Reconsidering stepped wedge cluster randomized trial designs with implementation periods: Fewer sequences or the parallel-group design with baseline and implementation periods are potentially more efficient

BACKGROUND/AIMS

When designing a cluster randomized trial, advantages and disadvantages of tentative designs must be weighed. The stepped wedge design is popular for multiple reasons, including its potential to increase power via improved efficiency relative to a parallel-group design. In many realistic settings, it will take time for clusters to fully implement the intervention. When designing the HEALing (Helping to End Addiction Long-termSM) Communities Study, implementation time was a major consideration, and we examined the efficiency and practicality of three designs. Specifically, a three-sequence stepped wedge design with implementation periods, a corresponding two-sequence modified design that is created by removing the middle sequence, and a parallel-group design with baseline and implementation periods. In this article, we study the relative efficiencies of these specific designs. More generally, we study the relative efficiencies of modified designs when the stepped wedge design with implementation periods has three or more sequences. We also consider different correlation structures.

METHODS

We compare efficiencies of stepped wedge designs with implementation periods consisting of three to nine sequences with a variety of corresponding designs. The three-sequence design is compared to the two-sequence modified design and to the parallel-group design with baseline and implementation periods analysed via analysis of covariance. Stepped wedge designs with implementation periods consisting of four or more sequences are compared to modified designs that remove all or a subset of 'middle' sequences. Efficiencies are based on the use of linear mixed effects models.

RESULTS

In the studied settings, the modified design is more efficient than the three-sequence stepped wedge design with implementation periods. The parallel-group design with baseline and implementation periods with analysis of covariance-based analysis is often more efficient than the three-sequence design. With respect to stepped wedge designs with implementation periods that are comprised of more sequences, there are often corresponding modified designs that improve efficiency. However, use of only the first and last sequences has the potential to be either relatively efficient or inefficient. Relative efficiency is impacted by the strength of the statistical correlation among outcomes from the same cluster; for example, the relative efficiencies of modified designs tend to be greater for smaller cluster auto-correlation values.

CONCLUSION

If a three-sequence stepped wedge design with implementation periods is being considered for a future cluster randomized trial, then a corresponding modified design using only the first and last sequences should be considered if sole focus is on efficiency. However, a parallel-group design with baseline and implementation periods and analysis of covariance-based analysis can be a practical, efficient alternative. For stepped wedge designs with implementation periods and a larger number of sequences, modified versions that remove 'middle' sequences should be considered. Due to the potential sensitivity of design efficiencies, statistical correlation should be carefully considered.

Cluster Randomized Trials with a Pretest and Posttest: Equivalence of Three-, Two- and One-Level Analyses, and Sample Size Calculation

In a cluster randomized trial clusters of persons, for instance, schools or health centers, are assigned to treatments, and all persons in the same cluster get the same treatment. Although less powerful than individual randomization, cluster randomization is a good alternative if individual randomization is impossible or leads to severe treatment contamination (carry-over). Focusing on cluster randomized trials with a pretest and post-test of a quantitative outcome, this paper shows the equivalence of four methods of analysis: a three-level mixed (multilevel) regression for repeated measures with as levels cluster, person, and time, and allowing for unstructured between-cluster and within-cluster covariance matrices; a two-level mixed regression with as levels cluster and person, using change from baseline as outcome; a two-level mixed regression with as levels cluster and time, using cluster means as data; a one-level analysis of cluster means of change from baseline. Subsequently, similar equivalences are shown between a constrained mixed model and methods using the pretest as covariate. All methods are also compared on a cluster randomized trial on mental health in children. From these equivalences follows a simple method to calculate the sample size for a cluster randomized trial with baseline measurement, which is demonstrated step-by-step.

Data-driven decision-making for district health management: a cluster-randomised study in 24 districts of Ethiopia

BACKGROUND

Use of local data for health system planning and decision-making in maternal, newborn and child health services is limited in low-income and middle-income countries, despite decentralisation and advances in data gathering. An improved culture of data-sharing and collaborative planning is needed. The Data-Informed Platform for Health is a system-strengthening strategy which promotes structured decision-making by district health officials using local data. Here, we describe implementation including process evaluation at district level in Ethiopia, and evaluation through a cluster-randomised trial.

METHODS

We supported district health teams in 4-month cycles of data-driven decision-making by: (a) defining problems using a health system framework; (b) reviewing data; (c) considering possible solutions; (d) value-based prioritising; and (e) a consultative process to develop, commit to and follow up on action plans. 12 districts were randomly selected from 24 in the North Shewa zone of Ethiopia between October 2020 and June 2022. The remaining districts formed the trial's comparison arm. Outcomes included health information system performance and governance of data-driven decision-making. Analysis was conducted using difference-in-differences.

RESULTS

58 4-month cycles were implemented, four or five in each district. Each focused on a health service delivery challenge at district level. Administrators' practice of, and competence in, data-driven decision-making showed a net increase of 77% (95% CI: 40%, 114%) in the regularity of monthly reviews of service performance, and 48% (95% CI: 9%, 87%) in data-based feedback to health facilities. Statistically significant improvement was also found in administrators' use of information to appraise services. Qualitative findings also suggested that district health staff reported enhanced data use and collaborative decision-making.

CONCLUSIONS

This study generated robust evidence that 20 months' implementation of the Data-Informed Platform for Health strengthened health management through better data use and appraisal practices, systemised problem analysis to follow up on action points and improved stakeholder engagement.

TRIAL REGISTRATION NUMBER

NCT05310682.

Impact of topical applications of sunflower seed oil on neonatal mortality and morbidity in southern Nepal: a community-based, cluster-randomised trial

INTRODUCTION

Hospital-based studies have demonstrated topical applications of sunflower seed oil (SSO) to skin of preterm infants can reduce nosocomial infections and improve survival. In South Asia, replacing traditional mustard with SSO might have similar benefits.

METHODS

340 communities in Sarlahi, Nepal were randomised to use mustard oil (MO) or SSO for community practice of daily newborn massage. Women were provided oil in late pregnancy and the first month post partum, and visited daily through the first week of life to encourage massage practice. A separate data collection team visited on days 1, 3, 7, 10, 14, 21 and 28 to record vital status and assess serious bacterial infection.

RESULTS

Between November 2010 and January 2017, we enrolled 39 479 pregnancies. 32 114 live births were analysed. Neonatal mortality rates (NMRs) were 31.8/1000 (520 deaths, 16 327 births) and 30.5/1000 (478 deaths, 15 676 births) in control and intervention, respectively (relative risk (RR)=0.95, 95% CI: 0.84, 1.08). Among preterm births, NMR was 90.4/1000 (229 deaths, 2533 births) and 79.2/1000 (188 deaths, 2373 births) in control and intervention, respectively (RR=0.88; 95% CI: 0.74, 1.05). Among preterm births <34 weeks, the RR was 0.83 (95% CI: 0.67, 1.02). No statistically significant differences were observed in incidence of serious bacterial infection.

CONCLUSIONS

We did not find any neonatal mortality or morbidity benefit of using SSO instead of MO as emollient therapy in the early neonatal period. Further studies examining whether very preterm babies may benefit are warranted.

TRIAL REGISTRATION NUMBER

ClinicalTrials.gov Registry (NCT01177111).

Examining local smoke-free coalitions in Armenia and Georgia: context and outcomes of a matched-pairs community-randomised controlled trial

INTRODUCTION

Local coalitions can advance public health initiatives such as smoke-free air but have not been widely used or well-studied in low-income and middle-income countries.

METHODS

We conducted a matched-pairs community-randomised controlled trial in 28 communities in Armenia and Georgia (N=14/country) in which we helped establish local coalitions in 2019 and provided training and technical assistance for coalition activity promoting smoke-free policy development and enforcement (2019-2021). Surveys of ~1450 households (Fall 2018, May-June 2022) were conducted to evaluate coalition impact on smoke-free policy support, smoke-free home adoption, secondhand smoke exposure (SHSe), and coalition awareness and activity exposure, using multivariable mixed modelling.

RESULTS

Bivariate analyses indicated that, at follow-up versus baseline, both conditions reported greater smoke-free home rates (53.6% vs 38.5%) and fewer days of SHSe on average (~11 vs ~12 days), and that intervention versus control condition communities reported greater coalition awareness (24.3% vs 12.2%) and activity exposure (71.2% vs 64.5%). Multivariable modelling indicated that intervention (vs control) communities reported greater rates of complete smoke-free homes (adjusted Odds Ratio [aOR] 1.55, 95% confiedence interval [CI] 1.11 to 2.18, p=0.011) and coalition awareness (aOR 2.89, 95% CI 1.44 to 8.05, p=0.043) at follow-up. However, there were no intervention effects on policy support, SHSe or community-based activity exposure.

CONCLUSIONS

Findings must be considered alongside several sociopolitical factors during the study, including national smoke-free policies implementation (Georgia, 2018; Armenia, 2022), these countries' participation in an international tobacco legislation initiative, the COVID-19 pandemic and regional/local war). The intervention effect on smoke-free homes is critical, as smoke-free policy implementation provides opportunities to accelerate smoke-free home adoption via local coalitions.

TRIAL REGISTRATION NUMBER

NCT03447912.

Practical and methodological challenges when conducting a cluster randomized trial: Examples and recommendations

The use of cluster randomized trial design to answer research questions is increasing. This design and associated variants such as the cluster randomized crossover and stepped wedge are useful to assess complex interventions in a pragmatic way but when adopting such designs, one may face specific implementation challenges. This article summarizes common challenges faced when conducting cluster randomized trials, cluster randomized crossover trials, and stepped wedge trials, and provides recommendations.

An integrated newborn care kit (iNCK) to save newborn lives and improve health outcomes in Gilgit Baltistan (GB), Pakistan: study protocol for a cluster randomized controlled trial

BACKGROUND

Ongoing high neonatal mortality rates (NMRs) represent a global challenge. In 2021, of the 5 million deaths reported worldwide for children under five years of age, 47% were newborns. Pakistan has one of the five highest national NMRs in the world, with an estimated 39 neonatal deaths per 1,000 live births. Reducing newborn deaths requires sustainable, evidence-based, and cost-effective interventions that can be integrated within existing community healthcare infrastructure across regions with high NMR.

METHODS

This pragmatic, community-based, parallel-arm, open-label, cluster randomized controlled trial aims to estimate the effect of Lady Health Workers (LHWs) providing an integrated newborn care kit (iNCK) with educational instructions to pregnant women in their third trimester, compared to the local standard of care in Gilgit-Baltistan, Pakistan, on neonatal mortality and other newborn and maternal health outcomes. The iNCK contains a clean birth kit, 4% chlorhexidine topical gel, sunflower oil emollient, a ThermoSpot™ temperature monitoring sticker, a fleece blanket, a click-to-heat reusable warmer, three 200 μg misoprostol tablets, and a pictorial instruction guide and diary. LHWs are also provided with a handheld scale to weigh the newborn. The primary study outcome is neonatal mortality, defined as a newborn death in the first 28 days of life.

