This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on http://publichealth.jmir.org, as well as this copyright and license information must be included.
Young adults often browse the internet for self-triage and diagnosis. More sophisticated digital platforms such as symptom checkers have recently become pervasive; however, little is known about their use.
The aim of this study was to understand young adults’ (18-34 years old) perspectives on the use of the Google search engine versus a symptom checker, as well as to identify the barriers and enablers for using a symptom checker for self-triage and self-diagnosis.
A qualitative descriptive case study research design was used. Semistructured interviews were conducted with 24 young adults enrolled in a university in Ontario, Canada. All participants were given a clinical vignette and were asked to use a symptom checker (WebMD Symptom Checker or Babylon Health) while thinking out loud, and were asked questions regarding their experience. Interviews were audio-recorded, transcribed, and imported into the NVivo software program. Inductive thematic analysis was conducted independently by two researchers.
Using the Google search engine was perceived to be faster and more customizable (ie, ability to enter symptoms freely in the search engine) than a symptom checker; however, a symptom checker was perceived to be useful for a more personalized assessment. After having used a symptom checker, most of the participants believed that the platform needed improvement in the areas of accuracy, security and privacy, and medical jargon used. Given these limitations, most participants believed that symptom checkers could be more useful for self-triage than for self-diagnosis. Interestingly, more than half of the participants were not aware of symptom checkers prior to this study and most believed that this lack of awareness about the existence of symptom checkers hindered their use.
Awareness related to the existence of symptom checkers and their integration into the health care system are required to maximize benefits related to these platforms. Addressing the barriers identified in this study is likely to increase the acceptance and use of symptom checkers by young adults.
Seeking online health information through search engines is common [
In assessing symptom checker benefits, it is important to understand user perspectives on the platform after actual use rather than simply assessing their accuracy in fictitious situations [
A common theme that emerged from most studies examining symptom checkers is the importance of gathering user perspectives on the use of the platform to enable high acceptance (and use) as well as to prevent lost investments [
This qualitative analysis represents a subset of findings that emerged from a larger mixed-methods study that seeks to understand the factors associated with the behavioral intention of using symptom checkers for self-triage and self-diagnosis. A qualitative descriptive case study research design was used and is differentiated by other research study designs by its focus on a bounded system or case [
To allow for a broad range of perspectives to be gathered, university students between the ages of 18 and 34 years across faculties in all levels of education and year of study were eligible to participate. Following ethics approval from the Research Ethics Board at the University of Waterloo (41366), university students were notified of the study through emails from the administrative assistant of their faculty; as such, the number of students who received and opened the email is unknown. Interested individuals were asked to contact the principal investigator (SA) to schedule an interview. Participants were recruited between November 2019 and May 2020. A total of 24 participants were included in the study based on a first-come, first-served basis and time of data saturation. There were no dropouts in this study. All participants were provided with an information letter prior to the interview outlining the study objectives. Informed consent was obtained from all participants. One-on-one interviews took place on the university campus or virtually through a digital university-approved platform, Whereby. All participants were provided with a Can $10 (US $7.80) coffee shop gift card as a token of appreciation for taking the time to participate in the study.
The main sources of data were a preinterview questionnaire (
To ensure that all participants were familiar with symptom checkers, they were provided with a clinical vignette and, based on a draw, were asked to use one of two web-based symptom checkers: WebMD [
After having read the clinical vignette and accessed the symptom checker, participants were guided by the first author (SA) to conduct the think-aloud exercise, which involved the participants thinking out loud while they performed a task without synthesizing or interpreting their thoughts [
Data analysis was conducted independently by two authors (SA and SM) using the thematic analysis steps outlined by Castleberry and Nolen [
The third step of reassembling consisted of grouping the codes into main themes; in NVivo, this consists of creating nodes (themes) and child nodes (codes under those themes)—the hierarchy can contain many levels depending on the level of detail required. In this work, a hierarchy was used to represent how themes are subordinate or superordinate to each other [
Most of the participants had a high score on the four health domains measured (see
Participant information (N=24).
