scholarly journals On the estimation of intracluster correlation for time-to-event outcomes in cluster randomized trials

2016 ◽  
Vol 35 (30) ◽  
pp. 5551-5560 ◽  
Author(s):  
Sumeet Kalia ◽  
Neil Klar ◽  
Allan Donner
Keyword(s):  
Randomized Trials ◽  
Cluster Randomized Trials ◽  
Intracluster Correlation ◽  
Time To Event ◽  
Cluster Randomized

scholarly journals A Review of Assumed and Reported Intracluster Correlations in Cluster Randomized Trials

2019 ◽  
Author(s):  
Xiaoran Han ◽  
Jiaye Lin ◽  
Jinjing Xu ◽  
Maggie Wang ◽  
Benny Zee ◽  
...  
Keyword(s):  
Sample Size ◽  
Randomized Trials ◽  
Intraclass Correlation ◽  
Cluster Effect ◽  
Size Estimation ◽  
Cluster Randomized Trials ◽  
Intracluster Correlation ◽  
Eligibility Criteria ◽  
Sample Size Planning ◽  
Cluster Randomized

Abstract Background Cluster randomized trials (CRTs) are widely adopted in health and primary care research. However, the cluster effect needs to be taken into account appropriately in the design and analysis of CRTs. The objectives of this study were (i) to review the reporting of intracluster correlations in CRTs; and (ii) to evaluate whether the assumed intracluster correlation measures in sample size planning are consistent with those obtained in the analysis. Methods The Aggregate Analysis of ClinicalTrials.gov database was searched to identify CRTs registered between January 1, 2004 and March 27, 2016. The selected CRTs with accessible publications were screened according to eligibility criteria. Results Of the 281 CRTs identified, the percentage of studies accounting for cluster effect increased annually. A total of 183 studies accounted for clustering in sample size estimation, among them 43% of CRTs adopted the intraclass correlation coefficient (ICC) but the exact estimated value of ICC was provided in only 26% of the included studies. In different intervention types, there were no statistically significant differences between the assumed and reported values of ICC (all p-values >0.05). Conclusion Although the difference between the values of ICC assumed in sample size planning and that reported in the analysis was not statistically significant, deficiencies in CRTs are still common, such as low rates of considering cluster effect in sample size and reporting intracluster correlation estimates. We also suggest that researchers ought to be familiar with the properties of statistical approaches to improve the analysis of CRTs. Thus, more recommendations and guidelines such as the CONSORT statement for CRTs should be suggested to researchers.

Sample size in cluster-randomized trials with time to event as the primary endpoint

2012 ◽  
Vol 32 (5) ◽  
pp. 739-751 ◽  
Author(s):  
Antje Jahn-Eimermacher ◽  
Katharina Ingel ◽  
Astrid Schneider
Keyword(s):  
Sample Size ◽  
Primary Endpoint ◽  
Randomized Trials ◽  
Cluster Randomized Trials ◽  
Time To Event ◽  
Cluster Randomized

scholarly journals Intracluster correlation coefficients in the Greater Mekong Subregion for sample size calculations of cluster randomized malaria trials

2019 ◽  
Vol 18 (1) ◽  
Author(s):  
Pimnara Peerawaranun ◽  
Jordi Landier ◽  
Francois H. Nosten ◽  
Thuy-Nhien Nguyen ◽  
Tran Tinh Hien ◽  
...  
Keyword(s):  
Randomized Trial ◽  
Sample Size ◽  
Randomized Trials ◽  
Cluster Randomized Trial ◽  
Cluster Randomized Trials ◽  
Intracluster Correlation ◽  
Variable Approach ◽  
Sample Size Calculations ◽  
Cluster Randomized ◽  
Greater Mekong Subregion