DISCUSSION

This study will generate policy-relevant knowledge on the effectiveness of integrating evidence-based maternal and newborn interventions and delivering them directly to pregnant women via existing community health infrastructure, for reducing neonatal mortality and morbidity, in a remote, mountainous area with a high NMR.

TRIAL REGISTRATION

NCT04798833, March 15, 2021.

Eliminating Monitor Overuse (EMO) type III effectiveness-deimplementation cluster-randomized trial: Statistical analysis plan

Background

Deimplementing overused health interventions is essential to maximizing quality and value while minimizing harm, waste, and inefficiencies. Three national guidelines discourage continuous pulse oximetry (SpO2) monitoring in children who are not receiving supplemental oxygen, but the guideline-discordant practice remains prevalent, making it a prime target for deimplementation. This paper details the statistical analysis plan for the Eliminating Monitor Overuse (EMO) SpO2 trial, which compares the effect of two competing deimplementation strategies (unlearning only vs. unlearning plus substitution) on the sustainment of deimplementation of SpO2 monitoring in children with bronchiolitis who are in room air.

Methods

The EMO Trial is a hybrid type 3 effectiveness-deimplementation trial with a longitudinal cluster-randomized design, conducted in Pediatric Research in Inpatient Settings Network hospitals. The primary outcome is deimplementation sustainment, analyzed as a longitudinal difference-in-differences comparison between study arms. This analysis will use generalized hierarchical mixed-effects models for longitudinal clustering outcomes. Secondary outcomes include the length of hospital stay and oxygen supplementation duration, modeled using linear mixed-effects regressions. Using the well-established counterfactual approach, we will also perform a mediation analysis of hospital-level mechanistic measures on the association between the deimplementation strategy and the sustainment outcome.

Discussion

We anticipate that the EMO Trial will advance the science of deimplementation by providing new insights into the processes, mechanisms, and likelihood of sustained practice change using rigorously designed deimplementation strategies. This pre-specified statistical analysis plan will mitigate reporting bias and support data-driven approaches.

Trial registration

ClinicalTrials.gov NCT05132322. Registered on 24 November 2021.

Statistical analysis plan for the LAKANA trial: a cluster-randomized, placebo-controlled, double-blinded, parallel group, three-arm clinical trial testing the effects of mass drug administration of azithromycin on mortality and other outcomes among 1-11-month-old infants in Mali

BACKGROUND

The Large-scale Assessment of the Key health-promoting Activities of two New mass drug administration regimens with Azithromycin (LAKANA) trial in Mali aims to evaluate the efficacy and safety of azithromycin (AZI) mass drug administration (MDA) to 1-11-month-old infants as well as the impact of the intervention on antimicrobial resistance (AMR) and mechanisms of action of azithromycin. To improve the transparency and quality of this clinical trial, we prepared this statistical analysis plan (SAP).

METHODS/DESIGN

LAKANA is a cluster randomized trial that aims to address the mortality and health impacts of biannual and quarterly AZI MDA. AZI is given to 1-11-month-old infants in a high-mortality setting where a seasonal malaria chemoprevention (SMC) program is in place. The participating villages are randomly assigned to placebo (control), two-dose AZI (biannual azithromycin-MDA), and four-dose AZI (quarterly azithromycin-MDA) in a 3:4:2 ratio. The primary outcome of the study is mortality among the intention-to-treat population of 1-11-month-old infants. We will evaluate relative risk reduction between the study arms using a mixed-effects Poisson model with random intercepts for villages, using log link function with person-years as an offset variable. We will model outcomes related to secondary objectives of the study using generalized linear models with considerations on clustering.

CONCLUSION

The SAP written prior to data collection completion will help avoid reporting bias and data-driven analysis for the primary and secondary aims of the trial. If there are deviations from the analysis methods described here, they will be described and justified in the publications of the trial results.

TRIAL REGISTRATION

ClinicalTrials.gov ID NCT04424511 . Registered on 11 June 2020.

A cluster-randomised controlled equivalence trial of the Surprise Soap handwashing intervention among older children living in a refugee settlement in Sudan

INTRODUCTION

Increasing handwashing with soap (HWWS) among older children in emergency settings can have a large public health impact, however, evidence on what works is limited. One promising approach is the 'Surprise Soap' intervention in which a novel soap with an embedded toy is delivered to children in a short, participatory household session that includes a glitter game and HWWS practice. Here, we evaluate this intervention against a standard intervention in a complex emergency setting.

METHODS

A cluster-randomised controlled equivalence trial was conducted in Naivasha refugee settlement, Sudan. Blinding was not possible. 203 randomly selected households, with at least one child aged 5-12, were randomised to receive the Surprise Soap intervention (n=101) or a standard intervention comprising a short household session with health messaging and plain soap distribution (n=102). The primary outcome was the proportion of prespecified potential HWWS events observed for children aged 5-12, accompanied by HWWS, at baseline, 4, 12 and 16 weeks post intervention delivery.

RESULTS

200 households were included in the analyses: 101 intervention and 99 control. No difference in intervention effectiveness was observed at any follow-up (4 weeks: adjusted rate ratio (RR) 1.2, 95% CI 0.8 to 1.7; 12 weeks: RR 0.8, 95% CI 0.5 to 1.1; 16 weeks: RR 1.1, 95% CI 0.8 to 1.5). However, we observed increased HWWS in both arms at 4 weeks (27 and 23 percentage point increase in the intervention and control arm, respectively) that was sustained at 16 weeks.

CONCLUSIONS

We find that the Surprise Soap intervention is no more effective at increasing older children's HWWS than a standard, household-level, health-based intervention in this complex humanitarian emergency. There appears to be no marginal benefit in terms of HWWS that would justify the additional cost of implementing the Surprise Soap intervention. Further trials that include a passive control arm are needed to determine the independent effects of each intervention and guide future intervention design.

Restricted mean survival time to estimate an intervention effect in a cluster randomized trial

For time-to-event outcomes, the difference in restricted mean survival time is a measure of the intervention effect, an alternative to the hazard ratio, corresponding to the expected survival duration gain due to the intervention up to a predefined time t*. We extended two existing approaches of restricted mean survival time estimation for independent data to clustered data in the framework of cluster randomized trials: one based on the direct integration of Kaplan-Meier curves and the other based on pseudo-values regression. Then, we conducted a simulation study to assess and compare the statistical performance of the proposed methods, varying the number and size of clusters, the degree of clustering, and the magnitude of the intervention effect under proportional and non-proportional hazards assumption. We found that the extended methods well estimated the variance and controlled the type I error if there was a sufficient number of clusters (≥ 50) under both proportional and non-proportional hazards assumption. For cluster randomized trials with a limited number of clusters (< 50), a permutation test for pseudo-values regression was implemented and corrected the type I error. We also provided a procedure to estimate permutation-based confidence intervals which produced adequate coverage. All the extended methods performed similarly, but the pseudo-values regression offered the possibility to adjust for covariates. Finally, we illustrated each considered method with a cluster randomized trial evaluating the effectiveness of an asthma-control education program.

Protocol for an economic evaluation of a tele-neurologic intervention alongside a stepped wedge randomised controlled trial (NeTKoH)

BACKGROUND

A significant and growing portion of the global burden of diseases is caused by neurological disorders. Tele-neurology has the potential to improve access to health care services and the quality of care, particularly in rural and underserved areas. The economic evaluation of the stepped wedge randomised controlled trial NeTKoH aims to ascertain the cost-effectiveness and cost-utility regarding the effects of a tele-neurologic intervention in primary care in a rural area in Germany.

METHODS

This protocol outlines the methods used when conducting the trial-based economic evaluation of NeTKoH. The outcomes used in our economic analysis are all prespecified endpoints of the NeTKoH trial. Outcomes considered for the cost-utility and cost-effectiveness analyses will be quality-adjusted life years (QALYs) derived from the EQ-5D-5L, proportion of neurologic problems being solved at the GP's office (primary outcome), hospital length-of-stay and number of hospital stays. Costs will be prospectively collected during the trial by the participating statutory health insurances, and will be analysed from a statutory health insurance perspective within the German health care system. This economic evaluation will be reported complying with the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist.

DISCUSSION

This within-trial economic evaluation relaying the costs and outcomes of an interdisciplinary tele-consulting intervention will provide high-quality evidence for cost-effectiveness and policy implications of a tele-neurological programme, including the potential for application in other rural areas in Germany or other jurisdictions with a comparable health system.

TRIAL REGISTRATION

German Clinical Trials Register (DRKS00024492), date registered: September 28, 2021.

Cluster randomized trials in nursing homes should better be planned as open-cohort than as closed-cohort

OBJECTIVES

Two designs are frequently used in cluster randomized trials in nursing homes: closed cohort and open cohort. The former design includes residents at the beginning of the trial and then follows them. In the latter design, participants are enrolled at the beginning of the trial or although it is ongoing; at dates of assessment, all residents present in the nursing home are assessed. The open-cohort design is much less used than the closed-cohort design, but it offers several advantages such as less exposure to individual attrition. Objective was to assess whether an open-cohort design could have been feasible in trials with a closed-cohort design.

STUDY DESIGN AND SETTING

Twenty-two closed-cohort trials in nursing homes.

RESULTS

An open-cohort design was considered a relevant alternative for 20 trials. For 16 trials, a resident newly admitted could not opt out of the intervention, and for all trials, the resident could benefit from an intervention effect if it existed. For two trials, newly admitted residents could not benefit from the intervention effect, if it existed.

CONCLUSION

The open-cohort design is well-adapted for most of the interventions assessed in nursing homes by means of a cluster randomized trial and should be considered more often.

Community Walks: a cluster randomized controlled trial of a multilevel physical activity intervention for low income public housing residents

BACKGROUND

Physical activity behavioral interventions to change individual-level drivers of activity, like motivation, attitudes, and self-efficacy, are often not sustained beyond the intervention period. Interventions at both environmental and individual levels might facilitate durable change. This community-based study seeks to test a multilevel, multicomponent intervention to increase moderate intensity physical activity among people with low incomes living in U.S. public housing developments, over a 2 year period.

