Aspects of statistical design for the Community Intervention Trial for Smoking Cessation (COMMIT)

Improving cancer prevention and control through implementing academic-local public health department partnerships - protocol for a cluster-randomized implementation trial using a positive deviance approach

BACKGROUND

Local public health departments in the United States are responsible for implementing cancer-related programs and policies in their communities; however, many staff have not been trained to use evidence-based processes, and the organizational climate may be unsupportive of evidence-based processes. A promising approach to address these gaps is through academic-public health department (AHD) partnerships, in which practitioners and academics collaborate to improve public health practice and education through joint research projects and educational opportunities. Prior research has demonstrated the benefits of AHD partnerships to public health practice and education. However, knowledge about how AHD partnerships should be structured to support implementation of programs and policies is sparse.

METHODS

This is a mixed methods, two-phase study, guided by the Exploration, Preparation, Implementation, and Sustainment (EPIS) Framework, in which AHD partnerships are a relational type of bridging factor. A positive deviance approach will be used to understand how AHD partnerships are best structured and supported. In the formative phase, we will survey academics and local health department staff (n = 500) to characterize AHD partnerships and understand contextual influences. We will conduct in-depth interviews with eight AHD partnerships (four high and four low engagement), to identify differences between high and low engagement partnerships. The second, experimental phase will be a paired group randomized trial with 28 AHD partnerships (n = 14 randomized to implementation arm and n = 14 to the control arm). A menu of strategies will be refined through survey and interview findings, literature, and our team's previous work. The trial will assess whether these strategies can be used to strengthen partnerships and improve adoption of cancer prevention and control programs and policies. We will evaluate changes in AHD partnership engagement and implementation of evidence-based programs and policies.

DISCUSSION

This first-of-its-kind study will focus on collaborations that leverage complementary expertise of health department staff and academics to improve public health practice. Our results can impact the field by identifying new, sustainable models for how public health practitioners and academics can work together to meet common goals, increase the use of evidence-based programs and policies, and expand our understanding of bridging factors within the EPIS framework.

TRIAL REGISTRATION

Prospective registered on 9/17/2024 at clinicaltrials.gov no. NCT06605196 ( https://clinicaltrials.gov/study/NCT06605196 ).

A multilevel approach for promoting physical activity in rural communities: a cluster randomized controlled trial

Background

Physical activity (PA) has demonstrated a decreased risk in various cancers and other chronic diseases; however, rural residents are less likely to attain recommended levels of PA compared to urban and suburban counterparts. Given rural residents make up 15% of the United States population, there is a need for novel approaches to increase PA among this population. The goal of the present study is to investigate the effectiveness of a multilevel intervention to increase PA rates among rural residents.

Methods/design

Guided by an ecological framework, a group-randomized design will be used to evaluate the effects of a three-level intervention for increasing PA among adult residents residing in 6 rural communities (n = 600) along with 6 control communities (n = 600). The intervention includes components at the individual (short message service [SMS] text messages), interpersonal (social support in walking groups), and community levels (events at existing trails). Innovative methods to encourage participation will be employed as well as a focus on life priorities (family, recreation, hobbies) other than health. Aim 1 includes a literature review and key informant interviews to determine the local contexts for intervention adaptation. Aim 2 will employ a set of interventions at the individual, interpersonal, and community-levels to evaluate their impact on moderate-to-vigorous PA as measured by self-reported (telephone survey) and objectively assessed (accelerometry) measures. These data are supplemented by location based on Global Positioning System and community audits, which provide information on recreational amenities, programs/policies, and street segments.

Discussion

This study is among the first of its kind to test a multilevel intervention in a rural setting, address life priorities that compliment health outcomes, and examine moderation between behavioral interventions and the natural environments where people are physically active. Our results will influence the field by enhancing the ability to scale-up innovative, PA interventions with the potential to reach high-risk, rural populations.

Trial registration

Clinical Trials NCT03683173, September 25, 2018.

Implementing a larviciding efficacy or effectiveness control intervention against malaria vectors: key parameters for success

During the last decade, scale-up of vector control tools such as long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) contributed to the reduction of malaria morbidity and mortality across the continent. Because these first line interventions are now affected by many challenges such as insecticide resistance, change in vector feeding and biting behaviour, outdoor malaria transmission and adaptation of mosquito to polluted environments, the World Health Organization recommends the use of integrated control approaches to improve, control and elimination of malaria. Larviciding is one of these approaches which, if well implemented, could help control malaria in areas where this intervention is suitable. Unfortunately, important knowledge gaps remain in its successful application. The present review summarises key parameters that should be considered when implementing larviciding efficacy or effectiveness trials.

The use of permutation tests for the analysis of parallel and stepped-wedge cluster-randomized trials

We investigate the use of permutation tests for the analysis of parallel and stepped-wedge cluster-randomized trials. Permutation tests for parallel designs with exponential family endpoints have been extensively studied. The optimal permutation tests developed for exponential family alternatives require information on intraclass correlation, a quantity not yet defined for time-to-event endpoints. Therefore, it is unclear how efficient permutation tests can be constructed for cluster-randomized trials with such endpoints. We consider a class of test statistics formed by a weighted average of pair-specific treatment effect estimates and offer practical guidance on the choice of weights to improve efficiency. We apply the permutation tests to a cluster-randomized trial evaluating the effect of an intervention to reduce the incidence of hospital-acquired infection. In some settings, outcomes from different clusters may be correlated, and we evaluate the validity and efficiency of permutation test in such settings. Lastly, we propose a permutation test for stepped-wedge designs and compare its performance with mixed-effect modeling and illustrate its superiority when sample sizes are small, the underlying distribution is skewed, or there is correlation across clusters. Copyright © 2017 John Wiley & Sons, Ltd.

A mixed model formulation for designing cluster randomized trials with binary outcomes

Cluster randomized trials (CRTs) are unlike traditional individually randomized trials because observations within the same cluster are positively correlated and the sample size (number of clusters) is relatively small. Although formulae for sample size and power estimates of CRT designs do exist, these formulae rely upon first-order asymptotic approximations for the distribution of the average intervention effect and are inaccurate for CRTs that have a small number of clusters. These formulae also assume that the intracluster correlation (ICC) is the same for each cluster in the CRT. However, for CRTs in which the clusters are classrooms or medical practices, the degree of ICC is often a factor of how many students are in each classroom or how many patients are in each practice. Specifically, smaller clusters are expected to have larger ICC than larger clusters. A weighted sum of the cluster means, D, is the statistic often used to estimate the average intervention effect in a CRT. Therefore, we propose that a saddlepoint approximation is a natural choice to approximate the distributions of the cluster means more precisely than a standard large-sample approximation. We parameterize the ICC for each cluster as a random effect with a predefined prior distribution that is dependent upon the size of each cluster. After integrating over the range of the random effect, we use Monte Carlo methods to generate sample cluster means, which are in turn used to approximate the distribution of D with saddlepoint methods. Through numerical examples and an actual application, we show that our method has accuracy that is equal to or better than that of existing methods. Futhermore, our method accommodates CRTs in which the correlation within cluster is expected to diminish with the cluster size.

