This is the first infodemiological study to examine the online search behavior for the keyword “autoimmune encephalitis” using GT as a data collection tool. Similar studies explored the use and role of the internet, specifically web searches in relation to HISB. The use of search volume data as a surrogate of online HISB has been suggested to provide information that is acquired traditionally from large-scale data collection via survey questionnaires and interviews that may be time-consuming and costly. The only caveat is that interpreting online search trend data requires context.
The uptrend in online search interest for AE may be explained by the increased awareness on the condition by the public and the medical community. The first increase was noted a year after Dr. Dalmau and his colleagues described four young females with neuropsychiatric symptoms, all diagnosed with encephalitis and ovarian teratoma. The succeeding increase in November 2007 was the same year when the antibody, anti-NMDA (N-methyl-D-aspartic acid) receptor was identified [1]. Because user characteristics querying the web engine cannot be identified through GT, it is uncertain whether the upward tendency of search interest may be attributed to increased HISB of patients and their relatives or clinicians.
The maximum SVI was recorded in July 2018. A Google search for peak-related news headlines using a custom date range and “news” filter showed the release of a biographical film on AE entitled “Brain on Fire” in an online streaming service on June 22, 2018, which can probably account for this peak in search interest. This is consistent with the results of previous studies that showed the effect of news headlines or mass media events on online search volumes [13, 14]. Notably, “brain on fire” was also included in the returned search queries for the keyword, further suggesting the increased search interest driven by mass media.
Aside from long-term longitudinal trends, infodemiology reveals temporal patterns in health information-seeking behavior within a single year. In this study, no annual seasonal variation in search interest for AE was observed, suggesting the possible lack of seasonality of the disease. This is in contrast to the findings of prior research demonstrating a cyclical trend of interest in relation to seasonal variation of a particular disease, e.g., influenza, multiple sclerosis [13]. Analyzing temporal trends would help determine optimal times for public health education and promotion campaigns. The impact of public health education activities may be gauged by evaluating the change in online search interest in relation to campaign launch. This effect of awareness campaigns to increase online HISB was suggested in infodemiological studies on epilepsy and colorectal cancer [9, 20].
Higher SVIs for AE were recorded from developed countries. There is no available epidemiology study on the geographical distribution of AE to suggest a possible disease-specific explanation for the regional trend. Better internet access and literacy may still be a plausible reason for the greater online search interest in developed countries. Interestingly, the Philippines, a developing country, emerged with a relatively high SVI. A possible explanation would be the relatively high percentage of internet users in the country at 46.88%, based on the most recent statistical report by the World Bank in 2019 [21].
With regard to spatial bias, a Google search returned news articles on AE published online from countries with high search volumes, namely Singapore, Australia, United States, Ireland, United Kingdom, and India. This mirrors the results of previous studies of increased public interest driven by mass media events [13, 14]. However, this influence on search interest based on the geographic origin of the news report is only implied, as further analysis of the temporal relationship between increased search volumes and release of news reports from the same country of origin was not done in this study.
Most terms associated with the search queries for autoimmune encephalitis were related to causes and symptoms. In accordance with a similar publication, our results confirm that most people look for definitions and symptomatology regardless of the chronicity of the condition [13]. We also hypothesize that the heterogeneity of AE presentation is another reason for this need for clarification of the definition and symptoms. The top related search query was “autoimmune encephalitis treatment” with immunoglobulin therapy and rituximab emerging in the list. These are used as first and second-line treatment options for AE, which may explain the frequency by which these are queried.
This study has some limitations inherent to infodemiological studies using web search engine trends. One limitation is intrinsic to the online search analysis tool used, that is, the data provided GT are normalized or relative, and thus are non-absolute search volumes. The online search queries sampled by GT are based only on data collected through Google. Although Google is the most popular search engine worldwide, the online search activity using other web search engines was not included in the analysis. Another limitation is the lack of demographic information regarding the users of the web search engine and their purpose for their search queries. This study may not reflect real-world epidemiologic data due to the lack of precision of the use of internet metrics and traffic, but it could serve as indirect evidence of information demand and health information gaps. While causality testing cannot be done using data from GT, they may still be used for hypothesis generation and descriptive analysis. Despite these limitations, the use of infodemiology as a research method allows a quick and easy way to study the online behavior of millions of internet users worldwide.