Assessment of the validity and consequences of different methods of expressing the severity of dental fluorosis in a subject.

The aim was to assess the validity and consequences of different methods of expressing severity of dental fluorosis in a subject. The analyses were based on Ugandan children (n = 481), aged 10-14 years, with life-long consumption of drinking water with either 0.5 or 2.5 mg fluoride per liter. Fluorosis was assessed using the Thylstrup and Fejerskov (TF) index. All children (n = 219) with 28 teeth and fluorosis on at least 1 tooth pair were selected to test methods of expressing fluorosis, e.g. the TF score on the most severely affected tooth, a maxillary central incisor, the median and the mean scores. A test group (n = 40), the 10 most and the 10 least severely affected children in each district, was used to evaluate the methods and a reference group (n = 179) to confirm or refute the findings in the test group. To evaluate consequences of the different methods of expressing severity of fluorosis in a subject, children from the low (n = 130) and the high fluoride (n = 132) districts not included in the test or reference group formed the community comparison groups. Comparison between the median (gold standard) and mean scores showed a significant deviation in the reference group only. Most of the partial recording methods, such as the score for the most severely affected tooth, were significantly and systematically higher than the median. While for all recording methods the median score was zero in both communities, the distribution of the subjects according to severity differed significantly between the communities.