DOI: https://doi.org/10.21203/rs.3.rs-1398923/v1
Background
Patients and their caregivers, including clinicians and educators, use web-based search engines to access healthcare-related information from the internet. Online search behavior analysis has been used to obtain insights on health information demand.
Objectives
We aimed to describe the online search behavior for autoimmune encephalitis worldwide over time through the analysis of search volumes made on Google.
Methods
In this infodemiological study, we retrieved search volume indices for the keyword “autoimmune encephalitis (disease)” based on worldwide search data from January 01, 2004 to October 31, 2021, using Google Trends. We performed a descriptive analysis of search volume patterns, including related topics and queries.
Results
There was a progressive increase in search volume numbers over time for the keyword “autoimmune encephalitis (disease)” with no annual seasonal variation. Peak search volume was recorded in July 2018. The greatest search volume was recorded in Singapore, followed by Australia, the United States of America, the Philippines, and New Zealand. The most searched topics were related to autoimmune encephalitis definition, causes, symptoms, diagnosis, and treatment. All related topics and queries increased in volume by more than 5000-fold over time.
Conclusions
This study showed an uptrend in the online search interest on autoimmune encephalitis over time, which may reflect the increased awareness on the condition by the public and the medical community. Information on online health information-seeking behavior may be obtained from Google Trends data despite its limitations.
Since the discovery of neural tissue-specific autoantibodies by Dalmau and colleagues in 2007, autoimmune encephalitis (AE) has been increasingly recognized as a neurologic disease entity [1]. It presents with a subacute progression of neuropsychiatric symptoms, a clinical picture similar to infectious encephalitis. Advances in the diagnosis and treatment of AE resulted in an increased awareness of the condition among clinicians and patients alike [2]. A population-based study showed an increase in the incidence rate of AE from 0.4 to 1.2 per 100,000 person-years when comparing data from 1995–2005 versus 2006–2015. Previously thought of as rare, it is now considered as common as the infectious type, with the discovery of the autoantibody and increased diagnostic ability as major contributors to the uptrend [3]. While it is emphasized that early diagnosis and appropriate treatment lead to improved outcomes, some patients are left with the chronic neuropsychiatric sequelae of the disease, whereas others experience relapses based on long-term follow-up studies [4]. Given the debilitating nature of the condition, often with delayed diagnosis and treatment, patients and their caregivers may turn to other information sources such as the internet to help them in decision-making.
The internet has become a significant source of healthcare-related information that is easily accessible and widely available to the public through the use of web-based databases and search engines [5]. Patients with chronic neurological disease resort to web-based materials and social media platforms to look up information about the symptoms, diagnostic work-up and available treatments, to find support groups, and to improve their coping strategies [6]. Clinicians and other medical practitioners have also been found to use the internet as a source to help them in patient management and education. Likewise, educators use the internet to improve medical education.
In 2004, Google Trends (GT) was released, providing an online search data mining tool that made real-time and archived web-based search data accessible to the public. It has been increasingly used as a methodology tool over the past years in what is now called infodemiology, or the “science of distribution and determinants of information in an electronic medium, specifically the Internet, or in a population, with the ultimate aim to inform public health and public policy” [7, 8].
Infodemiology was first used in predicting outbreaks of influenza and other infectious diseases and has since been applied to various medical fields [7]. Several infodemiological studies on neurologic disorders and on various fields have been published, including multiple sclerosis, epilepsy, status epilepticus, stroke, meningitis, polio, movement disorders, Alzheimer disease, teleneurology and telerehabilitation [9–18]. Online search trends for neurologic disorders were found to be related to a number of factors such as real-world epidemiologic data, seasonality of the disease, news about celebrities and other prominent figures or mass-media events, patients’ emotions and perceptions about their disease, and public awareness campaigns [5]. Results of infodemiologic research may provide clinicians and health educators insights on health information-seeking behavior (HISB) as an indirect method of user-based needs assessment. Accordingly, such information could serve as a basis for health information exchange program design and implementation to ultimately improve care. To date, there is no published infodemiological study on autoimmune encephalitis. Therefore, the aim of this study was to describe the online search behavior for autoimmune encephalitis worldwide over time through the evaluation of trends of search volumes made on Google.
