Comparing completely and stratified randomized designs in cluster randomized trials when the stratifying factor is cluster size: a simulation study

Sample size considerations for stratified cluster randomization design with binary outcomes and varying cluster size

Stratified cluster randomization trials (CRTs) have been frequently employed in clinical and healthcare research. Comparing with simple randomized CRTs, stratified CRTs reduce the imbalance of baseline prognostic factors among different intervention groups. Due to the popularity, there has been a growing interest in methodological development on sample size estimation and power analysis for stratified CRTs; however, existing work mostly assumes equal cluster size within each stratum and uses multilevel models. Clusters are often naturally formed with random sizes in CRTs. With varying cluster size, commonly used ad hoc approaches ignore the variability in cluster size, which may underestimate (overestimate) the required number of clusters for each group per stratum and lead to underpowered (overpowered) clinical trials. We propose closed-form sample size formulas for estimating the required total number of subjects and for estimating the number of clusters for each group per stratum, based on Cochran-Mantel-Haenszel statistic for stratified cluster randomization design with binary outcomes, accounting for both clustering and varying cluster size. We investigate the impact of various design parameters on the relative change in the required number of clusters for each group per stratum due to varying cluster size. Simulation studies are conducted to evaluate the finite-sample performance of the proposed sample size method. A real application example of a pragmatic stratified CRT of a triad of chronic kidney disease, diabetes, and hypertension is presented for illustration.

Efficacy of a Community-Based Physical Activity Program KM2H2 for Stroke and Heart Attack Prevention among Senior Hypertensive Patients: A Cluster Randomized Controlled Phase-II Trial

Objective

To evaluate the efficacy of the program Keep Moving toward Healthy Heart and Healthy Brain (KM2H²) in encouraging physical activities for the prevention of heart attack and stroke among hypertensive patients enrolled in the Community-Based Hypertension Control Program (CBHCP).

Design

Cluster randomized controlled trial with three waves of longitudinal assessments at baseline, 3 and 6 months post intervention.

Setting

Community-based and patient-centered self-care for behavioral intervention in urban settings of China.

Participants

A total of 450 participants diagnosed with hypertension from 12 community health centers in Wuhan, China were recruited, and were randomly assigned by center to receive either KM2H² plus standard CBHCP care (6 centers and 232 patients) or the standard care only (6 centers and 218 patients).

Intervention

KM2H² is a behavioral intervention guided by the Transtheoretical Model, the Model of Personalized Medicine and Social Capital Theory. It consists of six intervention sessions and two booster sessions engineered in a progressive manner. The purpose is to motivate and maintain physical activities for the prevention of heart attack and stroke.

Outcome Measures

Heart attack and stroke (clinically diagnosed, primary outcome), blood pressure (measured, secondary outcome), and physical activity (self-report, tertiary outcome) were assessed at the individual level during the baseline, 3- and 6-month post-intervention.

Results

Relative to the standard care, receiving KM2H² was associated with significant reductions in the incidence of heart attack (3.60% vs. 7.03%, p < .05) and stroke (5.11% vs. 9.90%, p<0.05), and moderate reduction in blood pressure (-3.72mmHg in DBP and -2.92 mmHg in DBP) at 6-month post-intervention; and significant increases in physical activity at 3- (d = 0.53, 95% CI: 0.21, 0.85) and 6-month (d = 0.45, 95% CI: 0.04, 0.85) post-intervention, respectively.

Conclusion

The program KM2H² is efficacious to reduce the risk of heart attack and stroke among senior patients who are on anti-hypertensive medication. Findings of this study provide solid data supporting a formal phase-III trial to establish the effectiveness of KM2H² for use in community settings for prevention.

Trial Registration

ISRCTN Register ISRCTN12608966

Evidence-based care of older people with suspected cognitive impairment in general practice: protocol for the IRIS cluster randomised trial

BACKGROUND

Dementia is a common and complex condition. Evidence-based guidelines for the management of people with dementia in general practice exist; however, detection, diagnosis and disclosure of dementia have been identified as potential evidence-practice gaps. Interventions to implement guidelines into practice have had varying success. The use of theory in designing implementation interventions has been limited, but is advocated because of its potential to yield more effective interventions and aid understanding of factors modifying the magnitude of intervention effects across trials. This protocol describes methods of a randomised trial that tests a theory-informed implementation intervention that, if effective, may provide benefits for patients with dementia and their carers.