Characteristics | Value | |
|
||
|
Female | 14 (58) |
|
Male | 9 (38) |
|
Nonbinary | 1 (4) |
|
||
|
White | 9 (38) |
|
Asian | 6 (25) |
|
Chinese | 3 (13) |
|
Arab | 2 (8) |
|
Indian | 2 (8) |
|
Black | 2 (8) |
|
||
|
High school | 2 (8) |
|
Undergraduate degree | 14 (58) |
|
Master’s degree | 8 (33) |
|
||
|
Engineering | 8 (33) |
|
Sciences | 6 (25) |
|
Applied health sciences | 3 (13) |
|
Environment | 3 (13) |
|
Arts | 3 (13) |
|
Mathematics | 1 (4) |
|
||
|
Excellent | 2 (8) |
|
Very good | 13 (54) |
|
Good | 5 (21) |
|
Fair | 4 (17) |
|
Poor | 0 (0) |
|
||
|
Feeling understood by health care providersa | 2.92 (1.5-4.0) |
|
Actively managing my healtha | 3.05 (1.8-4.0) |
|
Ability to actively engage with health care providersb | 3.64 (2.6-4.6) |
|
Ability to find good health informationb | 3.81 (1.8-5.0) |
|
||
|
WebMD | 11 (46) |
|
Babylon Health | 13 (54) |
aMaximum possible average is 4.
bMaximum possible average is 5.
Data related to the use of a search engine or symptom checker were grouped into positive or negative themes. Positive themes suggest a desirable attribute, function, or experience related to a platform, whereas negative themes encompass themes that suggest the opposite. An overview of these themes is provided in
Provides information without claiming a diagnosis
More customizable
Allows entry of all symptoms in the search engine
More personalized
More interactive due to chatbot feature
Good for those who do not know how to use Google
Straightforward design
Easy to use
Real-time output
Makes the correlation between symptoms and potential conditions
More intuitive
More reliable
More specific
More structured
Absence of chatbot feature
Symptom checkers
Accuracy is questionable
Limits the number of symptoms that can be inputted
Not widely known
Thought process of the platform is unclear
User more vulnerable when using this platform
Text input is insufficient
Suboptimal reliability
Participants perceived the Google search engine and symptom checkers to be useful for mild symptoms; however, some perceived that using the Google search engine was faster than having to answer questions in a symptom checker: “If you’re Googling something quick then it’s easy, quick, and straightforward, you don’t have to take 10 minutes to answer all these questions…” [P2].
Positive themes related to the use of the internet mainly pertained to the perspective that a Google search engine allows users to input as many symptoms as needed, enabling a more comprehensive search of potential conditions that may be relevant to their health context. Some users also mentioned that they prefer that the platform does not claim that this is the condition they may have.
On the other hand, I think it may be easier to get accurate results on symptoms through a Google search because I can type multiple symptoms and see how they fit, I may get more garbage results but I can use my judgment to decide what is true and not true. Whereas the symptom checker has only one piece of information which is fever. The symptom checker did not give me the opportunity to put in more from what I can recall.
Nonetheless, some participants mentioned that the absence of a chatbot feature in the Google search engine limits the platform’s ability to ask follow-up questions based on symptoms inputted. As such, some users who many not be able to identify all symptoms experienced may omit certain symptoms or may not elaborate on symptoms, which hinders the quality and comprehensiveness of results.
…if I were to Google my symptom, I would just put in a fever and rashes that could be a million things. But with a symptom checker, I would put in fever and it asked me for a specific temperature and other specific questions which I would not know to search on my own.
Symptom checkers were perceived by some participants to be a good option for individuals who are less proficient in using the internet for information retrieval. Some had a positive attitude toward the symptom checkers because the platform asked questions regarding age and gender, giving the impression that it is more personalized and in turn, in their perception, more accurate.
…surfing through the internet and coming through a particular diagnosis takes a lot of time although it might give you more information about other diseases that have similar symptoms, but this is not what I am looking for, I am looking for what I am suffering from. So, for which, I think a personalized software is helpful.
Some participants believed that the symptom checker “had more structure,” “provided a greater level of detail,” “was more interactive,” and “was more reliable” than using the Google search engine.
