Understanding the cluster randomised crossover design: a graphical illustraton of the components of variation and a sample size tutorial

Optimal designs using generalized estimating equations in cluster randomized crossover and stepped wedge trials

Cluster randomized crossover and stepped wedge cluster randomized trials are two types of longitudinal cluster randomized trials that leverage both the within- and between-cluster comparisons to estimate the treatment effect and are increasingly used in healthcare delivery and implementation science research. While the variance expressions of estimated treatment effect have been previously developed from the method of generalized estimating equations for analyzing cluster randomized crossover trials and stepped wedge cluster randomized trials, little guidance has been provided for optimal designs to ensure maximum efficiency. Here, an optimal design refers to the combination of optimal cluster-period size and optimal number of clusters that provide the smallest variance of the treatment effect estimator or maximum efficiency under a fixed total budget. In this work, we develop optimal designs for multiple-period cluster randomized crossover trials and stepped wedge cluster randomized trials with continuous outcomes, including both closed-cohort and repeated cross-sectional sampling schemes. Local optimal design algorithms are proposed when the correlation parameters in the working correlation structure are known. MaxiMin optimal design algorithms are proposed when the exact values are unavailable, but investigators may specify a range of correlation values. The closed-form formulae of local optimal design and MaxiMin optimal design are derived for multiple-period cluster randomized crossover trials, where the cluster-period size and number of clusters are decimal. The decimal estimates from closed-form formulae can then be used to investigate the performances of integer estimates from local optimal design and MaxiMin optimal design algorithms. One unique contribution from this work, compared to the previous optimal design research, is that we adopt constrained optimization techniques to obtain integer estimates under the MaxiMin optimal design. To assist practical implementation, we also develop four SAS macros to find local optimal designs and MaxiMin optimal designs.

Challenges in Evaluating a Community-Level Intervention to Address Root Causes of Youth Violence

Violence disproportionately impacts Black American youth, representing a major health disparity. Addressing the possible root causes of structural inequities to reduce violence may increase the impact of prevention strategies. However, efforts to evaluate the impact of such interventions pose numerous methodological challenges, particularly around selecting an effective evaluation design to detect change at the community level, with adequate power and sampling, and appropriate constructs and measurement strategies. We propose a multiple baseline experimental design to evaluate the impact of a community-level youth violence and suicidality prevention strategy. A multiple baseline experimental design with multiple community units balances the need for scientific rigor with practical and values-based considerations. It includes randomization and plausible counterfactuals without requiring large samples or placing some communities in the position of not receiving the intervention. Considerations related to the conceptualization of the logic model, mechanisms of change, and health disparity outcomes informed the development of the measurement strategy. The strengths and weaknesses of a multiple baseline experimental design are discussed in comparison to versions of randomized clinical trials. Future health disparity intervention evaluation research will benefit from (1) building a shared sense of urgent public need to promote health; (2) respecting the validity of values- and partnership-based decision-making; and (3) promoting community-based and systems-level partnerships in scientific grant funding. The described study has been registered prospectively at clinicaltrials.gov, Protocol Record 21-454.

Sample size and power calculation for testing treatment effect heterogeneity in cluster randomized crossover designs

The cluster randomized crossover design has been proposed to improve efficiency over the traditional parallel-arm cluster randomized design. While statistical methods have been developed for designing cluster randomized crossover trials, they have exclusively focused on testing the overall average treatment effect, with little attention to differential treatment effects across subpopulations. Recently, interest has grown in understanding whether treatment effects may vary across pre-specified patient subpopulations, such as those defined by demographic or clinical characteristics. In this article, we consider the two-treatment two-period cluster randomized crossover design under either a cross-sectional or closed-cohort sampling scheme, where it is of interest to detect the heterogeneity of treatment effect via an interaction test. Assuming a patterned correlation structure for both the covariate and the outcome, we derive new sample size formulas for testing the heterogeneity of treatment effect with continuous outcomes based on linear mixed models. Our formulas also address unequal cluster sizes and therefore allow us to analytically assess the impact of unequal cluster sizes on the power of the interaction test in cluster randomized crossover designs. We conduct simulations to confirm the accuracy of the proposed methods, and illustrate their application in two real cluster randomized crossover trials.

Video Laryngoscopy vs Direct Laryngoscopy for Endotracheal Intubation in the Operating Room: A Cluster Randomized Clinical Trial

Importance

Endotracheal tubes are typically inserted in the operating room using direct laryngoscopy. Video laryngoscopy has been reported to improve airway visualization; however, whether improved visualization reduces intubation attempts in surgical patients is unclear.

Objective

To determine whether the number of intubation attempts per surgical procedure is lower when initial laryngoscopy is performed using video laryngoscopy or direct laryngoscopy.

Design, Setting, and Participants

Cluster randomized multiple crossover clinical trial conducted at a single US academic hospital. Patients were adults aged 18 years or older having elective or emergent cardiac, thoracic, or vascular surgical procedures who required single-lumen endotracheal intubation for general anesthesia. Patients were enrolled from March 30, 2021, to December 31, 2022. Data analysis was based on intention to treat.

Interventions

Two sets of 11 operating rooms were randomized on a 1-week basis to perform hyperangulated video laryngoscopy or direct laryngoscopy for the initial intubation attempt.

Main Outcomes and Measures

The primary outcome was the number of operating room intubation attempts per surgical procedure. Secondary outcomes were intubation failure, defined as the responsible clinician switching to an alternative laryngoscopy device for any reason at any time, or by more than 3 intubation attempts, and a composite of airway and dental injuries.

Results

Among 8429 surgical procedures in 7736 patients, the median patient age was 66 (IQR, 56-73) years, 35% (2950) were women, and 85% (7135) had elective surgical procedures. More than 1 intubation attempt was required in 77 of 4413 surgical procedures (1.7%) randomized to receive video laryngoscopy vs 306 of 4016 surgical procedures (7.6%) randomized to receive direct laryngoscopy, with an estimated proportional odds ratio for the number of intubation attempts of 0.20 (95% CI, 0.14-0.28; P < .001). Intubation failure occurred in 12 of 4413 surgical procedures (0.27%) using video laryngoscopy vs 161 of 4016 surgical procedures (4.0%) using direct laryngoscopy (relative risk, 0.06; 95% CI, 0.03-0.14; P < .001) with an unadjusted absolute risk difference of -3.7% (95% CI, -4.4% to -3.2%). Airway and dental injuries did not differ significantly between video laryngoscopy (41 injuries [0.93%]) vs direct laryngoscopy (42 injuries [1.1%]).

Conclusion and Relevance

In this study among adults having surgical procedures who required single-lumen endotracheal intubation for general anesthesia, hyperangulated video laryngoscopy decreased the number of attempts needed to achieve endotracheal intubation compared with direct laryngoscopy at a single academic medical center in the US. Results suggest that video laryngoscopy may be a preferable approach for intubating patients undergoing surgical procedures.

Trial Registration

ClinicalTrials.gov Identifier: NCT04701762.

Sepsis-associated acute kidney injury: recent advances in enrichment strategies, sub-phenotyping and clinical trials

Acute kidney injury (AKI) often complicates sepsis and is associated with high morbidity and mortality. In recent years, several important clinical trials have improved our understanding of sepsis-associated AKI (SA-AKI) and impacted clinical care. Advances in sub-phenotyping of sepsis and AKI and clinical trial design offer unprecedented opportunities to fill gaps in knowledge and generate better evidence for improving the outcome of critically ill patients with SA-AKI. In this manuscript, we review the recent literature of clinical trials in sepsis with focus on studies that explore SA-AKI as a primary or secondary outcome. We discuss lessons learned and potential opportunities to improve the design of clinical trials and generate actionable evidence in future research. We specifically discuss the role of enrichment strategies to target populations that are most likely to derive benefit and the importance of patient-centered clinical trial endpoints and appropriate trial designs with the aim to provide guidance in designing future trials.