METHODS

The study design is a prospective, cluster randomized controlled trial, with housing developments (n=12) as the units of randomization. In a four-group, factorial trial, we will compare an environmental intervention (E) alone (3 developments), an individual intervention (I) alone (3 developments), an environmental plus individual (E+I) intervention (3 developments), against an assessment only control group (3 developments). The environmental only intervention consists of community health workers leading walking groups and indoor activities, a walking advocacy program for residents, and provision of walking maps/signage. The individual only intervention consists of a 12-week automated telephone program to increase physical activity motivation and self-efficacy. All residents are invited to participate in the intervention activities being delivered at their development. The primary outcome is change in moderate intensity physical activity measured via an accelerometer-based device among an evaluation cohort (n=50 individuals at each of the 12 developments) from baseline to 24-month follow up. Mediation (e.g., neighborhood walkability, motivation) and moderation (e.g., neighborhood stress) of our interventions will be assessed. Lastly, we will interview key informants to assess factors from the Consolidated Framework for Implementation Research domains to inform future implementation.

DISCUSSION

We hypothesize participants living in developments in any of the three intervention groups (E only, I only, and E+I combined) will increase minutes of moderate intensity physical activity more than participants in control group developments. We expect delivery of an intervention package targeting environmental and social factors to become active, combined with the individual level intervention, will improve overall physical activity levels to recommended guidelines at the development level. If effective, this trial has the potential for implementation through other federal and state housing authorities.

TRIAL REGISTRATION

Clinical Trails.gov PRS Protocol Registration and Results System, NCT05147298 . Registered 28 November 2021.

[French version of The Ottawa Statement on the Ethical Design and Conduct of Cluster Randomized Trials, in a French law context]

INTRODUCTION

There is growing evidence on the ethical challenges raised by cluster randomized trials. This specificity is not reflected in the legal texts regulating research, which creates difficulties for researchers implementing these experimental designs. The Ottawa Statement (Weijer et al. 2012) aims to provide detailed guidance on the ethical design, conduct and assessment of cluster trials. More broadly aims to help research stakeholders and decision-makers to make informed ethical decisions regarding the particularity of these experimental designs. It seems that this international statement, written in English, is not sufficiently accessible to all of the French professionals involved in health research. The aim of this article is to provide these professionals with a contextualized and illustrated French translation of the "Ottawa statement".

METHOD

. The "complex design" working group of the RECaP network (Research in Clinical Epidemiology and Public Health), carried out this work. A first version was discussed by the authors in several meetings. It was completed by contextual explanations and examples of French studies currently conducted by the authors. The final version was obtained by consensus and validated by the group.

RESULTS

. This work reports 15 recommendations grouped into 7 key questions: How to justify cluster design? How to submit an article to an ethics committee? How to identify research participants? How and when to obtain informed consent? Who are the gatekeepers? How to assess benefits and harm? How to protect vulnerable participants? Each of these recommendations is specific to cluster trials. The recommendations are explained and detailed through concrete examples.

CONCLUSION

Without interfering with current French laws, this work provides a framework for the organization, conduct and ethical assessment of cluster randomized trials in France. In the present-day context, it is essential that all concerned groups can base their decisions on recommendations in line with the elementary principles of health research ethics.

Does ignoring transmission dynamics lead to underestimation of the impact of interventions against mosquito-borne disease?

New vector-control technologies to fight mosquito-borne diseases are urgently needed, the adoption of which depends on efficacy estimates from large-scale cluster-randomised trials (CRTs). The release of Wolbachia-infected mosquitoes is one promising strategy to curb dengue virus (DENV) transmission, and a recent CRT reported impressive reductions in dengue incidence following the release of these mosquitoes. Such trials can be affected by multiple sources of bias, however. We used mathematical models of DENV transmission during a CRT of Wolbachia-infected mosquitoes to explore three such biases: human movement, mosquito movement and coupled transmission dynamics between trial arms. We show that failure to account for each of these biases would lead to underestimated efficacy, and that the majority of this underestimation is due to a heretofore unrecognised bias caused by transmission coupling. Taken together, our findings suggest that Wolbachia-infected mosquitoes could be even more promising than the recent CRT suggested. By emphasising the importance of accounting for transmission coupling between arms, which requires a mathematical model, we highlight the key role that models can play in interpreting and extrapolating the results from trials of vector control interventions.

An overview of optimal designs under a given budget in cluster randomized trials with a binary outcome

Cluster randomized trial design may raise financial concerns because the cost to recruit an additional cluster is much higher than to enroll an additional subject in subject-level randomized trials. Therefore, it is desirable to develop an optimal design. For local optimal designs, optimization means the minimum variance of the estimated treatment effect under the total budget. The local optimal design derived from the variance needs the input of an association parameter ρ in terms of a "working" correlation structure R ( ρ ) in the generalized estimating equation models. When the range of ρ instead of an exact value is available, the parameter space is defined as the range of ρ and the design space is defined as enrollment feasibility, for example, the number of clusters or cluster size. For any value ρ within the range, the optimal design and relative efficiency for each design in the design space is obtained. Then, for each design in the design space, the minimum relative efficiency within the parameter space is calculated. MaxiMin design is the optimal design that maximizes the minimum relative efficiency among all designs in the design space. Our contributions are threefold. First, for three common measures (risk difference, risk ratio, and odds ratio), we summarize all available local optimal designs and MaxiMin designs utilizing generalized estimating equation models when the group allocation proportion is predetermined for two-level and three-level parallel cluster randomized trials. We then propose the local optimal designs and MaxiMin designs using the same models when the group allocation proportion is undecided. Second, for partially nested designs, we develop the optimal designs for three common measures under the setting of equal number of subjects per cluster and exchangeable working correlation structure in the intervention group. Third, we create three new Statistical Analysis System (SAS) macros and update two existing SAS macros for all the optimal designs. We provide two examples to illustrate our methods.

Improving health literacy responsiveness to reduce ethnic and social disparity in stillbirth and infant health: A cluster randomized controlled effectiveness trial of the MAMAACT intervention

OBJECTIVE

The MAMAACT intervention aims to reduce ethnic and social disparities in stillbirth and infant death by improving communication between pregnant women and midwives regarding warning signs of pregnancy complications. This study evaluates the effect of the intervention on pregnant women's health literacy (two domains from the Health Literacy Questionnaire) and complication management - interpreted as improved health literacy responsiveness among midwives.

DESIGN

Cluster randomized controlled trial, 2018-2019.

SETTING

19 of 20 Danish maternity wards.

PARTICIPANTS

Cross-sectional survey data were collected using telephone interviews (n = 4150 pregnant women including 670 women with a non-Western immigrant background).

INTERVENTION

A six-hour training session for midwives in intercultural communication and cultural competence, two follow-up dialog meetings, and health education materials for pregnant women on warning signs of pregnancy complications - in six languages.

MAIN OUTCOME MEASURES

Differences in mean scores at post-implementation of the domains Active engagement with healthcare providers (Active engagement) and Navigating the healthcare system from the Health Literacy Questionnaire, and differences in the certainty of how to respond to pregnancy complication signs between women in the intervention and control group.

RESULTS

No difference was observed in women's level of Active engagement or Navigating the healthcare system. Women from the intervention group were more certain of how to respond to complication signs: Redness, swelling, and heat in one leg: 69.4 % vs 59.1 %; aOR 1.57 (95 % CI 1.32-1.88), Severe headache: 75.6 % vs 67.3 %; aOR 1.50 (95 % CI 1.24-1.82), and Vaginal bleeding: 97.3 % vs 95.1 %; aOR 1.67 (95 % CI 1.04-2.66).

CONCLUSION

The intervention improved women's certainty of how to respond to complication signs, but was unable to improve pregnant women's health literacy levels of Active engagement and Navigating the healthcare system, likely due to barriers related to the organization of antenatal care. A reorganization of antenatal care and a care model sensitive to diversity within the entire healthcare system might help reduce disparities in perinatal health.

TRIAL REGISTRATION

ClinicalTrials.gov identifier: NCT03751774.

Evaluating an app for digital medical history taking in urgent care practices: study protocol of the cluster-randomized interventional trial 'DASI'

BACKGROUND

In out-of-hours urgent care practices in Germany, physicians of different specialties care for a large number of patients, most of all unknown to them, resulting in a high workload and challenging diagnostic decision-making. As there is no common patient file, physicians have no information about patients' previous conditions or received treatments. In this setting, a digital tool for medical history taking could improve the quality of medical care. This study aims to implement and evaluate a software application (app) that takes a structured symptom-oriented medical history from patients in urgent care settings.

METHODS

We conduct a time-cluster-randomized trial in two out-of-hours urgent care practices in Germany for 12 consecutive months. Each week during the study defines a cluster. We will compare participants with (intervention group) and without app use (control group) prior to consultation and provision of the self-reported information for the physician. We expect the app to improve diagnostic accuracy (primary outcome), reduce physicians' perceived diagnostic uncertainty, and increase patients' satisfaction and the satisfaction with communication of both physician and patient (secondary outcomes).

DISCUSSION

While similar tools have only been subject to small-scale pilot studies surveying feasibility and usability, the present study uses a rigorous study design to measure outcomes that are directly associated with the quality of delivered care.

TRIAL REGISTRATION

The study was registered at the German Clinical Trials Register (No. DRKS00026659 registered Nov 03 2021. World Health Organization Trial Registration Data Set, https://trialsearch.who.int/Trial2.aspx? TrialID = DRKS00026659.

Accounting for expected attrition in the planning of cluster randomized trials for assessing treatment effect heterogeneity

BACKGROUND

Detecting treatment effect heterogeneity is an important objective in cluster randomized trials and implementation research. While sample size procedures for testing the average treatment effect accounting for participant attrition assuming missing completely at random or missing at random have been previously developed, the impact of attrition on the power for detecting heterogeneous treatment effects in cluster randomized trials remains unknown.

METHODS

We provide a sample size formula for testing for a heterogeneous treatment effect assuming the outcome is missing completely at random. We also propose an efficient Monte Carlo sample size procedure for assessing heterogeneous treatment effect assuming covariate-dependent outcome missingness (missing at random). We compare our sample size methods with the direct inflation method that divides the estimated sample size by the mean follow-up rate. We also evaluate our methods through simulation studies and illustrate them with a real-world example.

RESULTS

Simulation results show that our proposed sample size methods under both missing completely at random and missing at random provide sufficient power for assessing heterogeneous treatment effect. The proposed sample size methods lead to more accurate sample size estimates than the direct inflation method when the missingness rate is high (e.g., ≥ 30%). Moreover, sample size estimation under both missing completely at random and missing at random is sensitive to the missingness rate, but not sensitive to the intracluster correlation coefficient among the missingness indicators.

CONCLUSION

Our new sample size methods can assist in planning cluster randomized trials that plan to assess a heterogeneous treatment effect and participant attrition is expected to occur.