Proactive telephone counseling for adolescent smokers: Comparing regular smokers with infrequent and occasional smokers on treatment receptivity, engagement, and outcomes

BACKGROUND

Adolescent smoking cessation efforts to date have tended to focus on regular smokers. Consequently, infrequent and occasional smokers' receptivity and response to smoking cessation interventions is unknown. To address this gap, this study examines data from the Hutchinson Study of High School Smoking-a randomized trial that examined the effectiveness of a telephone-delivered smoking cessation intervention for a large, population-based cohort of adolescent smokers proactively recruited in an educational setting.

METHODS

The study population included 1837 proactively identified high school smokers. Intervention receptivity, engagement, and outcomes were examined among adolescent infrequent (1-4days/month) and occasional (5-19days/month) smokers and compared with regular smokers (20 or more days/month).

RESULTS

With regard to treatment receptivity, intervention recruitment did not differ by smoking frequency. For engagement, intervention completion rates were higher for infrequent smokers (80.5%) compared with occasional (63.8%) and regular smokers (61.5%, p<0.01). Intervention effect sizes were not statistically different across groups.

CONCLUSIONS

Adolescent infrequent and occasional smokers are at least as receptive to a proactively delivered smoking cessation intervention as regular smokers and can benefit just as much from it. Including these adolescent smokers in cessation programs and research-with the goal of interrupting progression of smoking before young adulthood-should help reduce the high smoking prevalence among young adults.

Does Effectiveness of Adolescent Smoking-Cessation Intervention Endure Into Young Adulthood? 7-Year Follow-Up Results from a Group-Randomized Trial

BACKGROUND

The Hutchinson Study of High School Smoking was the first randomized trial to show effectiveness of a smoking cessation intervention on 6-months prolonged smoking abstinence at one year post-intervention in a large population-based sample of adolescent smokers. An important question remains: Do the positive effects from teen smoking cessation interventions seen at up to 12 months post-intervention endure into young adulthood? This study examines for the first time whether such positive early effects from teen smoking cessation intervention can endure into young adulthood in the absence of additional intervention.

METHODS

High school smokers (n = 2,151) were proactively recruited into the trial from fifty randomly selected Washington State high schools randomized to the experimental (Motivational Interviewing + Cognitive Behavioral Skills Training telephone counseling intervention) or control (no intervention) condition. These smokers were followed to 7 years post high school to ascertain rates of six-year prolonged smoking abstinence in young adulthood. All statistical tests are two-sided.

RESULTS

No evidence of intervention impact at seven years post high school was observed for the main endpoint of six-year prolonged abstinence, neither among all smokers (14.2% in the experimental condition vs. 13.1% in the control condition, difference = +1.1%, 95% confidence interval (CI) = -3.4 to 5.8, p = .61), nor among the subgroups of daily smokers and less-than-daily smokers, nor among other a priori subgroups. But, observed among males was some evidence of an intervention impact on two endpoints related to progress towards quitting: reduction in number of days smoked in the past month, and increase in the length of the longest quit attempt in the past year.

CONCLUSIONS

There was no evidence from this trial among adolescent smokers that positive effectiveness of the proactive telephone intervention for smoking abstinence, observed previously at one year post-intervention, was sustained for the long-term into young adulthood. In light of the positive short-term effectiveness consistently observed from this and other trials for teen smokers, together with the lack of evidence from this study that such short-term impact can endure into young adulthood, sustained interventions that continue into young adulthood should be developed and tested for long-term impact.

TRIAL REGISTRATION

ClinicalTrials.gov NCT00115882.

Sample size calculation for a stepped wedge trial

BACKGROUND

Stepped wedge trials (SWTs) can be considered as a variant of a clustered randomised trial, although in many ways they embed additional complications from the point of view of statistical design and analysis. While the literature is rich for standard parallel or clustered randomised clinical trials (CRTs), it is much less so for SWTs. The specific features of SWTs need to be addressed properly in the sample size calculations to ensure valid estimates of the intervention effect.

METHODS

We critically review the available literature on analytical methods to perform sample size and power calculations in a SWT. In particular, we highlight the specific assumptions underlying currently used methods and comment on their validity and potential for extensions. Finally, we propose the use of simulation-based methods to overcome some of the limitations of analytical formulae. We performed a simulation exercise in which we compared simulation-based sample size computations with analytical methods and assessed the impact of varying the basic parameters to the resulting sample size/power, in the case of continuous and binary outcomes and assuming both cross-sectional data and the closed cohort design.

RESULTS

We compared the sample size requirements for a SWT in comparison to CRTs based on comparable number of measurements in each cluster. In line with the existing literature, we found that when the level of correlation within the clusters is relatively high (for example, greater than 0.1), the SWT requires a smaller number of clusters. For low values of the intracluster correlation, the two designs produce more similar requirements in terms of total number of clusters. We validated our simulation-based approach and compared the results of sample size calculations to analytical methods; the simulation-based procedures perform well, producing results that are extremely similar to the analytical methods. We found that usually the SWT is relatively insensitive to variations in the intracluster correlation, and that failure to account for a potential time effect will artificially and grossly overestimate the power of a study.

CONCLUSIONS

We provide a framework for handling the sample size and power calculations of a SWT and suggest that simulation-based procedures may be more effective, especially in dealing with the specific features of the study at hand. In selected situations and depending on the level of intracluster correlation and the cluster size, SWTs may be more efficient than comparable CRTs. However, the decision about the design to be implemented will be based on a wide range of considerations, including the cost associated with the number of clusters, number of measurements and the trial duration.

Stochastic variation in network epidemic models: implications for the design of community level HIV prevention trials

Important sources of variation in the spread of HIV in communities arise from overlapping sexual networks and heterogeneity in biological and behavioral risk factors in populations. These sources of variation are not routinely accounted for in the design of HIV prevention trials. In this paper, we use agent-based models to account for these sources of variation. We illustrate the approach with an agent-based model for the spread of HIV infection among men who have sex with men in South Africa. We find that traditional sample size approaches that rely on binomial (or Poisson) models are inadequate and can lead to underpowered studies. We develop sample size and power formulas for community randomized trials that incorporate estimates of variation determined from agent-based models. We conclude that agent-based models offer a useful tool in the design of HIV prevention trials.

Sample size considerations in the design of cluster randomized trials of combination HIV prevention

Background Cluster randomized trials have been utilized to evaluate the effectiveness of HIV prevention strategies on reducing incidence. Design of such studies must take into account possible correlation of outcomes within randomized units. Purpose To discuss power and sample size considerations for cluster randomized trials of combination HIV prevention, using an HIV prevention study in Botswana as an illustration. Methods We introduce a new agent-based model to simulate the community-level impact of a combination prevention strategy and investigate how correlation structure within a community affects the coefficient of variation - an essential parameter in designing a cluster randomized trial. Results We construct collections of sexual networks and then propagate HIV on them to simulate the disease epidemic. Increasing level of sexual mixing between intervention and standard-of-care (SOC) communities reduces the difference in cumulative incidence in the two sets of communities. Fifteen clusters per arm and 500 incidence cohort members per community provide 95% power to detect the projected difference in cumulative HIV incidence between SOC and intervention communities (3.93% and 2.34%) at the end of the third study year, using a coefficient of variation 0.25. Although available formulas for calculating sample size for cluster randomized trials can be derived by assuming an exchangeable correlation structure within clusters, we show that deviations from this assumption do not generally affect the validity of such formulas. Limitations We construct sexual networks based on data from Likoma Island, Malawi, and base disease progression on longitudinal estimates from an incidence cohort in Botswana and in Durban as well as a household survey in Mochudi, Botswana. Network data from Botswana and larger sample sizes to estimate rates of disease progression would be useful in assessing the robustness of our model results. Conclusion Epidemic modeling plays a critical role in planning and evaluating interventions for prevention. Simulation studies allow us to take into consideration available information on sexual network characteristics, such as mixing within and between communities as well as coverage levels for different prevention modalities in the combination prevention package.

