Clinical Validation Of 8 Parameters For CMD - Screening M.O. Ahlers*, U. Pichlmeier, M. Maghsudi, U. Platzer (Universitäts-Krankenhaus Eppendorf, Klinik Und Poliklinik Für ZMK-Krankheiten, Abteilung Zahnerhaltung Und Parodontologie, Martinistr. 52, 20246 Hamburg, Germany) Prevention-oriented dentistry requires routine dental examinations on a regular basis. Apart from history taking and careful inspection, these examinations should cover screening tests for caries risk and periodontal disease as well as for latent CMD and other oral diseases. As for CMD-screening, the Clinical Dysfunctions Index (Helkimo 1974) served as a scoring tool for years. Yet, a current review questions its validity (Goulet and Glark 1990). Therefore, the aim of this study was to determine, if an abbreviated clinical functional analysis based on merely 8 parameters is able to disriminate between healthy subjects and CMD-patients in routine examination.   The design of this in-vivo diagnostic study involved two groups:   - a group of 102 volunteers (Age ø 22,2 years), who anamnestically did not suffer from dental or orofacial pain within the last 6 months;   - a group of 80 patients with a history of orofacial pain and therefore referred to the CMD-ambulance of the School of Dentistry for further diagnosis and therapy.   In addition to the complete clinical functional analysis as the gold standard, a screening involving 8 parameters adapted from Krogh-Poulsen´s suggestion was performed.   Sensitivity and specificity revealed that none of the parameters alone was able to discriminate healthy and CMD-patients.   Instead, the number of positive parameters provided a highly significant separation between both groups (Mann-Whitney-U-Test, p<0.01). Assuming a CMD-prevalence of 12% (Dworkin et al. 1990) and denoting patients with more than one positive parameter as "CMD-positive“, negative and positive predictive values of 99% and 32 % were reached. If patients with more than two parameters were declared „CMD-positive“, respective values were 95% and 100%.   Exploratory classification tree analyses suggested that the eight parameters evaluated could be reduced to merely two parameters (joint sounds, painful muscle palpation). If a least one of these symptoms were observed, positive and negative predictive values reached 87% and 98%, respectively. Thus, a further reduction of the eight parameters in routine application seems feasible. As these results were obtained based on an exploratory data analysis, further validation should be performed via European multicentre approach.