PubMed1978
Sixty-six dental students, randomly assigned to one of three examiners, were scored initially for gingivitis using the Columbia gingival index (GI). After four weekly prophylaxes, gingival scores for all teeth were recorded at day 0. Following prophylaxes, all subjects were randomly assigned one of three mouthwash formulations. Gingival scores were taken on the full mouth after three weeks and again after six weeks. Pearson product moment correlation coefficients (r) were obtained for the gingivitis index from six selected teeth (partial score) with the full mouth score, usually based on 24 to 28 teeth. GI scores obtained from partial recording on the six selected teeth correlated very well with scores obtained from full mouth examinations. A special study was set up to obtain independent GI readings for the six target teeth and for the full mouth. GI scores obtained by independent examinations of the six selected teeth also correlated well with the full mouth scores, although not as well as when the partial scores were derived from the full mouth scores. Partial scores should only be compared with partial scores obtained from evaluating the same teeth and not with full mouth scores, in clinical trials or epidemiologic surveys.