Abstract WMP20: Intracluster Correlation Coefficients and Reliability of Randomised Multicentre Stroke Trials within VISTA

Abstract

Background/Purpose: Reliable estimates of intracluster correlation coefficients (ICCs) for specific outcome measures are crucial for sample size calculations of future cluster randomised trials. ICCs indicate the proportion of data variability that is explained by defined levels of clustering. Methods: ICCs were estimated from linear and generalised linear mixed models using maximum likelihood estimation for common measures used in stroke research, including modified Rankin Skale (mRs), National Institutes of Health Stroke Scale (NIHSS) and Barthel Index (BI). Results: Data were available for 11841 patients with ischaemic stroke from 11 randomised trials. After adjusting for age, thrombolysis, and baseline NIHSS, the median ICC for follow-up data, using centre as the level of clustering, ranged from 0.007 to 0.041. The ICCs for follow-up data using trial, continent or year of enrolment as level of clustering were distinctly lower. Less than 1% of the variability of mRS, NIHSS, and BI was explained by any of these three cluster levels. Conclusion: These estimates of relevant ICCs should assist trial planning. For example the sample size for a cluster trial with 150 patients per centre using ordinal analysis of mRS should be inflated by 2.0 due to the ICC of 0.007; whereas the ICC of 0.031 using mRS dichotomised above mRS 0-1, requires inflation by 5.6. The low contribution of trials, year or continent of enrolment to overall variation in outcome offers reassurance that analyses using pooled data from multiple trials in VISTA are unlikely to suffer from bias from these sources. Table: Adjusted ICCs with various levels of clustering for outcome at 90 days.