So I think having that more structured approach to inputting symptoms and figuring out what is likely wrong with you would be a lot nicer for the user and the user would have more faith in the result rather than just going on Google that brings up a whole bunch of results and the user thinking that they could have anything.
Although having a more structured approach to symptom input was favored, some participants were unable to enter all symptoms in the platform, which led them to question the accuracy and reliability of the platform; this also hindered trust toward the platform.
I feel like I don’t like the symptom checker as much because it limits the number of symptoms. I did not have the chance to mention the thing with the red bumps; it just asked me a lot of questions about the one “symptom that was bothering me the most.”
There was also a sense that participants would feel more vulnerable using a symptom checker due to the more personalized nature of the questions asked. Interestingly, some participants believed that their judgment and thought process to identify potential diagnoses was superior than using a symptom checker due to lack of knowledge about how symptom checkers work.
It feels more vulnerable and personal to put my symptoms into a list or generator of some kind. It feels like I am just looking at a series of articles I feel there's more of a distance…. If I am typing in a symptom checker and it comes back at me with answers, I don’t know how it came to that conclusion and I don’t know what the process was to decide that “yes, this is what you have,” whereas if I am the one doing the analysis through a bunch of articles that I deem legitimate—whether or not they truly are legitimate—at least I know what the thought process was and I feel like I can trust that.
Despite various shortcomings that were mentioned related to the use of symptom checkers, some participants believed that an important issue is the lack of awareness about the existence of the platform: “But the issue is that we don’t know about symptom checkers so making them widely available would be super helpful.” [P13]
In addition, there was a consensus that consulting a primary care provider or nurse was superior than searching the internet or using a symptom checker to assess the severity of symptoms; this was especially the case when certain symptoms required a physical examination and text input was insufficient. Reliability of the Google search engine and symptom checker was also questionable and was perceived negatively by some participants.
I think seeing a provider face to face is better than both options. I feel that you can’t accurately portray all your symptoms and general health by text input. You need someone looking at you and take measurements and touch injured areas, I think that’s far superior.
I think Google is a very wide platform so it’s very hard to analyze the reliability or the source. In this case, it depends on the reliability of the symptom checker as well.
Factors that would hinder the use of a symptom checker were identified as barriers, whereas factors that would enable an individual to use a symptom checker were identified as enablers. Participants enumerated many enablers and barriers for using symptom checkers, which were mainly related to the (1) individual, (2) disease, (3) health care system, or (4) symptom checker. An overview of all identified barriers and enablers is provided in
Internet access
Low health literacy
Trust in the platform
Younger age
Lack of time
Convenience
Lack of trust in doctors
Curiosity
Embarrassing topic
Increase empowerment
Aversion to medical professionals
Having pre-existing conditions
Unable to discuss the topic with a health provider
Uncertain about care required
Worried about health of oneself
Lack of internet access
Low health literacy
Lack of trust in the platform
Low technology literacy
Older age
Social influence
Not wanting to know
Previous bad experience
Mild symptoms
A “broad category of illness”
Symptoms can be easily described
Severe condition
Need for a physical examination
Approved by doctors
Lack of access to health services
Cost of health services
Public education
Increased awareness
Long wait times for health services
Reputable organizations recommend it
Authoritarianism in health care
Increased advertisement
Easy interface
Data privacy
Free of charge
Short to complete
Precision
Use of artificial intelligence
Gamification
Good source of information
Integrated with an electronic health record
Useful in identifying potential conditions
Information about the creators of the platform
Interactive platform
Reliability
Lack of awareness
Poor design
Asking identifiable questions
Cost of the platform
Time to complete
Lack of inclusivity measures
Lack of language options
Lack of credibility
Lack of human interaction
Disclaimer
Inability to obtain elaboration on a question
Liability
Concerns about using data for profits
Internet access, health literacy, trust toward the platform, technology literacy, and age were factors mentioned to be either enablers or barriers for using a symptom checker for self-triage and self-diagnosis. Younger age, internet access, high technology literacy, and trust toward the platform were perceived to enable the use of symptom checkers. Low health literacy was perceived to be both an enabler and barrier for using a symptom checker. Although some participants believed that individuals with low health literacy are more likely to use a symptom checker because they may be less critical, others perceived low health literacy to be a barrier due to inability to understand and input symptoms into the platform.