Parental perspectives on a trial using waived informed consent at birth

OBJECTIVES

To determine parental perspectives in a trial with waived consent.

STUDY DESIGN

Anonymous survey of birth parents with term infants who were randomized using a waiver of consent, administered after infant discharge.

RESULTS

121 (11%) survey responses were collected. Of the 121 responding parents 111 (92%) reported that this form of consent was acceptable and 116 (96%) reported feeling comfortable having another child participate in a similar study. 110 (91%) respondents reported that they both understood the information provided in the consent process and had enough time to consider participation. Four percent had a negative opinion on the study's effect on their child's health.

CONCLUSIONS

Most responding parents reported both acceptability of this study design in the neonatal period and that the study had a positive effect on their child's health. Future work should investigate additional ways to involve parents and elicit feedback on varied methods of pediatric consent.

Practical considerations for sample size calculation for cluster randomized trials

Cluster randomized trials are an essential design in public health and medical research, when individual randomization is infeasible or undesirable for scientific or logistical reasons. However, the correlation among observations within clusters leads to a decrease in statistical power compared to an individually randomised trial with the same total sample size. This correlation - often quantified using the intra-cluster correlation coefficient - must be accounted for in the sample size calculation to ensure that the trial is adequately powered. In this paper, we first describe the principles of sample size calculation for parallel-arm CRTs, and explain how these calculations can be extended to CRTs with cross-over designs, with a baseline measurement and stepped-wedge designs. We introduce tools to guide researchers with their sample size calculation and discuss methods to inform the choice of the a priori estimate of the intra-cluster correlation coefficient for the calculation. We also include additional considerations with respect to anticipated attrition, a small number of clusters, and use of covariates in the randomisation process and in the analysis.

Effects of Cognitive Orientation to Daily Occupational Performance and Conductive Education Treatment Approaches on Fine Motor Skills, Activity and Participation Limitations in Children with Down Syndrome: A Randomised Controlled Trial

This study aiming to compare the effectiveness of Cognitive Orientation to Daily Occupational Performance (CO-OP) and Conductive Education (CE) approaches on motor skills, activity limitation and participation restrictions in children with Down Syndrome (DS). Twelwe children were randomly assigned into two groups. Twelve-week CO-OP or CE intervention (period-1) followed by a 12-week washout period. Same interventions were crossed over for another 12 weeks (period-2). The Performance Quality Rating Scale (PQRS), Canadian Occupational Performance Measure (COPM) and the Bruininks Oseretsky Test of Motor Proficiency Second Edition-Brief Form (BOT2-BF) were used for outcome measurements. CO-OP was effective in the improvement of task-specific activity performance, while both approaches have similar effects on the improvement of perceived performance, satisfaction, and motor skills performance.

National ICU Registries as Enablers of Clinical Research and Quality Improvement

OBJECTIVES

Clinical quality registries (CQRs) have been implemented worldwide by several medical specialties aiming to generate a better characterization of epidemiology, treatments, and outcomes of patients. National ICU registries were created almost 3 decades ago to improve the understanding of case-mix, resource use, and outcomes of critically ill patients. This narrative review describes the challenges, proposed solutions, and evidence generated by National ICU registries as facilitators for research and quality improvement.

DATA SOURCES

English language articles were identified in PubMed using phrases related to ICU registries, CQRs, outcomes, and case-mix.

STUDY SELECTION

Original research, review articles, letters, and commentaries, were considered.

DATA EXTRACTION

Data from relevant literature were identified, reviewed, and integrated into a concise narrative review.

DATA SYNTHESIS

CQRs have been implemented worldwide by several medical specialties aiming to generate a better characterization of epidemiology, treatments, and outcomes of patients. National ICU registries were created almost 3 decades ago to improve the understanding of case-mix, resource use, and outcomes of critically ill patients. The initial experience in European countries and in Oceania ensured that through locally generated data, ICUs could assess their performances by using risk-adjusted measures and compare their results through fair and validated benchmarking metrics with other ICUs contributing to the CQR. The accomplishment of these initiatives, coupled with the increasing adoption of information technology, resulted in a broad geographic expansion of CQRs as well as their use in quality improvement studies, clinical trials as well as international comparisons, and benchmarking for ICUs.

CONCLUSIONS

ICU registries have provided increased knowledge of case-mix and outcomes of ICU patients based on real-world data and contributed to improve care delivery through quality improvement initiatives and trials. Recent increases in adoption of new technologies (i.e., cloud-based structures, artificial intelligence, machine learning) will ensure a broader and better use of data for epidemiology, healthcare policies, quality improvement, and clinical trials.

Using a Video Game Intervention to Increase Hospitalists' Advance Care Planning Conversations with Older Adults: a Stepped Wedge Randomized Clinical Trial

BACKGROUND

Guidelines recommend Advance Care Planning (ACP) for seriously ill older adults to increase the patient-centeredness of end-of-life care. Few interventions target the inpatient setting.

OBJECTIVE

To test the effect of a novel physician-directed intervention on ACP conversations in the inpatient setting.

DESIGN

Stepped wedge cluster-randomized design with five 1-month steps (October 2020-February 2021), and 3-month extensions at each end.

SETTING

A total of 35/125 hospitals staffed by a nationwide physician practice with an existing quality improvement initiative to increase ACP (enhanced usual care).

PARTICIPANTS

Physicians employed for ≥ 6 months at these hospitals; patients aged ≥ 65 years they treated between July 2020-May 2021.

INTERVENTION

Greater than or equal to 2 h of exposure to a theory-based video game designed to increase autonomous motivation for ACP; enhanced usual care.

MAIN MEASURE

ACP billing (data abstractors blinded to intervention status).

RESULTS

A total of 163/319 (52%) invited, eligible hospitalists consented to participate, 161 (98%) responded, and 132 (81%) completed all tasks. Physicians' mean age was 40 (SD 7); most were male (76%), Asian (52%), and reported playing the game for ≥ 2 h (81%). These physicians treated 44,235 eligible patients over the entire study period. Most patients (57%) were ≥ 75; 15% had COVID. ACP billing decreased between the pre- and post-intervention periods (26% v. 21%). After adjustment, the homogeneous effect of the game on ACP billing was non-significant (OR 0.96; 95% CI 0.88-1.06; p = 0.42). There was effect modification by step (p < 0.001), with the game associated with increased billing in steps 1-3 (OR 1.03 [step 1]; OR 1.15 [step 2]; OR 1.13 [step 3]) and decreased billing in steps 4-5 (OR 0.66 [step 4]; OR 0.95 [step 5]).

CONCLUSIONS

When added to enhanced usual care, a novel video game intervention had no clear effect on ACP billing, but variation across steps of the trial raised concerns about confounding from secular trends (i.e., COVID).

TRIAL REGISTRATION

Clinicaltrials.gov; NCT04557930, 9/21/2020.

Parental Perspectives on a Trial Using Waived Informed Consent at Birth

Objectives

To determine parental perspectives in a trial with waived consent.

Study Design

Biological parents of non-vigorous term infants randomized using a waiver of consent for a delivery room intervention completed an anonymous survey after discharge.

Results

121 survey responses were collected. Most responding parents reported that this form of consent was acceptable (92%) and that they would feel comfortable having another child participate in a similar study (96%). The majority (> 90%) also reported that the information provided after randomization was clear to understand future data collection procedures. Four percent had a negative opinion on the study's effect on their child's health.

Conclusions

The majority of responding parents reported both acceptability of this study design in the neonatal period and that the study had a positive effect on their child's health. Future work should investigate additional ways to involve parents and elicit feedback on varied methods of pediatric consent.

