Intraclass Correlation Coefficients for Planning Cluster Randomized Trials in Community-Dwelling Older Adults

Area of Science:

• Gerontology
• Biostatistics
• Clinical Trials

Background:

• Cluster randomized trials (CRTs) are increasingly used for older adults, randomizing facilities instead of individuals.
• The intraclass correlation coefficient (ICC), measuring within-facility similarity, is crucial for CRTs but often unavailable.
• Understanding ICCs is vital for accurate sample size calculations and interpreting results in geriatric CRTs.

Purpose of the Study:

• To document intraclass correlation coefficients (ICCs) for common measures in community-dwelling older adults.
• To assess the impact of ICCs on the planning and analysis of cluster randomized trials in this population.

Main Methods:

• Secondary analysis of baseline data from the 'On the Move' cluster randomized trial.
• Calculation of ICCs using a linear mixed-effects model for various health and functional measures.

Main Results:

• Self-reported disability and sense of community measures showed the highest ICCs (>0.10).
• Mobility performance measures had ICCs ranging from 0.05 to 0.10.
• Cognitive measures exhibited an ICC of 0.11.

Conclusions:

• The documented ICCs for older adult measures are substantial enough to significantly influence study sample sizes.
• These ICCs can affect the credibility of findings in cluster randomized trials.
• Researchers must consider ICCs during the planning and data analysis phases of geriatric CRTs.