Abstract Background Sample size calculations for cluster randomized trials are a recognized methodological challenge for malaria research in pre-elimination settings. Positively correlated responses from the participants in the same cluster are a key feature in the estimated sample size required for a cluster randomized trial. The degree of correlation is measured by the intracluster correlation coefficient (ICC) where a higher coefficient suggests a closer correlation hence less heterogeneity within clusters but more heterogeneity between clusters. Methods Data on uPCR-detected Plasmodium falciparum and Plasmodium vivax infections from a recent cluster randomized trial which aimed at interrupting malaria transmission through mass drug administrations were used to calculate the ICCs for prevalence and incidence of Plasmodium infections. The trial was conducted in four countries in the Greater Mekong Subregion, Laos, Myanmar, Vietnam and Cambodia. Exact and simulation approaches were used to estimate ICC values for both the prevalence and the incidence of parasitaemia. In addition, the latent variable approach to estimate ICCs for the prevalence was utilized. Results The ICCs for prevalence ranged between 0.001 and 0.082 for all countries. The ICC from the combined 16 villages in the Greater Mekong Subregion were 0.26 and 0.21 for P. falciparum and P. vivax respectively. The ICCs for incidence of parasitaemia ranged between 0.002 and 0.075 for Myanmar, Cambodia and Vietnam. There were very high ICCs for incidence in the range of 0.701 to 0.806 in Laos during follow-up. Conclusion ICC estimates can help researchers when designing malaria cluster randomized trials. A high variability in ICCs and hence sample size requirements between study sites was observed. Realistic sample size estimates for cluster randomized malaria trials in the Greater Mekong Subregion have to assume high between cluster heterogeneity and ICCs. This work focused on uPCR-detected infections; there remains a need to develop more ICC references for trials designed around prevalence and incidence of clinical outcomes. Adequately powered trials are critical to estimate the benefit of interventions to malaria in a reliable and reproducible fashion. Trial registration: ClinicalTrials.govNCT01872702. Registered 7 June 2013. Retrospectively registered. https://clinicaltrials.gov/ct2/show/NCT01872702

Implementation Research on Shared Decision Making in Primary Care: Inventory of Intracluster Correlation Coefficients

2019 ◽  
Vol 39 (6) ◽  
pp. 661-672
Author(s):  
Ali Ben Charif ◽  
Jordie Croteau ◽  
Rhéda Adekpedjou ◽  
Hervé Tchala Vignon Zomahoun ◽  
Evehouenou Lionel Adisso ◽  
...  
Keyword(s):  
Primary Care ◽  
Decision Making ◽  
Shared Decision Making ◽  
Randomized Trials ◽  
Correlation Coefficients ◽  
Shared Decision ◽  
Cluster Randomized Trials ◽  
Intracluster Correlation ◽  
Convenience Sample ◽  
Cluster Randomized

Background. Cluster randomized trials are important sources of information on evidence-based practices in primary care. However, there are few sources of intracluster correlation coefficients (ICCs) for designing such trials. We inventoried ICC estimates for shared decision-making (SDM) measures in primary care. Methods. Data sources were studies led by the Canada Research Chair in Shared Decision Making and Knowledge Transition. Eligible studies were conducted in primary care, included at least 2 hierarchical levels, included SDM measures for individual units nested under any type of cluster (area, clinic, or provider), and were approved by an ethics committee. We classified measures into decision antecedents, decision processes, and decision outcomes. We used Bayesian random-effect models to estimate mode ICCs and the 95% highest probability density interval (HPDI). We summarized estimates by calculating median and interquartile range (IQR). Results. Six of 14 studies were included. There were 97 ICC estimates for 17 measures. ICC estimates ranged from 0 to 0.5 (median, 0.03; IRQ, 0–0.07). They were higher for process measures (median, 0.03; IQR, 0–0.07) than for antecedent measures (0.02; 0–0.07) or outcome measures (0.02; 0–0.06), for which, respectively, “decisional conflict” (mode, 0.48; 95% HPDI, 0.39–0.57), “reluctance to disclose uncertainty to patients” (0.5; 0.11–0.89), and “quality of the decision” (0.45; 0.14–0.84) had the highest ICCs. ICCs for provider-level clustering (median, 0.06; IQR, 0–0.13) were higher than for other levels. Limitations. This convenience sample of studies may not reflect all potential ICC ranges for primary care SDM measures. Conclusions. Our inventory of ICC estimates for SDM measures in primary care will improve the ease and accuracy of power calculations in cluster randomized trials and inspire its further expansion in SDM contexts.

Sign in / Sign up

Export Citation Format

Share Document