Best (but oft forgotten) practices: Efficient sample sizes for commonly used trial designs

Designing studies such that they have a high level of power to detect an effect or association of interest is an important tool to improve the quality and reproducibility of findings from such studies. Since resources (research subjects, time, and money) are scarce, it is important to obtain sufficient power with minimum use of such resources. For commonly used randomized trials of the treatment effect on a continuous outcome, designs are presented that minimize the number of subjects or the amount of research budget when aiming for a desired power level. This concerns the optimal allocation of subjects to treatments and, in case of nested designs such as cluster-randomized trials and multicenter trials, also the optimal number of centers versus the number of persons per center. Since such optimal designs require knowledge of parameters of the analysis model that are not known in the design stage, in particular outcome variances, maximin designs are presented. These designs guarantee a prespecified power level for plausible ranges of the unknown parameters and minimize research costs for the worst-case values of these parameters. The focus is on a 2-group parallel design, the AB/BA crossover design, and cluster-randomized and multicenter trials with a continuous outcome. How to calculate sample sizes for maximin designs is illustrated for examples from nutrition. Several computer programs that are helpful in calculating sample sizes for optimal and maximin designs are discussed as well as some results on optimal designs for other types of outcomes.

Norepinephrine versus phenylephrine for treating hypotension during general anaesthesia in adult patients undergoing major noncardiac surgery: a multicentre, open-label, cluster-randomised, crossover, feasibility, and pilot trial

BACKGROUND

Intraoperative hypotension is associated with postoperative complications. The use of vasopressors is often required to correct hypotension but the best vasopressor is unknown.

METHODS

A multicentre, cluster-randomised, crossover, feasibility and pilot trial was conducted across five hospitals in California. Phenylephrine (PE) vs norepinephrine (NE) infusion as the first-line vasopressor in patients under general anaesthesia alternated monthly at each hospital for 6 months. The primary endpoint was first-line vasopressor administration compliance of 80% or higher. Secondary endpoints were acute kidney injury (AKI), 30-day mortality, myocardial injury after noncardiac surgery (MINS), hospital length of stay, and rehospitalisation within 30 days.

RESULTS

A total of 3626 patients were enrolled over 6 months; 1809 patients were randomised in the NE group, 1817 in the PE group. Overall, 88.2% received the assigned first-line vasopressor. No drug infiltrations requiring treatment were reported in either group. Patients were median 63 yr old, 50% female, and 58% white. Randomisation in the NE group vs PE group did not reduce readmission within 30 days (adjusted odds ratio=0.92; 95% confidence interval, 0.6-1.39), 30-day mortality (1.01; 0.48-2.09), AKI (1.1; 0.92-1.31), or MINS (1.63; 0.84-3.16).

CONCLUSIONS

A large and diverse population undergoing major surgery under general anaesthesia was successfully enrolled and randomised to receive NE or PE infusion. This pilot and feasibility trial was not powered for adverse postoperative outcomes and a follow-up multicentre effectiveness trial is planned.

CLINICAL TRIAL REGISTRATION

NCT04789330 (ClinicalTrials.gov).

Rethinking children's physical activity interventions at school: A new context-specific approach

Physical activity is important for children's health. However, evidence suggests that many children and adults do not meet international physical activity recommendations. Current school-based interventions have had limited effect on physical activity and alternative approaches are needed. Context, which includes school setting, ethos, staff, and sociodemographic factors, is a key and largely ignored contributing factor to school-based physical activity intervention effectiveness, impacting in several interacting ways.

Conceptualization

Current programs focus on tightly-constructed content that ignores the context in which the program will be delivered, thereby limiting effectiveness. We propose a move away from uniform interventions that maximize internal validity toward a flexible approach that enables schools to tailor content to their specific context.

Evaluation designs

Evaluation of context-specific interventions should explicitly consider context. This is challenging in cluster randomized controlled trial designs. Thus, alternative designs such as natural experiment and stepped-wedge designs warrant further consideration.

Primary outcome

A collective focus on average minutes of moderate-to-vigorous intensity physical activity may not always be the most appropriate choice. A wider range of outcomes may improve children's physical activity and health in the long-term. In this paper, we argue that greater consideration of school context is key in the design and analysis of school-based physical activity interventions and may help overcome existing limitations in the design of effective interventions and thus progress the field. While this focus on context-specific interventions and evaluation is untested, we hope to stimulate debate of the key issues to improve future physical activity intervention development and implementation.

Sample size calculations for continuous outcomes in clinical nutrition

In nutrition research, sample size calculations for continuous outcomes are important for the planning phase of many randomized trials and could also be relevant for some observational studies such as cohort and cross-sectional studies. However, only little literature dedicated to this topic exists within nutritional science. This article reviews the most common methods for sample size calculations in nutrition research. Approximate formulas are used for explaining concepts and requirements and for working through examples from the literature. Sample size calculations for the various study designs, which are covered, may all be seen as extensions of the sample size calculation for the basic two-group comparison through the application of suitable scaling factors and, possibly, modification of the significance level. The latter is needed for sample size calculations for multi-group designs and designs involving multiple primary outcomes. Like cluster-randomized designs, these types of study designs may be more challenging than standard sample size calculations. In such non-standard scenarios, there may be a need for consulting a biostatistician. Finally, it should be stressed that there may be many ways to plan a study. The final sample size calculation provided for a grant applicant, study protocol, or publication will often not only depend on considerations and input information as described in this article but will also involve restrictions in terms of logistics and/or resources.

Optimizing the Design and Analysis of Future AKI Trials

AKI is a complex clinical syndrome associated with an increased risk of morbidity and mortality, particularly in critically ill and perioperative patient populations. Most AKI clinical trials have been inconclusive, failing to detect clinically important treatment effects at predetermined statistical thresholds. Heterogeneity in the pathobiology, etiology, presentation, and clinical course of AKI remains a key challenge in successfully testing new approaches for AKI prevention and treatment. This article, derived from the "AKI" session of the "Kidney Disease Clinical Trialists" virtual workshop held in October 2021, reviews barriers to and strategies for improving the design and implementation of clinical trials in patients with, or at risk of, developing AKI. The novel approaches to trial design included in this review span adaptive trial designs that increase the knowledge gained from each trial participant; pragmatic trial designs that allow for the efficient enrollment of sufficiently large numbers of patients to detect small, but clinically significant, treatment effects; and platform trial designs that use one trial infrastructure to answer multiple clinical questions simultaneously. This review also covers novel approaches to clinical trial analysis, such as Bayesian analysis and assessing heterogeneity in the response to therapies among trial participants. We also propose a road map and actionable recommendations to facilitate the adoption of the reviewed approaches. We hope that the resulting road map will help guide future clinical trial planning, maximize learning from AKI trials, and reduce the risk of missing important signals of benefit (or harm) from trial interventions.

Use of information criteria for selecting a correlation structure for longitudinal cluster randomised trials

BACKGROUND

When designing and analysing longitudinal cluster randomised trials, such as the stepped wedge, the similarity of outcomes from the same cluster must be accounted for through the choice of a form for the within-cluster correlation structure. Several choices for this structure are commonly considered for application within the linear mixed model paradigm. The first assumes a constant intra-cluster correlation for all pairs of outcomes from the same cluster (the exchangeable/Hussey and Hughes model); the second assumes that correlations of outcomes measured in the same period are higher than outcomes measured in different periods (the block exchangeable model) and the third is the discrete-time decay model, which allows the correlation between pairs of outcomes to decay over time. Currently, there is limited guidance on how to select the most appropriate within-cluster correlation structure.

METHODS

We simulated continuous outcomes under each of the three considered within-cluster correlation structures for a range of design and parameter choices, and, using the ASReml-R package, fit each linear mixed model to each simulated dataset. We evaluated the performance of the Akaike and Bayesian information criteria for selecting the correct within-cluster correlation structure for each dataset.

