[Scientific validity of epidemiological knowledge based on data from the Brazilian Oral Health Survey (SB Brazil 2003)].

Indicators and analyses that used the database from SB Brazil 2003 (the most recent nationwide oral health survey) have been criticized as unreliable due to sampling problems. The current study countered that this critique was based solely on statistical concepts, unsupported by empirical evidence. The critique's essentially epistemic approach leads to peremptory reductionism that denies other forms of knowledge and fails to recognize the multidisciplinary nature of epidemiology. The current study retrieves information on the implementation of the oral health survey and its impact on knowledge output in the field. The article draws an analogy between science and art, demonstrating the multifaceted images obtained by both. Thus, recognition of validity requires a full grasp of the field and appropriate use of value criteria. The current article concludes that use of the SB Brazil 2003 database is a reliable and relevant application of epidemiology to oral health.