Accounting for twin births in sample size calculations for randomised trials

BACKGROUND Including twins in randomised trials leads to non-independence or clustering in the data. Clustering has important implications for sample size calculations, yet few trials take this into account. Estimates of the intracluster correlation coefficient (ICC), or the correlation between outcomes of twins, are needed to assist with sample size planning. Our aims were to provide ICC estimates for infant outcomes, describe the information that must be specified in order to account for clustering due to twins in sample size calculations, and develop a simple tool for performing sample size calculations for trials including twins. METHODS ICCs were estimated for infant outcomes collected in four randomised trials that included twins. The information required to account for clustering due to twins in sample size calculations is described. A tool that calculates the sample size based on this information was developed in Microsoft Excel and in R as a Shiny web app. RESULTS ICC estimates ranged between -0.12, indicating a weak negative relationship, and 0.98, indicating a strong positive relationship between outcomes of twins. Example calculations illustrate how the ICC estimates and sample size calculator can be used to determine the target sample size for trials including twins. CONCLUSIONS Clustering among outcomes measured on twins should be taken into account in sample size calculations to obtain the desired power. Our ICC estimates and sample size calculator will be useful for designing future trials that include twins. Publication of additional ICCs is needed to further assist with sample size planning for future trials.

[1]  B. Leroux,et al.  Efficiency of regression estimates for clustered data. , 1996, Biometrics.

[2]  P. Davis,et al.  Docosahexaenoic Acid and Bronchopulmonary Dysplasia in Preterm Infants , 2017, The New England journal of medicine.

[3]  Owen D Williamson,et al.  Determining the sample size in a clinical trial , 2003, The Medical journal of Australia.

[4]  S. Gates,et al.  How should randomised trials including multiple pregnancies be analysed? , 2004, BJOG : an international journal of obstetrics and gynaecology.

[5]  Robert W Platt,et al.  Regression models for clustered binary responses: implications of ignoring the intracluster correlation in an analysis of perinatal mortality in twin gestations. , 2005, Annals of epidemiology.

[6]  B. Thilaganathan,et al.  Elective birth at 37 weeks of gestation versus standard care for women with an uncomplicated twin pregnancy at term: the Twins Timing of Birth Randomised Trial , 2012, BJOG : an international journal of obstetrics and gynaecology.

[7]  Maria Makrides,et al.  Accounting for multiple births in randomised trials: a systematic review , 2014, Archives of Disease in Childhood: Fetal and Neonatal Edition.

[8]  Mercedes Onis,et al.  WHO Child Growth Standards based on length/height, weight and age , 2006, Acta paediatrica (Oslo, Norway : 1992). Supplement.

[9]  C. Roberts,et al.  Australian national birthweight percentiles by gestational age , 1999, The Medical journal of Australia.

[10]  C. Ananth,et al.  Epidemiology of twinning in developed countries. , 2012, Seminars in perinatology.

[11]  John B. Carlin,et al.  Statistics for clinicians: 7: Sample size , 2002 .

[12]  Maria Blettner,et al.  Sample size calculation in clinical trials: part 13 of a series on evaluation of scientific publications. , 2010, Deutsches Arzteblatt international.

[13]  H. Snieder,et al.  Testing the fetal origins hypothesis in twins: the Birmingham twin study , 2001, Diabetologia.

[14]  S. Seaman,et al.  Analysis of Randomised Trials Including Multiple Births When Birth Size Is Informative. , 2015, Paediatric and perinatal epidemiology.

[15]  L. Doyle,et al.  Neurodevelopmental outcomes of preterm infants fed high-dose docosahexaenoic acid: a randomized controlled trial. , 2009, JAMA.

[16]  Keming Yu,et al.  Comparing methods of analysing datasets with small clusters: case studies using four paediatric datasets. , 2009, Paediatric and perinatal epidemiology.

[17]  Katherine J. Lee,et al.  Sample size calculations for randomised trials including both independent and paired data , 2017, Statistics in medicine.

[18]  Ziyad Mahfoud,et al.  What Is an Intracluster Correlation Coefficient? Crucial Concepts for Primary Care Researchers , 2004, The Annals of Family Medicine.

[19]  B. Giraudeau.  Model mis‐specification and overestimation of the intraclass correlation coefficient in cluster randomized trials , 2006, Statistics in medicine.

[20]  C. Crowther,et al.  Elective birth at 37 weeks of gestation versus standard care for women with an uncomplicated twin pregnancy at term: the Twins Timing of Birth Randomised Trial , 2012, BJOG : an international journal of obstetrics and gynaecology.

[21]  David M. Murray,et al.  Methods To Reduce The Impact Of Intraclass Correlation In Group-Randomized Trials , 2003, Evaluation review.

[22]  G. Zou,et al.  A modified poisson regression approach to prospective studies with binary data. , 2004, American journal of epidemiology.

[23]  Lisa N Yelland,et al.  Performance of the modified Poisson regression approach for estimating relative risks from clustered prospective data. , 2011, American journal of epidemiology.

[24]  Allan Donner,et al.  Intracluster correlation coefficients from the 2005 WHO Global Survey on Maternal and Perinatal Health: implications for implementation research. , 2008, Paediatric and perinatal epidemiology.

[25]  G. Fitzmaurice,et al.  A caveat concerning independence estimating equations with multivariate binary data. , 1995, Biometrics.

[26]  B. Giraudeau,et al.  Bmc Medical Research Methodology Open Access Design Effect in Multicenter Studies: Gain or Loss of Power? , 2022 .

[27]  A. Nowacki,et al.  Multiples and parents of multiples prefer same arm randomization of siblings in neonatal trials , 2014, Journal of Perinatology.

[28]  J Carpenter,et al.  Bootstrap confidence intervals: when, which, what? A practical guide for medical statisticians. , 2000, Statistics in medicine.

[29]  O. Sauzet,et al.  Modelling the hierarchical structure in datasets with very small clusters: a simulation study to explore the effect of the proportion of clusters when the outcome is continuous , 2013, Statistics in medicine.

[30]  S. Zeger,et al.  Longitudinal data analysis using generalized linear models , 1986 .

[31]  J. Hanley,et al.  Statistical analysis of correlated data using generalized estimating equations: an orientation. , 2003, American journal of epidemiology.

[32]  J. Hiller,et al.  Effect of bottles, cups, and dummies on breast feeding in preterm infants: a randomised controlled trial , 2004, BMJ : British Medical Journal.

[33]  C. Roberts,et al.  National birthweight percentiles by gestational age for twins born in Australia , 1999, Journal of paediatrics and child health.

[34]  M. Makrides,et al.  Analysis of binary outcomes from randomised trials including multiple births: when should clustering be taken into account? , 2011, Paediatric and perinatal epidemiology.

[35]  M. Walsh,et al.  Accounting for multiple births in neonatal and perinatal trials: systematic review and case study. , 2010, The Journal of pediatrics.