RESULTS

For smaller total sample sizes, neither criteria performs particularly well in selecting the correct within-cluster correlation structure, with the simpler exchangeable model being favoured. Furthermore, in general, the Bayesian information criterion favours the exchangeable model. When the cluster auto-correlation (which defines the degree of dependence between observations in adjacent time periods) is large and number of periods is small, neither criteria is able to distinguish between the block exchangeable and discrete time decay models. However, for increasing numbers of clusters, periods, and subjects per cluster period, both the Akaike and Bayesian information criteria perform increasingly well in the detection of the correct within-cluster correlation structure.

CONCLUSIONS

With increasing amounts of data, be they number of clusters, periods or subjects per cluster period, both the Akaike and Bayesian information criteria are increasingly likely to select the correct correlation structure. We recommend that if there are sufficient data available when planning a trial, that the Akaike or Bayesian information criterion is used to guide the choice of within-cluster correlation structure in the absence of other compelling justifications for a specific correlation structure. We also suggest that researchers conduct supplementary analyses under alternate correlation structures to gauge sensitivity to the initial choice.

Promoting patient and nurse safety: testing a behavioural health intervention in a learning healthcare system: results of the DEMEANOR pragmatic, cluster, cross-over trial

Background

Based on clinical staff safety within a learning healthcare system, the purpose of this study was to test an innovative model of care for addressing disruptive behaviour in hospitalised patients to determine whether it should be scaled up at the system level.

Methods

The Disruptive bEhaviour manageMEnt ANd prevention in hospitalised patients using a behaviOuRal (DEMEANOR) intervention team was a pragmatic, cluster, cross-over trial. A behavioural intervention team (BIT) with a psychiatric mental health advanced practice nurse and a social worker, with psychiatrist consultation, switched between units each month and occurrences of disruptive behaviours (eg, documented violence control measures, violence risk) compared. Nursing surveys assessed self-perceived efficacy and comfort managing disruptive patient behaviour.

Results

A total of 3800 patients hospitalised on the two units met the criteria for inclusion. Of those, 1841 (48.4%) were exposed to the BIT intervention and 1959 (51.6%) were in the control group. A total of 11 132 individual behavioural issues associated with 203 patient encounters were documented. There were no differences in the use of behavioural interventions, violence risk or injurious behaviour or sitter use between patients exposed to BIT and those in the control group. Tracking these data did rely on nursing documentation of such events. Nurses (82 pre and 48 post) rated BIT as the most beneficial support they received to manage patients exhibiting disruptive, threatening or acting out behaviour.

Conclusions

The BIT intervention was perceived as beneficial by nurses in preparing them to provide care for patients exhibiting disruptive, threatening or acting out behaviour, but documented patient behaviour was not observed to change.

Trial registration number

NCT03777241.

Variable coverage in an Autocidal Gravid Ovitrap intervention impacts efficacy of Aedes aegypti control

Control of the arboviral disease vector Aedes aegypti has shown variable levels of efficacy around the globe. We evaluated an Autocidal Gravid Ovitrap (AGO) intervention as a stand-alone control tool for population suppression of A. aegypti in US communities bordering Mexico.We conducted a cluster randomized crossover trial with weekly mosquito surveillance of sentinel households from July 2017 to December 2018. The intervention took place from August to December of both years. Multilevel models (generalized linear and additive mixed models) were used to analyse the changes in population abundance of female A. aegypti.We observed that female populations were being suppressed 77% (2018) and four times lower outdoor female abundance when AGO coverage (number of intervention AGO traps that surrounded a sentinel home) was high (2.7 AGOs/house). However, we also observed that areas with low intervention AGO coverage resulted in no difference (2017) or slightly higher abundance compared to the control. These results suggest that coverage rate might play a critical role on how populations of female A. aegypti are being modulated in the field. The lack of larval source habitat reduction and the short duration of the intervention period might have limited the A. aegypti population suppression observed in this study. Synthesis and applications. The mosquito, A. aegypti, is a public health concern in most tropical and subtropical regions. With the rise of insecticide resistance, the evaluation of non-chemical tools has become pivotal in the fight against arboviral disease transmission. Our study shows that the AGO intervention, as a stand-alone control tool, is limited by its coverage in human settlements. Vector control programmes should consider, that if the target coverage rate is not achieved, measures will be ineffective unless coupled with other control approaches. Although our multilevel modelling was focused on A. aegypti and the AGO, the approach can be applied to other mosquito vector species.

A flexible sample size solution for longitudinal and crossover cluster randomized trials with continuous outcomes

Longitudinal cluster randomized trial (LCRT) and crossover cluster randomized trial (CCRT) are two variants of cluster randomized trials. In LCRTs, clusters of subjects are randomly assigned to different treatment groups and each subject has repeated measurements over the study period. In CCRTs, clusters of subjects are randomly assigned to different sequences. Within each sequence, clusters receive all treatments in a particular order. Both LCRTs and CCRTs lead to complicated correlation structures that involve longitudinal and intracluster correlations. Generalized linear mixed model (GLMM) and generalized estimating equation (GEE) approaches have been frequently employed in data analysis and sample size estimation. In this study we propose closed-form sample size and power formulas for LCRTs and CCRTs based on the GEE approach. These formulas are flexible to incorporate unbalanced randomization, different missing patterns, arbitrary correlation structures, and randomly varying cluster sizes, providing a practical yet robust sample size solution. Simulation studies show that the proposed methods achieve good performance with empirical powers and type I errors close to their nominal values.

Increasing vegetables purchase with a descriptive-norm message: A cluster randomized controlled intervention in two university canteens

Exposure to social norms is a popular way to foster healthy food behavior. Testing the robustness of this effect, we report a field study assessing the impact of a vegetable-related descriptive norm message on vegetables purchase. The first contribution was to rely on a cluster randomized crossover design: Two canteens were randomly selected to display either a vegetable-related or a neutral-behavior norm message. After a first period of data collection, the displays were reversed for a second period: The number of vegetable portions on the main plate were recorded before, during and after the message display (N = 12.994). The second contribution was to test the impact of a message describing vegetables as the normative choice beyond the mere selection of vegetables, on the quantity of vegetables purchased in lunches containing some. Results indicated that the vegetable-related norm message led to a sustained probability of choosing vegetables, contrary to a decrease observed in the control condition. Moreover, students who ordered vegetables ordered a higher quantity when exposed to a vegetable-related message than before whereas quantity declined in the control condition. By treating both canteens as experimental and control and by analyzing both the presence and the amount of vegetables, these results extend and strengthen those previously observed, bringing support for the effectiveness of a descriptive norm message in eliciting healthier food behavior.

Design and analysis of a 2-year parallel follow-up of repeated ivermectin mass drug administrations for control of malaria: Small sample considerations for cluster-randomized trials with count data

BACKGROUND

Cluster-randomized trials allow for the evaluation of a community-level or group-/cluster-level intervention. For studies that require a cluster-randomized trial design to evaluate cluster-level interventions aimed at controlling vector-borne diseases, it may be difficult to assess a large number of clusters while performing the additional work needed to monitor participants, vectors, and environmental factors associated with the disease. One such example of a cluster-randomized trial with few clusters was the "efficacy and risk of harms of repeated ivermectin mass drug administrations for control of malaria" trial. Although previous work has provided recommendations for analyzing trials like repeated ivermectin mass drug administrations for control of malaria, additional evaluation of the multiple approaches for analysis is needed for study designs with count outcomes.

METHODS

Using a simulation study, we applied three analysis frameworks to three cluster-randomized trial designs (single-year, 2-year parallel, and 2-year crossover) in the context of a 2-year parallel follow-up of repeated ivermectin mass drug administrations for control of malaria. Mixed-effects models, generalized estimating equations, and cluster-level analyses were evaluated. Additional 2-year parallel designs with different numbers of clusters and different cluster correlations were also explored.