Keep It Up! 3.0: Study protocol for a type III hybrid implementation-effectiveness cluster-randomized trial

BACKGROUND

Despite evidence that eHealth approaches can be effective in reducing HIV risk, their implementation requirements for public health scale up are not well established, and effective strategies to bring these programs into practice are still unknown. Keep It Up! (KIU!) is an online program proven to reduce HIV risk among young men who have sex with men (YMSM) and ideal candidate to develop and evaluate novel strategies for implementing eHealth HIV prevention programs. KIU! 3.0 is a Type III Hybrid Effectiveness-Implementation cluster randomized trial designed to 1) compare two strategies for implementing KIU!: community-based organizations (CBO) versus centralized direct-to-consumer (DTC) recruitment; 2) examine the effect of strategies and determinants on variability in implementation success; and 3) develop materials for sustainment of KIU! after the trial concludes. In this article, we describe the approaches used to achieve these aims.

METHODS

Using county-level population estimates of YMSM, 66 counties were selected and randomized 2:1 to the CBO and DTC approaches. The RE-AIM model was used to drive outcome measurements, which were collected from CBO staff, YMSM, and technology providers. Mixed-methods research mapped onto the domains of the Consolidated Framework for Implementation Research will examine determinants and their relationship with implementation outcomes.

DISCUSSION

In comparing our implementation recruitment models, we are examining two strategies which have shown effectiveness in delivering health technology interventions in the past, yet little is known about their comparative advantages and disadvantages in implementation. The results of the trial will further the understanding of eHealth prevention intervention implementation.

A cluster randomized controlled trial of an electronic decision-support system to enhance antenatal care services in pregnancy at primary healthcare level in Telangana, India: trial protocol

BACKGROUND

India contributes 15% of the total global maternal mortality burden. An increasing proportion of these deaths are due to Pregnancy Induced Hypertension (PIH), Gestational Diabetes Mellitus (GDM), and anaemia. This study aims to evaluate the effectiveness of a tablet-based electronic decision-support system (EDSS) to enhance routine antenatal care (ANC) and improve the screening and management of PIH, GDM, and anaemia in pregnancy in primary healthcare facilities of Telangana, India. The EDSS will work at two levels of primary health facilities and is customized for three cadres of healthcare providers - Auxiliary Nurse Midwifes (ANMs), staff nurses, and physicians (Medical Officers).

METHODS

This will be a cluster randomized controlled trial involving 66 clusters with a total of 1320 women in both the intervention and control arms. Each cluster will include three health facilities-one Primary Health Centre (PHC) and two linked sub-centers (SC). In the facilities under the intervention arm, ANMs, staff nurses, and Medical Officers will use the EDSS while providing ANC for all pregnant women. Facilities in the control arm will continue to provide ANC services using the existing standard of care in Telangana. The primary outcome is ANC quality, measured as provision of a composite of four selected ANC components (measurement of blood pressure, blood glucose, hemoglobin levels, and conducting a urinary dipstick test) by the healthcare providers per visit, observed over two visits. Trained field research staff will collect outcome data via an observation checklist.

DISCUSSION

To our knowledge, this is the first trial in India to evaluate an EDSS, targeted to enhance the quality of ANC and improve the screening and management of PIH, GDM, and anaemia, for multiple levels of health facilities and several cadres of healthcare providers. If effective, insights from the trial on the feasibility and cost of implementing the EDSS can inform potential national scale-up. Lessons learned from this trial will also inform recommendations for designing and upscaling similar mHealth interventions in other low and middle-income countries.

TRIAL REGISTRATION

CLINICALTRIALS

gov, NCT03700034, registered 9 Oct 2018, https://www.

CLINICALTRIALS

gov/ct2/show/NCT03700034 CTRI, CTRI/2019/01/016857, registered on 3 Mar 2019, http://www.ctri.nic.in/Clinicaltrials/pdf_generate.php?trialid=28627&EncHid=&modid=&compid=%27,%2728627det%27.

Response to: "Effect of behavioral activation on time and frequency domain heart rate variability in older adults with subthreshold depression: a cluster randomized controlled trial in Thailand"

Ayudhaya et al. examined the effect of Behavioral Activation on daily step count and heart rate variability among older adults with depression in a study labeled a cluster randomized controlled trial (cRCT). However, only one cluster was assigned to either of the study conditions. Such a design would have zero degrees of freedom for inferential testing, because the variation due to cluster membership cannot be estimated apart from the variation due to treatment assignment. Thus, the intervention effect is completely confounded with the cluster effect. The study should be labeled a quasi-experimental study, not a cRCT. Accordingly, the numerical results should be interpreted as associations but not evidence for causal relationships.

Implementation research of a cluster randomized trial evaluating the implementation and effectiveness of intermittent preventive treatment for malaria using dihydroartemisinin-piperaquine on reducing malaria burden in school-aged children in Tanzania: methodology, challenges, and mitigation

BACKGROUND

It has been more than 20 years since the malaria epidemiologic shift to school-aged children was noted. In the meantime, school-aged children (5-15 years) have become increasingly more vulnerable with asymptomatic malaria prevalence reaching up to 70%, making them reservoirs for subsequent transmission of malaria in the endemic communities. Intermittent Preventive Treatment of malaria in schoolchildren (IPTsc) has proven to be an effective tool to shrink this reservoir. As of 3^rd June 2022, the World Health Organization recommends IPTsc in moderate and high endemic areas. Even so, for decision-makers, the adoption of scientific research recommendations has been stifled by real-world implementation challenges. This study presents methodology, challenges faced, and mitigations used in the evaluation of the implementation of IPTsc using dihydroartemisinin-piperaquine (DP) in three councils (Handeni District Council (DC), Handeni Town Council (TC) and Kilindi DC) of Tanga Region, Tanzania so as to understand the operational feasibility and effectiveness of IPTsc on malaria parasitaemia and clinical malaria incidence.

METHODS

The study deployed an effectiveness-implementation hybrid design to assess feasibility and effectiveness of IPTsc using DP, the interventional drug, against standard of care (control). Wards in the three study councils were the randomization unit (clusters). Each ward was randomized to implement IPTsc or not (control). In all wards in the IPTsc arm, DP was given to schoolchildren three times a year in four-month intervals. In each council, 24 randomly selected wards (12 per study arm, one school per ward) were chosen as representatives for intervention impact evaluation. Mixed design methods were used to assess the feasibility and acceptability of implementing IPTsc as part of a more comprehensive health package for schoolchildren. The study reimagined an existing school health programme for Neglected Tropical Diseases (NTD) control include IPTsc implementation.

RESULTS

The study shows IPTsc can feasibly be implemented by integrating it into existing school health and education systems, paving the way for sustainable programme adoption in a cost-effective manner.

CONCLUSIONS

Through this article other interested countries may realise a feasible plan for IPTsc implementation. Mitigation to any challenge can be customized based on local circumstances without jeopardising the gains expected from an IPTsc programme. Trial registration clinicaltrials.gov, NCT04245033. Registered 28 January 2020, https://clinicaltrials.gov/ct2/show/NCT04245033.

Effectiveness of a community-based education and peer support led by women's self-help groups in improving the control of hypertension in urban slums of Kerala, India: a cluster randomised controlled pragmatic trial

BACKGROUND

With less than 20% of people with hypertension achieving their target blood pressure (BP) goals, uncontrolled hypertension remains a major public health problem in India. We conducted a study to assess the effectiveness of a community-based education and peer support programme led by women's self-help group (SHG) members in reducing the mean systolic BP among people with hypertension in urban slums of Kochi city, Kerala, India.

METHODS

A cluster randomised controlled pragmatic trial was conducted where 20 slums were randomised to either the intervention or the control arms. In each slum, participants who had elevated BP (>140/90) or were on antihypertensive medications were recruited. The intervention was delivered through women's SHG members (1 per 20-30 households) who provided (1) assistance in daily hypertension management, (2) social and emotional support to encourage healthy behaviours and (3) referral to the primary healthcare system. Those in the control arm received standard of care. The primary outcome was change in mean systolic BP (SBP) after 6 months.

RESULTS

A total of 1952 participants were recruited-968 in the intervention arm and 984 in the control arm. Mean SBP was reduced by 6.26 mm Hg (SE 0.69) in the intervention arm compared with 2.16 mm Hg (SE 0.70) in the control arm; the net difference being 4.09 (95% CI 2.15 to 4.09), p<0.001.

CONCLUSION

This women's SHG members led community intervention was effective in reducing SBP among people with hypertension compared with those who received usual care, over 6 months in urban slums of Kerala, India.

TRIAL REGISTRATION NUMBER

CTRI/2019/12/022252.

Randomized controlled trials in de-implementation research: a systematic scoping review

BACKGROUND

Healthcare costs are rising, and a substantial proportion of medical care is of little value. De-implementation of low-value practices is important for improving overall health outcomes and reducing costs. We aimed to identify and synthesize randomized controlled trials (RCTs) on de-implementation interventions and to provide guidance to improve future research.

METHODS

MEDLINE and Scopus up to May 24, 2021, for individual and cluster RCTs comparing de-implementation interventions to usual care, another intervention, or placebo. We applied independent duplicate assessment of eligibility, study characteristics, outcomes, intervention categories, implementation theories, and risk of bias.

RESULTS

Of the 227 eligible trials, 145 (64%) were cluster randomized trials (median 24 clusters; median follow-up time 305 days), and 82 (36%) were individually randomized trials (median follow-up time 274 days). Of the trials, 118 (52%) were published after 2010, 149 (66%) were conducted in a primary care setting, 163 (72%) aimed to reduce the use of drug treatment, 194 (85%) measured the total volume of care, and 64 (28%) low-value care use as outcomes. Of the trials, 48 (21%) described a theoretical basis for the intervention, and 40 (18%) had the study tailored by context-specific factors. Of the de-implementation interventions, 193 (85%) were targeted at physicians, 115 (51%) tested educational sessions, and 152 (67%) multicomponent interventions. Missing data led to high risk of bias in 137 (60%) trials, followed by baseline imbalances in 99 (44%), and deficiencies in allocation concealment in 56 (25%).

CONCLUSIONS

De-implementation trials were mainly conducted in primary care and typically aimed to reduce low-value drug treatments. Limitations of current de-implementation research may have led to unreliable effect estimates and decreased clinical applicability of studied de-implementation strategies. We identified potential research gaps, including de-implementation in secondary and tertiary care settings, and interventions targeted at other than physicians. Future trials could be improved by favoring simpler intervention designs, better control of potential confounders, larger number of clusters in cluster trials, considering context-specific factors when planning the intervention (tailoring), and using a theoretical basis in intervention design.

REGISTRATION

OSF Open Science Framework hk4b2.

