ABSTRACT
OBJECTIVE: To develop differential criteria on the tongue coating thickness (TCT), and especially propose the standard for judgment on thin and thick coating in tongue diagnosis. METHODS: Sixty oriental medical doctors evaluated the TCT (none, thin, and thick coating) in 50 realistic tongue photographs revealing from tip-to-root. The photographs were obtained with a digital tongue imaging system (DTIS) which measured the percentages of tongue coating on the tongue surface. We calculated the match rate which is the ratio in which the assessor's judgment is consistent with the decision standard on the tongue coating, and then 24 assessors (≥80% match rate) were selected to improve the reliability of the decision. The agreement level among 24 assessors was examined to assess the inter-rater reliability. The correlation between TCT judgments and DTIS-measured values was examined to ascertain the reliability of DTIS measurements. Finally, the assumption probability for the analysis of quantified characteristics of the tongue coating was calculated with a proportional odds model. RESULTS: The inter-rater reliability was assessed as moderate (κ=0.56) among 24 assessors, the level of correlation between TCT judgments and DTIS measurements in 24 assessors was relatively high (0.76, p<0.01). As the analysis of the proportional odds model, 29.06% was a cut-off point to separate no coating and thin coating, 63.51% was a cut-off point to separate thin and thick coating. CONCLUSIONS: The differential criteria for TCT in tongue diagnosis were suggested, and particularly thick coating is defined as one that tongue coating which tongue body is invisible, occupy approximately more than two third areas on the tongue surface.
