2025 Volume 7 Issue 6 Pages 419-425
Background: Because bystander performance of cardiopulmonary resuscitation in out-of-hospital cardiac arrest cases is influenced by the number of rescuers/bystanders, we assessed the relationship between the presence of other people and performance of rescue actions in an actual emergency situation.
Methods and Results: A cross-sectional study was performed using data from an anonymous self-administered questionnaire-based survey that included laypersons who had encountered emergency situations during the past 5 years. Based on their responses related to the presence of other people, laypersons were divided into 2 groups: “single-bystander” (absence of others) and “multiple-bystanders” (presence of others). The primary outcome was any rescue action(s) performed by laypersons during an emergency. A total of 1,219 laypersons were eligible for our analysis; 69 (5.7%) encountered emergencies in which others were absent (single-bystander group) and 1,150 (94.3%) encountered emergencies in which others were present (multiple-bystanders group). The proportion of laypersons who performed any rescue actions was 95.7% in the single-bystander group and 73.8% in the multiple-bystanders group.
Conclusions: The proportion of laypersons who performed any rescue actions was lower when other people were present compared with when they were absent.
The annual incidence of out-of-hospital cardiac arrest (OHCA) of cardiac origin is approximately 82,000 in Japan; however, the survival rate with favorable functional outcomes remains low.1 It is widely accepted that cardiac arrest is a condition that can result in death within a short timeframe. Hence, for successful resuscitation after OHCA,2–4 rescue actions, including cardiopulmonary resuscitation (CPR) and defibrillation by automated external defibrillator (AED), must be initiated by laypersons who encounter such emergency situations.
There are negative factors that prevent laypersons from initiating CPR, such as psychological barriers, the collapsed person’s symptoms, sex of the collapsed person, lack of confidence in ability to perform CPR, and difficulties in recognizing cardiac arrest.5–12 However, for a bystander to initiate CPR (bystander CPR), it is necessary to approach the collapsed person first. Therefore, it is essential to clarify the factors that influence initiating any rescue action regardless of CPR. Observations from our previous research suggest that the fear of approaching a collapsed person is itself a psychological barrier that needs to be overcome for initiating any rescue action.13 Although some resuscitation-related studies have indicated that the presence of multiple bystanders/rescuers promotes bystander CPR,9,14–16 it remains to be evaluated whether the presence of others influences taking the initiative to perform rescue actions regardless of CPR. Moreover, psychological studies have established a phenomenon, called the “bystander effect”, in which an individual’s likelihood of helping decreases when other people are present in emergency situations. This phenomenon has been explained by 3 processes: audience inhibition (bystanders are concerned about how others will perceive them), social influence (if no one else is acting, bystanders may interpret the situation as less urgent and feel less compelled to act), and diffusion of responsibility (bystanders can rationalize their inaction by convincing themselves that “somebody else must be doing something”), and it is believed that these phenomena, caused by others, make people hesitant to help. This situation may also be observed in situations related to resuscitation.17,18 In this study we aimed to assess the relationship between the presence of other people and the performance of rescue actions in an actual emergency situation.
The current study was a secondary analysis of the results of our previous study.13 This cross-sectional study was performed using data from an anonymous self-administered questionnaire-based survey conducted between August 1, 2018, and November 20, 2018, among the general public who participated in CPR training held by 4 fire departments and 3 organizations, or worked at a certain Japanese educational institution or company. We included adult laypersons aged over 18 years but excluded physicians, nurses, and paramedics.
Data Collection and Outcome MeasureDetails of the questionnaire have been previously described.13 The data related to the “presence of other people” was included in the questionnaire as a binary choice question (“Yes” or “No”). We used the following 3 major data points for this study: (1) layperson’s characteristics: including sex, age, previous experience of CPR training, and previous experience of performing any rescue actions; (2) emergency setting: including the location, relationship with the collapsed person, cause of the collapse, sex of the collapsed person, and age category of the collapsed person; and (3) laypersons’ rescue actions: such as checking patient’s consciousness, gathering people or giving instructions to other laypersons, making the emergency call, looking for or carrying an AED, performing chest compressions, providing rescue breaths, applying the AED pads and/or pushing the buttons of the AED, and performing any resuscitation actions other than the above.
