A roadmap to using randomization in clinical trials

Abstract Background Randomization is the foundation of any clinical trial involving treatment comparison. It helps mitigate selection bias, promotes similarity of treatment groups with respect to important known and unknown confounders, and contributes to the validity of statistical tests. Various restricted randomization procedures with different probabilistic structures and different statistical properties are available. The goal of this paper is to present a systematic roadmap for the choice and application of a restricted randomization procedure in a clinical trial. Methods We survey available restricted randomization procedures for sequential allocation of subjects in a randomized, comparative, parallel group clinical trial with equal (1:1) allocation. We explore statistical properties of these procedures, including balance/randomness tradeoff, type I error rate and power. We perform head-to-head comparisons of different procedures through simulation under various experimental scenarios, including cases when common model assumptions are violated. We also provide some real-life clinical trial examples to illustrate the thinking process for selecting a randomization procedure for implementation in practice. Results Restricted randomization procedures targeting 1:1 allocation vary in the degree of balance/randomness they induce, and more importantly, they vary in terms of validity and efficiency of statistical inference when common model assumptions are violated (e.g. when outcomes are affected by a linear time trend; measurement error distribution is misspecified; or selection bias is introduced in the experiment). Some procedures are more robust than others. Covariate-adjusted analysis may be essential to ensure validity of the results. Special considerations are required when selecting a randomization procedure for a clinical trial with very small sample size. Conclusions The choice of randomization design, data analytic technique (parametric or nonparametric), and analysis strategy (randomization-based or population model-based) are all very important considerations. Randomization-based tests are robust and valid alternatives to likelihood-based tests and should be considered more frequently by clinical investigators.

Download Full-text

A roadmap to using randomization in clinical trials

10.21203/rs.3.rs-135735/v1 ◽

2021 ◽

Author(s):

Vance Berger ◽

Louis Bour ◽

Kerstine Carter ◽

Jonathan Chipman ◽

Colin Everett ◽

...

Keyword(s):

Clinical Trial ◽

Selection Bias ◽

Linear Time ◽

Statistical Tests ◽

Real Life ◽

Small Sample ◽

Statistical Properties ◽

Type I ◽

Restricted Randomization ◽

Treatment Comparison

Abstract Background: Randomization is the foundation of any clinical trial involving treatment comparison. It helps mitigate selection bias, promotes similarity of treatment groups with respect to important known and unknown confounders, and contributes to the validity of statistical tests. Various restricted randomization procedures with different probabilistic structures and different statistical properties are available.Methods: We survey available restricted randomization procedures for sequential allocation of subjects in a randomized, comparative, parallel group clinical trial with equal (1:1) allocation. We explore statistical properties of these procedures, including balance/randomness tradeoff, type I error rate and power. We perform head-to-head comparisons of different procedures through simulation under various experimental scenarios, including cases when common model assumptions are violated. We also provide some real-life clinical trial examples to illustrate the thinking process for selecting a randomization procedure for implementation in practice. Results: Restricted randomization procedures targeting 1:1 allocation vary in the degree of balance/randomness they induce, and more importantly, they vary in terms of validity and efficiency of statistical inference when common model assumptions are violated (e.g. when outcomes are affected by a linear time trend; measurement error distribution is misspecified; or selection bias is introduced in the experiment). Some procedures are more robust than others. Covariate-adjusted analysis may be essential to ensure validity of the results. Special considerations are required when selecting a randomization procedure for a clinical trial with very small sample size.Conclusions: The choice of randomization design, data analytic technique (parametric or nonparametric), and analysis strategy (randomization-based or population model-based) are all very important considerations. Randomization-based tests are robust and valid alternatives to likelihood-based tests and should be considered more frequently by clinical investigators.

Download Full-text

Improved Small Sample Inference Methods for a Mixed-Effects Model for Repeated Measures Approach in Incomplete Longitudinal Data Analysis

Stats ◽

10.3390/stats2020013 ◽

2019 ◽

Vol 2 (2) ◽

pp. 174-188

Author(s):

Yoshifumi Ukyo ◽

Hisashi Noma ◽

Kazushi Maruo ◽

Masahiko Gosho

Keyword(s):

Clinical Trial ◽

Repeated Measures ◽

Type I Error ◽

Mixed Effects ◽

Error Rates ◽

Small Sample ◽

Mixed Effects Model ◽

Model Complexity ◽

Type I ◽

Inference Methods

The mixed-effects model for repeated measures (MMRM) approach has been widely applied for longitudinal clinical trials. Many of the standard inference methods of MMRM could possibly lead to the inflation of type I error rates for the tests of treatment effect, when the longitudinal dataset is small and involves missing measurements. We propose two improved inference methods for the MMRM analyses, (1) the Bartlett correction with the adjustment term approximated by bootstrap, and (2) the Monte Carlo test using an estimated null distribution by bootstrap. These methods can be implemented regardless of model complexity and missing patterns via a unified computational framework. Through simulation studies, the proposed methods maintain the type I error rate properly, even for small and incomplete longitudinal clinical trial settings. Applications to a postnatal depression clinical trial are also presented.

