[Assessment of municipal management of oral health in primary care: data collection instrument accuracy].

This validation study seeks to check the accuracy of an evaluation model. In an evaluation, it is necessary to validate the precision and reliability of the data collection instrument. In this study, the Management Assessment of Oral Health in Primary Care in Santa Catarina was used as a benchmark to calculate the indicators. Its model analyzes primary data, collected via an electronic form, and secondary data, available in the Unified Health System (SUS) database. For this study, the form was applied in the cities of Santa Catarina's Coal Region at two different moments to check its reproducibility, followed by a discussion over the answers with the researcher. The results obtained were analyzed and debated in a consensus workshop with specialists in the field, detecting inaccuracies relating to the concept, the source used and the profile of the respondents themselves. The gross agreement rate in the two data collections was 87%, and the inaccuracies amounted to 36% of the answers. Preferential source suggestions, question modifications and guidelines for the correct filling out of the form were some of the proposed changes, improving the original matrix and the data collection instrument.