Related Experiment Video
Updated: May 9, 2026
07:42
Quasistatic Mechanical Testing for Computer-Aided Design and Manufacturing Occlusal Veneers Cemented to Milled Dentin Analog Material
Published on: December 20, 2024
Cluster Effects in a National Dental PBRN restorative study
M S Litaker1, V V Gordan, D B Rindal
1Department of Clinical and Community Sciences, School of Dentistry, University of Alabama at Birmingham, USA. mlitaker@uab.edu
Journal of Dental Research
|July 17, 2013
Summary
Intraclass correlation coefficients (ICCs) in dental research significantly impact study precision and power. Understanding these ICCs is crucial for accurate sample size planning in clustered dental studies.
Area of Science:
- Dental Research
- Biostatistics
- Health Services Research
Background:
- Clustered data, where observations within groups (e.g., patients per clinician) are more similar than those between groups, are common in health research.
- This within-group similarity, quantified by the intraclass correlation coefficient (ICC), reduces statistical precision, leading to lower power and wider confidence intervals compared to non-clustered samples.
- Accounting for ICC is essential in the design and analysis of clustered studies to ensure valid and reliable results.
Purpose of the Study:
- To provide intraclass correlation coefficient (ICC) estimates from a large-scale dental practice-based study.
- To serve as a resource for sample size planning in restorative dental research.
- To highlight the impact of clustering on statistical power and precision in oral health studies.
Main Methods:
- Utilized data from 7,826 dental restorations on 4,672 patients treated by 222 clinicians in the National Dental Practice-Based Research Network.
- Calculated ICC estimates for various dental outcomes, including clinician-influenced choices (e.g., rubber dam use) and patient-level factors (e.g., caries lesion occurrence).
Main Results:
- Substantial ICCs were observed in practice-based dental research, significantly affecting precision and statistical power.
- ICCs for clinician-influenced characteristics were relatively large (e.g., 0.36 for rubber dam use).
- ICCs for patient-level outcomes, such as caries, ranged from 0.03 to 0.15, but were still large enough to impact statistical power.
Conclusions:
- Clustering effects, as measured by ICCs, are substantial in dental research and must be considered during study design.
- Accurate sample size and statistical power calculations for oral health studies require incorporating ICC estimates.
- These findings underscore the importance of accounting for clustering in the design and analysis of dental practice-based research.