The search strategy employed in this paper is summarized in Table 1 based on the methodology recommendations from a systematic review on the use of GT (Alphabet Inc, Mountain View CA, USA) in healthcare research [8]. We entered the keyword “autoimmune encephalitis” in the GT main page (available at: http://www.google.com/trends, accessed on November 6, 2021) to extract data on internet user search activities on AE. Using the GT built-in autocomplete function that suggests the most cost commonly used keyword, we selected “autoimmune encephalitis” as “disease” to encompass all search queries that fall within the category, including the terms that share the same concept in any language. Worldwide search data from January 01, 2004 to October 31, 2021 were retrieved and downloaded as comma-separated value (.csv) datasets after applying the “health” category search parameter and the “web search” filter. Results were returned as search volume indices (SVI), which represent the search volume number for a particular search term relative to the overall search volume over time expressed on a scale of 0 to 100.
| Section/Topic | Checklist Item |
|---|---|
| Search Variables | |
| Access Date | November 6, 2021 |
| Google Data Source | Google Trends |
| Time Period | January 1, 2004 to October 31, 2021 |
| Query Category | “Health” query category |
| Region | Worldwide |
| Countries with Low Search Volume | Included |
| Search Input | |
| Full Search Input | “Autoimmune encephalitis” as “disease” was entered into the search bar. |
| Combination | Only one search term was used; hence no combination of terms were applied in the search input. |
| Quotation Marks | Keywords were queried with quotation marks. |
| Rationale for Search Strategy | |
| For Search Input | We chose this keyword to capture all search queries related to AE regardless of language used. |
| For Settings Chosen | We chose January 1, 2004 as the start date to capture baseline interest prior to the discovery of the autoantibodies attributed to the condition (2007), “worldwide” to explore regional variation in search interest, and the “health” category to assess information demand in the context of health. |
We also retrieved and downloaded the .csv files of the “Top related topics,” “Rising related topics,” “Top related queries,” and “Rising related queries.” Top searches were defined as “top searches that are most frequently searched with the term entered in the same search session, within the chosen category, country, or region,” while Rising searches were “terms that were searched for with the keyword entered which had the most significant growth in volume in the requested time period” [19].
Search volume analysis revealed a progressive increase in search volume numbers over time for the keyword “autoimmune encephalitis” with no annual seasonal variation (see Fig. 1). The first increase in online search interest since the start date was in June 2006 (17/100), followed by another increase in November 2007 (8/100). A steadily increasing trend was then seen with the peak search volume observed in July 2018 (100/100). The minimum search volume index value was seen predominantly in 2004 to 2007 and finally in 2012 with a subsequent positive shift in background search volume.
Based on regional interest, the greatest search volume for the keyword “autoimmune encephalitis” (disease) was recorded in Singapore (100/100), followed by Australia (84/100), United States (83/100), Philippines (79/100), New Zealand (77/100), Ireland (70/100), United Kingdom (69/100), United Arab Emirates (53/100), India (52/100), and Canada (46/100) (see Table 2).
| Country | SVI |
|---|---|
| Singapore | 100 |
| Australia | 84 |
| United States | 83 |
| Philippines | 79 |
| New Zealand | 77 |
| Ireland | 70 |
| United Kingdom | 69 |
| United Arab Emirates | 53 |
| India | 52 |
| Canada | 46 |
| South Africa | 31 |
| Malaysia | 31 |
| Germany | 8 |
| Turkey | 4 |
| Indonesia | 3 |
The search queries returned for the keyword “autoimmune encephalitis” were related to AE definition, causes, symptoms, diagnosis, and treatment. In top related queries, “autoimmune encephalitis treatment” was the most related. Autoimmune encephalitis-related topics were terms corresponding to AE signs and symptoms (seizure and encephalopathy), diagnostics (cerebrospinal fluid, receptor, and Mayo Clinic), and treatment (immunoglobulin therapy and rituximab). Of note, the results returned for the keyword included only one specific type of antibody-related to AE, N-methyl-D-aspartic acid or NMDA receptor. All related topics and queries had a more than 5000-fold increase in search volumes over time (see Table 3).
