Sphere spectrophotometer versus human evaluation of tooth shade

The purpose of this study was to compare an objective to a subjective method of tooth shade evaluation using the SP78 sphere spectrophotometer and human visual analysis. Twenty extracted fully developed human maxillary teeth had their L* values (lightness) read by the SP78 on day 1 and then again on day 14. Using a blind method, five human evaluators then attempted to match shade tabs from a six-tab experimental Vita shade guide to the same teeth on days 1 and 14. The SP78 L* values read for each tooth were evaluated for a match between those obtained on day 1 and those read on day 14. The results of human evaluation of tooth shade were compared among the evaluators and then between themselves over time. Finally, the ability of human evaluators to match tooth shade was compared with the results obtained with the SP78. The SP78 reproduced L* readings within the standard error of the machine (< or = 1.0) in 16 of 20 (80%) teeth. In contrast, interevaluator agreement expressed as majority agreement (3, 4, or 5 of 5 evaluators agreeing) was only 10 of 20 (50%) teeth on day 1 and 13 of 20 (65%) teeth on day 14. Intraevaluator agreement over the experimental period ranged from 20 to 60%. The results of this study confirm that human evaluation of tooth shade is unreliable and that the SP78 sphere spectrophotometer can provide a more predictable and accurate method of evaluating tooth shade in vitro. This finding opens up a new avenue of investigation for testing the effectiveness of materials and techniques in bleaching discolored teeth.