RESULTS

Mixed-effects models with a small sample correction and unweighted cluster-level summaries yielded both high power and control of the Type I error rate. Generalized estimating equation approaches that utilized small sample corrections controlled the Type I error rate but did not confer greater power when compared to a mixed model approach with small sample correction. The crossover design generally yielded higher power relative to the parallel equivalent. Differences in power between analysis methods became less pronounced as the number of clusters increased. The strength of within-cluster correlation impacted the relative differences in power.

CONCLUSION

Regardless of study design, cluster-level analyses as well as individual-level analyses like mixed-effects models or generalized estimating equations with small sample size corrections can both provide reliable results in small cluster settings. For 2-year parallel follow-up of repeated ivermectin mass drug administrations for control of malaria, we recommend a mixed-effects model with a pseudo-likelihood approximation method and Kenward-Roger correction. Similarly designed studies with small sample sizes and count outcomes should consider adjustments for small sample sizes when using a mixed-effects model or generalized estimating equation for analysis. Although the 2-year parallel follow-up of repeated ivermectin mass drug administrations for control of malaria is already underway as a parallel trial, applying the simulation parameters to a crossover design yielded improved power, suggesting that crossover designs may be valuable in settings where the number of available clusters is limited. Finally, the sensitivity of the analysis approach to the strength of within-cluster correlation should be carefully considered when selecting the primary analysis for a cluster-randomized trial.

Provision of the progestogen-only pill by community pharmacies as bridging contraception for women receiving emergency contraception: the Bridge-it RCT

INTRODUCTION

Unless women start effective contraception after using emergency contraception, they remain at risk of unintended pregnancy. Most women in the UK obtain emergency contraception from community pharmacies that are unable to provide ongoing contraception (apart from barrier methods which have high failure rates). This means that women need an appointment with a general practitioner or at a sexual and reproductive health clinic. We conducted a pragmatic cluster randomised cohort crossover trial to determine whether or not pharmacist provision of a bridging supply of a progestogen-only pill plus the invitation to attend a sexual and reproductive health clinic resulted in increased subsequent use of effective contraception (hormonal or intrauterine).

METHODS

Twenty-nine pharmacies in three UK cities recruited women receiving emergency contraception (levonorgestrel). In the intervention, women received a 3-month supply of the progestogen-only pill (75 µg of desogestrel) plus a card that provided rapid access to a local sexual and reproductive health clinic. In the control arm, pharmacists advised women to attend their usual contraceptive provider. The primary outcome was reported use of an effective contraception (hormonal and intrauterine methods) at 4 months. Process evaluation was also conducted to inform any future implementation.

RESULTS

The study took place December 2017 and June 2019 and recruited 636 women to the intervention (n = 316) and control groups (n = 320). There were no statistically significant differences in demographic characteristics between the groups. Four-month follow-up data were available for 406 participants: 63% (198/315) of the control group and 65% (208/318) of the intervention group. The proportion of participants reporting use of effective contraception was 20.1% greater (95% confidence interval 5.2% to 35.0%) in the intervention group (58.4%, 95% confidence interval 48.6% to 68.2%) than in the control group (40.5%, 95% confidence interval 29.7% to 51.3%) (adjusted for recruitment period, treatment arm and centre; p = 0.011). The proportion of women using effective contraception remained statistically significantly larger, when adjusted for age, current sexual relationship and history of past use of effective contraception, and was robust to the missing data. There were no serious adverse events.

CONCLUSION

Provision of a bridging supply of the progestogen-only pill with emergency contraception from a pharmacist and the invitation to a sexual and reproductive health clinic resulted in a significant increase in self-reported subsequent use of effective contraception. This simple intervention has the potential to prevent more unintended pregnancies for women after emergency contraception.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN70616901.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 27. See the NIHR Journals Library website for further project information.

Rationale, Methodological Quality, and Reporting of Cluster-Randomized Controlled Trials in Critical Care Medicine: A Systematic Review

OBJECTIVE

Compared with individual-patient randomized controlled trials, cluster randomized controlled trials have unique methodological and ethical considerations. We evaluated the rationale, methodological quality, and reporting of cluster randomized controlled trials in critical care studies.

DATA SOURCES

Systematic searches of Medline, Embase, and Cochrane Central Register were performed.

STUDY SELECTION

We included all cluster randomized controlled trials conducted in adult, pediatric, or neonatal critical care units from January 2005 to September 2019.

DATA EXTRACTION

Two reviewers independently screened citations, reviewed full texts, protocols, and supplements of potentially eligible studies, abstracted data, and assessed methodology of included studies.

DATA SYNTHESIS

From 1,902 citations, 59 cluster randomized controlled trials met criteria. Most focused on quality improvement (24, 41%), antimicrobial therapy (9, 15%), or infection control (9, 15%) interventions. Designs included parallel-group (25, 42%), crossover (21, 36%), and stepped-wedge (13, 22%). Concealment of allocation was reported in 21 studies (36%). Thirteen studies (22%) reported at least one method of blinding. The median total sample size was 1,660 patients (interquartile range, 813-4,295); the median number of clusters was 12 (interquartile range, 5-24); and the median patients per cluster was 141 (interquartile range, 54-452). Sample size calculations were reported in 90% of trials, but only 54% met Consolidated Standards of Reporting Trials guidance for sample size reporting. Twenty-seven of the studies (46%) identified a fixed number of available clusters prior to trial commencement, and only nine (15%) prespecified both the number of clusters and patients required to detect the expected effect size. Overall, 36 trials (68%) achieved the total prespecified sample size. When analyzing data, 44 studies (75%) appropriately adjusted for clustering when analyzing the primary outcome. Only 12 (20%) reported an intracluster coefficient (median 0.047 [interquartile range, 0.01-0.13]).

CONCLUSIONS

Cluster randomized controlled trials in critical care typically involve a small and fixed number of relatively large clusters. The reporting of key methodological aspects of these trials is often inadequate.

Effect of an Appearance-Based vs. a Health-Based Sun-Protective Intervention on French Summer Tourists' Behaviors in a Cluster Randomized Crossover Trial: The PRISME Protocol

Background: Sun exposure has short- and long-term adverse effects on eyes, skin, and the immune system. The most serious effect, melanoma, is largely attributable to natural ultraviolet radiation. Its prevalence is steadily increasing in fair-skinned populations in most European countries. Despite annual prevention campaigns, the French population continues to be overexposed to the sun and under-protected. Social and psychosocial characteristics may play an important role in sun protection determinants. Overexposure is partially motivated by a desire to tan oneself for aesthetic reasons. During summer, intense exposure constitutes a major risk factor for melanoma, making tourists a particularly high-risk population. Literature reviews concluded that appearance-based interventions highlighting the aesthetic effects of sun exposure on skin photoaging showed promise in terms of improving sun-exposure and sun-protection behaviors, especially among younger people, but that more rigorous studies were needed. In this context, we implemented the PRISME study to: - identify the determinants, in particular social and psychosocial, of sun-protection of French summer tourists visiting the Mediterranean coastline; - design two prevention interventions grounded in psychosocial theories; - compare the impact of both interventions on tourists' sun-protection behaviors, and identify the determinants influencing this impact. This paper presents the methodology of the PRISME study. Methods: During summer 2019, we conducted a cluster randomized crossover trial to compare two prevention interventions, one based on health-related messages (health effects information, phototype calculation), the other on appearance-related messages (photoaging information, ultraviolet photography), among French tourists aged 12-55 years old in coastline campsites in the French region of Occitanie. Both interventions were anchored in the theory of planned behavior and in the transtheoretical model. The interventions' impact was measured using face-to-face questionnaires and skin color measurements both immediately before and 4 days after the interventions. A second follow-up, using an online questionnaire, will be conducted in September 2020 to measure the longer-term effects of both interventions. Discussion: Despite certain study limitations, PRISME take into consideration several known methodological gaps. The study's results will enable to evaluate the efficacy of the promising appearance-based approach in France, and to identify vulnerable sub-populations and mechanisms to improve sun-protection behaviors of French tourists.