AIMS

This trial aims to estimate the effectiveness of a theory-informed intervention to increase GPs' (in Victoria, Australia) adherence to a clinical guideline for the detection, diagnosis, and management of dementia in general practice, compared with providing GPs with a printed copy of the guideline. Primary objectives include testing if the intervention is effective in increasing the percentage of patients with suspected cognitive impairment who receive care consistent with two key guideline recommendations: receipt of a i) formal cognitive assessment, and ii) depression assessment using a validated scale (primary outcomes for the trial).

METHODS

The design is a parallel cluster randomised trial, with clusters being general practices. We aim to recruit 60 practices per group. Practices will be randomised to the intervention and control groups using restricted randomisation. Patients meeting the inclusion criteria, and GPs' detection and diagnosis behaviours directed toward these patients, will be identified and measured via an electronic search of the medical records nine months after the start of the intervention. Practitioners in the control group will receive a printed copy of the guideline. In addition to receipt of the printed guideline, practitioners in the intervention group will be invited to participate in an interactive, opinion leader-led, educational face-to-face workshop. The theory-informed intervention aims to address identified barriers to and enablers of implementation of recommendations. Researchers responsible for identifying the cohort of patients with suspected cognitive impairment, and their detection and diagnosis outcomes, will be blind to group allocation.

TRIAL REGISTRATION

Australian New Zealand Clinical Trials Registry: ACTRN12611001032943 (date registered 28 September, 2011).

A pragmatic cluster randomized trial evaluating the impact of a community pharmacy intervention on statin adherence: rationale and design of the Community Pharmacy Assisting in Total Cardiovascular Health (CPATCH) study

BACKGROUND

Traditional randomized controlled trials are considered the gold standard for evaluating the efficacy of a treatment. However, in adherence research, limitations to this study design exist, especially when evaluating real-world applicability of an intervention. Although adherence interventions by community pharmacists have been tested, problems with internal and external validity have limited the usefulness of these studies, and further well-designed and well-conducted research is needed. We aimed to determine the real-world effectiveness of a community pharmacy adherence intervention using a robust study design. This novel design integrates cluster randomization and an outcome evaluation of medication adherence using a population-based administrative data source in the province of Saskatchewan, Canada.

METHODS/DESIGN

Community pharmacies from across the province of Saskatchewan, Canada were randomized to deliver an adherence intervention to their patients or usual care. Intervention pharmacies were trained to employ a practical adherence strategy targeted at new users of statin medications. While randomization and implementation of the intervention occurred at the community pharmacy level, the outcome analysis will occur at the level of the individual subjects. The primary outcome is the mean statin adherence among all eligible new users of statin medications. Secondary outcomes include the proportion of new statin users who exhibit adherence>or=80%, and persistence with statin use.

DISCUSSION

This novel study design was developed to combine the rigor of a randomized trial with a pragmatic approach to implementing and capturing the results in a real-world fashion. We believe this approach can serve as an example for future study designs evaluating practice-based adherence interventions.

TRIAL REGISTRATION

ClinicalTrials.gov no. NCT00971412.

A comparison of the statistical power of different methods for the analysis of repeated cross-sectional cluster randomization trials with binary outcomes

Repeated cross-sectional cluster randomization trials are cluster randomization trials in which the response variable is measured on a sample of subjects from each cluster at baseline and on a different sample of subjects from each cluster at follow-up. One can estimate the effect of the intervention on the follow-up response alone, on the follow-up responses after adjusting for baseline responses, or on the change in the follow-up response from the baseline response. We used Monte Carlo simulations to determine the relative statistical power of different methods of analysis. We examined methods of analysis based on generalized estimating equations (GEE) and a random effects model to account for within-cluster homogeneity. We also examined cluster-level analyses that treated the cluster as the unit of analysis. We found that the use of random effects models to estimate the effect of the intervention on the change in the follow-up response from the baseline response had lower statistical power compared to the other competing methods across a wide range of scenarios. The other methods tended to have similar statistical power in many settings. However, in some scenarios, those analyses that adjusted for the baseline response tended to have marginally greater power than did methods that did not account for the baseline response.

Sample size considerations when groups are the appropriate unit of analysis

This article discusses issues to be considered by nurse researchers when groups should be used as the unit of randomization. Advantages and disadvantages are presented, with statistical calculations needed to determine the effective sample size. Examples of these concepts are presented using data from the Black Cosmetologists Promoting Health Program. Different hypothetical scenarios and their impact on sample size are also presented. Given the complexity of calculating the sample size when using groups as the unit of randomization, it is advantageous for researchers to work closely with statisticians when designing and implementing studies that anticipate the use of groups as the unit of randomization.

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.