Background: English has become the international language in medical literature. In keeping up to date physicians are expected to read and understand English language. However in non English speaking countries this assumption might convey more difficulties than usually accepted. On this assumption, the Iberoamerican Cochrane Centre has devoted efforts to translating Cochrane Reviews into Spanish.
Objectives: To determine whether reading a Cochrane Review abstract in English or Spanish affects the ability of physicians to understand medical information and to measure the time needed to achieve this understanding in each language.
Methods: After recording baseline characteristics of participants, particularly on self-assessed English language skills, physicians starting their Residency Programmes at the Pontificia Universidad Catolica of Chile were randomly assigned to read a Cochrane Review Abstract on a general medical topic in English or in Spanish. They all answered a questionnaire in Spanish with 5 open questions to assess their understanding, while recording time to complete the questionnaire. We encouraged them to answer according to what they read and not based on previous knowledge. Questionnaires were blindly reviewed based on a 0 to 19 points scale. A two tailed Mann-Whitney U was used to compare the groups.
Results: 97 physicians agree to participate and all answered the questionnaire. Baseline characteristics were similar for the 2 groups.
Mean score was 11,2 points (SD=3,4; range 1 to 18). Mean time to read the abstract and complete the questionnaire was 12,2 minutes (SD=2, range = 7,7 to 17,7). There was a statistically significant difference among the groups for both outcomes. The group reading in Spanish had a mean score of 11.9 (SD=2.8, range=5 to 18) compared to 10.5 (SD=3.8, range 1 to 17) for the group in English (p=0.04).
The time to complete the questionnaire was 11.8 min (SD=2, range = 7.7 to 17.2) for the group in Spanish and 12.6 min (SD=2, range = 8.8 to 17.7min) for the English group (p=0.047). 16.7% of the group reading in Spanish scored 9 or less compared to 34.7% of the group reading in English (p=0.042).
Conclusions: Language barrier should be taken into account when teaching evidence based health care in non-English speaking countries.