The primary outcome was any rescue action(s), defined as cases where the layperson performed at ≥1 of the 8 rescue actions mentioned above.
Statistical AnalysisWe included the same study population as in our previous work,13 and analyzed laypersons who had encountered emergency situations within the past 5 years, excluding those who encountered such situations at either hospitals or nursing homes.
Depending on the answers to the question related to the presence of other people, laypersons were divided into 2 groups: the single-bystander group that answered “No” (absence of others) and the multiple-bystanders group that answered “Yes” (presence of others). Laypersons’ characteristics (age, sex, location of the collapsed person, cause of collapse, relationship with the collapsed person, sex of the collapsed person, age category of the collapsed person, previous experience of CPR training, and previous experience of performing any rescue actions) were compared between groups using Mann-Whitney U tests for continuous variables and chi-square tests for categorical variables, and laypersons’ rescue actions were described for both groups. Furthermore, the proportion of the primary outcome was tabulated. Subgroup analysis stratified by relationship (family or acquaintance, stranger) with the collapsed person was also performed for laypersons’ characteristics and primary outcome. To consider the differences of the emergency situation, the same analysis was performed for laypersons excluding those who did not perform any rescue actions because other people had already initiated them. Furthermore, although there are limits to the interpretation, univariate and multivariate logistic regression analyses were conducted as exploratory analyses to assess the association between the presence of other people and the performance of rescue actions. Crude and adjusted odds ratios with 95% confidence intervals were calculated, and laypersons’ characteristics were included in the multivariable analysis as potential confounding factors.
All P values were two-tailed, and all statistical analyses were performed using SPSS v24.0J (IBM Corp., Armonk, NY, USA).
Ethical ConsiderationsThis study was conducted in accordance with the Declaration of Helsinki and the Ethical Guidelines for Medical and Health Research Involving Human Subjects of Japan. Based on these guidelines, we provided all participants with a written explanation form, and if they agreed to cooperate in this survey, they answered the questionnaire. This study was approved by the Ethics Committees of the Kyoto University Graduate School of Medicine (R1393).
The flow chart for this study is illustrated in the Figure. We distributed the questionnaire to 8,430 laypersons; of them, 7,827 (92.8%) responded and 7,744 (91.9%) answered ≥1 question (Supplementary Table 1 shows the characteristics of the laypersons who answered ≥1 question). After excluding 336 laypersons aged <18 years or whose age was unknown, and 400 laypersons holding medical qualifications or whose medical qualification was unknown, 7,008 (83.1%) laypersons were eligible for the study. Of them, 1,361 (16.1%) had encountered an emergency situation during the past 5 years, and 1,219 (14.5%) were eligible for our analysis after excluding those who encountered an emergency situation at either a hospital or nursing home or gave invalid answers.
Flow chart of the selection of respondents and analyses in this study.
The laypersons’ characteristics stratified by the presence of other people are outlined in Table 1. Among the eligible laypersons, 69 (5.7%) encountered an emergency situation in which others were absent, and 1,150 (94.3%) encountered an emergency where others were present. There was a statistically significant difference between the groups in terms of the location and the relationship of the layperson with the collapsed person.
