A procedure for evaluating the reliability of a gingivitis index

Area of Science:

• Clinical dentistry
• Biostatistics
• Dental public health

Background:

• Assessing the reliability of clinical indices is crucial for valid research.
• The Periodontal Maximal Gingival Index (PMGI) is used to evaluate gingival inflammation.
• Ensuring inter-examiner reliability is vital for consistent data in clinical trials.

Purpose of the Study:

• To present a methodology for assessing the reliability of a scaled index.
• To illustrate this methodology using data from a clinical trial on gingival inflammation.
• To provide procedures for training and calibrating examiners to achieve adequate reliability.

Main Methods:

• Utilized data from a clinical trial assessing gingival inflammation with the PMGI index.
• Employed 5 examiners (1 experienced, 4 newly trained) evaluating subjects initially and at study end.
• Estimated reliability using intraclass correlation coefficient and weighted kappa statistic.

Main Results:

• Reliability of average scores, maximum scores, and percentage of affected sites were estimated.
• Weighted kappa statistics were used to dissect examiner agreement patterns across sites and index components.
• The methodology demonstrated its utility in identifying areas for examiner calibration.

Conclusions:

• The presented methodology effectively assesses the reliability of the PMGI index.
• Statistical procedures can dissect examiner agreement, aiding in training and calibration.
• Implementing these procedures can help achieve adequate reliability levels for clinical studies.