Introduction
Being the initial care provided for an acute illness or injury until professional help arrives, first aid constitutes the foundation of public emergency response that has immense potential for protecting life and health in time-sensitive medical conditions. Whereas universal, high-quality, evidence-based first aid education plays a pivotal role in developing communities’ capacity to take appropriate action in health crises,Reference Singletary, Zideman and Bendall 1 its availability remains limited in many geographies,Reference Heard, Pearce and Rogers 2 hindering laypeople’s engagement in helping behaviors.
Currently, over two-thirds of the global population is actively using the Internet, 3 and the number of smartphone users exceeds 61% of the world’s inhabitants. 4 High availability and coverage of information, convenience and affordability of access, interactivity, and anonymity have made web search a preferred way to seek health information.Reference Jia, Pang and Liu 5 In the current highly digitalized era, where online health information seeking behavior has become a global trend, it is very likely that an individual will respond to their information needs, triggered by a health crisis, by attempting to obtain valuable content by surfing the web.Reference Yang, Zhang and Xiang 6 In particular, when faced with a real-life medical emergency, without appropriate first aid knowledge and skills, bystanders may seek instant “how-to-do” guidance online.Reference Bennett, Spasić and Filimonov 7 , Reference Birkun, Dyulicheva and Chuksina 8 Reliance on the internet-based advice could be especially high for people living in the developing world, where coverage of the public with first aid training is low,Reference Heard, Pearce and Rogers 2 and performance capabilities of emergency medical services (EMS) are lagging behind the rate of internet penetration into society. 3 In disasters and disease outbreaks, when EMS arrival is significantly delayed or impossible,Reference Scquizzato, Landoni and Paoli 9 first aid guidance found on the web can be the only possible solution for applying life-saving interventions by an untrained individual prior to professional assistance. Furthermore, the application of first aid following high-quality internet guidance may reduce health care services’ burden related to unnecessary EMS call-outs and emergency department visits for non-critical conditions.Reference Ismail, Gibbons and Gnani 10
On the other hand, public health information-seeking behavior could be influenced by the emerging threats of digital misinformation and disinformation,Reference Wang, McKee and Torbica 11 , Reference Swire-Thompson and Lazer 12 thereby exposing the individual to the risk of inappropriate and damaging interventions or delayed access to EMS. Considering the ever-growing global trend toward seeking health information online,Reference Jia, Pang and Liu 5 laypeople’s reliance on the Internet in medical emergencies may represent a serious public health concern that currently remains understudied and underappreciated despite its potential negative impact.
Although numerous prior studies have examined users’ interactions and receipt of support during major public health crises, such as COVID-19, particularly through social media,Reference Yang, Zhang and Xiang 6 , Reference Liu, Zhu and Xia 13 , Reference Terry, Yang and Yao 14 much less is known about laypeople’s online health information-seeking behavior aimed at obtaining immediate first aid guidance.Reference Bennett, Spasić and Filimonov 7 , Reference Birkun, Dyulicheva and Chuksina 8
The purpose of this study was to explore the public attitudes toward and experiences of using the Internet to get advice on first aid in real-life emergencies and to develop an understanding of the possible influence of the information-seeking behavior on community health.
Methods
This observational study employed 2 complementary methodological approaches, allowing us to explore the problem of online help-seeking in medical emergencies from different perspectives: an analysis of public comments posted on a social media platform, and a web-based questionnaire survey. The analysis of social media comments offered a large-scale, real-world snapshot of public discourse related to web searches for immediate first aid advice during emergencies. In contrast, the questionnaire survey provided targeted quantitative data on laypeople’s personal experiences and attitudes toward such help-seeking practices, complementing the qualitative insights from the social media analysis.
Analysis of Social Media Comments
YouTube (YouTube LLC, San Bruno, California, USA) online video-sharing platform was searched for English-language videos containing instructions on how to give first aid in case of foreign body airway obstruction (search query: “what to do choking”; search location: The United Kingdom). Videos lasting 60 seconds or less, videos without comments, presenting self-help only, first aid for animals, or real-life incident footage were excluded. Links to all eligible videos with at least 100,000 views were collected. Then, YouTube Data Tools, an online toolbox for extracting data from YouTube,Reference Rieder 15 was employed to retrieve all user comments posted to the selected videos.
