Quality evaluation of a Cochrane Plain language summary using the DISCERN tool: cross-sectional study

Session: 

Oral session: Knowledge translation and communicating evidence (1)

Date: 

Sunday 16 September 2018 - 11:40 to 11:50

Location: 

Sidlaw Auditorium

All authors in correct order:

Logullo P1, Riera R1, Torloni MR1
1 Universidade Federal de São Paulo (UNIFESP) and Cochrane Brazil, Brazil
Presenting author and contact person

Presenting author:

Patricia Logullo

Contact person:

Patricia Logullo
Abstract text
Background: Clear and accessible information on prevention and treatment choices help patients adhere to recommendations from healthcare professionals. The DISCERN instrument is a validated questionnaire designed to assess the quality of patient-targeted health information, and it is applicable for both printed and online texts. We worked on the translation of the first Brazilian Portuguese version of DISCERN and evaluated the psychometric properties of the translated tool using a Cochrane Plain language summary (PLS).

Objective: To report the average scores of the quality assessment of the PLS using DISCERN.

Methods: First we translated DISCERN using the best available guidelines for translation and validation studies (to be reported soon). We calculated Cronbach's alpha for all questions and for the whole questionnaire. Then we presented the Portuguese version to first-year journalism students and asked them to evaluate the quality of a PLS (CD009329, Figure 1) using DISCERN. They did it twice, with a four-week interval between, and we calculated the intraclass correlation coefficient (ICC) between the evaluations.

Results: Cronbach's alpha was above 0.8 for all questions and for the whole translated DISCERN. Average scores varied from 1.57 to 4.33 in average for each question, and 45.9 (± 12.2) for the whole questionnaire (from a possible total score of 80 points) and 48.9 (± 13.4) in the second evaluation. The reproducibility between the first and second evaluations was almost perfect (ICC = 0.845) for total DISCERN score (Table 1).

Conclusions: The reliable and reproducible Portuguese version of DISCERN revealed a low-quality score for a text intended to be understood by laypeople. Cochrane should start using validated tools such as DISCERN to evaluate the quality of PLS.

Attachments: 

Image icon Table 1.png
Image icon Figure 1.png
Image icon Figure 2.png

Relevance to patients and consumers: 

By using a validated text quality assessment tool, such as DISCERN, in English, Spanish, German and now Brazilian Portuguese, Cochrane can evaluate the plain language summaries that are published in the internet, and are available for lay people. DISCERN can be used by patients to evaluate Cochrane PLS and the scores for each question can reveal information gaps or pitfalls that can be corrected before publication. Also, DISCERN can be used by the Knowledge Translation Working Group to evaluate the impact of the new PLS Project.