A prediction model for periodontal disease: modelling and validation from a National Survey of 4061 Taiwanese adults. (22nd April 2015)

• Record Type: Journal Article
• Authors: Lai, Hongmin; Su, Chiu‐Wen; Yen, Amy Ming‐Fang; Chiu, Sherry Yueh‐Hsia; Fann, Jean Ching‐Yuan; Wu, Wendy Yi‐Ying; Chuang, Shu‐Lin; Liu, Hsing‐Chih; Chen, Hsiu‐Hsi; Chen, Li‐Sheng
• Is Part Of: Journal of clinical periodontology. Volume 42:Number 5(2015:May)
• Page Start: 413
• Page End: 421
• Publication Date: 2015-04-22
• Subjects: Periodontics -- Periodicals; 617.6
• DOI: 10.1111/jcpe.12389
• Languages: English
• ISSNs: 0303-6979

Abstract

<abstract abstract-type="main" id="jcpe12389-abs-0001"> <title>Abstract</title> <sec id="jcpe12389-sec-0001" sec-type="section"> <title>Aims</title> <p>The aim of this study was to predict periodontal disease (PD) with demographical features, oral health behaviour, and clinical correlates based on a national survey of periodontal disease in Taiwan.</p> </sec> <sec id="jcpe12389-sec-0002" sec-type="section"> <title>Materials and Methods</title> <p>A total of 4061 subjects who were enrolled in a cross‐sectional nationwide survey on periodontal conditions of residents aged 18 years or older in Taiwan between 2007 and 2008 were included. The community periodontal index (CPI) was used to measure the periodontal status at the subject and sextant levels. Information on demographical features and other relevant predictive factors for PD was collected using a questionnaire.</p> </sec> <sec id="jcpe12389-sec-0003" sec-type="section"> <title>Results</title> <p>In our study population, 56.2% of subjects had CPI grades ≥3. Periodontitis, as defined by CPI ≥3, was best predicted by a model including age, gender, education, brushing frequency, mobile teeth, gingival bleeding, smoking, and BMI. The area under the curve (AUC) for the final prediction model was 0.712 (0.690–0.734). The AUC was 0.702 (0.665–0.740) according to cross‐validation.</p> </sec> <sec id="jcpe12389-sec-0004" sec-type="section"> <title>Conclusions</title> <p>A prediction model for PD using information obtained from questionnaires was developed. The feasibility of its application to risk stratification of PD should be considered with regard to community‐based screening for asymptomatic PD.</p> </sec> </abstract>