Promoting state health department evidence-based cancer and chronic disease prevention: a multi-phase dissemination study with a cluster randomized trial component

Background

Cancer and other chronic diseases reduce quality and length of life and productivity, and represent a significant financial burden to society. Evidence-based public health approaches to prevent cancer and other chronic diseases have been identified in recent decades and have the potential for high impact. Yet, barriers to implement prevention approaches persist as a result of multiple factors including lack of organizational support, limited resources, competing emerging priorities and crises, and limited skill among the public health workforce. The purpose of this study is to learn how best to promote the adoption of evidence based public health practice related to chronic disease prevention.

Methods/design

This paper describes the methods for a multi-phase dissemination study with a cluster randomized trial component that will evaluate the dissemination of public health knowledge about evidence-based prevention of cancer and other chronic diseases. Phase one involves development of measures of practitioner views on and organizational supports for evidence-based public health and data collection using a national online survey involving state health department chronic disease practitioners. In phase two, a cluster randomized trial design will be conducted to test receptivity and usefulness of dissemination strategies directed toward state health department chronic disease practitioners to enhance capacity and organizational support for evidence-based chronic disease prevention. Twelve state health department chronic disease units will be randomly selected and assigned to intervention or control. State health department staff and the university-based study team will jointly identify, refine, and select dissemination strategies within intervention units. Intervention (dissemination) strategies may include multi-day in-person training workshops, electronic information exchange modalities, and remote technical assistance. Evaluation methods include pre-post surveys, structured qualitative phone interviews, and abstraction of state-level chronic disease prevention program plans and progress reports.

Trial registration

clinicaltrials.gov: NCT01978054.

Strategies to improve retention in randomised trials

BACKGROUND

Loss to follow-up from randomised trials can introduce bias and reduce study power, affecting the generalisability, validity and reliability of results. Many strategies are used to reduce loss to follow-up and improve retention but few have been formally evaluated.

OBJECTIVES

To quantify the effect of strategies to improve retention on the proportion of participants retained in randomised trials and to investigate if the effect varied by trial strategy and trial setting.

SEARCH METHODS

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, PreMEDLINE, EMBASE, PsycINFO, DARE, CINAHL, Campbell Collaboration's Social, Psychological, Educational and Criminological Trials Register, and ERIC. We handsearched conference proceedings and publication reference lists for eligible retention trials. We also surveyed all UK Clinical Trials Units to identify further studies.

SELECTION CRITERIA

We included eligible retention trials of randomised or quasi-randomised evaluations of strategies to increase retention that were embedded in 'host' randomised trials from all disease areas and healthcare settings. We excluded studies aiming to increase treatment compliance.

DATA COLLECTION AND ANALYSIS

We contacted authors to supplement or confirm data that we had extracted. For retention trials, we recorded data on the method of randomisation, type of strategy evaluated, comparator, primary outcome, planned sample size, numbers randomised and numbers retained. We used risk ratios (RR) to evaluate the effectiveness of the addition of strategies to improve retention. We assessed heterogeneity between trials using the Chi(2) and I(2) statistics. For main trials that hosted retention trials, we extracted data on disease area, intervention, population, healthcare setting, sequence generation and allocation concealment.

MAIN RESULTS

We identified 38 eligible retention trials. Included trials evaluated six broad types of strategies to improve retention. These were incentives, communication strategies, new questionnaire format, participant case management, behavioural and methodological interventions. For 34 of the included trials, retention was response to postal and electronic questionnaires with or without medical test kits. For four trials, retention was the number of participants remaining in the trial. Included trials were conducted across a spectrum of disease areas, countries, healthcare and community settings. Strategies that improved trial retention were addition of monetary incentives compared with no incentive for return of trial-related postal questionnaires (RR 1.18; 95% CI 1.09 to 1.28, P value < 0.0001), addition of an offer of monetary incentive compared with no offer for return of electronic questionnaires (RR 1.25; 95% CI 1.14 to 1.38, P value < 0.00001) and an offer of a GBP20 voucher compared with GBP10 for return of postal questionnaires and biomedical test kits (RR 1.12; 95% CI 1.04 to 1.22, P value < 0.005). The evidence that shorter questionnaires are better than longer questionnaires was unclear (RR 1.04; 95% CI 1.00 to 1.08, P value = 0.07) and the evidence for questionnaires relevant to the disease/condition was also unclear (RR 1.07; 95% CI 1.01 to 1.14). Although each was based on the results of a single trial, recorded delivery of questionnaires seemed to be more effective than telephone reminders (RR 2.08; 95% CI 1.11 to 3.87, P value = 0.02) and a 'package' of postal communication strategies with reminder letters appeared to be better than standard procedures (RR 1.43; 95% CI 1.22 to 1.67, P value < 0.0001). An open trial design also appeared more effective than a blind trial design for return of questionnaires in one fracture prevention trial (RR 1.37; 95% CI 1.16 to 1.63, P value = 0.0003).There was no good evidence that the addition of a non-monetary incentive, an offer of a non-monetary incentive, 'enhanced' letters, letters delivered by priority post, additional reminders, or questionnaire question order either increased or decreased trial questionnaire response/retention. There was also no evidence that a telephone survey was either more or less effective than a monetary incentive and a questionnaire. As our analyses are based on single trials, the effect on questionnaire response of using offers of charity donations, sending reminders to trial sites and when a questionnaire is sent, may need further evaluation. Case management and behavioural strategies used for trial retention may also warrant further evaluation.

AUTHORS' CONCLUSIONS

Most of the retention trials that we identified evaluated questionnaire response. There were few evaluations of ways to improve participants returning to trial sites for trial follow-up. Monetary incentives and offers of monetary incentives increased postal and electronic questionnaire response. Some other strategies evaluated in single trials looked promising but need further evaluation. Application of the findings of this review would depend on trial setting, population, disease area, data collection and follow-up procedures.

Interventions for tobacco use cessation in people in treatment for or recovery from substance abuse

This is the protocol for a review and there is no abstract. The objectives are as follows: To evaluate the effectiveness of tobacco cessation therapy offered concurrently with treatment for drug and alcohol addiction.

Interventions for tobacco use prevention in Indigenous youth

BACKGROUND

Tobacco use in Indigenous populations (people who have inhabited a country for thousands of years) is often double that in the non-Indigenous population. Addiction to nicotine usually begins during early adolescence and young people who reach the age of 18 as non-smokers are unlikely to become smokers thereafter. Indigenous youth in particular commence smoking at an early age, and a disproportionate burden of substance-related morbidity and mortality exists as a result.

