[Sample size and costs estimate in epidemiological survey of dental caries]

This study aimed to analyze how the prevalence and the distribution of dental caries influence the sample size in epidemiological surveys, and how much are the costs. Secondary data of oral health surveys in 12-year-old schoolchildren from Bauru in 1976, 1984, 1990, 1994, and 2001, and from Piracicaba in 2001 and 2005 were studied. Sample sizes were estimated taking into account the mean DMFT and standard deviation of each survey, establishing sampling errors of 1%, 2%, 5%, and 10%. Costs were estimated considering permanent material, consumption material and human resources. The sample size in both towns needed to be increased, ranging from 119 in 1976 to 1,118 in 2001 in Bauru, and from 954 in 2001 to 1,252 in 2005 in Piracicaba, when a sampling error of 10% was considered. The cost of dental caries surveys was verified considering different sampling errors. This cost depends on how acceptable is the margin of difference between the true mean and the one found in the survey. In conclusion, the reduction in the prevalence of dental caries has determined the need for increase in sample size and in costs for conducting the surveys.