The comments corpus was manually cleared of non-English text, meaningless comments, URLs and emails, hashtags, and user mentions. Further, the OpenRefine application 16 was applied to automatically detect and delete all emojis, special characters, punctuations, and redundant whitespaces, and to change all text to lowercase.
The resulting cleared corpus (see DatasetReference Birkun 17) was imported into Voyant Tools softwareReference Sinclair and Rockwell 18 for performing text analysis involving automated topic modeling: generation of term clusters using a machine learning algorithm called latent Dirichlet allocation. The algorithm assumes that the documents (i.e., social media comments) are represented as random mixtures over latent topics, where a topic is characterized by a distribution over words.Reference Jelodar, Wang and Yuan 19 Voyant Tools generated term clusters by determining which terms were best suited to which clusters based on co-occurrence in the text. The automatically produced term clusters were subsequently manually interpreted and described as distinct topics, with consideration given to the contextual meaning within the respective social media comments.
Questionnaire Survey
A survey was conducted among the subjects who completed the massive open online course on Basic Life Support (available at: https://stepik.org/course/13222/info). This course represents a web-based multimedia program covering key aspects of management of adult and pediatric cardiac arrest by lay rescuers. The course is provided in Russian language. Running in a web browser environment, it is accessible from a personal computer or a mobile device to anyone interested, without time restrictions for completing the curriculum or limitations on the user’s geographic location.
After completing the course, all users were asked to complete a Google Forms-based (Google LLC, Mountain View, California, USA) online questionnaire in Russian to assess their attitudes toward, and experience of, using the Internet as a source of advice on first aid in real-life medical emergencies, and collect additional information, including age, gender, professional medical education background, prior first aid training, number of hours per day spent online, and whether they had searched for health-related information on the Internet within the previous year (see Appendix A for the original questionnaire form translated into English).
Ethical Considerations
The dataset of YouTube commentsReference Birkun 17 was fully anonymized, preventing any possibility of identifying specific comment authors.
Participation in the survey was completely voluntary and anonymous. As such, no formal consent to participate was necessary, and subjects confirmed their agreement to take part by completing the questionnaire. Potential participants were informed about the anonymous nature and research purpose of the survey before taking part in it. The study was granted institutional review board exemption as it involved no identifiable or sensitive personal information.
Data Analysis
The results were analyzed using descriptive statistics, the Pearson chi-squared test, and the Spearman correlation coefficient. A 2-tailed P-value < 0.05 was considered statistically significant.
Results
Social Media Comments
The YouTube search revealed 49 eligible videos on first aid for foreign body airway obstruction which had a total of 44,123,955 views (min-max: 104,641-9,219,600, median [interquartile range (IQR)]: 300,955 [164,916-1,084,695]), 344,237 likes (min-max: 175-51,473, median [IQR]: 1,868 [1,016-6,869]), and 11,402 comments (min-max: 1-2,523, median [IQR]: 57 [28-231]).
As a result of the comments corpus clearing, 4,616 entries were excluded, leaving 6,786 comments for text analysis, with 144,958 total words and 7,313 unique word forms.
After applying the English stopword list (a set of function words that should be excluded from the results, including determiners, coordinating conjunctions, prepositions, etc.) manually supplemented with the prevalent corpus-specific terms (“choked,” “choking,” “I’m,” “it’s,” “just,” “like,” “thank,” “thanks”), the following 10 terms showed the highest frequency of occurrence: video (n = 1,245), ads (an umbrella term that was applied to the corpus to join all advertisement-related terms, n = 942), know (n = 522), life (n = 519), baby (n = 493), CPR (n = 394), help (n = 375), time (n = 368), Heimlich (n = 368), and save (n = 351).
