Validation of a questionnaire to assess the quality of health information in Argentinian newspapers

The daily press is an important source of health information and may influence health care utilization. However, medical reports published in newspapers from developed countries have shown a poor quality. The reliability of the questionnaire Index of Scientific Quality was evaluated by using it to measure the scientific quality of health information published by Argentinian newspapers between 2000 and 2002. It assessed the readability of the texts in grade levels and explored the relationship between quality and other factors. The Spanish adaptation of the instrument consisted in translation, back-traslation and pilot study. The reliability was assessed by applying the instrument to newspaper articles with more than 300 words that discussed therapy, diagnosis, prevention, lifestyle effects, and hazardous exposure. Two physicians independently graded 129 articles. Inter-observer and intra-observer concordance was quantified for each item with the intra-class correlation coefficient (CI95%). To measure scientific quality, a randomized sample of 210 articles was assessed. Each received a mark that ranged from 0 to 100. Readability was determined by the FRY graph method. The relationship between quality and other variables was explored with multiple linear regression analysis. The inter-rater concordance varied between 0.48 (0.34-0.61) and 0.67 (0.56-0.75). Intra-rater concordance varied from 0.51 (0.37-0.63) to 0.95 (0.93-0.96). The internal consistency (Cronbach's alpha) was 0.88. The quality-of-health-information was rated at 25 points (16.7, 33.3) [median (intercuartil range)]. The reading level was assessed to be at the 10.4 grade (10.2-10.6) [mean (CI 95%)]. Quality of the health information was greatly deficient; however, no specific factors were related with quality. Readability was probably a key barrier for access to the health information.