Download Full-text

Randomization

Critical Thinking in Clinical Research ◽

10.1093/med/9780199324491.003.0005 ◽

2018 ◽

pp. 87-104

Author(s):

Juliana C Ferreira ◽

Ben M. W. Illigens ◽

Felipe Fregni

Keyword(s):

Clinical Trial ◽

Clinical Trials ◽

Randomized Clinical Trial ◽

Selection Bias ◽

Allocation Concealment ◽

Small Sample ◽

Adaptive Randomization ◽

Entire Process ◽

Advantages And Disadvantages ◽

Small Sample Sizes

Chapter 5 gives the reader an overview of a major important feature in every randomized clinical trial: the process of randomization. This chapter describes the main features of randomization, its importance, advantages, and disadvantages. It also discusses the most common methods of randomization (simple randomization, blocked randomization, stratified randomization, adaptive randomization), as well as what the investigator should take into consideration when choosing among these options. In this scenario, the challenges of defining a method in clinical trials with small sample sizes are also discussed. Additionally, the chapter explores the consequences that may arise from lack of randomization, such as selection bias. It also focuses on defining allocation concealment and its importance to the appropriate conduction of a study. In this chapter, the reader is taken through the entire process—from choosing a suitable randomization option to ensuring the appropriate implementation of the selected method.

Download Full-text

Clinical Trial Design and Statistics

10.2310/vasc.2189 ◽

2016 ◽

Author(s):

Julie Ann Sosa

Keyword(s):

Clinical Trial ◽

Clinical Trials ◽

Sample Size ◽

Statistical Tests ◽

Critical Role ◽

Clinical Trial Design ◽

Type I ◽

Comparable Group ◽

Power Strategies ◽

Clinical Trial Results

Clinical trials are planned experiments that measure the effectiveness of an intervention by comparing outcomes in a group of subjects treated with the test intervention with those observed in a comparable group of subjects receiving another intervention. As a result, clinical trials require investigators to assume significant responsibility and often consume significant human and financial resources. This chapter describes implementation of the study, including formulating the protocol from the study question; inclusion and exclusion criteria; pretesting and quality control; missing data; standardization, blinding, and randomization; planning for data management, monitoring, and audits; interim analyses; and the components of the research team. Because biostatistics plays a critical role in the interpretation of clinical trial results, this chapter discusses hypotheses and underlying principles, Type I and II errors, estimating sample size and power, strategies for minimizing sample size, statistical tests of significant, and relative risk. This review contains 32 references.

Download Full-text

Statistical properties of type I intermittency

Journal de Physique ◽

10.1051/jphys:01982004304058500 ◽

1982 ◽

Vol 43 (4) ◽

pp. 585-589 ◽

Cited By ~ 12

Author(s):

M. N. Bussac ◽

C. Meunier

Keyword(s):

Statistical Properties ◽

Type I

Download Full-text

An influence of social isolation on the level of physical activity as well as on well-being and mental state of people during the coronavirus COVID-19 pandemic

Polish Journal of Public Health ◽

10.2478/pjph-2019-0029 ◽

2019 ◽

Vol 129 (4) ◽

pp. 127-131

Author(s):

Agnieszka Parfin ◽

Krystian Wdowiak ◽

Marzena Furtak-Niczyporuk ◽

Jolanta Herda

Keyword(s):

Physical Activity ◽

Social Isolation ◽

Small Sample Size ◽

Statistical Tests ◽

Population Group ◽

Well Being ◽

Small Sample ◽

Subjective Well Being ◽

Tract Infections ◽

The Impact

AbstractIntroduction. The COVID-19 is the name of an infectious disease caused by a new strain of coronavirus SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus 2). It was first diagnosed in December 2019 in patients in Wuhan City, Hubei Province, China. The symptoms are dominated by features of respiratory tract infections, in some patients with a very severe course leading to respiratory failure and, in extreme cases to death. Due to the spread of the infection worldwide, the WHO declared a pandemic in March 2020.Aim. An investigation of the impact of social isolation introduced due to the coronavirus pandemic on selected aspects of life. The researchers focused on observing changes in habits related to physical activity and their connections with people’s subjective well-being and emotional state.Material and methods. The study was carried out within the international project of the group „IRG on COVID and exercise”. The research tool was a standardized questionnaire.Results. Based on the data collected and the analysis of the percentage results, it can be observed that the overwhelming majority of people taking up physical activity reported a better mood during the pandemic. However, statistical tests do not confirm these relationships due to the small sample size.Conclusions. Isolation favours physical activity. Future, in-depth studies, by enlarging the population group, are necessary to confirm the above observations.

