The search for studies for inclusion in a systematic review should be reported in detail. This ensures transparency, which is important for assessing the strengths and weaknesses of a systematic review. It also enables the searches to be reproduced when conducting an update review. Web searching using search engines and websites is sometimes carried out as a supplementary search method and should be reported alongside other search methods.
This study reviews the reporting of web searching in a sample of Cochrane Reviews published between August 2016 and January 2017. Recommendations for improving practice are made where appropriate based on guidance in the Cochrane Handbook, the MECIR standards, other prominent systematic review reporting guidance and suggestions developed by the author.
Cochrane Reviews published between August 2016 and January 2017 were identified via the Cochrane Database of Systematic Reviews by searching using the asterisk symbol (*) in the ‘Search All Text’ field and using the ‘Online Publication Date’ feature to date limit the results. All Cochrane Reviews identified were downloaded. The search methods section and appendices of all reviews were manually inspected for detail about web searching and the Control-F search feature was used to search for ‘web’, ‘internet’ and ‘online’. Details reported about web searching were recorded including the 1) name of the resource; 2) URL; 3) search date; 4) search strategy; and 5) number of results.
In total, 423 Cochrane Reviews were identified from the period August 2016 to January 2017. Of these, 61 reviews reported web searching using one or more search engine or website. The name of a resource was the most frequently reported item for search engines (96%) and websites (64%). The search terms and total number of results were the least frequently reported details (less than 25% for each detail).
In the majority of reviews the reporting of web searching was found to lack essential detail for ensuring transparency and reproducibility, such as the search terms. Recommendations are made with detailed examples for improving the reporting of web searching.
Patient or healthcare consumer involvement:
Patients can benefit from transparent reporting of review methods by being able to better scrutinize what was done.