The true “first responder” is believed to be the layperson who most immediately recognizes and responds to an emergency, with a bystander commonly defined as a “person at the scene of an event who voluntarily steps forward to help.”Reference Faul, Aikman and Sasser 1 Bystanders exert a tremendous effect on the ability to direct the balance of the response. As such, there is a growing impetus to train and assess the role of interventions performed by bystanders during acute situations. Indeed, bystander intervention training has successfully been applied to cardiopulmonary resuscitation (CPR),Reference Nakahara, Tomio and Ichikawa 2 –Reference Stiell, Nichol and Wells 4 the Stop the Bleed campaign,Reference Van Winkle, DiBrito and Amini 5 –Reference Pasley, Parker and Levy 8 sexual assault prevention,Reference Mahoney, Gielen and Bailey 9 –Reference Marcantonio, Willis and Schisler 13 and general medical emergencies, among others.Reference Iversen, Blauenfeldt and Johnsen 14
Bystander intervention, nevertheless, remains highly variable. The intervention rate prior to the arrival of emergency medical services (EMS) is estimated at only 11% of emergencies.Reference Faul, Aikman and Sasser 15 The collegiate environment has risen in prominence in the bid to cultivate higher rates of bystander intervention. Undergraduate students, themselves at varying levels of emotional, physical, and intellectual development, are frequently placed in proximity for the first time in living and studying environments when they enroll in university. New freedoms compounded by rapidly fluctuating social dynamics and ready access to alcohol and other intoxicantsReference Daniel, Haddad, McConaha and Lunney 16 –Reference McCabe, West and Teter 18 create a platform with a seemingly greater natural predilection for emergencies than would typically be observed in the general population. Binge drinking is common on college campuses. Ngo et al. reported an overall incidence rate of 59/10 000 person-years for at least 1 emergency department visit related to alcohol use among first-year college students.Reference Ngo, Rege and Ait-Daoud 19
In view of these specific considerations, early and immediate intervention by bystanders in the university setting may represent a critical opportunity to save lives and summon appropriately trained responders. Prior research in disaster preparedness has sought to explain why and how people react to various types of emergencies and has proposed methods to increase intervention. Based largely on the Extended Parallel Process Model (EPPM),Reference Popova 20 health care and public health professionals’ willingness to respond (WTR) largely depends on how threatening a disaster is perceived to be and the perceived self-efficacy of the potential responder in “their ability to perform the recommended response.”Reference Witte, Cameron and McKeon 21 Therefore, the most effective responders are those who believe the threat is both highly likely and their response would be highly effective. In the emergency response domain, the EPPM has been applied to various other types of responders and scenarios, including local health department workers,Reference Barnett, Thompson and Semon 22 hospital staff,Reference Balicer, Barnett and Thompson 23 –Reference Kaiser, Barnett and Hayanga 25 Medical Reserve Corps volunteers,Reference Errett, Barnett and Thompson 26 and EMS providers.Reference Barnett, Levine and Thompson 27 EPPM-based interventions have similarly been extended to college students in the context of COVID-19 social distancing,Reference Lin and Chen 28 , Reference Roberto, Zhou and Lu 29 other infectious disease,Reference Gore and Bracken 30 , Reference Jain, Hoffman and Beam 31 and the dangers of e-cigarettes.Reference Escoto, Watkins and Welter 32
To distinguish between traditional bystander intervention focused on sexual assault and substance misuse, the term “emergency onlooker response” in the collegiate setting more aptly describes the broader environment where bystanders observe and consider intervening prior to the arrival of trained first responders in a potential medical emergency. Given the success in shifting lifesaving interventions as far upstream as possible to the point of injury or illness,Reference Nakahara, Tomio and Ichikawa 2 universities represent an understudied area for greater emergency onlooker response intervention research. Little is known about the self-efficacy and perceived threats of undergraduate students in various common emergencies. To that end, this study sought to assess perceptions of knowledge, attitudes, and beliefs of undergraduate students to better understand their WTR in various potential emergency scenarios and to assess their thoughts regarding education concerning bystander intervention and emergency onlooker response.
