Tradução, adaptação cultural e validação da Patient - Practitioner Orientation Scale (PPOS) para a língua portuguesa do Brasil

Abstract

National curricular guidelines suggest a shift in healthcare model with the use of patient-centered practices and flexible physician behavior that meets patients\' preferences. This change requires appropriate evaluation of medical students\', physicians\' and patients\' attitudes with valid and reliable instruments. The objective of this study was to translate, culturally adapt and validate the Patient-Practitioner Orientation Scale (PPOS) to Portuguese spoken in Brazil. The PPOS is a measure of individual attitudes regarding medical/disease or patient-oriented practices. This instrument was translated according to international guidelines of translation. Two translated versions, a back-translation and author\'s observations were assessed by five reviewers through modified Delphi and the pre-final version of the PPOS was obtained. This pre-final version was then assessed by the author of the scale and pre-tested with 37 participants (12 resident physicians, 13 medical students and 12 patients). Pre-test results were analyzed by three evaluators (doctors) and two researchers to produce a final version of the Brazilian version of the PPOS - the EOMP. This final version was approved by the scale author and by a language coordinator. The EOMP was then applied to 360 participants (120 resident physicians, 120 medical students and 120 patients). We verified data quality (lost data, floor and ceiling effects), test-retest reliability (intraclass correlation reliability), internal consistency reliability (Cronbach\'s alpha coefficient) and construct validity (explanatory factor analysis with principal component analysis, two factors extraction, Oblimin rotation and Kaiser criterion; confirmatory factor analysis with maximum likelihood method and goodness-of-fit indexes: X2/df, RMSEA, SRMR, GFI, AGFI, NNFI and CFI). During independent review process (modified Delphi), only two items (04 and 17) did not reach a minimum of 80% agreement among the reviewers or produce answer stability throughout the different testing sessions. During pre-testing, seven items (38.9%) were modified according to pre-established criteria. The rate of lost data was 0.28%. Floor effect was observed in most of items answered by patients. Ceiling effect was detected in most of items answered by residents, students and patients. Internal consistency reliability (Cronbach s alpha = 0.605) and test-retest reliability (intraclass correlation coefficient = 0.670) were considered to be adequate. In the two-factor principal component analysis (a priori criteria), one item (item 09) did not achieve a loading factor higher than 0.3, one item (item 15) was considered factorially complex and two items (items 02 and 14) were inconsistent with a priori factors (explanatory factor analysis). A confirmatory factor analysis provided an acceptable adjustment for the observed variables (X2/df = 2.33; GFI= 0.91; AGFI= 0.89; CFI= 0.84; NFI= 0.75; NNFI= 0.81; RMSEA= 0.062 (p= 0.016) e SRMR= 0.065). These results show that the EOMP has adequate reliability and acceptable validity. This scale may be useful in evaluating physicians\', medical students\' and patients\' attitudes in Brazil.