| Related Topics | Related Queries | ||
|---|---|---|---|
| Top (SVI) | Rising (% increase) | Top (SVI) | Rising (% increase) |
| Autoimmunity (100) | Autoimmunity (> 5000%) | autoimmune encephalitis treatment (100) | autoimmune encephalitis treatment (> 5000%) |
| Autoimmune encephalitis (94) | Autoimmune encephalitis (> 5000%) | autoimmune disease (83) | autoimmune disease (> 5000%) |
| Encephalitis (26) | Encephalitis (> 5000%) | autoimmune encephalitis symptoms (77) | autoimmune encephalitis symptoms (> 5000%) |
| Autoimmune disease (11) | Autoimmune disease (> 5000%) | autoimmune encephalitis panel (76) | autoimmune encephalitis panel (> 5000%) |
| Anti-NMDA receptor encephalitis (6) | Anti-NMDA receptor encephalitis (> 5000%) | what is autoimmune encephalitis (55) | what is autoimmune encephalitis (> 5000%) |
| Antibody (5) | Antibody (> 5000%) | nmda receptor autoimmune encephalitis (54) | nmda receptor autoimmune encephalitis (> 5000%) |
| Brain (4) | Brain (> 5000%) | autoimmune encephalitis causes (46) | autoimmune encephalitis causes (> 5000%) |
| NMDA receptor (3) | NMDA receptor (> 5000%) | autoimmune encephalitis icd 10 (44) | autoimmune encephalitis icd 10 (> 5000%) |
| Seizure (2) | Seizure (> 5000%) | autoimmune encephalitis mri (37) | autoimmune encephalitis mri (> 5000%) |
| Immunoglobulin therapy (2) | Immunoglobulin therapy (> 5000%) | brain on fire (26) | brain on fire (> 5000%) |
| Encephalopathy (2) | Encephalopathy (> 5000%) | anti nmda receptor autoimmune encephalitis (24) | anti nmda receptor autoimmune encephalitis (> 5000%) |
| Rituximab (2) | Rituximab (> 5000%) | autoimmune encephalitis in children (24) | autoimmune encephalitis in children (> 5000%) |
| N-Methyl-D-aspartic acid (2) | N-Methyl-D-aspartic acid (> 5000%) | symptoms of autoimmune encephalitis (13) | symptoms of autoimmune encephalitis (> 5000%) |
| Cerebrospinal fluid (2) | Cerebrospinal fluid (> 5000%) | seronegative autoimmune encephalitis (9) | seronegative autoimmune encephalitis (> 5000%) |
| Immune system (2) | Immune system (> 5000%) | autoimmune encephalitis radiology (9) | autoimmune encephalitis radiology (> 5000%) |
| Hashimoto's thyroiditis (2) | Hashimoto's thyroiditis (> 5000%) | anti-nmda-receptor autoimmune encephalitis (7) | anti-nmda-receptor autoimmune encephalitis (> 5000%) |
| Receptor (1) | Receptor (> 5000%) | brain on fire disease (4) | brain on fire disease (> 5000%) |
| Mayo Clinic (1) | Mayo Clinic (> 5000%) | ||
| Limbic encephalitis (1) | Limbic encephalitis (> 5000%) | ||
| Limbic system (1) | Limbic system (> 5000%) |
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.
To our knowledge, this was the first descriptive analysis of online search behavior related to autoimmune encephalitis. This study showed an increasing online interest on autoimmune encephalitis over time, with most search queries related to symptoms, diagnosis, and treatment. Information on online health information-seeking behavior from GT data may provide insights on the unmet needs related to health information and may serve to inform health information exchange policies and awareness campaigns.
Statement of authorship
KTR: Conceptualization, data curation, formal analysis, interpretation of data, writing-original draft, writing-review, and editing. RDGJ: Conceptualization, data curation, formal analysis, interpretation of data, writing-original draft, writing-review, and editing. KMCM: Conceptualization, data curation, formal analysis, interpretation of data, writing-original draft, writing-review, and editing. AIE: Conceptualization, data curation, formal analysis, interpretation of data, writing-original draft, writing-review, and editing.
Sources of support
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interests
None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
GT Google Trends
SVI Search volume indices
AE Autoimmune encephalitis
HISB Health information-seeking behavior
CSV Comma separated values