Systematic review on the prevalence, frequency and comparative value of adverse events data in social media

E-ISSN： 1365-2125|80|4|878-888

ISSN： 0306-5251

Source： BRITISH JOURNAL OF CLINICAL PHARMACOLOGY (ELECTRONIC), Vol.80, Iss.4, 2015-10, pp. : 878-888

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

AimThe aim of this review was to summarize the prevalence, frequency and comparative value of information on the adverse events of healthcare interventions from user comments and videos in social media.MethodsA systematic review of assessments of the prevalence or type of information on adverse events in social media was undertaken. Sixteen databases and two internet search engines were searched in addition to handsearching, reference checking and contacting experts. The results were sifted independently by two researchers. Data extraction and quality assessment were carried out by one researcher and checked by a second. The quality assessment tool was devised in‐house and a narrative synthesis of the results followed.ResultsFrom 3064 records, 51 studies met the inclusion criteria. The studies assessed over 174 social media sites with discussion forums (71%) being the most popular. The overall prevalence of adverse events reports in social media varied from 0.2% to 8% of posts.Twenty‐nine studies compared the results from searching social media with using other data sources to identify adverse events. There was general agreement that a higher frequency of adverse events was found in social media and that this was particularly true for ‘symptom’ related and ‘mild’ adverse events.Those adverse events that were under‐represented in social media were laboratory‐based and serious adverse events.ConclusionsReports of adverse events are identifiable within social media. However, there is considerable heterogeneity in the frequency and type of events reported, and the reliability or validity of the data has not been thoroughly evaluated.