Community psychosocial music intervention (CHIME) to reduce antenatal common mental disorder symptoms in The Gambia: a feasibility trial

OBJECTIVES

Examine the feasibility of a Community Health Intervention through Musical Engagement (CHIME) in The Gambia to reduce common mental disorder (CMD) symptoms in pregnant women.

DESIGN

Feasibility trial testing a randomised stepped-wedge cluster design.

SETTING

Four local antenatal clinics.

PARTICIPANTS

Women who were 14-24 weeks pregnant and spoke Mandinka or Wolof were recruited into the intervention (n=50) or control group (n=74).

INTERVENTION

Music-based psychosocial support sessions designed and delivered by all-female fertility societies. Sessions lasted 1 hour and were held weekly for 6 weeks. Delivered to groups of women with no preselection. Sessions were designed to lift mood, build social connection and provide health messaging through participatory music making. The control group received standard antenatal care.

OUTCOMES

Demographic, feasibility, acceptability outcomes and the appropriateness of the study design were assessed. Translated measurement tools (Self-Reporting Questionnaire (SRQ-20); Edinburgh Postnatal Depression Scale (EPDS)) were used to assess CMD symptoms at baseline, post-intervention and 4-week follow-up.

RESULTS

All clinics and 82% of women approached consented to take part. A 33% attrition rate across all time points was observed. 72% in the intervention group attended at least three sessions. Audio and video analysis confirmed fidelity of the intervention and a thematic analysis of participant interviews demonstrated acceptability and positive evaluation. Results showed a potential beneficial effect with a reduction of 2.13 points (95% CI (0.89 to 3.38), p<0.01, n=99) on the SRQ-20 and 1.98 points (95% CI (1.06 to 2.90), p<0.01, n=99) on the EPDS at the post-intervention time point for the intervention group compared with standard care.

CONCLUSION

Results demonstrate that CHIME is acceptable and feasible in The Gambia. To our knowledge, CHIME is the first example of a music-based psychosocial intervention to be applied to perinatal mental health in a low- and middle-income country context.

TRIAL REGISTRATION NUMBER

Pan African Clinical Trials Registry (PACTR201901917619299).

Use of effective contraception following provision of the progestogen-only pill for women presenting to community pharmacies for emergency contraception (Bridge-It): a pragmatic cluster-randomised crossover trial

BACKGROUND

Unless women start effective contraception after oral emergency contraception, they remain at risk of unintended pregnancy. Most women in the UK obtain emergency contraception from community pharmacies. We hypothesised that pharmacist provision of the progestogen-only pill as a bridging interim method of contraception with emergency contraception plus an invitation to a sexual and reproductive health clinic, in which all methods of contraception are available, would result in increased subsequent use of effective contraception.

METHODS

We did a pragmatic cluster-randomised crossover trial in 29 UK pharmacies among women receiving levonorgestrel emergency contraception. Women aged 16 years or older, not already using hormonal contraception, not on medication that could interfere with the progestogen-only pill, and willing to give contact details for follow-up were invited to participate. In the intervention group, women received a 3-month supply of the progestogen-only pill (75 μg desogestrel) plus a rapid access card to a participating sexual and reproductive health clinic. In the control group, pharmacists advised women to attend their usual contraceptive provider. The order in which each pharmacy provided the intervention or control was randomly assigned using a computer software algorithm. The primary outcome was the use of effective contraception (hormonal or intrauterine) at 4 months. This study is registered, ISRCTN70616901 (complete).

FINDINGS

Between Dec 19, 2017, and June 26, 2019, 636 women were recruited to the intervention group (316 [49·6%], mean age 22·7 years [SD 5·7]) or the control group (320 [50·3%], 22·6 years [5·1]). Three women (one in the intervention group and two in the control group) were excluded after randomisation. 4-month follow-up data were available for 406 (64%) participants, 25 were lost to follow-up, and two participants no longer wanted to participate in the study. The proportion of women using effective contraception was 20·1% greater (95% CI 5·2-35·0) in the intervention group (mean 58·4%, 48·6-68·2), than in the control group (mean 40·5%, 29·7-51·3 [adjusted for recruitment period, treatment group, and centre]; p=0·011).The difference remained significant after adjusting for age, current sexual relationship, and history of effective contraception use, and was robust to the effect of missing data (assuming missingness at random). No serious adverse events occurred.

INTERPRETATION

Provision of a supply of the progestogen-only pill with emergency contraception from a community pharmacist, along with an invitation to a sexual and reproductive health clinic, results in a clinically meaningful increase in subsequent use of effective contraception. Widely implemented, this practice could prevent unintended pregnancies after use of emergency contraception.

FUNDING

National Institute for Health Research (Health Technology Assessment Programme project 15/113/01).

Extending the I-squared statistic to describe treatment effect heterogeneity in cluster, multi-centre randomized trials and individual patient data meta-analysis

Treatment effect heterogeneity is commonly investigated in meta-analyses to identify if treatment effects vary across studies. When conducting an aggregate level data meta-analysis it is common to describe the magnitude of any treatment effect heterogeneity using the I-squared statistic, which is an intuitive and easily understood concept. The effect of a treatment might also vary across clusters in a cluster randomized trial, or across centres in multi-centre randomized trial, and it can be of interest to explore this at the analysis stage. In cross-over trials and other randomized designs, in which clusters or centres are exposed to both treatment and control conditions, this treatment effect heterogeneity can be identified. Here we derive and evaluate a comparable I-squared measure to describe the magnitude of heterogeneity in treatment effects across clusters or centres in randomized trials. We further show how this methodology can be used to estimate treatment effect heterogeneity in an individual patient data meta-analysis.

Cluster identification, selection, and description in cluster randomized crossover trials: the PREP-IT trials

Background

In cluster randomized crossover (CRXO) trials, groups of participants (i.e., clusters) are randomly allocated to receive a sequence of interventions over time (i.e., cluster periods). CRXO trials are becoming more comment when they are feasible, as they require fewer clusters than parallel group cluster randomized trials. However, CRXO trials have not been frequently used in orthopedic fracture trials and represent a novel methodological application within the field. To disseminate the early knowledge gained from our experience initiating two cluster randomized crossover trials, we describe our process for the identification and selection of the orthopedic practices (i.e., clusters) participating in the PREP-IT program and present data to describe their key characteristics.

Methods

The PREP-IT program comprises two ongoing pragmatic cluster randomized crossover trials (Aqueous-PREP and PREPARE) which compare the effect of iodophor versus chlorhexidine solutions on surgical site infection and unplanned fracture-related reoperations in patients undergoing operative fracture management. We describe the process we used to identify and select orthopedic practices (clusters) for the PREP-IT trials, along with their characteristics.

Results

We identified 58 potential orthopedic practices for inclusion in the PREP-IT trials. After screening each practice for eligibility, we selected 30 practices for participation and randomized each to a sequence of interventions (15 for Aqueous-PREP and 20 for PREPARE). The majority of orthopedic practices included in the Aqueous-PREP and PREPARE trials were situated in level I trauma centers (100% and 87%, respectively). Orthopedic practices in the Aqueous-PREP trial operatively treated a median of 149 open fracture patients per year, included a median of 11 orthopedic surgeons, and had access to a median of 5 infection preventionists. Orthopedic practices in the PREPARE trial treated a median of 142 open fracture and 1090 closed fracture patients per year, included a median of 7.5 orthopedic surgeons, and had access to a median of 6 infection preventionists.

Conclusions

The PREP-IT trials provide an example of how to follow the reporting standards for cluster randomized crossover trials by providing a clear definition of the cluster unit, a thorough description of the cluster identification and selection process, and sufficient description of key cluster characteristics.