Characteristics of Laypersons Stratified by the Presence of Other People
Single-bystander (n=69) |
Missing | Multiple-bystanders (n=1,150) |
Missing | P value | |
---|---|---|---|---|---|
Sex, male, n (%) | 38 (55.1) | 0 (0.0) | 607 (52.8) | 1 (0.1) | 0.717 |
Age, years, median (IQR) | 40 (28–50) | 39 (27–49) | 0.525 | ||
Previous experience of CPR training, n (%) | 37 (53.6) | 7 (10.1) | 707 (61.5) | 79 (6.9) | 0.307 |
Previous experience of performing any rescue action, n (%) |
37 (53.6) | 1 (1.4) | 574 (49.9) | 26 (2.3) | 0.592 |
Location of collapsed person, n (%) | 1 (1.4) | 25 (2.2) | <0.001 | ||
Workplace | 9 (13.0) | 283 (24.6) | |||
Public area | 4 (5.8) | 173 (15.0) | |||
Residence | 26 (37.7) | 75 (6.5) | |||
Educational institution | 0 (0.0) | 78 (6.8) | |||
Sports facility | 0 (0.0) | 56 (4.9) | |||
Other locations | 29 (42.0) | 460 (40.0) | |||
Relationship with the collapsed person, n (%) | 1 (1.4) | 5 (0.4) | <0.001 | ||
Family | 24 (34.8) | 80 (7.0) | |||
Friend | 2 (2.9) | 55 (4.8) | |||
Associate | 1 (1.4) | 53 (4.6) | |||
Colleague | 2 (2.9) | 98 (8.5) | |||
Others | 39 (56.5) | 859 (74.7) | |||
Cause of the collapse, n (%) | 0 (0.0) | 0 (0.0) | 0.057 | ||
Non-cardiac arrest | 49 (71.0) | 818 (71.1) | |||
Cardiac arrest | 12 (17.4) | 112 (9.7) | |||
Unknown | 8 (11.6) | 220 (19.1) | |||
Collapsed person, n (%) | |||||
Male | 37 (53.6) | 2 (2.9) | 690 (60.0) | 26 (2.3) | 0.315 |
Adult | 64 (92.8) | 1 (1.4) | 1,025 (89.1) | 7 (0.6) | 0.237 |
CPR, cardiopulmonary resuscitation; IQR, interquartile range.
The proportion of each rescue action performed by laypersons is outlined in Table 2. The proportion of rescue actions such as checking the patient’s consciousness, making emergency calls, and performing chest compressions were higher when other people were absent than when they were present. Regarding the primary outcome, the proportion of laypersons who performed any rescue actions was higher in the single-bystander group (95.7%) than in the multiple-bystanders group (73.8%).
Laypersons’ Rescue Actions Stratified by the Presence of Other People
Single-bystander (n=69) |
Multiple-bystanders (n=1,150) |
|
---|---|---|
Checking patient’s consciousness, n (%) | ||
Performed | 60 (87.0) | 701 (61.0) |
Not performed | 5 (7.2) | 41 (3.6) |
Not performed because other people had already performed | – | 398 (34.6) |
Missing | 4 (5.8) | 10 (0.9) |
Gathering people or giving instructions to other laypersons, n (%) | ||
Performed | 11 (15.9) | 364 (31.7) |
Not performed | 49 (71.0) | 285 (24.8) |
Not performed because other people had already performed | – | 468 (40.7) |
Missing | 9 (13.0) | 33 (2.9) |
Making the emergency call, n (%) | ||
Performed | 48 (69.6) | 406 (35.3) |
Not performed | 14 (20.3) | 179 (15.6) |
Not performed because other people had already performed | – | 527 (45.8) |
Missing | 7 (10.1) | 38 (3.3) |
Looking for or carrying an AED, n (%) | ||
Performed | 1 (1.4) | 98 (8.5) |
Not performed | 57 (82.6) | 550 (47.8) |
Not performed because other people had already performed | – | 428 (37.2) |
Missing | 11 (15.9) | 74 (6.4) |
Chest compressions, n (%) | ||
Performed | 11 (15.9) | 69 (6.0) |
Not performed | 50 (72.5) | 702 (61.0) |
Not performed because other people had already performed | – | 326 (28.3) |
Missing | 8 (11.5) | 53 (4.6) |
Rescue breaths, n (%) | ||
Performed | 2 (2.9) | 23 (2.0) |
Not performed | 57 (82.6) | 762 (66.3) |
Not performed because other people had already performed | – | 305 (26.5) |
Missing | 10 (14.5) | 60 (5.2) |
Applying the AED pads and/or pushing the buttons of the AED, n (%) | ||
Performed | 2 (2.9) | 54 (4.7) |
Not performed | 57 (82.6) | 740 (64.3) |
Not performed because other people had already performed | – | 305 (26.5) |
Missing | 10 (14.5) | 51 (4.4) |
Performing any rescue action other than the above, n (%) | ||
Performed | 37 (53.6) | 448 (39.0) |
Not performed | 20 (29.0) | 280 (24.3) |
Not performed because other people had already performed | – | 264 (23.0) |
Missing | 12 (17.3) | 158 (13.7) |
Performing any rescue actions, n (%) | 66 (95.7) | 849 (73.8) |
AED, automated external defibrillator.
