Readability assessment of plain language summaries in Cochrane diagnostic test accuracy reviews

ID: 

210

Session: 

Poster session 2

Date: 

Monday 17 September 2018 - 12:30 to 14:00

All authors in correct order:

Civile VT1, Rocha A1, Nunes AC1, Teixeira K1, Puga ME1, Trevisani VFM1
1 Cochrane Brazil, Brazil
Presenting author and contact person

Presenting author:

Vinicius Civile

Contact person:

Vinicius Civile
Abstract text
Background: The content of text published in unlimited and often free access medical journals should be clearly and succinctly written to facilitate understanding and optimise the main purpose of any informative paper: multiplying knowledge. Cochrane has therefore developed standards that are considered mandatory or highly desirable for the reporting of plain language summaries (PLS). All Cochrane PLS are expected to adhere these methodological standards and use the suggested tools to improve readability and accessibility to non-expert readers. Diagnostic test accuracy reviews (DTAR) have relatively new, but rapidly evolving methods. The Cochrane Screening and Diagnostic Test Methods (SDTM) group has also developed draft guidance for the elaboration of PLS for DTAR.

Objectives: To assess the readability of PLS in Cochrane DTAR and to estimate their level of difficulty compared to Cochrane DTAR abstracts.

Methods: We included all 84 Cochrane DTAR published until 28 February 2018 available on the Cochrane SDTM Group website and one remaining review obtained from the diagnosis topic in the Cochrane Library. We included DTAR that had PLS and excluded protocols and DTAR without PLS. We used the free online tool 'readability formulas' and estimated the degree of difficulty using the Simple Measure of Gobbledygook (SMOG) Index, as recommended by the SDTM group.

Results: Surprisingly, PLS had a mean SMOG index of 12.4 ± 1.7, whereas abstracts had a mean of 12.1 ± 1.9, with no statistically significant difference (P = 0.2797). We observed that 20.9% of the PLS and 25.3% of the abstracts were composed of polysyllabic words. Although PLS had fewer polysyllabic words and were shorter (mean of 447 ± 165, whereas abstracts were 780 ± 189 words, P < 0.0001), PLS sentences were longer (mean of 22 ± 4.2, whereas abstracts were 18 ± 4.7 words/sentence, P < 0.0001).

Conclusions: We found similar readability for abstracts and PLS in Cochrane DTAR. Readers should have at least 12 years of education to understand PLS and abstracts.

Patient or healthcare consumer involvement: Although the audience for Cochrane DTAR may be different from intervention reviews, PLS should help to convey the results of DTAR to a broad audience and they should be easier to be read than abstracts. Implementation of the recommendations in the draft guidance for the elaboration of PLS for DTAR is necessary.

Relevance to patients and consumers: 

Plain language summaries (PLS) help patients and healthcare consumers to access the systematic review information and text readability is still a challenge to be overcome by researchers. In this research, we discuss the importance of a straightforward style that can be comprehended by healthcare consumers, and the use of specific softwares by researchers to help in this task.