Trial registration

Both trials are registered at ClinicalTrials.gov (A-PREP: NCT03385304 December 28, 2017, and PREPARE: NCT03523962 May 14, 2018).

A tutorial on sample size calculation for multiple-period cluster randomized parallel, cross-over and stepped-wedge trials using the Shiny CRT Calculator

It has long been recognized that sample size calculations for cluster randomized trials require consideration of the correlation between multiple observations within the same cluster. When measurements are taken at anything other than a single point in time, these correlations depend not only on the cluster but also on the time separation between measurements and additionally, on whether different participants (cross-sectional designs) or the same participants (cohort designs) are repeatedly measured. This is particularly relevant in trials with multiple periods of measurement, such as the cluster cross-over and stepped-wedge designs, but also to some degree in parallel designs. Several papers describing sample size methodology for these designs have been published, but this methodology might not be accessible to all researchers. In this article we provide a tutorial on sample size calculation for cluster randomized designs with particular emphasis on designs with multiple periods of measurement and provide a web-based tool, the Shiny CRT Calculator, to allow researchers to easily conduct these sample size calculations. We consider both cross-sectional and cohort designs and allow for a variety of assumed within-cluster correlation structures. We consider cluster heterogeneity in treatment effects (for designs where treatment is crossed with cluster), as well as individually randomized group-treatment trials with differential clustering between arms, for example designs where clustering arises from interventions being delivered in groups. The calculator will compute power or precision, as a function of cluster size or number of clusters, for a wide variety of designs and correlation structures. We illustrate the methodology and the flexibility of the Shiny CRT Calculator using a range of examples.

The Disruptive bEhavior manageMEnt ANd prevention in hospitalized patients using a behaviORal intervention team (DEMEANOR) study protocol: a pragmatic, cluster, crossover trial

BACKGROUND

Disruptive behavior in hospitalized patients has become a priority area of safety concern for clinical staff, and also has consequences for patient management and hospital course. Proactive screening and intervention of patients with behavioral comorbidities has been reported to reduce disruptive behavior in some settings, but it has not been studied in a rigorous way.

METHODS

The Disruptive bEhavior manageMEnt ANd prevention in hospitalized patients using a behaviORal intervention team (DEMEANOR) study is a pragmatic, cluster, crossover trial that is being conducted. Each month, the behavioral intervention team, comprising a psychiatric-mental health advanced practice nurse and a clinical social worker, with psychiatrist consultation as needed, rotates between an adult medicine unit and a mixed cardiac unit at Vanderbilt University Medical Center in Nashville, TN, USA. The team proactively screens patients upon admission, utilizing a protocol which includes a comprehensive chart review and, if indicated, a brief interview, seeking to identify those patients who possess risk factors indicative of either a potential psychological barrier to their own clinical progress or a potential risk for exhibiting disruptive, aggressive, or self-injurious behavior during their hospitalization. Once identified, the team provides interventions aimed at mitigating these risks, educates and supports the patient care teams (nurses, physicians, and others), and assists non-psychiatric staff in the management of patients who require behavioral healthcare. Patients who are both admitted to and discharged from either unit are included in the study. Anticipated enrollment is approximately 1790 patients. The two primary outcomes are (1) a composite of objective measures related to the patients' disruptive, threatening, or acting out behaviors, and (2) staff self-reported comfort with and confidence in their ability to manage patients exhibiting disruptive, threatening, or acting out behavior. Secondary outcomes include patient length of stay, patient attendant (sitter) use, and the unit nursing staff retention.

DISCUSSION

This ongoing trial will provide evidence on the real-world effectiveness of a proactive behavioral intervention to prevent disruptive, threatening, or acting out events in adult hospitalized patients.

TRIAL REGISTRATION

ClinicalTrials.gov: NCT03777241. Registered on 14 December 2018.

The cluster randomized crossover trial: The effects of attrition in the AB/BA design and how to account for it in sample size calculations

Background/Aims:

This article studies the effect of attrition in the cluster randomized crossover trial. The focus is on the two-treatment two-period AB/BA design where attrition occurs during the washout period. Attrition may occur at either the subject level or the cluster level. In the latter case, clusters drop out entirely and provide no measurements in the second period. Subject attrition can only occur in the cohort design, where each subject receives both treatments. Cluster attrition can also occur in the cross-sectional design, where different subjects are measured in the two time periods. Furthermore, this article explores two different strategies to account for potential levels of attrition: increasing sample size and replacing those subjects who drop out by others.

Methods:

The statistical model that takes into account the nesting of subjects within clusters, and the nesting of repeated measurements within subjects is presented. The effect of attrition is evaluated on the basis of the efficiency of the treatment effect estimator. Matrix algebra is used to derive the relation between efficiency, the degree of attrition, cluster size and the intraclass correlations: the within-cluster within-period correlation, the within-cluster between-period correlation and (in the case of a cohort design) the within-subject correlation. The methodology is implemented in two Shiny Apps.

Results:

Attrition in a cluster randomized crossover trial implies a loss of efficiency. Efficiency decreases with an increase of the attrition rate. The loss of efficiency due to attrition of subjects in a cohort design is largest for small number of subjects per cluster-period, but it may be repaired to a large degree by increasing the number of subjects per cluster-period or by replacing those subjects who drop out by others. Attrition of clusters results in a larger loss of efficiency, but this loss does not depend on the number of subjects per cluster-period. Repairing for this loss requires a large increase in the number of subjects per cluster-period. The methodology of this article is illustrated by an example on the effect of lavender scent on dental patients’ anxiety.

Conclusion:

This article provides the methodology of exploring the effect of attrition in cluster randomized crossover trials, and to repair for attrition. As such, it helps researchers plan their trial in an appropriate way and avoid underpowered trials. To use the methodology, prior estimates of the degree of attrition and intraclass correlation coefficients are needed. It is advocated that researchers clearly report the estimates of these quantities to help facilitate planning future trials.

Sample size and power calculations for open cohort longitudinal cluster randomized trials

When calculating sample size or power for stepped wedge or other types of longitudinal cluster randomized trials, it is critical that the planned sampling structure be accurately specified. One common assumption is that participants will provide measurements in each trial period, that is, a closed cohort, and another is that each participant provides only one measurement during the course of the trial. However some studies have an "open cohort" sampling structure, where participants may provide measurements in variable numbers of periods. To date, sample size calculations for longitudinal cluster randomized trials have not accommodated open cohorts. Feldman and McKinlay (1994) provided some guidance, stating that the participant-level autocorrelation could be varied to account for the degree of overlap in different periods of the study, but did not indicate precisely how to do so. We present sample size and power formulas that allow for open cohorts and discuss the impact of the degree of "openness" on sample size and power. We consider designs where the number of participants in each cluster will be maintained throughout the trial, but individual participants may provide differing numbers of measurements. Our results are a unification of closed cohort and repeated cross-sectional sample results of Hooper et al (2016), and indicate precisely how participant autocorrelation of Feldman and McKinlay should be varied to account for an open cohort sampling structure. We discuss different types of open cohort sampling schemes and how open cohort sampling structure impacts on power in the presence of decaying within-cluster correlations and autoregressive participant-level errors.

Challenges in Implementing Hospice Clinical Trials: Preserving Scientific Integrity While Facing Change

BACKGROUND/AIMS

Numerous changes can occur between the original design plans for clinical trials, the submission of funding proposals, and the implementation of the clinical trial. In the hospice setting, environmental changes can present significant obstacles, which require changes to the original plan designs, recruitment, and staffing. The purpose of the study was to share lessons and problem-solving strategies that can assist in future hospice trials.

METHODS

This study uses one hospice clinical trial as an exemplar to demonstrate challenges for clinical trial research in this setting. Using preliminary data collected during the first months of a trial, the research team details the many ways their current protocol reflects changes from the originally proposed plans. Experiences are used as an exemplar to address the following questions: 1) How do research environments change between the initial submission of a funding proposal and the eventual award? 2) How can investigators maintain the integrity of the research and accommodate unexpected changes in the research environment?