Methods
We developed a cross-sectional web-based survey and distributed it to all full-time undergraduate students at West Virginia University (WVU), a public, land-grant university with its main campus in Morgantown, West Virginia. In the Fall 2019 semester, the institution reported a total undergraduate student count of 21 086. 33 The custom survey instrument was developed iteratively using the EPPM as a framework for assessing self-efficacy and perceived behaviors in real-world fear-inducing situations. Based on the key domains of response-efficacy and self-efficacy, the survey sought to evaluate attitudes and beliefs as well as WTR to 3 representative potential emergency scenarios: Scenario 1, a college student “found down” at a large (200+ person) party; Scenario 2, a new associate member of a fraternity/sorority collapses at an initiation event; and Scenario 3, a community member “found down” at a grocery store. When a person is found unresponsive with no clear etiology, the emergency department often triages the patient diagnosis as “found down.”Reference Howard, Kornblith and Conroy 34
The final survey tool was built using Qualtrics (Qualtrics, Provo, Utah, USA) and included 88 questions. After providing basic demographic data, respondents addressed the 3 representative vignettes and then answered general questions about emergency preparedness. The survey instrument assessed self-reported perceptions of knowledge and skills, as well as attitudes and beliefs regarding the following: education and training, obligation, safety, psychological readiness, efficacy, personal preparedness, and WTR. The survey tool used a combination of a 9-point Likert scale, binary questions, a single-open-answer question and a select all that apply question. The full survey tool is available in Supplemental 1.
A link to the survey was sent via the student news platform “U-News” operated by University Relations and Enrollment Management and was intended for all undergraduate students in the fall semester of 2019. Similarly, invitations to participate were sent electronically via the WVU Engage student organization platform (operated by CampusLabs) to student email accounts associated with the Office of Student Engagement and Leadership. The survey was available online from September to December 2019 and all responses were submitted anonymously. Fifty 20-dollar gift cards were offered for completion of the survey. The WVU Institutional Review Board approved this study and deemed it exempt from the informed consent requirement.
Statistical Analysis
Continuous variables were summarized by their means or medians. Categorical variables were summarized as proportions, often represented as a percentage. Likert-style questions were broken into binary responses, with 1-4 representing agreement, 5-9 denoting disagreement, and an option to indicate being unsure. Incomplete surveys were utilized to the extent they were completed and “unsure” responses to specific questions were excluded from that analysis. Attitudes and beliefs between scenarios were compared using McNemar’s Exact test when the respondent answered both questions. Additionally, 95% Clopper-Pearson exact confidence intervals were calculated for percent agreement. The relationship between self-described WTR and involvement in student activities, athletics, fraternity/sorority/Greek life, or desire to pursue a health care career was also assessed and differences were tabulated using Chi-square tests. Logistic regression models were run for each of the WTR questions in order to measure the association by EPPM quadrant (i.e., low perceived threat/low perceived efficacy; low perceived threat/high perceived efficacy; high perceived threat/low perceived efficacy; high perceived threat/high perceived efficacy). This article adheres to best practices in conducting and reporting survey research.Reference Magid, Ranney and Risica 36 Data analysis was performed using SAS, Version 9.4 (SAS, Cary, NC, USA).
Results
At the close of the survey period, 970 student responses were received, with 75.3% (730/970) fully completing the survey. The median age of respondents was 20 years old (IQR 18-21), with 65% of respondents identifying as female and 95% as full-time students (enrolled in 12 or more credit hours). Students were roughly evenly distributed among year of education, with 28.8% being freshmen, 23.2% sophomores, 21.3% juniors, and 17.8% seniors. Intended majors were distributed throughout the university’s departments, with an overrepresentation of health and biological sciences at 28.4%. Students reported a variety of activities, including no student groups (35.7%), multiple activities (12.6%), and fraternity/sorority/Greek life participation (13.6%). Students predominantly lived off-campus (54.8%) and 88.7% and 77.1% responded that they had received training in emergency procedures during orientation and during their undergraduate education, respectively. The full student respondent demographics are displayed in Table 1.