Implementing a text-messaging intervention for unhealthy alcohol use in emergency departments: protocol for implementation strategy development and a pilot cluster randomized implementation trial

BACKGROUND

Unhealthy alcohol use (UAU) is a leading cause of premature mortality among adults in the USA. Emergency departments (EDs) are key intervention settings for UAU but often have limited time and resources. One low-burden, scalable approach to address UAU is text-messaging interventions. Despite strong research support and promise for scalability, there is little research on how to implement such interventions in healthcare settings. The process of providers making them available to patients in an efficient way within already busy and overburdened ED workflows and patients adopting them remains a new area of research. The purpose of this three-phase study is to develop and test an implementation strategy for UAU text-messaging interventions in EDs.

METHOD

Our first aim is to examine barriers and facilitators to staff offering and patients accepting a text-messaging intervention in the ED using an explanatory, sequential mixed methods approach. We will examine alcohol screening data in the electronic health records of 17 EDs within a large integrated health system in the Northeast and conduct surveys among chairpersons in each. This data will be used to purposively sample 4 EDs for semi-structured interviews among 20 clinical staff, 20 patients, and 4 chairpersons. Our second aim is to conduct a stakeholder-engaged intervention mapping process to develop a multi-component implementation strategy for EDs. Our third aim is to conduct a mixed method 2-arm cluster randomized pilot study in 4 EDs that serve ~11,000 UAU patients per year to assess the feasibility, acceptability, and preliminary effectiveness of the implementation strategy. The Integrated Promoting Action on Research Implementation in Health Services framework will guide study activities.

DISCUSSION

Low-burden technology, like text messaging, along with targeted implementation support and strategies driven by identified barriers and facilitators could sustain large-scale ED-based alcohol screening programs and provide much needed support to patients who screen positive while reducing burden on EDs. The proposed study would be the first to develop and test this targeted implementation strategy and will prepare for a larger, fully powered hybrid effectiveness-implementation trial. Findings may also be broadly applicable to implementation of patient-facing mobile health technologies.

TRIAL REGISTRATION

This study was registered at ClinicalTrials.gov (NCT05350878) on 4/28/2022.

The ring vaccination trial design for the estimation of vaccine efficacy and effectiveness during infectious disease outbreaks

The ring vaccination trial is a recently developed approach for evaluating the efficacy and effectiveness of vaccines, modeled after the surveillance and containment strategy of ring vaccination. Contacts and contacts of contacts of a newly identified disease case form a ring, and these rings are randomized as part of a cluster-randomized trial or with individual randomization within rings. Key advantages of the design include its flexibility to follow the epidemic as it progresses and the targeting of high-risk participants to increase power. We describe the application of the design to estimate the efficacy and effectiveness of an Ebola vaccine during the 2014-2016 West African Ebola epidemic. The design has several notable statistical features. Because vaccination occurs around the time of exposure, the design is particularly sensitive to the choice of per protocol analysis period. If incidence wanes before the per protocol analysis period begins (due to a slow-acting vaccine or a fast-moving pathogen), power can be substantially reduced. Mathematical modeling is valuable for exploring the suitability of the approach in different disease settings. Another statistical feature is zero inflation, which can occur if the chain of transmission does not take off within a ring. In the application to Ebola, the majority of rings had zero subsequent cases. The ring vaccination trial can be extended in several ways, including the definition of rings (e.g. contact-based, spatial, and occupational). The design will be valuable in settings where the spatio-temporal spread of the pathogen is highly focused and unpredictable.

Improving Chronic Illness Self-Management with the Apsáalooke Nation: The Báa nnilah Project, a cluster randomized trial protocol

Chronic illness (CI) is a major cause of morbidity and mortality for Indigenous people. In Montana, Indigenous communities disproportionately experience CI, a legacy of settler colonialism. For over two decades, Messengers for Health, an Apsáalooke (Crow Indian) non-profit, and Montana State University have partnered to improve community health using a community-based participatory research (CBPR) approach. We developed Báa nnilah, an intervention utilizing community strengths, to improve CI self-management. This manuscript describes the protocol for a cluster randomized trial with two arms: an intervention group and a wait list control group, who both participated in the Báa nnilah program. Enrollment occurred through family/clan networks and community outreach and attended to limitations of existing CI self-management interventions by using an approach and content that were culturally consonant. Participants received program materials, attended seven gatherings focused on improving CI management, and received and shared health information through storytelling based on a conceptual framework from the Apsáalooke culture and incorporating CI self-management strategies. Participant support occurred within partnership dyads during and between gatherings, from community mentors, and by program staff. The study used mixed methods to evaluate the intervention, with qualitative measures including the Short Form Health Survey (SF-12), Patient Health Questionnaire (PHQ-9), Patient Activation Measure (PAM), and a suite of PROMIS measures, various physical tests and qualitative survey responses, semi-structured interviews, and outcomes shared by participants with program staff. We hypothesized that Báa nnilah would significantly improve participant health outcome measures across multiple dimensions with quality of life (QoL) as the primary outcome. TRIAL REGISTRATION: ClinicalTrials.gov: NCT03036189. Registered on 30 January 2017. (From https://clinicaltrials.gov/ct2/show/NCT03036189).

Effect of home-based childcare on childhood mortality in rural Maharashtra, India: a cluster randomised controlled trial

Background

Melghat, an impoverished rural area in Maharashtra state, India; has scarce hospital services and low health-seeking behaviour. At baseline (2004) the under-five mortality rate (U5MR) (number of deaths in children aged 0–5 years/1000 live births) was 147.21 and infant mortality rate (IMR) (number of deaths of infants aged under 1 year/1000 live births) was 106.6 per 1000 live births. We aimed at reducing mortality rates through home-based child care (HBCC) using village health workers (VHWs).

Methods

A cluster-randomised control trial was conducted in 34 randomly assigned clusters/villages of Melghat, Maharashtra state, between 2004 and 2009. Participants included all under-five children and their parents. Interventions delivered through VHWs were patient–public involvement, newborn care, disease management and behaviour change communications. Primary outcome indicators were U5MR and IMR. Secondary outcome indicators were neonatal mortality rate (NMR) (number of neonatal deaths aged 0–28 days/1000 live births) and perinatal mortality rate (PMR) (number of stillbirths and early neonatal deaths/1000 total births). Analysis was by intention-to-treat at the individual level. This trial was extended to a service phase (2010–2015) in both arms and a government replication phase (2016–2019) only for the intervention clusters/areas (IA).

Findings

There were 18 control areas/clusters (CA) allocated and analysed with 4426 individuals, and 16 of 18 allocated IA, analysed with 3230 individuals. The IMR and U5MR in IA were reduced from 106.60 and 147.21 to 32.75 and 50.38 (reduction by 69.28% and 65.78%, respectively) compared with increases in CA from 67.67 and 105.3 to 86.83 and 122.8, respectively, from baseline to end of intervention. NMR and PMR in IA showed reductions from 50.76 to 22.67 (by 55.34%) and from 75.06 to 24.94 (by 66.77%) respectively. These gains extended to villages in the service and replication phases.

Interpretation

This socio-culturally contextualised model for HBCC through VHWs backed up with institutional support is effective for significant reduction of U5MR, IMR and NMR in impoverished rural areas. This reduction was maintained in the study area during the service phase, indicating feasibility of implementation in large-scale public health programmes. Replicability of the model was demonstrated by a linear decline in all the mortality rates in 20 new villages during the government phase.

Trial registration number

NCT02473796.

Systems analysis and improvement approach to optimize outpatient mental health treatment cascades in Mozambique (SAIA-MH): study protocol for a cluster randomized trial

BACKGROUND

Significant investments are being made to close the mental health (MH) treatment gap, which often exceeds 90% in many low- and middle-income countries (LMICs). However, limited attention has been paid to patient quality of care in nascent and evolving LMIC MH systems. In system assessments across sub-Saharan Africa, MH loss-to-follow-up often exceeds 50% and sub-optimal medication adherence often exceeds 60%. This study aims to fill a gap of evidence-based implementation strategies targeting the optimization of MH treatment cascades in LMICs by testing a low-cost multicomponent implementation strategy integrated into routine government MH care in Mozambique.

METHODS

Using a cluster-randomized trial design, 16 clinics (8 intervention and 8 control) providing primary MH care will be randomized to the Systems Analysis and Improvement Approach for Mental Health (SAIA-MH) or an attentional placebo control. SAIA-MH is a multicomponent implementation strategy blending external facilitation, clinical consultation, and provider team meetings with system-engineering tools in an overall continuous quality improvement framework. Following a 6-month baseline period, intervention facilities will implement the SAIA-MH strategy for a 2-year intensive implementation period, followed by a 1-year sustainment phase. Primary outcomes will be the proportion of all patients diagnosed with a MH condition and receiving pharmaceutical-based treatment who achieve functional improvement, adherence to medication, and retention in MH care. The Consolidated Framework for Implementation Research (CFIR) will be used to assess determinants of implementation success. Specific Aim 1b will include the evaluation of mechanisms of the SAIA-MH strategy using longitudinal structural equation modeling as well as specific aim 2 estimating cost and cost-effectiveness of scaling-up SAIA-MH in Mozambique to provincial and national levels.

DISCUSSION

This study is innovative in being the first, to our knowledge, to test a multicomponent implementation strategy for MH care cascade optimization in LMICs. By design, SAIA-MH is a low-cost strategy to generate contextually relevant solutions to barriers to effective primary MH care, and thus focuses on system improvements that can be sustained over the long term. Since SAIA-MH is integrated into routine government MH service delivery, this pragmatic trial has the potential to inform potential SAIA-MH scale-up in Mozambique and other similar LMICs.

TRIAL REGISTRATION

ClinicalTrials.gov; NCT05103033 ; 11/2/2021.

Smartphone-Based Telestroke Vs"Stroke Physician" led Acute Stroke Management (SMART INDIA): A Protocol for a Cluster-Randomized Trial

Background

One of the major challenges is to deliver adequate health care in rural India, where more than two-thirds of India's population lives. There is a severe shortage of specialists in rural areas with one of the world's lowest physician/population ratios. There is only one neurologist per 1.25 million population. Stroke rehabilitation is virtually nonexistent in most district hospitals. Two innovative solutions include training physicians in district hospitals to diagnose and manage acute stroke ('Stroke physician model') and using a low-cost Telestroke model. We will be assessing the efficacy of these models through a cluster-randomized trial with a standard of care database maintained simultaneously in tertiary nodal centers with neurologists.

Methods

SMART INDIA is a multicenter, open-label cluster-randomized trial with the hospital as a unit of randomization. The study will include district hospitals from the different states of India. We plan to enroll 22 district hospitals where a general physician manages the emergency without the services of a neurologist. These units (hospitals) will be randomized into either of two interventions using computer-generated random sequences with allocation concealment. Blinding of patients and clinicians will not be possible. The outcome assessment will be conducted by the blinded central adjudication team. The study includes 12 expert centers involved in the Telestroke arm by providing neurologists and telerehabilitation round the clock for attending calls. These centers will also be the training hub for "stroke physicians" where they will be given intensive short-term training for the management of acute stroke. There will be a preintervention data collection (1 month), followed by the intervention model implementation (3 months).