RESULTS

The changing environment within the hospice setting required design, sampling, and recruitment changes within the first year. The decision-making process resulted in a stronger design with greater generalization. As a result of necessary protocol changes, the study results are positioned to be translational following the study conclusion.

CONCLUSION

Researchers would do well to review their protocol and statistics early in a clinical trial. They should be prepared for adjustments to accommodate market and environmental changes outside their control. Ongoing data monitoring, specifically related to recruitment, is advised.

Pragmatic cluster randomised cohort cross-over trial to determine the effectiveness of bridging from emergency to regular contraception: the Bridge-It study protocol

INTRODUCTION

Oral emergency contraception (EC) can prevent unintended pregnancy but it is important to start a regular method of contraception. Women in the UK usually access EC from a pharmacy but then need a subsequent appointment with a general practitioner or a sexual and reproductive health (SRH) service to access regular contraception. Unintended pregnancies can occur during this time.

METHODS AND ANALYSIS

Bridge-It is a pragmatic cluster randomised cohort cross-over trial designed to determine whether pharmacist provision of a bridging supply of a progestogen-only pill (POP) plus rapid access to a local SRH clinic, results in increased uptake of effective contraception and prevents more unintended pregnancies than provision of EC alone. Bridge-It involves 31 pharmacies in three UK regions (London, Lothian and Tayside) aiming to recruit 626-737 women. Pharmacies will give EC (levonorgestrel) according to normal practice and recruit women to both intervention and the control phases of the study. In the intervention phase, pharmacists will provide the POP (desogestrel) and offer rapid access to an SRH clinic. In the control phase, pharmacists will advise women to attend a contraceptive provider for contraception (standard care).Women will be asked 4 months later about contraceptive use. Data linkage to abortion registries will provide abortion rates over 12 months. The sample size is calculated on the primary outcome of effective contraception use at 4 months (yes/no) with 90% power and a 5% level of significance. Abortion rates will be an exploratory secondary analysis. Process evaluation includes interviews with pharmacists, SRH clinicians and women. Cost-effectiveness analysis will use a healthcare system perspective and be expressed as incremental cost-effectiveness ratio.

ETHICS AND DISSEMINATION

Ethical approval was received from South East Scotland REC June 2017. Results will be published in peer-reviewed journals and conference presentations.

TRIAL REGISTRATION NUMBER

ISRCTN70616901.

A note on sample size calculations for cluster randomised crossover trials with a fixed number of clusters

Girardeau, Ravaud and Donner in 2008 presented a formula for sample size calculations for cluster randomised crossover trials, when the intracluster correlation coefficient, interperiod correlation coefficient and mean cluster size are specified in advance. However, in many randomised trials, the number of clusters is constrained in some way, but the mean cluster size is not. We present a version of the Girardeau formula for sample size calculations for cluster randomised crossover trials when the number of clusters is fixed. Formulae are given for the minimum number of clusters, the maximum cluster size and the relationship between the correlation coefficients when there are constraints on both the number of clusters and the cluster size. Our version of the formula may aid the efficient planning and design of cluster randomised crossover trials.

Communicating benign biopsy results by telephone in the NHS Breast Screening Programme: a protocol for a cluster randomised crossover trial

INTRODUCTION

One of the main harms from breast cancer screening is the anxiety caused by false positive results. Various factors may be associated with false-positive anxiety. One modifiable factor may be the method of communication used to deliver results. The aim of this study is to measure the effect on anxiety of receiving benign biopsy results in-person or by telephone.

METHODS AND ANALYSIS

This is a multi-centre cluster randomised crossover trial in the English National Health Service Breast Screening Programme (NHSBSP) involving repeated survey measures at four time points. Participants will be women of screening age who have a biopsy following a suspicious mammography result, who ultimately receive a benign or normal (B1) result. Centres will trial both telephone and in-person results on a month-by-month basis, being randomised to which communication method will be trialled first. Women will be blinded to the method of communication they will receive. The analysis will compare women who have received telephone results and women who have received in-person results. The primary outcome measure will be anxiety (measured by the Psychological Consequences Questionnaire) after receiving results, while controlling for baseline anxiety. Secondary outcome measures will include anxiety at 3 and 6 months post-results, understanding of results and patient preferences for how results are communicated. Qualitative telephone interviews will also be conducted to further explore women's reasons for communication preferences. Qualitative and quantitative data will be integrated after initial separate analysis using the pillar integration process.

ETHICS AND DISSEMINATION

This study has been approved by the Public Health England Breast Screening Programme Research Advisory Committee, (BSPRAC_0013, ODR1718_040) and the National Health Service Health Research Authority (HRA) West Midlands-Coventry & Warwickshire Research Ethics Committee (17/WM/0313). The findings from this study will be disseminated to key stakeholders within the NHSBSP and via academic publications.

TRIAL REGISTRATION NUMBER

ISRCTN36997684 TRIAL SPONSOR: This research is part of a PhD award and is funded by the Economic and Social Research Council Doctoral Training Centre at the University of Warwick and Public Health England. The sponsor for this research is Jane Prewett (sponsorship@warwick.ac.uk).

Design, implementation, and analysis considerations for cluster-randomized trials in infection control and hospital epidemiology: A systematic review

BACKGROUND

In cluster-randomized trials (CRT), groups rather than individuals are randomized to interventions. The aim of this study was to present critical design, implementation, and analysis issues to consider when planning a CRT in the healthcare setting and to synthesize characteristics of published CRT in the field of healthcare epidemiology.

METHODS

A systematic review was conducted to identify CRT with infection control outcomes.

RESULTS

We identified the following 7 epidemiological principles: (1) identify design type and justify the use of CRT; (2) account for clustering when estimating sample size and report intraclass correlation coefficient (ICC)/coefficient of variation (CV); (3) obtain consent; (4) define level of inference; (5) consider matching and/or stratification; (6) minimize bias and/or contamination; and (7) account for clustering in the analysis. Among 44 included studies, the most common design was CRT with crossover (n = 15, 34%), followed by parallel CRT (n = 11, 25%) and stratified CRT (n = 7, 16%). Moreover, 22 studies (50%) offered justification for their use of CRT, and 20 studies (45%) demonstrated that they accounted for clustering at the design phase. Only 15 studies (34%) reported the ICC, CV, or design effect. Also, 15 studies (34%) obtained waivers of consent, and 7 (16%) sought consent at the cluster level. Only 17 studies (39%) matched or stratified at randomization, and 10 studies (23%) did not report efforts to mitigate bias and/or contamination. Finally, 29 studies (88%) accounted for clustering in their analyses.

CONCLUSIONS

We must continue to improve the design and reporting of CRT to better evaluate the effectiveness of infection control interventions in the healthcare setting.

Power and sample size requirements for GEE analyses of cluster randomized crossover trials

The cluster randomized crossover design has been proposed to improve efficiency over the traditional parallel cluster randomized design, which often involves a limited number of clusters. In recent years, the cluster randomized crossover design has been increasingly used to evaluate the effectiveness of health care policy or programs, and the interest often lies in quantifying the population-averaged intervention effect. In this paper, we consider the two-treatment two-period crossover design, and develop sample size procedures for continuous and binary outcomes corresponding to a population-averaged model estimated by generalized estimating equations, accounting for both within-period and interperiod correlations. In particular, we show that the required sample size depends on the correlation parameters through an eigenvalue of the within-cluster correlation matrix for continuous outcomes and through two distinct eigenvalues of the correlation matrix for binary outcomes. We demonstrate that the empirical power corresponds well with the predicted power by the proposed formulae for as few as eight clusters, when outcomes are analyzed using the matrix-adjusted estimating equations for the correlation parameters concurrently with a suitable bias-corrected sandwich variance estimator.