…maybe if they just did not know a ton about health in general maybe they would be less critical than me.
Also, sometimes it’s hard to articulate to have the proper term of how you feel. For example, in the fever or the lymph node, you don’t know of things like that unless you have specific knowledge about it. So, it is hard for someone who does not have medical terminology to input what they have in there.
Given that a disease does not define the individual, disease-related factors were considered separately. Having a “broad category of illness,” an embarrassing issue, or an issue perceived to be mild were perceived to be enablers, whereas experiencing severe symptoms and needing a physical examination were mentioned to be barriers. Participants seemed to be more willing to use a symptom checker if they were experiencing nonspecific symptoms (eg, fatigue) due to the perceived notion that a symptom checker would allow the user to narrow down on a health condition. They are also more willing to use symptom checkers for issues perceived to be “embarrassing” such as conditions related to mental health.
If something was serious, people would not want to use it, they would want to go to a doctor. Not just physically but also emotionally, I could see them go to the doctor right away.
Lack of access to and cost of health services were perceived to enable the use of a symptom checker. Having the platform approved by reputable organizations and approved by doctors were also mentioned to be important factors for enabling individuals to use the platform. In contrast, some believed that some primary care providers may not be accepting of the technology, thus limiting its use.
First of all, they have to somehow not only advertise but maybe if the website is promoted by the health care organization that is reliable for people then I can make sure that the platform is trusted by an authentic organization so for sure I would use it.
An important factor related to the symptom checker was advertisement. More than half of participants were not aware of symptom checkers, thus limiting their use: “But the issue is that we don’t know about symptom checkers so making them widely available would be super helpful.” [P13]
Developing an easy interface, guaranteeing data privacy, offering the platform free of charge, and ensuring that the platform’s questionnaire is short to complete were all mentioned as potential enablers for using a symptom checker.
It sounds very interesting and it is very easy to use. Definitely I will use it again, I had a good user experience.
And if it was short—I think if there were options “hey, do you want to take the shorter version and it might not be as accurate or do you want to take the longer one that will take more time but will be more accurate.” I think people want something quick but quick won’t be as accurate.
The data suggest that the main barriers for using symptom checkers are the lack of transparency on how the data collected are used; some participants mentioned that they would not have an issue with the data being used by governmental institutions to improve health services but did not want their data to be used to generate profits. Although most of the participants understood the medical terms that were used by the digital platform, some believed that the average person may not understand some of the questions asked. Providing a brief description of medical terms would allow users to interact with the platform in a more informed manner.
I would not want my data to be used to anything that would harm me. I don’t know what it could be used for but if it is being used to find out the prevalence of a certain disease or whatever that is helpful for the health care system, I am fine with that but anything that would encourage the business part of it or pharmaceutical side of it or anything that is business related or goes back to making money, I would not like it.
Participants also stressed the importance for the digital platform to elaborate on why certain questions were being asked. In contrast to seeing a health professional, users are unable to interject and ask the platform questions for further elaboration. Moreover, most platforms use a disclaimer that they do not provide medical advice, which undermines the platform’s credibility.
If they know not to take it seriously, they won’t feel encouraged to do the test at all. If the disclaimer says this is not really a diagnosis, then what am I doing? I should just go to the doctor.
The platforms that were used during this study were in English; however, some mentioned the importance of having these platforms available in various languages to ensure that they are accessible to those who are less proficient in English. Lack of inclusivity measures does not allow persons with disabilities to use the platform and was also mentioned as a barrier to use: “…or various disabilities being able to use the screen or use computers or any type of access issues would be a problem.” [P6]
Approximately half of the participants were not aware about the existence of symptom checkers until their participation in this study. Most of the participants preferred consulting a health professional to address their health needs rather than researching using the internet’s search engine or using a symptom checker. Nonetheless, symptom checkers were generally preferred over the internet’s search engine due to their personalized approach; however, some perceived that the latter is faster to use than having to answer questions. There was also an acknowledgment that the results provided by symptom checkers can only be as good as the data that informed them.