Table 1. Basic demographics of undergraduate students responding to an emergency onlooker response survey
Overall, most students responding to the survey reported that each scenario would have severe life consequences for the person in question. However, Scenarios 2 (new associate member of a fraternity/sorority collapses at an initiation event) and 3 (community member “found down” at a grocery store) garnered a greater degree of concern for the long-term consequences with 75%, 82%, and 83% agreeing with that sentiment for each scenario, respectively, with P < 0.0001 comparing Scenario 1 (college student “found down” at a large party) to 2 and 3, individually. However, pre-event training and preparedness for such a situation were considerably lower with only 56%, 60%, and 58% believing their university adequately prepared them for each scenario, respectively. This was the case despite 84%, 82%, and 82%, respectively, reporting that their university should provide pre-event training for a similar scenario (Figure 1). Notably, only 67% were willing to respond to a student “found down” at a large party if that was their primary role in the response (Scenario 1), while 78% were willing to take on the same role for a new associate member of a fraternity/sorority collapsing at an initiation event (Scenario 2) and 83% were willing to assist a community member “found down” at a grocery store (Scenario 3). Finally, students were nearly equally willing to respond to each scenario regardless of any potential negative consequences for the patient. The full results of student attitudes and beliefs are displayed in Table 2.
Figure 1. College students attitudes and beliefs towards various emergency scenarios.
Table 2. Undergraduate students’ attitudes and beliefs and willingness to respond in various emergency scenarios (n = 970)
P values calculated via McNemar’s Exact test and restricted to respondents who answered both questions.
In subgroup analysis, students’ interests and activities were stratified to compare WTR and attitudes and beliefs. The full results of the subgroup analysis are displayed in Table 3. Overall, students involved in at least 1 student activity or interested in a career in health care were more likely to respond in whatever capacity they might be needed (P = 0.0478 and P = 0.0266, respectively). Students participating in fraternity/sorority/Greek life reported a greater obligation to respond to a student “found down” at a large party compared to their non-fraternity/sorority/Greek counterparts (97.6% vs 89.0%, P = 0.0158).
Table 3. Undergraduate students agreement with emergency onlooker response questions stratified by student activities
Finally, Table 4 applies the EPPM to compare high and low efficacy to high and low threats. Respondents with self-perceived high self-efficacy were significantly more likely than those with low self-efficacy to be willing to respond in whatever capacity needed across all 3 scenarios (Scenario 1: OR 29.7 [95% CI: 7.2, 122.7]; Scenario 2: OR 66.6 [95% CI: 9.18, 483.7]; Scenario 3: OR 19.4 [95% CI: 4.65, 81.2]). Similarly, those with high efficacy and high threat levels were significantly more likely to feel obligated to respond to someone “found down” than comparable low efficacy and low threat students, across all 3 scenarios (Scenario 1: OR 37.7 [95% CI: 5.17, 275.2]; Scenario 2: OR 34.6 [95% CI: 8.34, 143.4]; Scenario 3: OR 49.3 [95% CI: 6.74, 361]).
Table 4. Willingness to respond by perceived threat and self-efficacy for undergraduate students in various scenarios
HE = High Efficacy; LE = Low Efficacy; HT = High Threat; LT = Low Threat; OR = Odds Ratio; CI = Confidence Interval.
A non-response bias analysis was performed with a standardized difference greater than 10% considered a significant difference. Those who completed the survey were more likely to be female and younger than was observed within the WVU campus that same year with a standardized difference of 32.4% and 12.3%, respectively, as displayed in Table 5.
