Cost Assessment of Epidemiologic Surveys in Dentistry

Abstract

Objective: The aim was to assess the relationship between the variability of data, sample size (n) and costs involved in epidemiologic surveys of dental caries. Research design and settings: In order to conduct this study, simulations of the variation in costs of hypothetic epidemiologic surveys were made and studied. Thus, all costs with reference to a survey were described and divided into two categories: fixed and variable. Outcome measures: The following margins of sampling errors were analyzed; 5%, 10%, 12% and 15% and the coefficients of variation (CV) of sampled data evaluated were, 50%, 80%, 100% and 120%. Results: The required sample size increased with the reduction in the margin of error. For a CV of 50%, considering an error of 5%, the sample size was 384; for the same CV and error of 10%, n was 96. Thus it was observed that the relationship of sample size between the errors of 5% and 10 % was 4 times higher. Whereas with regard to cost, when an error of 5% was adopted, this was approximately three times higher when compared with the error of 10%. Conclusion: Thus, when planning sample calculation, it is important to consider the Coefficient of Variation and the coherent errors with the variables under study, thus avoiding overestimating the sample and, consequently, increasing the costs involved in the research. It is fundamental to consider the possibility of working with other margins of error, thereby maintaining scientific strictness and establishing adequate costs.