variations in study design (e.g. repeated measures, cohort versus cross-section). Many presented methods for sample size calculations. Others described how these designs could be applied to a new area (e.g. dissemination and implementation research). Among the reports with the highest relative citation ratios were the CONSORT statements for each design. Conclusions. Collectively, the influential reports address topics of great interest to investigators who might consider using one of these designs and need guidance on selecting the most appropriate design for their research question and on the best methods for design, analysis, and sample size.

A survey of patients with laryngotracheal stenosis on future clinical trial design

Background/Aims Laryngotracheal stenosis is a rare but devastating proximal airway fibrosis that restricts a patient’s ability to breathe. Treatment is primarily surgical and to date, there has never been a multi-institutional, randomized, prospective, and interventional clinical trial for a medical therapy to treat laryngotracheal stenosis. Therefore, we aimed to obtain patient feedback to guide successful trial design, recruitment, retention, and for identifying potential barriers to study participation. Methods Over 1000 members of an international laryngotracheal stenosis online support community (the Living with Idiopathic Subglottic Stenosis Facebook group) were sent two questionnaires for a proposed interventional double-blinded, randomized, placebo-controlled clinical trial. Results A total of 317 and 558 participants responded to the first and second surveys, respectively. The majority of participants (77%) were willing to consider enrollment, regardless of having a 50% chance of receiving placebo versus treatment (78%). The majority (84%) of participants were willing to travel 200 miles to participate for up to six in-person visits over 50 days. Specific side effects, including anemia/thrombocytopenia (72%) or risk of infection (69.3%) had the greatest impact on clinical trial participation with other side effects (peripheral edema (53%), oral ulcers (51%), and gastrointestinal side effects (41%)) having less impact. Conclusion Patients with laryngotracheal stenosis possess nuanced insight into their disease and treatment options. As a group, they are extremely motivated for better therapies. Future laryngotracheal stenosis clinical trials should focus on providing excellent side effect -related education and utilizing feedback from online advocacy groups to optimize recruitment and retention.

Contrasting approaches for addressing non-adherence in randomized controlled trials: An illustration from the REFLUX trial

Background: In many randomized controlled trials, a substantial proportion of patients do not comply with the treatment protocol to which they have been randomly assigned. Randomized controlled trials are required to report results according to the intention-to-treat estimand, but recent methodological guidance recognizes the importance of estimating other causal quantities. Methods: This article outlines an analytical framework for randomized controlled trials with non-compliance. We apply the ICH E9 (R1) addendum and combine it with the potential outcomes framework to define key estimands, outline the major assumptions for identification of each estimand, and highlight the assumptions that cannot be verified from the randomized controlled trial data. We contrast the assumptions and estimates in a re-analysis of the REFLUX trial. We report alternative estimates for the effectiveness of receipt of laparoscopic surgery versus medical management for patients with gastro-intestinal reflux disease. Results: The article finds that adjusted as-treated and per-protocol estimates were similar in magnitude to those based intention-to-treat methods. Instrumental variable estimates of the complier average causal effect were larger, with wider confidence intervals. Conclusion: We recommend that in randomized controlled trials with non-compliance, studies should outline which estimand is most relevant to the study context, evaluate key assumptions, and present estimates from a range of methods as a sensitivity analysis.

Clarifying selection bias in cluster randomized trials

Background In cluster randomized trials, patients are typically recruited after clusters are randomized, and the recruiters and patients may not be blinded to the assignment. This often leads to differential recruitment and consequently systematic differences in baseline characteristics of the recruited patients between intervention and control arms, inducing post-randomization selection bias. We aim to rigorously define causal estimands in the presence of selection bias. We elucidate the conditions under which standard covariate adjustment methods can validly estimate these estimands. We further discuss the additional data and assumptions necessary for estimating causal effects when such conditions are not met. Methods Adopting the principal stratification framework in causal inference, we clarify there are two average treatment effect (ATE) estimands in cluster randomized trials: one for the overall population and one for the recruited population. We derive analytical formula of the two estimands in terms of principal-stratum-specific causal effects. Furthermore, using simulation studies, we assess the empirical performance of the multivariable regression adjustment method under different data generating processes leading to selection bias. Results When treatment effects are heterogeneous across principal strata, the average treatment effect on the overall population generally differs from the average treatment effect on the recruited population. A naïve intention-to-treat analysis of the recruited sample leads to biased estimates of both average treatment effects. In the presence of post-randomization selection and without additional data on the non-recruited subjects, the average treatment effect on the recruited population is estimable only when the treatment effects are homogeneous between principal strata, and the average treatment effect on the overall population is generally not estimable. The extent to which covariate adjustment can remove selection bias depends on the degree of effect heterogeneity across principal strata. Conclusion There is a need and opportunity to improve the analysis of cluster randomized trials that are subject to post-randomization selection bias. For studies prone to selection bias, it is important to explicitly specify the target population that the causal estimands are defined on and adopt design and estimation strategies accordingly. To draw valid inferences about treatment effects, investigators should (1) assess the possibility of heterogeneous treatment effects, and (2) consider collecting data on covariates that are predictive of the recruitment process, and on the non-recruited population from external sources such as electronic health records.

Power and sample size calculations for cluster randomized trials with binary outcomes when intracluster correlation coefficients vary by treatment arm

Background/Aims Generalized estimating equations are commonly used to fit logistic regression models to clustered binary data from cluster randomized trials. A commonly used correlation structure assumes that the intracluster correlation coefficient does not vary by treatment arm or other covariates, but the consequences of this assumption are understudied. We aim to evaluate the effect of allowing variation of the intracluster correlation coefficient by treatment or other covariates on the efficiency of analysis and show how to account for such variation in sample size calculations. Methods We develop formulae for the asymptotic variance of the estimated difference in outcome between treatment arms obtained when the true exchangeable correlation structure depends on the treatment arm and the working correlation structure used in the generalized estimating equations analysis is: (i) correctly specified, (ii) independent, or (iii) exchangeable with no dependence on treatment arm. These formulae require a known distribution of cluster sizes; we also develop simplifications for the case when cluster sizes do not vary and approximations that can be used when the first two moments of the cluster size distribution are known. We then extend the results to settings with adjustment for a second binary cluster-level covariate. We provide formulae to calculate the required sample size for cluster randomized trials using these variances. Results We show that the asymptotic variance of the estimated difference in outcome between treatment arms using these three working correlation structures is the same if all clusters have the same size, and this asymptotic variance is approximately the same when intracluster correlation coefficient values are small. We illustrate these results using data from a recent cluster randomized trial for infectious disease prevention in which the clusters are groups of households and modest in size (mean 9.6 individuals), with intracluster correlation coefficient values of 0.078 in the control arm and 0.057 in an intervention arm. In this application, we found a negligible difference between the variances calculated using structures (i) and (iii) and only a small increase (typically [Formula: see text]) for the independent correlation structure (ii), and hence minimal effect on power or sample size requirements. The impact may be larger in other applications if there is greater variation in the ICC between treatment arms or with an additional covariate. Conclusion The common approach of fitting generalized estimating equations with an exchangeable working correlation structure with a common intracluster correlation coefficient across arms likely does not substantially reduce the power or efficiency of the analysis in the setting of a large number of small or modest-sized clusters, even if the intracluster correlation coefficient varies by treatment arm. Our formulae, however, allow formal evaluation of this and may identify situations in which variation in intracluster correlation coefficient by treatment arm or another binary covariate may have a more substantial impact on power and hence sample size requirements.