Multifactorial assessment of measurement errors affecting intraoral quantitative sensory testing reliability

Area of Science:

• Oral medicine
• Neuroscience
• Biostatistics

Background:

• Traditional reliability assessments for intraoral quantitative sensory testing (QST) often focus on single error sources.
• Existing methods, like intraclass correlation coefficients (ICCs), may not fully capture measurement error complexity.
• Previous studies typically used limited examiner numbers for inter-examiner reliability.

Purpose of the Study:

• To assess the reliability of intraoral QST by simultaneously considering multiple sources of measurement error.
• To provide a more comprehensive understanding of variability in QST measurements.
• To inform the design of future studies utilizing intraoral QST.

Main Methods:

• Employed a complex design with four examiners and 12 healthy participants over two visits.
• Utilized seven QST procedures: cold detection, warmth detection, cold pain, heat pain, mechanical detection, mechanical pain, and pressure pain.
• Applied mixed linear models for variance component estimation and dependability coefficients for scenario simulation.

Main Results:

• Participant differences (8.8-30.5%), visit-to-visit differences (4.6-52.8%), and residual error (13.3-28.3%) were primary sources of QST variability.
• Increasing visits with a single examiner improved dependability for most QST procedures.
• Reported a wide range of ICCs for both inter- (0.39-0.80) and intra-examiner (0.10-0.62) reliability.

Conclusions:

• Simultaneous assessment of multiple error sources offers a superior method for evaluating sensory testing reliability.
• Experimental settings require large participant numbers for accurate QST-based treatment effect estimation.
• Clinical use of QST can de-emphasize inter-participant variability, focusing on individual changes.