Outcomes

The primary outcome will be the composite score (percentage) of performance of acute stroke care bundle assessed at 1 and 3 months after the intervention. The highest score (100%) will be achieved if all the eligible patients receive the standard stroke care bundle. The study will have an open-label extension for 3 more months.

Conclusion

SMART INDIA assesses whether the low-cost Telestroke model is superior to the stroke physician model in achieving acute stroke care delivery. The results of this study can be utilized in national programs for stroke and can be a role model for stroke care delivery in low- and middle-Income countries. (CTRI/2021/11/038196).

Comparison of statistical models for estimating intervention effects based on time-to-recurrent-event in stepped wedge cluster randomized trial using open cohort design

Background

There are currently no methodological studies on the performance of the statistical models for estimating intervention effects based on the time-to-recurrent-event (TTRE) in stepped wedge cluster randomised trial (SWCRT) using an open cohort design. This study aims to address this by evaluating the performance of these statistical models using an open cohort design with the Monte Carlo simulation in various settings and their application using an actual example.

Methods

Using Monte Carlo simulations, we evaluated the performance of the existing extended Cox proportional hazard models, i.e., the Andersen-Gill (AG), Prentice-Williams-Peterson Total-Time (PWP-TT), and Prentice-Williams-Peterson Gap-time (PWP-GT) models, using the settings of several event generation models and true intervention effects, with and without stratification by clusters. Unidirectional switching in SWCRT was represented using time-dependent covariates.

Results

Using Monte Carlo simulations with the various described settings, in situations where inter-individual variability do not exist, the PWP-GT model with stratification by clusters showed the best performance in most settings and reasonable performance in the others. The only situation in which the performance of the PWP-TT model with stratification by clusters was not inferior to that of the PWP-GT model with stratification by clusters was when there was a certain amount of follow-up period, and the timing of the trial entry was random within the trial period, including the follow-up period. In situations where inter-individual variability existed, the PWP-GT model consistently underperformed compared to the PWP-TT model. The AG model performed well only in a specific setting. By analysing actual examples, it was found that almost all the statistical models suggested that the risk of events during the intervention condition may be somewhat higher than in the control, although the difference was not statistically significant.

Conclusions

When estimating the TTRE-based intervention effects of SWCRT in various settings using an open cohort design, the PWP-GT model with stratification by clusters performed most reasonably in situations where inter-individual variability was not present. However, if inter-individual variability was present, the PWP-TT model with stratification by clusters performed best.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-022-01552-6.

Evaluating the performance of Bayesian and restricted maximum likelihood estimation for stepped wedge cluster randomized trials with a small number of clusters

Background

Stepped wedge trials are an appealing and potentially powerful cluster randomized trial design. However, they are frequently implemented with a small number of clusters. Standard analysis methods for these trials such as a linear mixed model with estimation via maximum likelihood or restricted maximum likelihood (REML) rely on asymptotic properties and have been shown to yield inflated type I error when applied to studies with a small number of clusters. Small-sample methods such as the Kenward-Roger approximation in combination with REML can potentially improve estimation of the fixed effects such as the treatment effect. A Bayesian approach may also be promising for such multilevel models but has not yet seen much application in cluster randomized trials.

Methods

We conducted a simulation study comparing the performance of REML with and without a Kenward-Roger approximation to a Bayesian approach using weakly informative prior distributions on the intracluster correlation parameters. We considered a continuous outcome and a range of stepped wedge trial configurations with between 4 and 40 clusters. To assess method performance we calculated bias and mean squared error for the treatment effect and correlation parameters and the coverage of 95% confidence/credible intervals and relative percent error in model-based standard error for the treatment effect.

Results

Both REML with a Kenward-Roger standard error and degrees of freedom correction and the Bayesian method performed similarly well for the estimation of the treatment effect, while intracluster correlation parameter estimates obtained via the Bayesian method were less variable than REML estimates with different relative levels of bias.

Conclusions

The use of REML with a Kenward-Roger approximation may be sufficient for the analysis of stepped wedge cluster randomized trials with a small number of clusters. However, a Bayesian approach with weakly informative prior distributions on the intracluster correlation parameters offers a viable alternative, particularly when there is interest in the probability-based inferences permitted within this paradigm.

Supplementary Information

The online version contains supplementary material available at (10.1186/s12874-022-01550-8).

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

BACKGROUND/AIMS

This work is motivated by the HEALing Communities Study, which is a post-test only cluster randomized trial in which communities are randomized to two different trial arms. The primary interest is in reducing opioid overdose fatalities, which will be collected as a count outcome at the community level. Communities range in size from thousands to over one million residents, and fatalities are expected to be rare. Traditional marginal modeling approaches in the cluster randomized trial literature include the use of generalized estimating equations with an exchangeable correlation structure when utilizing subject-level data, or analogously quasi-likelihood based on an over-dispersed binomial variance when utilizing community-level data. These approaches account for and estimate the intra-cluster correlation coefficient, which should be provided in the results from a cluster randomized trial. Alternatively, the coefficient of variation or R coefficient could be reported. In this article, we show that negative binomial regression can also be utilized when communities are large and events are rare. The objectives of this article are (1) to show that the negative binomial regression approach targets the same marginal regression parameter(s) as an over-dispersed binomial model and to explain why the estimates may differ; (2) to derive formulas relating the negative binomial overdispersion parameter k with the intra-cluster correlation coefficient, coefficient of variation, and R coefficient; and (3) analyze pre-intervention data from the HEALing Communities Study to demonstrate and contrast models and to show how to report the intra-cluster correlation coefficient, coefficient of variation, and R coefficient when utilizing negative binomial regression.

METHODS

Negative binomial and over-dispersed binomial regression modeling are contrasted in terms of model setup, regression parameter estimation, and formulation of the overdispersion parameter. Three specific models are used to illustrate concepts and address the third objective.

RESULTS

The negative binomial regression approach targets the same marginal regression parameter(s) as an over-dispersed binomial model, although estimates may differ. Practical differences arise in regard to how overdispersion, and hence the intra-cluster correlation coefficient is modeled. The negative binomial overdispersion parameter is approximately equal to the ratio of the intra-cluster correlation coefficient and marginal probability, the square of the coefficient of variation, and the R coefficient minus 1. As a result, estimates corresponding to all four of these different types of overdispersion parameterizations can be reported when utilizing negative binomial regression.

CONCLUSION

Negative binomial regression provides a valid, practical, alternative approach to the analysis of count data, and corresponding reporting of overdispersion parameters, from community randomized trials in which communities are large and events are rare.

Promoting deceased organ and tissue donation registration in family physician waiting rooms (RegisterNow-1): a pragmatic stepped-wedge, cluster randomized controlled registry trial

BACKGROUND

The shortage of available organs for life-saving transplants persists worldwide. While a majority support donating their organs or tissue when they die, many have not registered their wish to do so. When registered, next of kin are much more likely to follow-through with the decision to donate. In many countries, most people visit their family physician office each year and this setting is a promising, yet underused, site where more people could register for deceased organ donation. Our primary aim was to evaluate the effectiveness of an intervention to promote organ donation registration in family physician's offices.

METHODS

We developed an intervention to address barriers and enablers to organ donation registration that involved physician office reception staff inviting patients to register on a tablet in the waiting room while they waited for their appointment. We conducted a cross-sectional stepped-wedge cluster randomized controlled registry trial to evaluate the intervention. We recruited six family physician offices in Canada. All offices began with usual care and then every two weeks, one office (randomly assigned) started the intervention until all offices delivered the intervention. The primary outcome was registration for deceased organ donation in the provincial organ registration registry, assessed within the 7 days of the physician visit. At the end of the trial, we also conducted interviews with clinic staff to assess any barriers and enablers to delivering the intervention.

RESULTS

The trial involved 24,616 patient visits by 13,562 unique patients: 12,484 visits in the intervention period and 12,132 in the control period. There was no statistically significant difference in the percentage of patients registered for deceased organ donation in the intervention versus control period (48.0% vs 46.2%; absolute difference after accounting for the secular trend: 0.12%; 95% CI: - 2.30, 2.54; p=0.92). Interviews with clinic staff indicated location of the tablet within a waiting room, patient rapport, existing registration, confidence and motivation to deliver the intervention and competing priorities as barriers and enablers to delivery.

CONCLUSIONS

Our intervention did not increase donor registration. Nonetheless, family physician offices may still remain a promising setting to develop and evaluate better interventions to increase organ donation registration.

TRIAL REGISTRATION

NCT03213171.

Randomization: Beyond the closurization principle

Many cluster randomized trials have relatively few numbers of clusters to be randomized. When baseline cluster-level covariates are available prior to randomization, the set of potential allocations can be restricted so as to ensure balance across study arms. This article discusses why and how restrictions can be made, and the ramifications of so doing. The Fisher-Bailey validity is explained, and examples are given regarding the tradeoff between balance and validity.

Cluster Randomized Trials: Lessons for Inflammatory Bowel Disease Trials

Randomized clinical trials are the gold standard for the development of new drugs. Clinical trials have become increasingly complex, in particular in the field of inflammatory bowel diseases. For this reason, a new study design called 'cluster randomized trial' has been introduced. Cluster randomized trials randomly assign groups of individuals within a population of interest, such as people in a particular research site, clinic or hospital, and measure the effects of an intervention at the individual level. In contrast to individual randomization, cluster randomization permits the evaluation of a cluster-level intervention, may be logistically simpler and less expensive than a conventional randomized trial, and can help reduce the effect of treatment contamination. An important aspect requiring particular attention in cluster randomized trials is the risk for imbalance of covariates at baseline which decreases the statistical power and precision of the results, and reduces face validity and credibility of the trial results. A variety of restricted randomization methods have been proposed to minimize risk of imbalance. Our aim here is to clarify the advantages and drawbacks of cluster randomized trials in order to properly interpret study results and to identify their role in upcoming inflammatory bowel disease trials.

Feasibility of a randomized clinical trial evaluating a community intervention for household tuberculosis child contact management in Cameroon and Uganda

Background

One of the main barriers of the management of household tuberculosis child contacts is the necessity for parents to bring healthy children to the facility. We assessed the feasibility of a community intervention for tuberculosis (TB) household child contact management and the conditions for its evaluation in a cluster randomized controlled trial in Cameroon and Uganda.

Methods

We assessed three dimensions of feasibility using a mixed method approach: (1) recruitment capability using retrospective aggregated data from facility registers; (2) acceptability of the intervention using focus group discussions with TB patients and in-depth interviews with healthcare providers and community leaders; and (3) adaptation, integration, and resources of the intervention in existing TB services using a survey and discussions with stakeholders.