OBJECTIVES

To evaluate the effectiveness of intervention programmes to prevent tobacco use initiation or progression to regular smoking amongst young Indigenous populations and to summarise these approaches for future prevention programmes and research.

SEARCH METHODS

The Cochrane Tobacco Addiction Group Specialised Register was searched in November 2011, with additional searches run in MEDLINE. Online clinical trial databases and publication references were also searched for potential studies.

SELECTION CRITERIA

We included randomized and non-randomized controlled trials aiming to prevent tobacco use initiation or progression from experimentation to regular tobacco use in Indigenous youth. Interventions could include school-based initiatives, mass media, multi-component community level interventions, family-based programmes or public policy.

DATA COLLECTION AND ANALYSIS

Data pertaining to methodology, participants, interventions and outcomes were extracted by one reviewer and checked by a second, whilst information on risk of bias was extracted independently by a combination of two reviewers. Studies were assessed by qualitative narrative synthesis, as insufficient data were available to conduct a meta-analysis. The review process was examined by an Indigenous (Aboriginal) Australian for applicability, acceptability and content.

MAIN RESULTS

Two studies met all of the eligibility criteria for inclusion within the review and a third was identified as ongoing. The two included studies employed multi-component community-based interventions tailored to the specific cultural aspects of the population and were based in Native American populations (1505 subjects in total). No difference was observed in weekly smoking at 42 months follow-up in the one study assessing this outcome (skills-community group versus control: risk ratio [RR] 0.95, 95% CI 0.78 to 1.14; skills-only group versus control: RR 0.86, 95% CI 0.71 to 1.05). For smokeless tobacco use, no difference was found between the skills-community arm and the control group at 42 weeks (RR 0.93, 95% CI 0.67 to 1.30), though a significant difference was observed between the skills-only arm and the control group (RR 0.57, 95% CI 0.39 to 0.85). Whilst the second study found positive changes for tobacco use in the intervention arm at post test (p < 0.05), this was not maintained at six month follow-up (change score -0.11 for intervention and 0.07 for control). Both studies were rated as high or unclear risk of bias in seven or more domains (out of a total of 10).

AUTHORS' CONCLUSIONS

Based on the available evidence, a conclusion cannot be drawn as to the efficacy of tobacco prevention initiatives tailored for Indigenous youth. This review highlights the paucity of data and the need for more research in this area. Smoking prevalence in Indigenous youth is twice that of the non-Indigenous population, with tobacco experimentation commencing at an early age. As such, a significant health disparity exists where Indigenous populations, a minority, are over-represented in the burden of smoking-related morbidity and mortality. Methodologically rigorous trials are needed to investigate interventions aimed at preventing the uptake of tobacco use amongst Indigenous youth and to assist in bridging the gap between tobacco-related health disparities in Indigenous and non-Indigenous populations.

Interventions for smoking cessation in Indigenous populations

BACKGROUND

Tobacco use in Indigenous populations (people who have inhabited a country for thousands of years) is often double that of the non-Indigenous population. A disproportionate burden of substance-related morbidity and mortality exists as a result.

OBJECTIVES

To evaluate the effectiveness of smoking cessation interventions in Indigenous populations and to summarise these approaches for future cessation programmes and research.

SEARCH METHODS

The Cochrane Tobacco Addiction Group Specialised Register of Trials was searched (April 2011), with additional searches of MEDLINE (May 2011). Online clinical trial databases and publication references were also searched for potential studies.

SELECTION CRITERIA

We included randomized and non-randomized controlled trials for smoking cessation interventions in Indigenous populations. Interventions could include pharmacotherapies, cognitive and behavioural therapies, alternative therapies, public policy and combination therapies. No attempts were made to re-define Indigenous status for the purpose of including a study in this review.

DATA COLLECTION AND ANALYSIS

Data pertaining to methodology, participants, interventions and outcomes were extracted by one reviewer and checked by a second, whilst methodological quality was extracted independently by two reviewers. Studies were assessed by qualitative narrative synthesis and where possible meta-analysis. The review process was examined by an Indigenous (Aboriginal) Australian for applicability, acceptability and content.

MAIN RESULTS

Four studies met all of the eligibility criteria for inclusion within the review. Two used combination therapies consisting of a pharmacotherapy combined with cognitive and behavioural therapies, whilst the remaining two used cognitive and behavioural therapy through counselling, one via text message support and the other delivered via clinic doctors trained in smoking cessation techniques. Smoking cessation data were pooled across all studies producing a statistically and clinically significant effect in favour of the intervention (risk ratio 1.43, 95%CI 1.03 to 1.98, p=0.032), however following sensitivity analysis a statistically non-significant but clinically significant effect was observed in favour of the intervention (risk ratio 1.33, 95%CI 0.95 to 1.85, p=NS) .

AUTHORS' CONCLUSIONS

A significant health disparity exists, whereby Indigenous populations, a minority, are over-represented in the burden of smoking-related morbidity and mortality. This review highlights the paucity of evidence available to evaluate the effectiveness of smoking cessation interventions, despite the known success of these interventions in non-Indigenous populations. Due to this lack of published investigations, the external validity of this review is limited, as is the ability to draw reliable conclusions from the results. The limited but available evidence reported does indicate that smoking cessation interventions specifically targeted at Indigenous populations can produce smoking abstinence. However this evidence base is not strong with a small number of methodologically sound trials investigating these interventions. More rigorous trials are now required to assist in bridging the gap between tobacco related health disparities in Indigenous and non-Indigenous populations.

Community interventions for preventing smoking in young people

BACKGROUND

Cigarette smoking is one of the leading causes of preventable death in the world. Decisions to smoke are often made within a broad social context and therefore community interventions using coordinated, multi-component programmes may be effective in influencing the smoking behaviour of young people.

OBJECTIVES

To determine the effectiveness of multi-component community based interventions in influencing smoking behaviour, which includes preventing the uptake of smoking in young people.

SEARCH STRATEGY

The Tobacco Addiction group's specialised register, Medline and other health, psychology and public policy electronic databases were searched, the bibliographies of identified studies were checked and raw data was requested from study authors. Searches were updated in August 2010.

SELECTION CRITERIA

Randomized and non randomized controlled trials that assessed the effectiveness of multi-component community interventions compared to no intervention or to single component or school-based programmes only. Reported outcomes had to include smoking behaviour in young people under the age of 25 years.

DATA COLLECTION AND ANALYSIS

Information relating to the characteristics and the content of community interventions, participants, outcomes and methods of the study was extracted by one reviewer and checked by a second. Studies were combined in a meta-analysis where possible and reported in narrative synthesis in text and table.

MAIN RESULTS

Twenty-five studies were included in the review and sixty-eight studies did not meet all of the inclusion criteria. All studies used a controlled trial design, with fifteen using random allocation of schools or communities. One study reported a reduction in short-term smoking prevalence (twelve months or less), while nine studies detected significant long-term effects. Two studies reported significantly lower smoking rates in the control population while the remaining thirteen studies showed no significant difference between groups. Improvements were seen in secondary outcomes for intentions to smoke in six out of eight studies, attitudes in five out of nine studies, perceptions in two out of six studies and knowledge in three out of six studies, while significant differences in favour of the control were seen in one of the nine studies assessing attitudes and one of six studies assessing perceptions.