The automated topic modeling (purposefully limited to 5 topics composed of 10 terms each; Table 1) revealed that the comments were primarily discussing techniques and maneuvers of first aid (topic I); sharing personal experience related to foreign body airway obstruction and application of first aid (topic II); indicating the importance of having knowledge and skills of first aid (topics III and IV); and expressing user dissatisfaction with appearance of unskippable advertisements preceding the first aid videos (topic V). The latter topic implies laypeople’s collective perception of the online videos as a source of immediate advice on first aid in real-life emergencies.
Results of the topic modeling analysis and examples of representative comments
Notes.
a Preceding terms in each topic contribute more to the topic than subsequent ones.
b The quotes were left unedited for spelling or grammar.
Abbreviations: CPR, Cardiopulmonary resuscitation.
A manual review of the comments’ corpus revealed additional evidence related to this phenomenon that was not identified by the topic modeling, presumably because of a lower occurrence of related comment pieces. Multiple users commented on the videos by posting timestamps to indicate the exact moment where the first aid maneuver is shown in a video so that people can skip to the action part in case of a real-life emergency (e.g., “If your baby is choking right now skip to 1:18,” “For those who have someone actively choking at the moment go to 0:41,” “If it’s an emergency skip to 0:46”). Other users endorsed such comments by putting “likes” and proposing to pin the timestamps above all other comments (e.g., “Whoever runs this channel, pin the ‘Skip to 3:57’ comment with 1.7k upvotes!! It’s the most vital information!,” “This comment needs to stay on top!,” “Admin should pin this comment”; see DatasetReference Birkun 17).
Survey Results
The web-based survey complemented the qualitative insights obtained from the analysis of social media comments by generating quantitative estimates of public attitudes and personal experiences related to online help-seeking in medical emergencies. In total, 731 completed questionnaires were collected from May 2023 to August 2024 (Table 2). The respondents were mostly female (n = 421, 57.6%), under the age of 40 years (n = 468, 64.0%), and without professional medical education (n = 592, 81.0%). Approximately 68.7% (n = 502) had previous training on resuscitation or first aid besides the completed online course, and 54.2% (n = 396) self-rated their general knowledge and skills in first aid as moderate to high-level (4-5 points on a 5-point scale).
Distribution of responses to the questionnaire survey (N = 731)
Time spent on the Internet was less than 2 hours per day for 12.6% of the respondents (n = 92), 2 to 5 hours per day for 63.6% (n = 465), and more than 5 hours per day for 23.8% (n = 174). The vast majority (n = 603, 82.5%) searched for some health-related information online at least once over the past year.
About 14.6% of respondents (n = 107) affirmed that they personally searched for brief instructions on first aid on the Internet in a real-life emergency. Another 8.9% (n = 65) reported such experience among their friends or relatives. Of those respondents who searched for first aid advice online themselves, 90.7% (n = 97) managed to find some instructions and followed them completely or partially.
Most respondents expressed moderate to high level of agreement (4-5 points on a 5-point scale) that they would attempt to find brief advice on first aid on the Internet in a real emergency if they or other bystanders would not have sufficient knowledge and skills to give help (n = 454, 62.1%; median [interquartile range (IQR)]: 4.0 [3.0-5.0]); that online search for brief first aid advice is important because it is potentially lifesaving (n = 508, 69.5%, median [IQR]: 5.0 [3.0-5.0]); and that, if necessary, they would be able to easily find such instructions (n = 435, 59.5%; median [IQR]: 4.0 [3.0-5.0]). Less than half of the respondents agreed (4-5 points) that brief instructions on first aid available online are sound enough to be relied upon in a real emergency (n = 315, 43.1%; median [IQR]: 3.0 [3.0-4.0]).