Table 3 shows the results of the subgroup analysis. If the layperson was a family member or acquaintance of the collapsed person, all laypersons in the single-bystander group and 83.9% in the multiple-bystanders group performed any rescue action. In the single-bystander group, the higher proportion of collapses occurred in “Residence” and the collapsed person was “Family”. In contrast, if the layperson was a stranger to the collapsed person, the proportion of laypersons who performed any rescue action was similar to that in the main analysis (92.3% in the single-bystander group and 70.3% in the multiple-bystander group). According to the location of the collapsed person, the proportion of “Other locations” was high and that of “Residence” was low. The proportion of males in the single-bystander group was higher than in the main analysis.
Subgroup Analysis: Outcomes and Characteristics of Laypersons Stratified by the Presence of Other People
Family or associate (n=315) | Stranger (n=898) | |||||||
---|---|---|---|---|---|---|---|---|
Single- bystander (n=29) |
Missing | Multiple- bystanders (n=286) |
Missing | Single- bystander (n=39) |
Missing | Multiple- bystanders (n=859) |
Missing | |
Characteristics | ||||||||
Sex, male, n (%) | 13 (44.8) | 0 (0.0) | 145 (50.7) | 0 (0.0) | 25 (64.1) | 0 (0.0) | 459 (53.4) | 1 (0.1) |
Age, years, median (IQR) |
41.0 (26.5–50.5) |
37.0 (24–48.3) |
39.0 (27.0–49.0) |
40.0 (28–49) |
||||
Previous experience of CPR training, n (%) |
15 (51.7) | 3 (10.3) | 173 (60.5) | 13 (4.5) | 22 (56.4) | 4 (10.3) | 531 (61.8) | 65 (7.6) |
Previous experience of performing any rescue action, n (%) |
15 (51.7) | 1 (3.4) | 126 (44.1) | 6 (2.1) | 21 (53.8) | 0 (0.0) | 445 (51.8) | 20 (2.3) |
Location of collapsed person, n (%) |
0 (0.0) | 6 (2.1) | 0 (0.0) | 17 (2.0) | ||||
Workplace | 2 (6.9) | 81 (28.3) | 7 (17.9) | 200 (23.3) | ||||
Public area | 0 (0.0) | 17 (5.9) | 4 (10.3) | 156 (18.2) | ||||
Residence | 25 (86.2) | 71 (24.8) | 1 (2.6) | 4 (0.5) | ||||
Educational institution |
0 (0.0) | 40 (14.0) | 0 (0.0) | 38 (4.4) | ||||
Sports facility | 0 (0.0) | 18 (6.3) | 0 (0.0) | 38 (4.4) | ||||
Other locations | 2 (6.9) | 53 (18.5) | 27 (69.2) | 406 (47.3) | ||||
Relationship with the collapsed person, n (%) |
0 (0.0) | 0 (0.0) | 1 (1.5) | 5 (0.5) | ||||
Family | 24 (82.8) | 80 (28.0) | – | – | ||||
Friend | 2 (6.9) | 55 (19.2) | – | – | ||||
Associate | 1 (3.4) | 53 (18.5) | – | – | ||||
Colleague | 2 (6.9) | 98 (34.3) | – | – | ||||
Others | – | – | 39 (100.0) | 859 (100.0) | ||||
Cause of the collapse, n (%) |
0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | ||||
Non-cardiac arrest | 22 (75.9) | 226 (79.0) | 26 (66.7) | 590 (68.7) | ||||
Cardiac arrest | 7 (24.1) | 27 (9.4) | 5 (12.8) | 84 (9.8) | ||||
Unknown | 0 (0.0) | 33 (11.5) | 8 (20.5) | 185 (21.5) | ||||
Collapsed person, n (%) | ||||||||
Male | 11 (37.9) | 1 (3.4) | 153 (53.5) | 2 (0.7) | 25 (64.1) | 1 (2.6) | 534 (62.2) | 24 (2.8) |
Adult | 27 (93.1) | 1 (3.4) | 241 (84.3) | 0 (0.0) | 36 (92.3) | 1 (1.4) | 781 (90.9) | 6 (0.7) |
Outcome | ||||||||
Performed any rescue action, n (%) |
29 (100.0) | 240 (83.9) | 36 (92.3) | 604 (70.3) |
CPR, cardiopulmonary resuscitation; IQR, interquartile range.
