Dental fluorosis prevalence and diagnostic test using Deans index based on 6 teeth, and on 28 teeth

Abstract To contrast the sensitivity, specificity, and positive negative predictive values between dental fluorosis prevalence scored on 28 (DF28) and on six permanent teeth (DF6), we undertook a cross-sectional study on 1,538 adolescents (12 and 15 years old) residing in Hidalgo State, Mexico, a naturally fluoridated (>0.7 ppm) area at an elevated altitude (>2,500 m above sea level). Dental fluorosis was scored using Dean_s modified index. Using the scores obtained for all teeth present (DF28) as a gold standard, we calculated the sensitivity, specificity, positive negative predictive values, and receiver operating characteristic and concordance index pertaining to the scores based on six teeth (upper incisors and canines). DF28 fluorosis prevalence was 81.7%; based on DF6, it was 58.7% (23% difference). Among 12 year olds, the difference between DF28 and DF6 was 20.1% (84.5 vs. 64.4%); among 15 year olds, it was 25.4% (79.4 vs. 54%). Among girls, it was 23.2% (81.1 vs. 57.9%) and among boys, 22.8% (82.2 vs. 59.4%). The fluorosis community indices were 1.75 (DF28) and 1.11 (DF6). All positive predictive values reached 100% while negative predictive values were below 45%. Concordance between DF28 and DF6 was 53.9%, and kappa coefficient was 0.40. Partial scoring of fluorosis based on esthetically important permanent teeth underestimated prevalence, compared to full-mouth scoring. The decision to use an abridged Dean_s index protocol must take into account the number of teeth examined, and which specific teeth are examined, to appraise the benefit of reduced data collection effort against possible information loss.