Background/Purpose: Online search engines are widely used to seek information about disease and management strategies. The aim of this study was to understand what terms people seeking information about gout use most frequently in their online searches and to explore the psychological and emotional tone of these searches using a linguistic analysis of their search histories. 

Methods: In cooperation with Microsoft Research, a large de-identified dataset of search histories from 200,000 consenting individuals was obtained from ComScore, a web analytics company, covering a two year period (2011 to 2013).  Time-stamped search terms were logged from all major search engines (e.g., Google, Bing).  From the larger dataset, three groups were identified: participants who searched for gout at least once (n = 1,388), participants who had searched for arthritis (arthritis control group, n = 2,289, matched to the gout group on age and sex), and a random set of participants matched on age and sex (general control group, n = 2,150).  Meaning Extraction Helper (Boyd, 2016), a word frequency software, was used to calculate search term frequencies from the search history of each participant.  Group membership was correlated with individual word frequencies.  Search terms were further analyzed using Linguistic Inquiry and Word Count (LIWC; Pennebaker, Boyd, Jordan, & Blackburn, 2015), a text analysis software from which psychological processes can be inferred from the words people use. 

Results: The most frequent unique searches in the gout group included gout-related and food-related terms (including uric, kidney, meats, purine, and atkins).  Those who searched for gout were more likely to search for words related to eating or avoidance.  In contrast, those who searched for arthritis were more likely to search for disease or health-related words. In the LIWC analysis, compared with the general control group, total word count was higher for the gout and arthritis groups, indicating higher information seeking by both groups.  Both the gout and arthritis groups searched more for health (e.g., clinic, flu, pill) and ingest (e.g., dish, eat, pizza) words, and fewer social (e.g. family, talk, they), leisure (e.g., cook, chat, movie), and sexual (e.g. love, sex) words.  Compared with the general control group, the searches of both the gout and arthritis groups were lower in positive emotion tone and higher in sadness words.  There were very few differences between the gout and arthritis groups in the LIWC analysis, with the exception of higher use of health words by the arthritis group and higher use of insight words (e.g., know, learn, and means) by the gout group.

Conclusion: People searching about gout or arthritis have high levels of information seeking.  The perception of gout as a condition managed by dietary strategies aligns with online information-seeking about the disease and its management.  In contrast, people searching about arthritis are more focused on searching about medical strategies. Linguistic analysis reflects greater disability in social and leisure activities and lower positive emotion for those searching for medical conditions such as gout or arthritis.