AUTHORS' CONCLUSIONS

There is some evidence to support the effectiveness of community interventions in reducing the uptake of smoking in young people, but the evidence is not strong and contains a number of methodological flaws.

Group-randomized trial of a proactive, personalized telephone counseling intervention for adolescent smoking cessation

BACKGROUND

The Hutchinson Study of High School Smoking randomized trial was designed to rigorously evaluate a proactive, personalized telephone counseling intervention for adolescent smoking cessation.

METHODS

Fifty randomly selected Washington State high schools were randomized to the experimental or control condition. High school junior smokers were proactively identified (N = 2151). Trained counselors delivered the motivational interviewing plus cognitive behavioral skills training telephone intervention to smokers in experimental schools during their senior year of high school. Participants were followed up, with 88.8% participation, to outcome ascertainment more than 1 year after random assignment. The main outcome was 6-months prolonged abstinence from smoking. All statistical tests were two-sided.

RESULTS

The intervention increased the percentage who achieved 6-month prolonged smoking abstinence among all smokers (21.8% in the experimental condition vs 17.7% in the control condition, difference = 4.0%, 95% confidence interval [CI] = -0.2 to 8.1, P = .06) and in particular among daily smokers (10.1% vs 5.9%, difference = 4.1%, 95% CI = 0.8 to 7.1, P = .02). There was also generally strong evidence of intervention impact for 3-month, 1-month, and 7-day abstinence and duration since last cigarette (P = .09, .015, .01, and .03, respectively). The intervention effect was strongest among male daily smokers and among female less-than-daily smokers.

CONCLUSIONS

Proactive identification and recruitment of adolescents via public high schools can produce a high level of intervention reach; a personalized motivational interviewing plus cognitive behavioral skills training counseling intervention delivered by counselor-initiated telephone calls is effective in increasing teen smoking cessation; and both daily and less-than-daily teen smokers participate in and benefit from telephone-based smoking cessation intervention.

An application of a modified constrained randomization process to a practice-based cluster randomized trial to improve colorectal cancer screening

BACKGROUND

When designing cluster randomized trials, it is important for researchers to be familiar with strategies to achieve valid study designs given limited resources. Constrained randomization is a technique to help ensure balance on pre-specified baseline covariates.

METHODS

The goal was to develop a randomization scheme that balanced 16 intervention and 16 control practices with respect to 7 factors that may influence improvement in study outcomes during a 4-year cluster randomized trial to improve colorectal cancer screening within a primary care practice-based research network. We used a novel approach that included simulating 30,000 randomization schemes, removing duplicates, identifying which schemes were sufficiently balanced, and randomly selecting one scheme for use in the trial. For a given factor, balance was considered achieved when the frequency of each factor's sub-classifications differed by no more than 1 between intervention and control groups. The population being studied includes approximately 32 primary care practices located in 19 states within the U.S. that care for approximately 56,000 patients at least 50 years old.

RESULTS

Of 29,782 unique simulated randomization schemes, 116 were determined to be balanced according to pre-specified criteria for all 7 baseline covariates. The final randomization scheme was randomly selected from these 116 acceptable schemes.

CONCLUSIONS

Using this technique, we were successfully able to find a randomization scheme that allocated 32 primary care practices into intervention and control groups in a way that preserved balance across 7 baseline covariates. This process may be a useful tool for ensuring covariate balance within moderately large cluster randomized trials.

Imputation strategies for missing continuous outcomes in cluster randomized trials

In cluster randomized trials, intact social units such as schools, worksites or medical practices - rather than individuals themselves - are randomly allocated to intervention and control conditions, while the outcomes of interest are then observed on individuals within each cluster. Such trials are becoming increasingly common in the fields of health promotion and health services research. Attrition is a common occurrence in randomized trials, and a standard approach for dealing with the resulting missing values is imputation. We consider imputation strategies for missing continuous outcomes, focusing on trials with a completely randomized design in which fixed cohorts from each cluster are enrolled prior to random assignment. We compare five different imputation strategies with respect to Type I and Type II error rates of the adjusted two-sample t -test for the intervention effect. Cluster mean imputation is compared with multiple imputation, using either within-cluster data or data pooled across clusters in each intervention group. In the case of pooling across clusters, we distinguish between standard multiple imputation procedures which do not account for intracluster correlation and a specialized procedure which does account for intracluster correlation but is not yet available in standard statistical software packages. A simulation study is used to evaluate the influence of cluster size, number of clusters, degree of intracluster correlation, and variability among cluster follow-up rates. We show that cluster mean imputation yields valid inferences and given its simplicity, may be an attractive option in some large community intervention trials which are subject to individual-level attrition only; however, it may yield less powerful inferences than alternative procedures which pool across clusters especially when the cluster sizes are small and cluster follow-up rates are highly variable. When pooling across clusters, the imputation procedure should generally take intracluster correlation into account to obtain valid inferences; however, as long as the intracluster correlation coefficient is small, we show that standard multiple imputation procedures may yield acceptable type I error rates; moreover, these procedures may yield more powerful inferences than a specialized procedure, especially when the number of available clusters is small. Within-cluster multiple imputation is shown to be the least powerful among the procedures considered.

A simple approach to test for interaction between intervention and an individual-level variable in community randomized trials

OBJECTIVE

To develop a simple and robust approach for the test of interaction between community intervention and an individual-level variable suitable for use in typical situations of community randomized trials (CRTs), i.e. small number of communities but large number of subjects per community.

METHODS

We propose a method based on taking the difference between summary statistics from groups of individuals with and without an attribute within each community, then applying a two-sample t-test or Wilcoxon test to compare the distribution of within-community differences between trial arms. The method is evaluated using simulations and illustrated using data from a CRT of a health education intervention. Approximate sample size formulas are derived.

RESULTS

Analyses based on the t-test give power very close to expected level in a variety of situations, including when the summary statistics are not symmetrically distributed across communities, the covariate is not distributed as planned, and the number of communities per intervention arm ranges from 8 to 20. Even in the situation with as few as four communities per arm, the power is only slightly lower than expected. Type I error rates always closely follow 5% as required, whether the distributional assumption is correct or not. The application of the Wilcoxon test appears too conservative.

CONCLUSIONS

The proposed approach to test for interaction is valid and easy to use. The application of the t-test in this setting is robust.

Addressing challenges in adolescent smoking cessation: design and baseline characteristics of the HS Group-Randomized trial

OBJECTIVE

Well-documented challenges have hampered both intervention development and research in teen smoking cessation. Addressing these challenges, the Hutchinson Study of High School Smoking (HS Study), the largest group-randomized trial in adolescent smoking cessation to date, incorporates several design innovations to investigate the effect of a counselor-initiated, individually tailored telephone counseling smoking cessation intervention for older adolescents. This paper presents and discusses these innovative design features, and baseline findings on the resulting study population.

METHOD

The trial used a population-based survey to proactively identify and recruit all high school juniors who had smoked in the past month - potentially expanding intervention reach to all smokers, even those who smoked less than daily and those not motivated to quit. For ethical and intervention reasons, some nonsmokers were enrolled in the intervention, also. Other important design features included the random allocation of schools into experimental conditions (intervention vs. no-intervention control) and a multi-wave design.