In sum, it was perceived that individuals who are younger, have low health literacy, and high technology literacy were more likely to use a symptom checker. Lack of time, convenience, having symptoms that were perceived to be minor or “embarrassing” were other factors that would result in the use of a symptom checker. Enablers that were related to the health system were lack of access to care, having the technology approved by credible associations, and having symptom checkers integrated as part of the public health system. Nonetheless, participants mentioned many improvements that would have to be made to the symptom checker to enable its use, including improving accuracy, ensuring that the platform is freely accessible, and ensuring privacy and anonymity. Although the participants appreciated the personalized approach of the platform, they would not want to use a symptom checker that asks too many questions or that takes too long to complete. Barriers for using symptom checkers included lack of access to the internet, medical jargon, and lack of trust. Despite the barriers and shortcomings of the platform, participants believed that symptom checkers would be useful if they were tested and validated. Some believed that these platforms must have a positive influence on their health due to the perception that these platforms are designed by medical doctors.
Although this study has some strengths, various limitations warrant mention. First, given that all participants had previously used the Google search engine for seeking health information, we did not ask participants to use the Google search engine during the interview, meaning that they had to rely on their previous Google search experiences to answer questions. Second, we asked the participants if they had previously used a symptom checker, but we did not ask about the frequency of use, which limited our ability to assess whether responses differed based on this potentially important factor. Third, we did not distinguish responses based on the digital platform used, as the main focus of this work was to understand perspectives on the use of symptom checkers in general; however, these perspectives may have differed if participants used another symptom checker than those used in this study (ie, WebMD or Babylon Health). In line with this, we did not examine whether participants chose the correct diagnosis based on the clinical vignette as the focus of the study was on the process of obtaining the list of diagnoses and getting participants familiar with the platform. Last but not least, the sample was comprised of highly educated individuals who were perceived to have a good health status; as such, this may limit the transferability of findings to other populations. Future studies should explore perspectives of other user groups.
Findings from this study are in line with the literature, which suggests that using the Google search engine for health information has many limitations, including the vast amount of information available and lack of quality requirements for publishing content [
This study highlights various factors associated with the use of symptom checkers that could be used as a starting point for future investigations studying the acceptance of such technology in other population groups. Lack of time and convenience were important enabling factors for using symptom checkers; these factors also explain the use of the internet’s search engine for health information [
Importantly, despite the participants reporting positive aspects of the question-answer format used by the symptom checker, most would have favored a more interactive platform that provides more information to the user regarding why certain questions are being asked. Participants also mentioned the importance of being able to ask questions to the platform—this is something that could be easily done during a conversation with a medical provider or a Google search; however, given the more rigid nature of most symptom checkers, this feature is not yet readily available. This is important for symptom checker developers to consider as patient-centered communication has been shown to be important for patient outcomes [
Lack of internet access is also a critical element that hinders access to any web-based information platform. Although lack of internet access is more prevalent in developing countries [
In line with perspectives from experts in the field [
Symptom checkers are promising tools and seem to be more accepted for self-triage rather than for self-diagnosis. To maximize acceptance and use among young adults, it is important to address the various barriers identified in this study, including those that seek to improve the user experience. Importantly, awareness related to the existence of symptom checkers and their integration into the health care system are required to maximize the benefits related to these platforms. Future studies targeting other group segments are needed to understand perspectives of symptom checker use among the wider population.
Preinterview questionnaire.
Interview questionnaire.
Think-aloud exercise protocol and clinical vignette.
Example quotes of enablers and barriers for using symptom checkers.
Health Literacy Questionnaire
National Health Service
The authors would like to thank the university students who agreed to participate in the study as well as the administration staff at the University of Waterloo for aiding with participant recruitment.
SA conceptualized the study and collected the data. SA and SM analyzed the data. SM, JW, and AC contributed to the interpretation of data. SA wrote the initial draft of the manuscript. SM, JW, and AC reviewed the manuscript. All authors have read and approved the final version of the manuscript.
None declared.