PubMed2002
BACKGROUND: While partial-mouth recording is often used in epidemiological studies of periodontal disease because of the efficiencies afforded in data collection, there has been no investigation of the extent to which information is lost in studies of young adults using the combined attachment loss (CAL) approach. METHODS: A random sample of 25- and 26-year-olds was periodontally examined at 3 sites per tooth in all 4 quadrants. The analysis obtained full-mouth prevalence estimates for gingival recession (GR), probing depth (PD), and CAL. The half-mouth analyses took 3 forms: 1) estimates from each of the left and right sides were obtained and compared; 2) estimates were obtained separately and compared for quadrants 1 and 3 (upper right and lower left), and quadrants 2 and 4 (upper left and lower right); and 3) estimates were obtained from a diagonal half-mouth count, whereby quadrants 1 and 3 were analyzed for study participants whose identification number was odd, and quadrants 2 and 4 were analyzed for the remainder. The utility of the half-mouth and full-mouth approaches in analytical epidemiology was examined by estimating the strength of the association between periodontitis prevalence and smoking, male gender, and episodic use of dental care. RESULTS: Of the 169 participants examined, 100 (59.2%) were female, 54 (32.0%) were smokers, and 78 (46.2%) were episodic dental visitors. The difference in prevalence estimates obtained from the different methods was considerably greater for GR than for PD and CAL. The unadjusted odds ratio (OR) for the prevalence of 1 or more teeth with > or = mm of CAL among smokers was 2.3 (95% confidence interval [CI] 1.0, 5.3) using the full data set and 2.4 (95% CI 0.9, 6.1) using the diagonal half-mouth approach. Similarly close odds ratios were observed for males and, to a lesser extent, for episodic dental visitors. CONCLUSIONS: Wherever possible, full-mouth data should be collected for descriptive epidemiological studies of periodontal disease, but where resource and time constraints mean that half-mouth examinations must be used, analytical studies of periodontitis should not be unduly affected by the loss of information. However, where the primary focus of the latter is upon gingival recession, the full-mouth design should be used in order to capture all relevant information, and attention should be directed to making economies in other areas of the data collection process.