RESULTS AND CONCLUSION

The design innovations address problems and challenges identified in adolescent smoking cessation literature. The heterogeneous baseline characteristics of the study population, well-balanced between the two arms, have three significant implications: They (1) demonstrate the effectiveness of the trial's design features, (2) highlight several intervention-related issues, and (3) provide assurance that the trial's evaluation of intervention effectiveness will be unbiased.

The merits of breaking the matches: a cautionary tale

Matched-pair cluster randomization trials are frequently adopted as the design of choice for evaluating an intervention offered at the community level. However, previous research has demonstrated that a strategy of breaking the matches and performing an unmatched analysis may be more efficient than performing a matched analysis on the resulting data, particularly when the total number of communities is small and the matching is judged as relatively ineffective. The research concerning this question has naturally focused on testing the effect of intervention. However, a secondary objective of many community intervention trials is to investigate the effect of individual-level risk factors on one or more outcome variables. Focusing on the case of a continuous outcome variable, we show that the practice of performing an unmatched analysis on data arising from a matched-pair design can lead to bias in the estimated regression coefficient, and a corresponding test of significance which is overly liberal. However, for large-scale community intervention trials, which typically recruit a relatively small number of large clusters, such an analysis will generally be both valid and efficient. We also consider other approaches to testing the effect of an individual-level risk factor in a matched-pair cluster randomization design, including a generalized linear model approach that preserves the matching, a two-stage cluster-level analysis, and an approach based on generalized estimating equations.

A cluster-randomized trial of benchmarking and multimodal quality improvement to improve rates of survival free of bronchopulmonary dysplasia for infants with birth weights of less than 1250 grams

OBJECTIVE

We tested whether NICU teams trained in benchmarking and quality improvement would change practices and improve rates of survival without bronchopulmonary dysplasia in inborn neonates with birth weights of <1250 g.

METHODS

A cluster-randomized trial enrolled 4093 inborn neonates with birth weights of <1250 g at 17 centers of the National Institute of Child Health and Human Development Neonatal Research Network. Three centers were selected as best performers, and the remaining 14 centers were randomized to intervention or control. Changes in rates of survival free of bronchopulmonary dysplasia were compared between study year 1 and year 3.

RESULTS

Intervention centers implemented potentially better practices successfully; changes included reduced oxygen saturation targets and reduced exposure to mechanical ventilation. Five of 7 intervention centers and 2 of 7 control centers implemented use of high-saturation alarms to reduce oxygen exposure. Lower oxygen saturation targets reduced oxygen levels in the first week of life. Despite these changes, rates of survival free of bronchopulmonary dysplasia were all similar between intervention and control groups and remained significantly less than the rate achieved in the best-performing centers (73.3%).

CONCLUSIONS

In this cluster-randomized trial, benchmarking and multimodal quality improvement changed practices but did not reduce bronchopulmonary dysplasia rates.

Developments in cluster randomized trials and Statistics in Medicine

The design and analysis of cluster randomized trials has been a recurrent theme in Statistics in Medicine since the early volumes. In celebration of 25 years of Statistics in Medicine, this paper reviews recent developments, particularly those that featured in the journal. Issues in design such as sample size calculations, matched paired designs, cohort versus cross-sectional designs, and practical design problems are covered. Developments in analysis include modification of robust methods to cope with small numbers of clusters, generalized estimation equations, population averaged and cluster specific models. Finally, issues on presenting data, some other clustering issues and the general problem of evaluating complex interventions are briefly mentioned.

A matched-pair cluster design study protocol to evaluate implementation of the Canadian C-spine rule in hospital emergency departments: Phase III

BACKGROUND

Physicians in Canadian emergency departments (EDs) annually treat 185,000 alert and stable trauma victims who are at risk for cervical spine (C-spine) injury. However, only 0.9% of these patients have suffered a cervical spine fracture. Current use of radiography is not efficient. The Canadian C-Spine Rule is designed to allow physicians to be more selective and accurate in ordering C-spine radiography, and to rapidly clear the C-spine without the need for radiography in many patients. The goal of this phase III study is to evaluate the effectiveness of an active strategy to implement the Canadian C-Spine Rule into physician practice. Specific objectives are to: 1) determine clinical impact, 2) determine sustainability, 3) evaluate performance, and 4) conduct an economic evaluation.

METHODS

We propose a matched-pair cluster design study that compares outcomes during three consecutive 12-months "before," "after," and "decay" periods at six pairs of "intervention" and "control" sites. These 12 hospital ED sites will be stratified as "teaching" or "community" hospitals, matched according to baseline C-spine radiography ordering rates, and then allocated within each pair to either intervention or control groups. During the "after" period at the intervention sites, simple and inexpensive strategies will be employed to actively implement the Canadian C-Spine Rule. The following outcomes will be assessed: 1) measures of clinical impact, 2) performance of the Canadian C-Spine Rule, and 3) economic measures. During the 12-month "decay" period, implementation strategies will continue, allowing us to evaluate the sustainability of the effect. We estimate a sample size of 4,800 patients in each period in order to have adequate power to evaluate the main outcomes.

DISCUSSION

Phase I successfully derived the Canadian C-Spine Rule and phase II confirmed the accuracy and safety of the rule, hence, the potential for physicians to improve care. What remains unknown is the actual change in clinical behaviors that can be affected by implementation of the Canadian C-Spine Rule, and whether implementation can be achieved with simple and inexpensive measures. We believe that the Canadian C-Spine Rule has the potential to significantly reduce health care costs and improve the efficiency of patient flow in busy Canadian EDs.

Design and analysis of stepped wedge cluster randomized trials

Cluster randomized trials (CRT) are often used to evaluate therapies or interventions in situations where individual randomization is not possible or not desirable for logistic, financial or ethical reasons. While a significant and rapidly growing body of literature exists on CRTs utilizing a "parallel" design (i.e. I clusters randomized to each treatment), only a few examples of CRTs using crossover designs have been described. In this article we discuss the design and analysis of a particular type of crossover CRT - the stepped wedge - and provide an example of its use.

Randomization inference for balanced cluster-randomized trials

This paper discusses the choice of randomization tests for inferences from cluster-randomized trials that have been designed to ensure a balanced allocation of clusters to treatments. Methods for covariate-adjusted randomization tests are reviewed and their application to balanced cluster-randomized trials discussed. Two cluster-randomized trials with balanced designs are used to illustrate the choices that can be made in selecting a randomization test, and methods for obtaining confidence intervals for treatment effects are illustrated. The balance imposed by the randomization in these trials makes adjustment for covariates less beneficial than for an unbalanced design. However, the adjusted analyses do not appear generally to have worse properties than the unadjusted ones, and may provide protection against any imbalance that has not been controlled for in the design. The only case when adjustment for covariates may result in worse precision is when a large number of cluster-level covariates are included in the analysis. An expression is provided that allows the size of this effect to be calculated for any given set of cluster-level covariates.

