Rodrigo Lopez1, Vibeke Baelum. 1. Department of Community Oral Health and Pediatric Dentistry, Faculty of Health Sciences, University of Aarhus, Aarhus, Denmark. rlopez@odont.au.dk
Abstract
OBJECTIVES: To evaluate the performance of four clinical classification systems proposed for periodontitis in young subjects when applied to epidemiological data on clinical attachment loss. We assess the extent to which the use of different case definition systems may influence the outcome of descriptive and analytical epidemiological studies. METHODS: The data originate in a screening examination for periodontitis carried out among 9162 high school students. Each of four previously published classification systems was applied to the data. The prevalence of cases according to each system was estimated and the association between case status, as defined by each system, and a set of candidate determinant variables was assessed using multivariable logistic regression analyses. RESULTS: The four classification systems yielded rather different prevalence estimates. For localized periodontitis the estimates varied by a factor of 10, and for generalized periodontitis, these varied by a factor of 30. The results of the logistic regression analyses using the different case-definitions essentially confirmed the results of a population-based analysis. However, the precision of the estimates decreased with decreasing numbers of cases identified by the classification systems. CONCLUSIONS: From an epidemiological point of view there is little justification for the use of the complicated classification systems. An approach based on the simple definition of a case as a person with clinical attachment loss, e.g. >/=3 mm, is preferable.
OBJECTIVES: To evaluate the performance of four clinical classification systems proposed for periodontitis in young subjects when applied to epidemiological data on clinical attachment loss. We assess the extent to which the use of different case definition systems may influence the outcome of descriptive and analytical epidemiological studies. METHODS: The data originate in a screening examination for periodontitis carried out among 9162 high school students. Each of four previously published classification systems was applied to the data. The prevalence of cases according to each system was estimated and the association between case status, as defined by each system, and a set of candidate determinant variables was assessed using multivariable logistic regression analyses. RESULTS: The four classification systems yielded rather different prevalence estimates. For localized periodontitis the estimates varied by a factor of 10, and for generalized periodontitis, these varied by a factor of 30. The results of the logistic regression analyses using the different case-definitions essentially confirmed the results of a population-based analysis. However, the precision of the estimates decreased with decreasing numbers of cases identified by the classification systems. CONCLUSIONS: From an epidemiological point of view there is little justification for the use of the complicated classification systems. An approach based on the simple definition of a case as a person with clinical attachment loss, e.g. >/=3 mm, is preferable.
Authors: Mark I Ryder; Tzy-Jyun Yao; Jonathan S Russell; Anna-Barbara Moscicki; Caroline H Shiboski Journal: J Clin Periodontol Date: 2016-11-22 Impact factor: 8.728