Results

Reaching the sample size is feasible in all clusters in 15 months with the condition of regrouping 2 facilities in the same cluster in Uganda due to decentralization of TB services. Community health worker (CHW) selection and training and simplified tools for contact screening, tolerability, and adherence of preventive therapy were key elements for the implementation of the community intervention. Healthcare providers and patients found the intervention of child contact investigations and TB preventive treatment management in the household acceptable in both countries due to its benefits (competing priorities, transport cost) as compared to facility-based management. TB stigma was present, but not a barrier for the community intervention. Visit schedule and team conduct were identified as key facilitators for the intervention.

Conclusions

This study shows that evaluating a community intervention for TB child contact management in a cluster randomized trial is feasible in Cameroon and Uganda.

Trial registration

Clini calTr ials. gov NCT03832023. Registered on February 6^th 2019.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40814-022-00996-3.

Mortality in HIV and tuberculosis patients following implementation of integrated HIV-TB treatment: Results from an open-label cluster-randomized trial

BACKGROUND

HIV-TB treatment integration reduces mortality. Operational implementation of integrated services is challenging. This study assessed the impact of quality improvement (QI) for HIV-TB integration on mortality within primary healthcare (PHC) clinics in South Africa.

METHODS

An open-label cluster randomized controlled study was conducted between 2016 and 2018 in 40 rural clinics in South Africa. The study statistician randomized PHC nurse-supervisors 1:1 into 16 clusters (eight nurse-supervisors supporting 20 clinics per arm) to receive QI, supported HIV-TB integration intervention or standard of care (control). Nurse supervisors and clinics under their supervision, based in the study health districts were eligible for inclusion in this study. Nurse supervisors were excluded if their clinics were managed by municipal health (different resource allocation), did not offer co-located antiretroviral therapy (ART) and TB services, services were performed by a single nurse, did not receive non-governmental organisation (NGO) support, patient data was not available for > 50% of attendees. The analysis population consists of all patients newly diagnosed with (i) both TB and HIV (ii) HIV only (among patients previously treated for TB or those who never had TB before) and (iii) TB only (among patients already diagnosed with HIV or those who were never diagnosed with HIV) after QI implementation in the intervention arm, or enrolment in the control arm. Mortality rates was assessed 12 months post enrolment, using unpaired t-tests and cox-proportional hazards model. (Clinicaltrials.gov, NCT02654613, registered 01 June 2015, trial closed).

FINDINGS

Overall, 21 379 participants were enrolled between December 2016 and December 2018 in intervention and control arm clinics: 1329 and 841 HIV-TB co-infected (10·2%); 10 799 and 6 611 people living with Human Immunodeficiency Virus (HIV)/ acquired immunodeficiency syndrome (AIDS) (PLWHA) only (81·4%); 1 131 and 668 patients with TB only (8·4%), respectively. Average cluster sizes were 1657 (range 170-5782) and 1015 (range 33-2027) in intervention and control arms. By 12 months, 6529 (68·7%) and 4074 (70·4%) were alive and in care, 568 (6·0%) and 321 (5·6%) had completed TB treatment, 1078 (11·3%) and 694 (12·0%) were lost to follow-up, with 245 and 156 deaths occurring in intervention and control arms, respectively. Mortality rates overall [95% confidence interval (CI)] was 4·5 (3·4-5·9) in intervention arm, and 3·8 (2·6-5·4) per 100 person-years in control arm clusters [mortality rate ratio (MRR): 1·19 (95% CI 0·79-1·80)]. Mortality rates among HIV-TB co-infected patients was 10·1 (6·7-15·3) and 9·8 (5·0-18·9) per 100 person-years, [MRR: 1·04 (95% CI 0·51-2·10)], in intervention and control arm clusters, respectively.

INTERPRETATION

HIV-TB integration supported by a QI intervention did not reduce mortality in HIV-TB co-infected patients. Demonstrating mortality benefit from health systems process improvements in real-world operational settings remains challenging. Despite the study being potentially underpowered to demonstrate the effect size, integration interventions were implemented using existing facility staff and infrastructure reflecting the real-world context where most patients in similar settings access care, thereby improving generalizability and scalability of study findings.

FUNDING

Research reported in this publication was supported by South African Medical Research Council (SAMRC), and UK Government's Newton Fund through United Kingdom Medical Research Council (UKMRC).

PATHWEIGH, pragmatic weight management in adult patients in primary care in Colorado, USA: study protocol for a stepped wedge cluster randomized trial

Background

Despite the overwhelming prevalence and health implications of obesity, it is rarely adequately addressed in a health care setting. PATHWEIGH is a pragmatic approach to weight management that uses tools built into the electronic medical record to overcome barriers and guide care. Implementation strategies are employed to facilitate adoption and use of the PATHWEIGH tools and processes. The current study will compare the effectiveness of PATHWEIGH versus standard of care (SOC) on patient weight loss in primary care and explore factors for its successful implementation.

Methods

A stepped wedge cluster randomized trial design will be used within an effectiveness-implementation hybrid study. Adult patient weight loss and weight loss maintenance will be compared in PATHWEIGH versus SOC in 57 family and internal medicine clinics in a large health system in Colorado, USA. Effectiveness will be evaluated using generalized linear mixed models to determine statistical differences in weight loss and weight loss maintenance at 6, 12, and 18 months. Patient-, provider-, and clinic-level predictors will be identified using mediator and moderator analyses. Conceptually guided by the Practical, Robust, Implementation and Sustainability Model (PRISM), a mixed methods approach including quantitative (practice surveys, use tracking) and qualitative (interviews, observations) data collection will be used to determine factors impeding and facilitating adoption, implementation, and maintenance of PATHWEIGH and evaluate specified implementation strategies. A cost analysis of the practice and system costs and resources required by PATHWEIGH relative to the reimbursement collected will be performed.

Discussion

The effectiveness and implementation of PATHWEIGH, and their interrelatedness, for patient weight loss are collectively the focus of the current trial. Findings from this study are expected to serve as a blueprint for available and effective weight management in primary care medical practice.

Trial registration

ClinicalTrials.govNCT04678752. Registered on December 21, 2020.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13063-021-05954-7.

Use of metabolomics in refining the effect of an organic food intervention on biomarkers of exposure to pesticides and biomarkers of oxidative damage in primary school children in Cyprus: A cluster-randomized cross-over trial

BACKGROUND

Exposure to pesticides has been associated with oxidative stress in animals and humans. Previously, we showed that an organic food intervention reduced pesticide exposure and oxidative damage (OD) biomarkers over time; however associated metabolic changes are not fully understood yet.

OBJECTIVES

We assessed perturbations of the urine metabolome in response to an organic food intervention for children and its association with pesticides biomarkers [3-phenoxybenzoic acid (3-PBA) and 6-chloronicotinic acid (6-CN)]. We also evaluated the molecular signatures of metabolites associated with biomarkers of OD (8-iso-PGF2a and 8-OHdG) and related biological pathways.

METHODS

We used data from the ORGANIKO LIFE + trial (NCT02998203), a cluster-randomized cross-over trial conducted among primary school children in Cyprus. Participants (n = 149) were asked to follow an organic food intervention for 40 days and their usual food habits for another 40 days, providing up to six first morning urine samples (>850 samples in total). Untargeted GC-MS metabolomics analysis was performed. Metabolites with RSD ≤ 20% and D-ratio ≤ 50% were retained for analysis. Associations were examined using mixed-effect regression models and corrected for false-discovery rate of 0.05. Pathway analysis followed.

RESULTS

Following strict quality checks, 156 features remained out of a total of 610. D-glucose was associated with the organic food intervention (β = -0.23, 95% CI: -0.37,-0.10), aminomalonic acid showed a time-dependent increase during the intervention period (β_int = 0.012; 95% CI:0.002, 0.022) and was associated with the two OD biomarkers (β = -0.27, 95% CI:-0.34,-0.20 for 8-iso-PGF2a and β = 0.19, 95% CI:0.11,0.28 for 8-OHdG) and uric acid with 8-OHdG (β = 0.19, 95% CI:0.11,0.26). Metabolites were involved in pathways such as the starch and sucrose metabolism and pentose and glucuronate interconversions.

DISCUSSION

This is the first metabolomics study providing evidence of differential expression of metabolites by an organic food intervention, corroborating the reduction in biomarkers of OD. Further mechanistic evidence is warranted to better understand the biological plausibility of an organic food treatment on children's health outcomes.

Compassionate mind training for caregivers of residential youth care: Early findings of a cluster randomized trial

BACKGROUND

Compassion plays a significant role in caregiving and its benefits have been largely reported in different settings. Nonetheless, compassion-based interventions have not yet been delivered to Residential Youth Care (RYC) staff.

OBJECTIVE

This study presents early findings of a Cluster Randomized Trial on the efficacy of a Compassionate Mind Training program for caregivers of RYC (CMT-Care Homes) following CONSORT guidelines.

PARTICIPANTS AND SETTING

Participants were 85 caregivers (89.4% female), aged between 25 and 62 years old, working on a regular basis with adolescents in RYC.

METHOD

Eleven Portuguese residential care homes for at-risk adolescents were selected and randomly allocated to the treatment (n = 5) or control (n = 6) conditions. Caregivers were assessed at pre- and post-intervention (n = 41 treatment, n = 44 control) through self-report scales on compassion and emotional climate related outcomes.

RESULTS

To investigate CMT-Care Homes effects, a two-factor mixed MANOVA was performed. Multivariate tests showed a significant and large Time × Group interaction effect (Pillais' trace = 0.291, F = 2.719, p = .005, η_p² = 0.291). Univariate tests indicated significant and positive effects in compassion and fears of compassion (low and medium effect sizes), as well as in soothing related emotions (emotional climate) and social safeness (both with medium effect sizes), in favor of the treatment group.

CONCLUSIONS

Findings offer preliminary evidence of the effectiveness of the CMT-Care Homes program, suggesting that this training allows the development of an affiliative mentality in caregivers working within RYC settings.

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

BACKGROUND/AIMS

Generalized estimating equations are commonly used to fit logistic regression models to clustered binary data from cluster randomized trials. A commonly used correlation structure assumes that the intracluster correlation coefficient does not vary by treatment arm or other covariates, but the consequences of this assumption are understudied. We aim to evaluate the effect of allowing variation of the intracluster correlation coefficient by treatment or other covariates on the efficiency of analysis and show how to account for such variation in sample size calculations.