A comparison of permutation and mixed-model regression methods for the analysis of simulated data in the context of a group-randomized trial

Our first purpose was to determine whether, in the context of a group-randomized trial (GRT) with Gaussian errors, permutation or mixed-model regression methods fare better in the presence of measurable confounding in terms of their Monte Carlo type I error rates and power. Our results indicate that given a proper randomization, the type I error rate is similar for both methods, whether unadjusted or adjusted, even in small studies. However, our results also show that should the investigator face the unfortunate circumstance in which modest confounding exists in the only realization available, the unadjusted analysis risks a type I error; in this regard, there was little to distinguish the two methods. Finally, our results show that power is similar for the two methods and, not surprisingly, better for the adjusted tests. Our second purpose was to examine the relative performance of permutation and mixed-model regression methods in the context of a GRT when the normality assumptions underlying the mixed model are violated. Published studies have examined the impact of violation of this assumption at the member level only. Our findings indicate that both methods perform well when the assumption is violated so long as the ICC is very small and the design is balanced at the group level. However, at ICC>or=0.01, the permutation test carries the nominal type I error rate while the model-based test is conservative and so less powerful. Binomial group- and member-level errors did not otherwise change the relative performance of the two methods with regard to confounding.

Reliability, effect size, and responsiveness of health status measures in the design of randomized and cluster-randomized trials

BACKGROUND

New health status survey instruments are often described by their psychometric (measurement) properties, such as Validity, Reliability, Effect Size, and Responsiveness. For cluster-randomized trials, another important statistic is the Intraclass Correlation (ICC) for the instrument within clusters. Studies using better instruments can be performed with smaller sample sizes, but better instruments may be more expensive in terms of dollars, opportunity cost, or poorer data quality due to the response burden of longer instruments.

METHODS

We defined the psychometric statistics in terms of a mathematical model, and examined the power of a two-sample test as a function of the test-retest Reliability, Effect Size, Responsiveness, and Intraclass Correlation of the instrument. We examined the "cost-effectiveness" of using a one-item versus a five-item measure of mental health status.

FINDINGS

Under the standard model for measurement error, the psychometric statistics are all functions of the same error term. They are also functions of the setting in which they were estimated. In randomized trials, power is a function of Reliability and sample size, and a less reliable instrument can achieve the desired power if N is increased. In cluster-randomized trials, adequate power may be obtained by increasing the number of clusters per treatment group (and often the number of persons per cluster), as well as by choosing a more reliable instrument. The one-item measure of mental health status may be more cost-effective than the five-item measure in some situations.

CONCLUSION

If the goal is to diagnose or refer individual patients, an instrument with high Validity and Reliability is needed. In settings where the sample sizes are large or can be increased easily, any valid instrument may be cost-effective. It is likely that many published values of psychometric statistics are accurate only in settings similar to that in which they were estimated.

Using baseline data to design a group randomized trial

Group randomized trials (GRT) are often designed with relatively little preliminary data available to estimate key parameters. In this paper, however, the opposite situation is considered-very good baseline data are available on the primary outcome of interest. These data can then be used to inform key design and analysis decisions such as (i) should the trial be designed as an unmatched or pair-matched study, or stratified in some other fashion; (ii) is analysis of "change from baseline" preferable to using end-of-study data alone; and (iii) what power might be expected by pursuing these various strategies. The results are applied to a GRT for sexually transmitted diseases prevention in Peru.

Evaluation of Community-Intervention Trials via Generalized Linear Mixed Models

In community-intervention trials, communities, rather than individuals, are randomized to experimental arms. Generalized linear mixed models offer a flexible parametric framework for the evaluation of community-intervention trials, incorporating both systematic and random variations at the community and individual levels. We propose here a simple two-stage inference method for generalized linear mixed models, specifically tailored to the analysis of community-intervention trials. In the first stage, community-specific random effects are estimated from individual-level data, adjusting for the effects of individual-level covariates. This reduces the model approximately to a linear mixed model with the unit of analysis being community. Because the number of communities is typically small in community-intervention studies, we apply the small-sample inference method of Kenward and Roger (1997, Biometrics53, 983-997) to the linear mixed model of second stage. We show by simulation that, under typical settings of community-intervention studies, the proposed approach improves the inference on the intervention-effect parameter uniformly over both the linearized mixed-effect approach and the adaptive Gaussian quadrature approach for generalized linear mixed models. This work is motivated by a series of large randomized trials that test community interventions for promoting cancer preventive lifestyles and behaviors.

Mounting a community-randomized trial: sample size, matching, selection, and randomization issues in PRISM

This paper discusses some of the processes for establishing a large cluster-randomized trial of a community and primary care intervention in 16 local government areas in Victoria, Australia. The development of the trial in terms of design factors such as sample size estimates and the selection and randomization of communities to intervention or comparison is described. The intervention program to be implemented in Program of Resources, Information and Support for Mothers (PRISM) was conceived as a whole community approach to improving support for all mothers in the first 12 months after birth. A cluster-randomized trial was thus the design of choice from the outset. With a limited number of communities available, a matched-pair design with eight pairs was chosen. Sample size estimates, adjusting for the cluster randomization and the pair-matched design, showed that with eight pairs, on average, 800 women from each community would need to respond to provide sufficient power to determine a 3% reduction in the prevalence of maternal depression 6 months after birth-a reduction deemed to be a worthwhile impact of the intervention to be reliably detected at 80% power. The process of selecting suitable communities and matching them into pairs required careful collection of data on numbers of births, size of the local government areas (LGAs), and an assessment of the capacity of communities to implement the intervention. Ways of dealing with boundary issues associated with potential contamination are discussed. Methods for the selection of feasible configurations of sets of pairs and the ultimate allocation to intervention or comparison are provided in detail. Ultimately, all such studies are a balancing act between selecting the minimum number of communities to detect a meaningful outcome effect of an intervention and the maximum size budget and other resources allow.

Statistical and Substantive Inferences in Public Health: Issues in the Application of Multilevel Models

Multilevel statistical models have become increasingly popular among public health researchers over the past decade. Yet the enthusiasm with which these models are being adopted may obscure rather than solve some problems of statistical and substantive inference. We discuss the three most common applications of multilevel models in public health: (a) cluster-randomized trials, (b) observational studies of the multilevel etiology of health and disease, and (c) assessments of health care provider performance. In each area of investigation, we describe how multilevel models are being applied, comment on the validity of the statistical and substantive inferences being drawn, and suggest ways in which the strengths of multilevel models might be more fully exploited. We conclude with a call for more careful thinking about multilevel causal inference.

Design and analysis of group-randomized trials: a review of recent methodological developments

We review recent developments in the design and analysis of group-randomized trials (GRTs). Regarding design, we summarize developments in estimates of intraclass correlation, power analysis, matched designs, designs involving one group per condition, and designs in which individuals are randomized to receive treatments in groups. Regarding analysis, we summarize developments in marginal and conditional models, the sandwich estimator, model-based estimators, binary data, survival analysis, randomization tests, survey methods, latent variable methods and nonlinear mixed models, time series methods, global tests for multiple endpoints, mediation effects, missing data, trial reporting, and software. We encourage investigators who conduct GRTs to become familiar with these developments and to collaborate with methodologists who can strengthen the design and analysis of their trials.

Meta-analyses of cluster randomization trials. Power considerations

A commonly cited purpose for conducting a meta-analysis of randomized trials is to increase the statistical power for detecting the effect of an intervention on a specified set of endpoints. At the same time, it also has been noted by several authors that many large-scale cluster randomization trials have not had the power to detect small or even moderate effect sizes. The loss of efficiency associated with cluster randomization relative to individual randomization, and the frequent failure of investigators to take this loss of efficiency into account at the planning stage of a trial, undoubtedly contributes to this problem. In this article, the authors present an approach that may be used to estimate the power of a planned meta-analysis that includes trials that are cluster randomized. Two examples are presented.