For all questions related to rescue actions, 150 laypersons (12.3%) answered “Not performed because other people had already performed”. The characteristics of the remaining laypersons, excluding the 150 who did not perform rescue actions, are shown in Supplementary Table 2. The proportion of laypersons who performed any rescue action was 97.1% (66/68) in the single-bystander group and 84.8% (849/1,001) in the multiple-bystanders group, showing a similar trend to the main analysis (Supplementary Table 2).
In the results of the exploratory analyses, the odds ratios for the presence of other people and rescue actions performed by laypersons are outlined in Supplementary Table 3.
In this study, we assessed the relationship between the presence of other people and the performance of any rescue action in emergency situations where a person has collapsed; we found that the proportion of laypersons who performed any rescue action was lower when other people were present compared with when they were absent. Moreover, this result was consistent with that of the subgroup analysis in cases where the layperson was a stranger to the collapsed person and that of analysis for laypersons excluding those who did not perform any rescue actions because other people had already performed them.
Some resuscitation-related studies have focused on the relationship between the presence of other people and performing CPR. Nishi et al. reported that the presence of multiple rescuers was associated with a higher proportion of CPR performance.14 A study by Langlais et al. analyzed recordings of telephone CPR and reported that chest compressions were more likely to be initiated when multiple bystanders were present.9 Compared with those studies, we found a different tendency, the main reason for which may be the difference in study design. The previous studies examined whether rescuers/bystanders performed CPR when other people were present in an emergency situation where some intervention (e.g., emergency call) had already been performed.9,14 In other words, in situations where the first intervention, such as approaching a collapsed person and checking their consciousness, has already been initiated, the presence of other people is likely to affect the performance of chest compression or CPR. In this study, we focused on whether the presence of other people influenced the performance of any rescue action regardless of the performance of CPR. Moreover, in our previous study, we found that one of the psychological barriers to performing rescue actions was “the fear of approaching a collapsed person”.13 In other words, not performing any rescue action means not approaching a collapsed person, or checking for consciousness, thus indicating that no action was taken. This suggests that taking the initiative to perform any rescue action may be affected by the presence of others. Furthermore, this tendency is supported by our results in which the proportion of laypersons “checking patient’s consciousness,” (which is the first rescue action), was greater (87.0%) in the single-bystander group than in the multiple-bystanders group (61.0%).
In the subgroup analysis, when the layperson was a stranger to the collapsed person, the result was consistent with that of the main analysis. However, when the layperson was a family member or acquaintance of the collapsed person, all laypersons performed rescue actions when other people were absent. Generally, a family member is less likely to perform bystander “CPR” compared with a non-family member.19 However, in this study, the proportion of any rescue action was higher when the layperson was a family member or acquaintance of the collapsed person than when they were strangers, regardless of the presence of other people. This is a natural result, because the primary outcome of this study included all rescue actions regardless of the content, and it is difficult to imagine that laypersons would not perform any rescue action when they are alone. In the case of the layperson being a stranger to the collapsed person, nearly 70% of those in the single-bystander group encountered the collapsed person in “Other locations”, whereas a relatively large number of those in the multiple-bystanders group encountered them in the “Workplace” or a “Public area”. Although workplace and public areas are environments where laypersons are more likely to ask for help and cooperation from others, the proportion of laypersons who performed any rescue action was lower in the multiple-bystanders group than in the single-bystander group. It is likely that a certain number of laypersons did not perform any rescue action because they thought that someone else would do it.