METHODS

We develop formulae for the asymptotic variance of the estimated difference in outcome between treatment arms obtained when the true exchangeable correlation structure depends on the treatment arm and the working correlation structure used in the generalized estimating equations analysis is: (i) correctly specified, (ii) independent, or (iii) exchangeable with no dependence on treatment arm. These formulae require a known distribution of cluster sizes; we also develop simplifications for the case when cluster sizes do not vary and approximations that can be used when the first two moments of the cluster size distribution are known. We then extend the results to settings with adjustment for a second binary cluster-level covariate. We provide formulae to calculate the required sample size for cluster randomized trials using these variances.

RESULTS

We show that the asymptotic variance of the estimated difference in outcome between treatment arms using these three working correlation structures is the same if all clusters have the same size, and this asymptotic variance is approximately the same when intracluster correlation coefficient values are small. We illustrate these results using data from a recent cluster randomized trial for infectious disease prevention in which the clusters are groups of households and modest in size (mean 9.6 individuals), with intracluster correlation coefficient values of 0.078 in the control arm and 0.057 in an intervention arm. In this application, we found a negligible difference between the variances calculated using structures (i) and (iii) and only a small increase (typically <5%) for the independent correlation structure (ii), and hence minimal effect on power or sample size requirements. The impact may be larger in other applications if there is greater variation in the ICC between treatment arms or with an additional covariate.

CONCLUSION

The common approach of fitting generalized estimating equations with an exchangeable working correlation structure with a common intracluster correlation coefficient across arms likely does not substantially reduce the power or efficiency of the analysis in the setting of a large number of small or modest-sized clusters, even if the intracluster correlation coefficient varies by treatment arm. Our formulae, however, allow formal evaluation of this and may identify situations in which variation in intracluster correlation coefficient by treatment arm or another binary covariate may have a more substantial impact on power and hence sample size requirements.

A cluster randomized trial to reduce HIV risk from outside partnerships in Zambian HIV-Negative couples using a novel behavioral intervention, "Strengthening Our Vows": Study protocol and baseline data

BACKGROUND

Heterosexual couples contribute to most new HIV infections in areas of generalized HIV epidemics in sub-Saharan Africa. After Couples' Voluntary HIV Counseling and Testing (CVCT), heterosexual concordant HIV negative couples (CNC) in cohabiting unions contribute to approximately 47% of residual new infections in couples. These infections are attributed to concurrent sexual partners, a key driver of the HIV epidemic in Zambia.

METHODS/DESIGN

Ten Zambian government clinics in two of the largest cities were randomized in matched pairs to a Strengthening Our Vows (SOV) intervention or a Good Health Package (GHP) comparison arm. SOV addressed preventing HIV infection from concurrent partners and protecting spouses after exposures outside the relationship. GHP focused on handwashing; water chlorination; household deworming; and screening for hypertension, diabetes and schistosomiasis. CNC were referred from CVCT services in government clinics. Follow-up includes post-intervention questionnaires and outcome assessments through 60 months. Longitudinal outcomes of interest include self-report and laboratory markers of condomless sex with outside partners and reported sexual agreements. We present baseline characteristics and factors associated with study arm and reported risk using descriptive statistics.

RESULTS

The mean age of men was 32 and 26 for women. On average, couples cohabited for 6 years and had 2 children. Baseline analyses demonstrated some failures of randomization by study arm which will be considered in future primary analyses of longitudinal data. An HIV/STI risk factor composite was not different in the two study arms. Almost one-quarter of couples had an HIV risk factor at baseline.

DISCUSSION

In preparation for future biomedical and behavioral interventions in sub-Saharan Africa, it is critical to understand and decrease HIV risk within CNC.

Protocol for the models of primary osteoporosis screening in men (MOPS) cluster randomized trial

Current guidelines recommend primary osteoporosis screening for at-risk men to reduce the morbidity, mortality, and cost associated with osteoporotic fractures. However, analyses in a national Veterans Health Administration cohort of over 4,000,000 men demonstrated that primary osteoporosis screening as it is currently operationalized does not benefit most older Veterans due to inefficient targeting and low subsequent treatment and adherence rates. The overall objective of this study is to determine whether a new model of primary osteoporosis screening reduces fracture risk compared to usual care. We are conducting a pragmatic group randomized trial of 38 primary care teams assigned to usual care or a Bone Health Service (BHS) screening model in which screening and adherence activities are managed by a centralized expert team. The study will: 1) compare the impact of the BHS model on patient-level outcomes strongly associated with fracture rates (eligible proportion screened, proportion meeting treatment criteria who receive osteoporosis medications, medication adherence, and femoral neck bone mineral density); 2) quantify the impact on provider and facility-level outcomes including change in DXA volume, change in metabolic bone disease clinic volume, and PACT provider time and satisfaction; and 3) estimate the impact on health system and policy outcomes using Markov models of screening program cost per quality adjusted life year based from health system and societal perspectives.

Randomization-based confidence intervals for cluster randomized trials

In a cluster randomized trial (CRT), groups of people are randomly assigned to different interventions. Existing parametric and semiparametric methods for CRTs rely on distributional assumptions or a large number of clusters to maintain nominal confidence interval (CI) coverage. Randomization-based inference is an alternative approach that is distribution-free and does not require a large number of clusters to be valid. Although it is well-known that a CI can be obtained by inverting a randomization test, this requires testing a non-zero null hypothesis, which is challenging with non-continuous and survival outcomes. In this article, we propose a general method for randomization-based CIs using individual-level data from a CRT. This approach accommodates various outcome types, can account for design features such as matching or stratification, and employs a computationally efficient algorithm. We evaluate this method's performance through simulations and apply it to the Botswana Combination Prevention Project, a large HIV prevention trial with an interval-censored time-to-event outcome.

A youth empowerment intervention to prevent childhood obesity: design and methods for a cluster randomized trial of the H2GO! program

BACKGROUND

Reducing sugar-sweetened beverage (SSB) consumption is a promising dietary target for childhood obesity prevention. This paper describes the design and methods of a cluster randomized trial of H2GO!, a youth empowerment intervention to prevent childhood obesity through reducing SSB consumption among a low-income, ethnically diverse sample of youth.

METHODS

This cluster randomized controlled trial is an academic-community partnership with the Massachusetts Alliance of Boys and Girls Clubs (BGC). Ten BGC sites will be randomly assigned to the H2GO! intervention or a wait-list, usual care control. Eligible study participants will be N = 450 parent-child pairs (youth ages 9-12 years and their parents/caregivers) recruited from participating BGCs. The 6-week in-person H2GO! intervention consists of 12 group-based sessions delivered by BGC staff and youth-led activities. An innovative feature of the intervention is the development of youth-produced narratives as a strategy to facilitate youth empowerment and parental engagement. Child outcomes include measured body mass index z scores (zBMI), beverage intake, and youth empowerment. Parent outcomes include beverage intake and availability of SSBs at home. Outcomes will be measured at baseline and at 2, 6, and 12 months. With a 75% retention rate, the study is powered to detect a minimum group difference of 0.1 zBMI units over 12 months.

DISCUSSION

Empowering youth may be a promising intervention approach to prevent childhood obesity through reducing SSB consumption. This intervention was designed to be delivered through BGCs and is hypothesized to be efficacious, relevant, and acceptable for the target population of low-income and ethnically diverse youth.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04265794 . Registered 11 February 2020.

Simple compared to covariate-constrained randomization methods in balancing baseline characteristics: a case study of randomly allocating 72 hemodialysis centers in a cluster trial

Background and aim

Some parallel-group cluster-randomized trials use covariate-constrained rather than simple randomization. This is done to increase the chance of balancing the groups on cluster- and patient-level baseline characteristics. This study assessed how well two covariate-constrained randomization methods balanced baseline characteristics compared with simple randomization.

Methods

We conducted a mock 3-year cluster-randomized trial, with no active intervention, that started April 1, 2014, and ended March 31, 2017. We included a total of 11,832 patients from 72 hemodialysis centers (clusters) in Ontario, Canada. We randomly allocated the 72 clusters into two groups in a 1:1 ratio on a single date using individual- and cluster-level data available until April 1, 2013. Initially, we generated 1000 allocation schemes using simple randomization. Then, as an alternative, we performed covariate-constrained randomization based on historical data from these centers. In one analysis, we restricted on a set of 11 individual-level prognostic variables; in the other, we restricted on principal components generated using 29 baseline historical variables.

We created 300,000 different allocations for the covariate-constrained randomizations, and we restricted our analysis to the 30,000 best allocations based on the smallest sum of the penalized standardized differences. We then randomly sampled 1000 schemes from the 30,000 best allocations. We summarized our results with each randomization approach as the median (25th and 75th percentile) number of balanced baseline characteristics. There were 156 baseline characteristics, and a variable was balanced when the between-group standardized difference was ≤ 10%.

Results

The three randomization techniques had at least 125 of 156 balanced baseline characteristics in 90% of sampled allocations. The median number of balanced baseline characteristics using simple randomization was 147 (142, 150). The corresponding value for covariate-constrained randomization using 11 prognostic characteristics was 149 (146, 151), while for principal components, the value was 150 (147, 151).

Conclusion

In this setting with 72 clusters, constraining the randomization using historical information achieved better balance on baseline characteristics compared with simple randomization; however, the magnitude of benefit was modest.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13063-021-05590-1.

CRTpowerdist: An R package to calculate attained power and construct the power distribution for cross-sectional stepped-wedge and parallel cluster randomized trials

BACKGROUND

The attained power, calculated conditional on the realized allocation, of a clinical trial may differ from the expected power, obtained pre-randomization through averaging over all potential allocations that could be generated by the randomization algorithm (RA). For example, a two-arm trial using a RA that is expected to allocate 20 participants to each arm will attain less than the expected power if by chance it allocates 25 and 15 participants to the arms. Cluster randomized trials with unequal cluster sizes have elevated risk of realizing an allocation that yields an attained power much lower than the expected power when modest numbers of clusters are randomized.

METHOD

We developed the R package CRTpowerdist, which implements both simulations and approximate analytic formulae to calculate the attained powers associated with different realized allocations and constructs the pre-randomization power distribution associated with the RA to facilitate assessing the risk of obtaining inadequate power. The package covers unequal cluster-size, cross-sectional stepped-wedge and parallel cluster randomized trials, with or without stratification. Allowed outcome types are: continuous (Gaussian), binary (Binomial) and count (Poisson). The analytic formulae-based calculations are also implemented in a Shiny app.

RESULTS

The functionality of the CRTpowerdist is illustrated for each type of trial design. The examples show how to obtain the attained power, the power distribution, and the risk of low attained power, using both simulation and analytic formulae.

CONCLUSION

For cluster randomized trials with unequal cluster sizes, the CRTpowerdist package can assist users in identifying an appropriate randomization algorithm by enabling the user to assess the risk that a randomization algorithm will lead to an allocation with inadequate attained power. The Shiny app makes these assessments accessible to researchers who are unable or do not wish to use the CRTpowerdist package.