Table 5. Non-response bias analysis assessing demographic characteristics of those who responded to the survey versus all full-time undergraduate students, fall 2019
Discussion
Emergency onlooker response remains a critical aspect of ensuring timely medical care for those in need. The results of this study show that most undergraduate students at a single, large public university reported that they have a role in responding to a person found unresponsive in a variety of social settings. While a nearly uniform 81%-82% of respondents reported feeling sufficiently skilled to respond, depending on the scenario, only 56%-60% reported that their university had adequately prepared them for those situations. In order to maximize the likelihood of a response, past research suggests that the responder must feel the scenario is highly likely and they can be highly effective in their role.Reference Witte, Cameron and McKeon 21 This presents an opportunity for universities to tailor their trainings and interventions to both promote lifesaving skills, self-efficacy in those skills, and awareness about possible emergency situations. For example, prior research into college students’ intention to perform CPR postulated that the training would be more useful if it included both technical skills and an expectation that these skills should be utilized.Reference Kelley, Clark and Brown 35
Nationwide, a focus has existed toward bystander intervention training pertaining to alcohol misuse and sexual assault. While these topics are critical and timely, an opportunity remains to broaden the scope of training beyond these specific topics to include a more expansive scope of emergencies that can occur in the collegiate setting. The findings of this survey show students who reported that a threat is highly likely, and they would be highly efficacious in their response, had 34.6-49.3 higher odds of responding in any capacity needed compared to their low threat and low efficacy counterparts, depending on the scenario. Given the potential for a college student to be present when someone collapses at a large party, fraternity or sorority event, or even at a grocery store, promoting an understanding that such a threat is both likely and that the student would have the self-efficacy to respond is key to creating a vast first responder safety net. This is a natural extension of this survey’s results where a plurality of students indicated that 1 hour per month would be the ideal amount of time to spend on emergency training.
Medical Amnesty
In an effort across the United States to promote ensuring adequate medical care when needed, medical amnesty policies in society generally and collegiate settings specifically have been enacted. As of January 2019, 47 states and Washington D.C. have some variation of a medical amnesty law.Reference Griner, Strasser and Kemp 37 Medical amnesty laws, frequently referred to as “Good Samaritan laws,” are designed to encourage bystanders to contact authorities for assistance during emergencies by providing provisions such as immunity from criminal action.Reference Griner, Strasser and Kemp 37 While settings of higher education may have differing prevailing rules and consequences related to alcohol and substance misuse (e.g., administrative actions versus criminal sanctions), ensuring broad knowledge and trust in these policies is key to effective utilization.Reference Weaver, Kroshus and Milroy 38 In this study, WTR decreased across all 3 scenarios when asked to consider possible negative consequences for the student responder compared to the person in distress. In the context of only 55% of the sample reporting familiarity with their university’s amnesty policy, this may represent a key area for intervention. Past research into promoting safe alcohol practices among fraternity/sorority/Greek-life-involved college students has increased understanding about local medical amnesty policies as well as safer alcohol consumption practices, especially when these efforts are peer-led.Reference Abadi, Shamblen and Thompson 39
Limitations
This study is subject to several limitations. While the survey was distributed to 21 086 students over a 4-month period, only 970 partial or complete responses were submitted for analysis. Accordingly, a non-response bias analysis was conducted to compare survey respondents to the general student population. Compared to the general student body, females and younger students were more likely to have completed the survey. The non-response bias analysis, nevertheless, revealed that the results of the study were largely an accurate reflection of the broader university student body and legitimized our inferences. Second, this study was cross-sectional; thus, causal relationships could not be interpreted. Third, this study was conducted at a single site, potentially limiting generalizability to other centers of higher education and broader communities. Fourth, participation was voluntary which may introduce a self-selection bias.
Conclusion
Emergency onlooker response remains a key aspect and unmet need among undergraduate students. Further efforts are needed to promote the understanding that various community emergencies may occur near college students. With the acquisition of some basic lifesaving intervention training and education, they would possess the skills and intent to respond. In the collegiate environment, specific protections, such as medical amnesty policies, need to be further emphasized and straightforward to understand to further promote a bystander response.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/dmp.2024.308.
Competing interest
The authors report there are no competing interests to declare.
Open access