In the field of psychology, studies have established that the presence of other people will affect the performance of rescue action in emergency situations, a phenomenon known as the “bystander effect,”17,18 and which may have occurred in this study. An American Heart Association Scientific Statement has also discussed the “bystander effect” and pointed out the importance of utilizing knowledge from other disciplines, such as psychology, in assisting rescuers.20 Therefore, in order to encourage laypersons to perform any rescue action in emergency situations, it will be necessary not only to teach CPR skills and impart the required knowledge but also to provide information about the existence of the “bystander effect” through CPR training. Because previous studies have reported that the performance of rescue actions, including CPR, is associated with psychological barriers, it is necessary to create CPR training that incorporates psychological knowledge.
Study LimitationsFirst, this study could not identify whether the presence or absence of other people changed during the rescue process. We fully recognize that changes in the situation may influence layperson rescue actions, and we understand that this is also a point of high interest to readers. Additionally, the relationship between a layperson and “other people” was unclear, which may also influence the layperson’s rescue actions. Therefore, we have chosen to describe the results carefully in this study. Further study is needed to assess how people will perform any rescue action in relation to a collapsed person when there are multiple laypersons over time. Second, most respondents were participants in CPR training; therefore, this study population may have included laypersons interested in learning CPR, which might have influenced the results. However, even laypersons with a high awareness of rescue actions, such as those who participate in CPR training, may not necessarily take any action when others are present. Future research is needed to clarify the rescue response among people with lower awareness of rescue actions. Third, we were not able to distinguish between “they did not perform because the collapsed person did not need any rescue actions” and “they did not perform even though the collapsed person needed rescue action”. We did not set any questions that would distinguish between these situations because it is difficult for laypersons to distinguish between these situations. Moreover, we were also unable to set a question that would distinguish between “they did not perform because they felt that someone else would perform” and “they did not perform because other people had already performed”. However, these could be reasons held simultaneously, making it difficult to pick out just one as the cause. Fourth, we collected data on the laypersons’ rescue actions using a self-administered questionnaire, so recall of emergency experiences could be inaccurate. Fifth, rescue actions are recognized as socially desirable, so social desirability bias might have influenced our results. Finaly, the presence of others is one of the factors, both negative and positive, that influence a layperson’s performance in an emergency situation. Further studies are needed to assess the relationships between each factor including the presence of others.
The relationship between the presence of other people and the performance of rescue actions during emergencies was assessed. The proportion of laypersons who performed any rescue action was lower when other people were present compared with when they were absent. The presence of others might influence the performance of laypersons’ rescue actions. It may be beneficial to provide information and promote understanding of the psychological “bystander effect” through CPR training to help laypersons feel more confident in performing any rescue action in emergency situations.
We are deeply grateful to the people from the following institutions that cooperated with the conducting of this study: the Japanese Red Cross Society, Osaka Municipal Fire Department, Takasaki City Large Area Fire Bureau, Sakai City Fire Bureau, Kishiwada City Fire Department Headquarters, Osaka Life Support Association, Ibaraki PUSH, Kyoto University, and Dai-ichi Life Insurance Company Limited. We are also deeply grateful to the following experts who participated in the Delphi designing of the questionnaires: Dr. Osamu Yamaoka, Dr. Junichi Izawa, Dr. Tasuku Matsuyma, Dr. Takeyuki Kiguchi, Mr. Keiji Akatsuka, Mr. Akira Tago, Mr. Shingo Moriguchi, Ms. Makiko Sano, Ms. Miyuki Inoko, Ms. Izumi Chida, and the following laypersons who assessed the draft questionnaire: Mr. Hiroshi Kiribuchi, Mr. Kazuhiro Takahashi, Mr. Hiroshi Yamaguchi, Ms. Yasuko Matsuyama, Mr. Toshio Maeshige, Ms. Nao Maeshige, Ms. Mari Aida, and Mr. Shinsuke Wada.
The authors declare that they have no conflicts of interest.
This work was supported by a Grant-in-Aid for Health and Labour Sciences Research Grants (H29-Junkankitou-Ippan-009; 23FA1023) from the Japanese Ministry of Health, Labour and Welfare.
This study was approved by the Ethics Committees of the Kyoto University Graduate School of Medicine (R1393).
The deidentified participant data will not be shared.
Please find supplementary file(s);
https://doi.org/10.1253/circrep.CR-24-0166