Uma metodologia estatística para análise do instrumento de avaliação do Programa Nacional de Melhoria do Acesso e da Qualidade da Atenção Básica (PMAQ-AB)

Abstract

The National Program for the Improvement of Access and Quality of Primary Care (PMAQ-AB) is a program of the Ministry of Health aimed at improving access and quality in Primary Health Care. The third step of the Program is the external evaluation. It is a set of actions that verifies the conditions of access and quality of the Primary Care of the teams participating in the program. This phase presents an extensive questionnaire that deals with several dimensions of Primary Care. This work proposes, based on statistical methodologies, to analyze the evaluation (questionnaires) instrument indicating weaknesses and proposing improvements. The first approach is concerned with the excess of missing data that is handled through an indicator constructed to provide a cutoff point of missing data maximizing the amount of information maintained. The second approach seeks to validate the set of items (questions) that present a single construct, for which the Item Response Theory (IRT) is used to identify the items that present low discrimination in order to, if necessary, be reformulated or removed from the questionnaire. This Technique was used in a group formed by 57 variables from PMAQ-AB and results show 27 variables need to be reviewed, reformulated or even eliminated from the instrument due to their low discrimination.