Primary dental care evaluation in Brazil: an item response theory approach.

Abstract

Item response theory (IRT) is a method used to design, analyze, and score tests, questionnaires, and similar instruments measuring abilities, attitudes, or other variables. The aim of this study was to assess the psychometric properties of dental care questions in the "Programa Nacional de Melhoria do Acesso e Qualidade da Atenção Básica-PMAQ-AB," Brazil, using IRT. Dentists in primary health care units in Brazil (n = 12,403) were interviewed face-to-face using a structured questionnaire. The questions were primarily dichotomous, with a no answer/do not know option. The items about dental care from a Brazilian national evaluation survey were analyzed using the IRT model (20 items). Oral health teams (OHTs) received scores that varied from the lowest performance level to the highest performance level. Scores for the Brazilian states and the Federal District were calculated to evaluate the regional distribution of OHT performance. The questions about dental care exhibited higher discrimination power for OHTs with below average performance. In general, the teams, including those with low performance, performed the actions, and procedures included in the questionnaire. Actions such as making prostheses and tracking and monitoring oral cancer cases characterized the high-performing teams. The performance of the teams distributed throughout the Brazilian states indicated that OHTs in the south and southeast performed better than OHTs in the rest of the country. Although the analyzed items are insufficient to determine the performance of OHTs, the items related to prosthesis and oral cancer tend to discriminate high-performing OHTs from other OHTs.