Intraclass correlation coefficients for cluster randomized trials in care pathways and usual care: hospital treatment for heart failure

Abstract

Background: Cluster randomized trials are increasingly being used in healthcare evaluation to show the effectiveness of a specific intervention. Care pathways (CPs) are becoming a popular tool to improve the quality of health-care services provided to heart failure patients. In order to perform a well-designed cluster randomized trial to demonstrate the effectiveness of Usual care (UC) and CP in heart failure treatment, the intraclass correlation coefficient (ICC) should be available before conducting a trial to estimate the required sample size. This study reports ICCs for both demographical and outcome variables from cluster randomized trials of heart failure patients in UC and care pathways.

Methods: To calculate the degree of within-cluster dependence, the ICC and associated 95% confidence interval were calculated by a method based on analysis of variance. All analyses were performed in R software version 2.15.1.

Results: ICCs for baseline characteristics ranged from 0.025 to 0.058. The median value and interquartile range was 0.043 [0.026-0.052] for ICCs of baseline characteristics. Among baseline characteristics, the highest ICCs were found for admission by referral or admission from home (ICC = 0.058) and the disease severity at admission (ICC = 0.046). Corresponding ICCs for appropriateness of the stay, length of stay and hospitalization cost were 0.069, 0.063, and 0.001 in CP group and 0.203, 0.020, 0.046 for usual care, respectively.

Conclusion: Reported values of ICCs from present care pathway trial and UC results for some common outcomes will be helpful for estimating sample size in future clustered randomized heart failure trials, in particular for the evaluation of care pathways.