A comparison of confidence interval methods for the intraclass correlation coefficient in cluster randomized trials

This study compared different methods for assigning confidence intervals to the analysis of variance estimator of the intraclass correlation coefficient (rho). The context of the comparison was the use of rho to estimate the variance inflation factor when planning cluster randomized trials. The methods were compared using Monte Carlo simulations of unbalanced clustered data and data from a cluster randomized trial of an intervention to improve the management of asthma in a general practice setting. The coverage and precision of the intervals were compared for data with different numbers of clusters, mean numbers of subjects per cluster and underlying values of rho. The performance of the methods was also compared for data with Normal and non-Normally distributed cluster specific effects. Results of the simulations showed that methods based upon the variance ratio statistic provided greater coverage levels than those based upon large sample approximations to the standard error of rho. Searle's method provided close to nominal coverage for data with Normally distributed random effects. Adjusted versions of Searle's method to allow for lack of balance in the data generally did not improve upon it either in terms of coverage or precision. Analyses of the trial data, however, showed that limits provided by Thomas and Hultquist's method may differ from those of the other variance ratio statistic methods when the arithmetic mean differs markedly from the harmonic mean cluster size. The simulation results demonstrated that marked non-Normality in the cluster level random effects compromised the performance of all methods. Confidence intervals for the methods were generally wide relative to the underlying size of rho suggesting that there may be great uncertainty associated with sample size calculations for cluster trials where large clusters are randomized. Data from cluster based studies with sample sizes much larger than those typical of cluster randomized trials are required to estimate rho with a reasonable degree of precision.

Some design issues in a community intervention trial

We present statistical considerations for the design of a 20-community randomized trial. The community intervention aims at multiple cancer prevention health behaviors including reducing dietary fat, increasing fruit and vegetable intake, smoking cessation, and increasing colorectal cancer screening. To better measure the overall impact of the intervention, individual endpoints as well as a global test of multiple endpoints are used. The statistical power analysis takes into account the heterogeneity between communities, the correlation between health behaviors over time, and the correlation between multiple endpoints. The study is being conducted in a relatively confined geographic area. Several measures have been taken to account for potential contamination. These include collecting information in baseline and follow-up surveys on intervention dose and conducting surveys in three similar communities in another part of the state.

Methods for the analysis of incidence rates in cluster randomized trials

BACKGROUND

The published literature on cluster randomized trials focuses on outcomes that are either continuous or binary. In many trials, the outcome is an incidence rate, such as mortality, based on person-years data. In this paper we review methods for the analysis of such data in cluster randomized trials and present some simple approaches.

METHODS

We discuss the choice of the measure of intervention effect and present methods for confidence interval estimation and hypothesis testing which are conceptually simple and easy to perform using standard statistical software. The method proposed for hypothesis testing applies a t-test to cluster observations. To control confounding, a Poisson regression model is fitted to the data incorporating all covariates except intervention status, and the analysis is carried out on the residuals from this model. The methods are presented for unpaired data, and extensions to paired or stratified clusters are outlined.

RESULTS

The methods are evaluated by simulation and illustrated by application to data from a trial of the effect of insecticide-impregnated bednets on child mortality.

CONCLUSIONS

The techniques provide a straightforward approach to the analysis of incidence rates in cluster randomized trials. Both the unadjusted analysis and the analysis adjusting for confounders are shown to be robust, even for very small numbers of clusters, in situations that are likely to arise in randomized trials.

Clinical trials of behavioural interventions with heterogeneous teaching subgroup effects

Behaviour modification is often delivered to teaching subgroups. For example, experimental and control smoking cessation programmes may be given to 15 classes (subgroups) with 10 (otherwise independent) individuals. We present general statistical tests and power estimates to compare continuous outcomes from two interventions in settings where the magnitude of teaching subgroup heterogeneity, number of subgroups and subgroup size can differ between intervention arms. An application is made to data from a trial to reduce disease-transmitting sexual behaviour. The statistical impact of teaching subgroup heterogeneity effect increases as the (a) number of participants in a subgroup increases, and (b) ratio of 'averaged experimental and control subgroup effect variance' to study subject variance increases. If plausible levels of subgroup teaching effect heterogeneity are ignored, the true sizes of tests with nominal 0.05 two-sided type I errors range from 0.055 to 0.47, while when planning studies, estimated sample sizes are only 11.1-95.2 per cent of the true requirements.

Recruitment strategies in a cluster randomized trial--cost implications

The presence of a non-zero intracluster correlation coefficient in cluster randomized trial data has well-known statistical implications for trial design, in particular, inflating the required sample size for given specifications. However, problems in recruitment are common in such trials and there may be different costs of recruitment resulting from different recruitment strategies. Examples of how such differences arise are taken from cluster randomized trials and more intuitive methods of describing clustering are summarized which, in conjunction with such cost issues, may provide a framework that enables triallists to consider explicitly the trade-off between power and cost that is often inherent in such trials.

Statistical methods for the meta-analysis of cluster randomization trials

Cluster randomization trials have become a very attractive research strategy, particularly for the evaluation of health service interventions. The need to conduct meta-analyses of such trials is also becoming more common. However, as with cluster randomization trials in general, such analyses raise special methodologic challenges. In this paper, we discuss and illustrate several statistical approaches that might be applied to a meta-analysis of cluster randomization trials, each of which has a binary endpoint. Statistical methods for constructing inferences for a summary intervention odds ratio include those based on Mantel-Haenszel procedures, the ratio estimator approach, Woolf procedures and generalized estimating equations using robust variance estimation. The advantages and disadvantages of each method are discussed in the context of an example.

The impact of insecticide-treated bednets on malaria and anaemia in pregnancy in Kassena-Nankana district, Ghana: a randomized controlled trial

The impact of insecticide-treated bednet use on malaria and anaemia in pregnancy was assessed, as a supplementary study, in a major WHO/TDR-supported bednet trial in northern Ghana between July 1994 and April 1995. The study area was divided into 96 clusters of compounds, with 48 clusters being randomly allocated to intervention. All pregnant women were included in the study but the focus was on primigravidae and secundigravidae. 1961 pregnant women were recruited into the study--1033 (52.7%) in the treated bednet group and 928 (47.3%) in the no net group. 1806 (92.1%) had blood taken for malaria microscopy and haemoglobin determination in the third trimester. Pregnancy outcomes were reported for 847 women. The characteristics of women in intervention and control groups were comparable. The odds ratios, with 95% confidence interval (CI), for different study endpoints were, for Plasmodium falciparum parasitaemia--0.89 (0.73, 1.08), for anaemia--0.88 (0.70, 1.09), for low birthweight (LBW)--0.87 (0.63, 1.19), indicating no benefit for treated bednet use. Effective net use by parity varied from 42% in primigravidae to 63% in multigravidae, in spite of free nets and insecticide impregnation. The main reasons for not using a net were warm weather and perceived absence of mosquito biting. Chloroquine use in pregnancy was low and comparable in both groups. Implications of findings for malaria control in pregnancy and further research are discussed.