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

Pimnara Peerawaranun, Jordi Landier, Francois H Nosten, Thuy-Nhien Nguyen, Tran Tinh Hien, Rupam Tripura, Thomas J Peto, Koukeo Phommasone, Mayfong Mayxay, Nicholas P J Day, Arjen Dondorp, Nick White, Lorenz von Seidlein, Mavuto Mukaka, Pimnara Peerawaranun, Jordi Landier, Francois H Nosten, Thuy-Nhien Nguyen, Tran Tinh Hien, Rupam Tripura, Thomas J Peto, Koukeo Phommasone, Mayfong Mayxay, Nicholas P J Day, Arjen Dondorp, Nick White, Lorenz von Seidlein, Mavuto Mukaka

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://ichgcp.net/clinical-trials-registry/NCT01872702.

Keywords: Bootstrapping; Cluster randomized trial; ICC; Incidence; Malaria; P. falciparum; P. vivax; Prevalence; Sample size.

Conflict of interest statement

The authors declare that they have no competing interests.

Figures

Fig. 1
Fig. 1
Intracluster correlation coefficient (ICC) for prevalence of P. falciparum and P. vivax infection by country and by estimation methods
Fig. 2
Fig. 2
Required number of villages for varying village sizes for the different ICCs (rho) assuming to detect a 95% fall in prevalence of P. falciparum from a 10% initial prevalence (control groups) with 80% power and 0.05 probability of Type I error

References

    1. Donner A, Klar N. Design and analysis of cluster randomization trials in health research. London: Arnold Publishers; 2000.
    1. Hayes RJ, Moulton LH. Cluster randomised trials. Boca Raton: Chapman & Hall/CRC Press; 2009.
    1. Eldridge SM, Ashby D, Kerry S. Sample size for cluster randomized trials: effect of coefficient of variation of cluster size and analysis method. Int J Epidemiol. 2006;35:1292–1300. doi: 10.1093/ije/dyl129.
    1. Hayes RJ, Bennett S. Simple sample size calculation for cluster-randomized trials. Int J Epidemiol. 1999;28:319–326. doi: 10.1093/ije/28.2.319.
    1. Donner A, Birkett N, Buck C. Randomization by cluster. Sample size requirements and analysis. Am J Epidemiol. 1981;114:906–914. doi: 10.1093/oxfordjournals.aje.a113261.
    1. Adams G, Gulliford M, Ukoumunne O, Eldridge S, Chinn S, Campbell M. Patterns of intra-cluster correlation from primary care research to inform study design and analysis. J Clin Epidemiol. 2004;57:785–794. doi: 10.1016/j.jclinepi.2003.12.013.
    1. Campbell M, Thomson S, Ramsay C, MacLennan G, Grimshaw J. Sample size calculator for cluster randomized trials. Comput Biol Med. 2004;34:113–125. doi: 10.1016/S0010-4825(03)00039-8.
    1. Austin P, Stryhn H, Leckie G, Merlo J. Measures of clustering and heterogeneity in multilevel Poisson regression analyses of rates/count data. Stat Med. 2018;37:572–589. doi: 10.1002/sim.7532.
    1. Singh J, Liddy C, Hogg W, Taljaard M. Intracluster correlation coefficients for sample size calculations related to cardiovascular disease prevention and management in primary care practices. BMC Res Notes. 2015;8:89. doi: 10.1186/s13104-015-1042-y.
    1. Browne WJ, Subramanian SV, Jones K, Goldstein H. Variance partitioning in multilevel models that exhibit overdispersion. J R Stat Soc A. 2005;168:599–614. doi: 10.1111/j.1467-985X.2004.00365.x.
    1. Goldstein H, Browne WJ, Rasbach J. Partitioning variation in multilevel models. Underst Stat. 2002;1:223–232. doi: 10.1207/S15328031US0104_02.
    1. von Seidlein L, Peto TJ, Landier J, Nguyen T-N, Tripura R, Phommasone K, et al. The impact of targeted malaria elimination with mass drug administrations on falciparum malaria in Southeast Asia: a cluster randomised trial. PLoS Med. 2019;16:e1002745. doi: 10.1371/journal.pmed.1002745.
    1. Nguyen T-N, von Seidlein L, Nguyen T-V, Truong P-N, Do Hung S, Pham H-T, et al. The persistence and oscillations of submicroscopic Plasmodium falciparum and Plasmodium vivax infections over time in Vietnam: an open cohort study. Lancet Infect Dis. 2018;18:565–572. doi: 10.1016/S1473-3099(18)30046-X.
    1. Pongvongsa T, Phommasone K, Adhikari B, Henriques G, Chotivanich K, Hanboonkunupakarn B, et al. The dynamic of asymptomatic Plasmodium falciparum infections following mass drug administrations with dihydroarteminisin–piperaquine plus a single low dose of primaquine in Savannakhet Province, Laos. Malar J. 2018;17:405. doi: 10.1186/s12936-018-2541-9.
    1. Tripura R, Peto TJ, Chea N, Chan D, Mukaka M, Sirithiranont P, et al. A controlled trial of mass drug administration to interrupt transmission of multidrug-resistant falciparum malaria in Cambodian villages. Clin Infect Dis. 2018;67:817–826. doi: 10.1093/cid/ciy196.
    1. Landier J, Kajeechiwa L, Thwin MM, Parker DM, Chaumeau V, Wiladphaingern J, et al. Safety and effectiveness of mass drug administration to accelerate elimination of artemisinin-resistant falciparum malaria: a pilot trial in four villages of Eastern Myanmar. Wellcome Open Res. 2017;2:81. doi: 10.12688/wellcomeopenres.12240.1.
    1. Eldridge SM, Ukoumunne OC, Carlin JB. The intra-cluster correlation coefficient in cluster randomized trials: a review of definitions. Int Stat Rev. 2009;77:378–394. doi: 10.1111/j.1751-5823.2009.00092.x.
    1. Nakagawa S, Schielzeth H. Repeatability for Gaussian and non-Gaussian data: a practical guide for biologists. Biol Rev Camb Philos Soc. 2010;85:935–956.
    1. Woertman W, de Hoop E, Moerbeek M, Zuidema SU, Gerritsen DL, Teerenstra S. Stepped wedge designs could reduce the required sample size in cluster randomized trials. J Clin Epidemiol. 2013;66:752–758. doi: 10.1016/j.jclinepi.2013.01.009.

Source: PubMed

Sponsors en medewerkers

Medische omstandigheden

Geneesmiddelinterventies

3
Abonneren