Download Full-text

Lead-Free Metal Halide Perovskite Nanocrystals for Photocatalysis in Water

10.26434/chemrxiv.9794270 ◽

2019 ◽

Author(s):

Subhajit Bhattacharjee ◽

Sonu Pratap Chaudhary ◽

Sayan Bhattacharyya

Keyword(s):

Charge Carrier ◽

Wide Spectrum ◽

Real Life ◽

Metal Halide ◽

Lead Free ◽

Water Medium ◽

Type I ◽

Type Ii ◽

Perovskite Layer ◽

Halide Perovskites

Metal halide perovskites with high absorption coefficient, direct generation of free charge carriers, excellent ambipolar charge carrier transport properties, point-defect tolerance, compositional versatility and solution processability are potentially transforming the photovoltaics and optoelectronics industries. However their limited ambient stability, particularly those of iodide perovskites, obscures their use as photocatalysts especially in aqueous medium. In an unprecedented approach we have exploited the photo-absorption property of the less toxic lead-free Cs3Bi2X9 (X = Br, I) nanocrystals (NCs) to catalyse the degradation of water pollutant organic dye, methylene blue (MB) in presence of visible light at room temperature. After providing a proof-of-concept with bromide perovskites in isopropanol, the perovskites are employed as photocatalysts in water medium by designing perovskite/Ag2S and perovskite/TiO2 composite systems, with Type I (or quasi Type II) and Type II alignments, respectively. Ag2S and TiO2 coatings decelerate penetration of water into the perovskite layer while facilitating charge carrier extraction. With a minimal NC loading, Cs3Bi2I9/Ag2S degrades ~90% MB within an hour. Our approach has the potential to unravel the photocatalytic properties of metal halide perovskites for a wide spectrum of real-life applications.

Download Full-text

Correction: “Influence of Selection Bias on the Test Decision – A Simulation Study”

Methods of Information in Medicine ◽

10.3414/me11-01-0043e ◽

2014 ◽

Vol 53 (05) ◽

pp. 343-343

Keyword(s):

Selection Bias ◽

Simulation Study ◽

Error Rate ◽

Type I Error ◽

Block Size ◽

Error Rates ◽

Type I ◽

Type I Error Rate ◽

Representation Error ◽

Numeric Representation

We have to report marginal changes in the empirical type I error rates for the cut-offs 2/3 and 4/7 of Table 4, Table 5 and Table 6 of the paper “Influence of Selection Bias on the Test Decision – A Simulation Study” by M. Tamm, E. Cramer, L. N. Kennes, N. Heussen (Methods Inf Med 2012; 51: 138 –143). In a small number of cases the kind of representation of numeric values in SAS has resulted in wrong categorization due to a numeric representation error of differences. We corrected the simulation by using the round function of SAS in the calculation process with the same seeds as before. For Table 4 the value for the cut-off 2/3 changes from 0.180323 to 0.153494. For Table 5 the value for the cut-off 4/7 changes from 0.144729 to 0.139626 and the value for the cut-off 2/3 changes from 0.114885 to 0.101773. For Table 6 the value for the cut-off 4/7 changes from 0.125528 to 0.122144 and the value for the cut-off 2/3 changes from 0.099488 to 0.090828. The sentence on p. 141 “E.g. for block size 4 and q = 2/3 the type I error rate is 18% (Table 4).” has to be replaced by “E.g. for block size 4 and q = 2/3 the type I error rate is 15.3% (Table 4).”. There were only minor changes smaller than 0.03. These changes do not affect the interpretation of the results or our recommendations.

Download Full-text

FRI0612-HPR EFFECT OF THERMOFORMABLE ORTHOSES ON FOOT FUNCTION IN RHEUMATOID ARTHRITIS PATIENTS: PRELIMINARY RESULTS FROM AN OPEN CLINICAL TRIAL.

Annals of the Rheumatic Diseases ◽

10.1136/annrheumdis-2020-eular.6308 ◽

2020 ◽

Vol 79 (Suppl 1) ◽

pp. 911.2-912

Author(s):

I. Aachari ◽

H. Rkain ◽

F. Safaa ◽

L. Benzakour ◽

T. Latifa ◽

...