Readiness to attempt an online search of advice on first aid in an emergency was found to be positively associated with respondents’ real-life experience of seeking first aid instructions on the Internet in a health crisis (P < 0.001), female gender (P < 0.05), and search for health-related information online in the previous year (P < 0.001), but not related with prior first aid training and professional medical education. Further, it showed a strong positive correlation with the perceived importance of searching first aid instructions online as potentially lifesaving means (rS = 0.702; P < 0.001), a moderate positive correlation with the respondents’ confidence in the reliability of the instructions available on the Internet (rS = 0.479; P < 0.001), weak positive correlation with self-evaluated level of first aid knowledge and skills (rS = 0.122; P = 0.001) and number of hours per day spent online (rS = 0.114; P < 0.01), and negligible negative correlation with age (rS = −0.088; P < 0.05).
Discussion
The findings reported here shed new light on the public information-seeking behavior in health crises. In particular, the study further provides evidence for the hypothesis that laypeople commonly consider and apply internet resources as a means of instant informational support when providing first aid in medical emergencies.Reference Bennett, Spasić and Filimonov 7 , Reference Birkun, Dyulicheva and Chuksina 8 According to the social media data analysis, the public strongly believes that online video demonstrations of first aid maneuvers should be immediately available to everyone so that instructions shown on the video could be repeated by an untrained bystander right away to give necessary help to a victim in a real life-threatening situation. The survey supplemented the results of the social media comment analysis by demonstrating how public beliefs about online help-seeking are translated into practice. At least 2 out of 10 persons have sought internet advice on first aid in emergencies themselves or had someone among their friends or relatives who reported similar experiences. Moreover, the majority showed their readiness to use first aid instructions presented online in future as they believe this could be lifesaving in the absence of proper knowledge on how to give help, and previous first aid training or professional medical education did not alter this intent.
Respondents’ readiness to attempt an online search of first aid advice in an emergency was found to be positively associated with real-life experience of seeking first aid advice on the web during a health crisis, searching for health-related information online within the previous year, and being female. The latter supports previous observations that females are more likely than males to seek health information online, although males have traditionally been more active internet users.Reference Hallyburton and Evarts 20
Public reliance on internet advice on first aid has important implications for public health, as it carries apparent risks. Firstly, the search for information on the web may prevent or delay a call for EMS, which constitutes the priority for time-sensitive, life-threatening conditions.Reference Berg, Cheng and Panchal 21 Besides promptly dispatching an ambulance team, in many EMS systems, emergency telecommunicators may offer bystanders professional telephone advice on how to provide first aid and support them until the arrival of EMS to the scene,Reference Kurz, Bobrow and Buckingham 22 so doubtful random self-search of instructions online by a bystander would be unreasonable.
Another major threat of the dependence on internet advice in emergencies lies in the poor quality of information on first aid published online. Widely available open-access internet resources, including websites,Reference Burgess, Cameron and Cuttle 23 online videos,Reference Yilmaz Ferhatoglu and Kudsioglu 24 , Reference O’Leary, O’Connor and Gibson 25 web search engine question-answering systems,Reference Birkun and Gautam 26 voice-based conversational digital assistants,Reference Picard, Smith and Picard 27 mobile apps,Reference Metelmann, Metelmann and Schuffert 28 and modern artificial intelligence-powered chatbots,Reference Birkun and Gautam 29 , Reference Knebel, Priglinger and Scherer 30 very often provide incomplete and inaccurate information that contradicts the current guidelines on first aid. Also, the web resources frequently include directions on first aid inapplicable for an untrained person (e.g., to apply a tourniquet in bleeding or to give rescue breaths in cardiac arrest) and occasionally contain incorrect, non-evidence-based advice that, if followed by an unsuspecting bystander, may cause additional harm to the victim.Reference Burgess, Cameron and Cuttle 23 , Reference Birkun and Gautam 26 , Reference Birkun and Gautam 29
The current study showed that over 40% of the survey respondents believed that instructions on first aid available on the web are sufficiently reliable to follow in emergency situations, and 9 out of 10 persons who searched for the instructions online in a real medical emergency have implemented the advice they found. However, implementation does not necessarily imply that the advice was appropriate, accurate, or safe, and it may carry potential health risks. These findings, along with the high prevalence of digital misinformation concerning first aid and the fact that online search for instructions on first aid in real emergencies is common, raise concerns about the continuous and widespread application of inadequate and damaging practices, carrying a considerable burden on public health.
