Agreement between two common tooth wear indices: Guidance on indications for use

Abstract

AimTo assess the level of agreement between the simplified Tooth Wear Index (S-TWI) and the Basic Erosive Tooth Wear Examination (BEWE). MethodsData from 477 adult participants in the Adult Dental Health Survey were analysed. They were examined at home using the S-TWI and the BEWE, in that order. Agreement in the highest score at person, segment and sextant levels was measured with Gwek's agreement coefficient. The impact of the threshold used to define tooth wear (any [mild/moderate/severe] vs no wear, moderate/severe vs no/mild wear and severe vs no/mild/moderate wear) on their agreement was also evaluated. ResultsThe prevalence of any, moderate/severe and severe tooth wear was 93.1%, 36.1% and 4.6% according to the S-TWI and it was 98.9%, 70.9% and 21.4% according to the BEWE, respectively. The agreement in the highest score between both indices was 0.745 (95% CI: 0.715, 0.775) at person level, 0.771 (95% CI: 0.746, 0.796) at the anterior segment level, 0.795 (95% CI: 0.766–0.824) for the upper anterior sextant and 0.905 (95% CI: 0.895–0.915) for the lower anterior sextant. The agreement between indices was higher for any wear than for severe wear at person, segment and sextant level. However, it was the weakest for moderate/severe wear regardless of the assessment level. ConclusionThe agreement between indices was high, especially when focusing on specific parts of the mouth (lower anterior teeth) and when using specific thresholds of tooth wear severity (any wear and severe wear). Agreement was constantly low when using the moderate/severe wear threshold. Clinical significanceWhen used as an epidemiological tool, the BEWE may report higher levels of tooth wear than the S-TWI. The BEWE may be a better screening tool whereas the S-TWI may be a better tool to use when determining secondary or tertiary care referral.