Background Package inserts (PIs) should promote the safe and effective use of medicines. The comprehension of PIs is related to socio-demographic features, such as education. Objectives To evaluate the participants' comprehension of a sample of PIs and to build an explanatory model of subjects' understanding of the content of these documents. Setting The data were collected from municipalities, city halls, firefighters, the military, schools and charities from two Portuguese regions. Methods Cross-sectional descriptive survey: 503 participants, homogeneously distributed by education and gender. The self-administered tool comprised questions on socio-demographic data, literacy tasks and comprehension evaluation of 12 purposively selected PIs. A logistic regression analysis was used. Main outcome measures Scores of numeracy tasks and comprehension. Results The average comprehension score for the PIs was 63 % (±32 %), with 48 % (n = 239) of the participants scoring <75 %. The most important predictors in explaining a comprehension score ≥75 % were having >12 years of education and correctly performing a numeracy task [respectively, OR 49.6 (CI 95 %: 22.8-108) and OR 2.48 (CI 95 %: 1.5-4.2)]. Conclusion An explanatory model of subjects' knowledge about the content of the tested PIs was built. Given that a high level of education and literacy were found to be the most relevant predictors for acceptable comprehension rates, PIs should be clearly written to assure that they are understood by all potential users, including the less educated. The evaluated PIs may thus need to be simplified.
Keywords: Comprehension questionnaire; Health literacy; Logistic regression; Package inserts; Portugal; Readability.