On the other hand, the presented results bear evidence of existing high community demand for instant digital informational support in emergencies, which is supposedly maintained by a combination of limited availability of public first aid education,Reference Heard, Pearce and Rogers 2 underdevelopment of EMS systems,Reference Suryanto, Plummer and Boyle 31 and poor digital health literacy (particularly, in terms of the ability of the general public to understand related risks),Reference Arias López, Ong and Borrat Frigola 32 on the background of continuous penetration of the Internet and mobile technologies into the society. 3 , 4 It must be admitted that instant digital informational support on first aid, on the premise of high accessibility and quality, has significant health-protecting potential for medically underserved population groups (e.g., in rural or wilderness settings) and communities inadequately covered with first aid training. When EMS response is delayed and dispatcher telephone assistance is unavailable, the provision of first aid by an untrained bystander following instructions obtained from a mobile device (via online search or preinstalled app) could be a better-than-nothing solution and, at times, the only possible way to protect the victim’s health and life.
In summary, the existing public demand for and continuous actual use of the Internet as a source of advice on first aid, the risks arising from laypeople’s reliance on the information of questionable quality, and the potential usefulness of authoritative digital first aid guidance for instances where other means of immediate assistance are unavailable, constitute an intricate public health issue that requires a multifaceted solution. A global call for action is needed to initiate a coordinated response of international professional medical bodies, governments, researchers, mass media, and content developers, aimed at designing and implementing effective strategies to enhance public digital health literacy,Reference Nutbeam and Lloyd 33 , Reference Sørensen, Levin-Zamir and Duong 34 mitigate the misinformation spread,Reference Swire-Thompson and Lazer 12 , Reference Geers, Swire-Thompson and Lorenz-Spreen 35 promote high-quality community education on first aid,Reference Hewett Brumberg, Douma and Alibertis 36 , Reference Nabecker, de Raad and Abelairas-Gomez 37 and provide the public with reliable open digital resources for immediate informational support in medical emergencies. The latter necessitates the creation of an expert-developed knowledge base that would serve as the sole and mandatory evidence-based source of standardized instructions on first aid adapted for laypeople’s use. Emergency and disaster medicine professionals might consider integrating approved digital guidance into public education and awareness initiatives, thereby ensuring accessibility of trusted, actionable, and safe resources and proper layperson behavior during high-stress emergencies.
Limitations
This study has limitations to consider. First, the social media data analysis involved one social media platform, and the collected comments were restricted to those posted to popular English-language instructional videos concerning first aid for choking. Repeated studies involving other social media platforms, different content selection criteria, and additional emergency scenarios (e.g., bleeding, cardiac arrest, or seizures) may yield dissimilar results and reveal new insights into the issue, and therefore should be encouraged. Second, the questionnaire survey sample was not representative of the general population as it consisted of the completers of the online course, mainly young people and active internet users, who probably have higher motivation to use the Internet for making health-related decisions in emergencies when compared to the general public. Also, the data collection and data analysis were carried out by a single researcher, which may have introduced bias, although all data were double-checked to ensure the reliability of the presented results.
Conclusions
Laypeople perceive the Internet as an important source of instant guidance on first aid that could be life-saving in a medical emergency. A large proportion has personal experience of searching and implementing first aid instructions found online in a real emergency situation. Risks related to reliance on inappropriate internet advice in emergencies, on the one hand, and community demand for immediate informational support in health crises, on the other, constitute a major public health concern that necessitates the development of effective strategies for enhancing digital health literacy, mitigating harmful impacts of misinformation, and satisfying laypeople’s health information needs in emergencies.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/dmp.2025.10292.
Data availability statement
The dataset generated and analyzed during the current study is available in the Mendeley Data repository.Reference Birkun 17
Author contribution
AAB is the sole author of the manuscript.
Funding statement
This research did not receive any funding.
Competing interests
None.