Keyword(s):

Rheumatoid Arthritis ◽

Clinical Trial ◽

Small Sample ◽

Activity Limitation ◽

Foot Orthoses ◽

Functional Impact ◽

Pain Disability ◽

Before And After ◽

Foot Function ◽

The Impact

Background:Orthoses and footwear can play an important role in managing foot pathology in patients whose systemic disease is controlled. Foot orthoses are frequently prescribed in clinical practice as an intervention for people with rheumatoid arthritis (RA).Objectives:The aim of our study is to evaluate the impact of thermoformable orthoses on the functional index of the foot (FFI) in patients with rheumatoid arthritis.Methods:We conducted an open clinical trial, having consecutively included 14 patients (85.7% female, average age 54.8 ± 10 years) suffering from rheumatoid arthritis (median progression time of 9 years [5 - 12]). The average DAS28 was 2.7 ± 1.2 and the functional impact objectified by the Health Assessment Questionnaire (HAQ) was on average 0.9 ± 0.7.The median deadline from the start of RA and the onset of the foot problem was 3 years [0 – 7,75]. The foot problem was bilateral in 100% of the cases and inaugural in 85.7% of the cases.We evaluated the functional impact of foot injury for all our patients at baseline and 8 weeks after the use of thermoformable orthoses, based on the FFI (Foot function Index) measuring the impact of foot pathology on function in terms of pain, disability and activity limitation.The comparison of the FFI domains before and after the use of orthoses was carried out using parametric or nonparametric paired tests using The SPSS statistical software.Results:With the use of foot orthoses, FFI values decreased in all subscales (p=0,024) (pain, disability and activity limitation). This reduction was significant for disability (0,011) but not for pain and activity limitation.There were no significant correlations between the global FFI and the progression of RA, the duration of foot damage and the functional impact measured by the HAQ.Table 1. The comparison of the FFI domains before and after the use of orthoses.psignificatif if< 0,05; Test used: Non-parametric test for two linked samples.Conclusion:Foot orthoses were effective as an adjuvant in the management of rheumatoid foot. They significantly reduced disability as measured by the FFI. The absence of factors associated with pain and limitation of activity could possibly be related to the small sample size.Disclosure of Interests:None declared

Download Full-text

Plasma interferon-alpha is associated with double-positivity for autoantibodies but is not a predictor of remission in early rheumatoid arthritis—a spin-off study of the NORD-STAR randomized clinical trial

Arthritis Research & Therapy ◽

10.1186/s13075-021-02556-1 ◽

2021 ◽

Vol 23 (1) ◽

Author(s):

Marit Stockfelt ◽

Anna-Carin Lundell ◽

Merete Lund Hetland ◽

Mikkel Østergaard ◽

Till Uhlig ◽

...

Keyword(s):

Rheumatoid Arthritis ◽

Clinical Trial ◽

Disease Activity ◽

Randomized Clinical Trial ◽

Early Rheumatoid Arthritis ◽

Treatment Initiation ◽

Certolizumab Pegol ◽

Type I ◽

Protein Levels ◽

Early Ra

Abstract Background The type I interferon (IFN) gene signature is present in a subgroup of patients with early rheumatoid arthritis (RA). Protein levels of IFNα have not been measured in RA and it is unknown whether they associate with clinical characteristics or treatment effect. Methods Patients with early untreated RA (n = 347) were randomized to methotrexate combined with prednisone, certolizumab-pegol, abatacept, or tocilizumab. Plasma IFNα protein levels were determined by single molecular array (Simoa) before and 24 weeks after treatment initiation and were related to demographic and clinical factors including clinical disease activity index, disease activity score in 28 joints, swollen and tender joint counts, and patient global assessment. Results IFNα protein positivity was found in 26% of the patients, and of these, 92% were double-positive for rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA). IFNα protein levels were reduced 24 weeks after treatment initiation, and the absolute change was similar irrespective of treatment. IFNα protein positivity was associated neither with disease activity nor with achievement of CDAI remission 24 weeks after randomization. Conclusion IFNα protein positivity is present in a subgroup of patients with early RA and associates with double-positivity for autoantibodies but not with disease activity. Pre-treatment IFNα positivity did not predict remission in any of the treatment arms, suggesting that the IFNα system is distinct from the pathways of TNF, IL-6, and T-cell activation in early RA. A spin-off study of the NORD-STAR randomized clinical trial, NCT01491815 (ClinicalTrials), registered 12/08/2011, https://clinicaltrials.gov/ct2/show/NCT01491815.

Download Full-text