Background

In the UK it is estimated that around 60,000 out of hospital cardiac arrests (OHCA) occur each year [1]. For the estimated 30,000 arrests where emergency medical services (EMS) initiate resuscitation, rate of survival is still low at around 9% [2]. However, there is a significant body of evidence showing that bystander cardiopulmonary resuscitation (CPR) significantly improves outcomes after OHCA [3, 4]. Moreover, rapid defibrillation by bystanders using an automated external defibrillator (AED) has shown improved survival rate to hospital discharge [5,6,7] and discharge with favourable functional outcome [7], as well as being a cost-effective OHCA intervention [6]. Moreover, studies have shown that the earlier the defibrillation following an OHCAthe better the survival outcomes, most notably within the first 3 min after collapse [8], and each 1-min delay of first defibrillation has been associated with a reduction in the likelihood of good neurological recovery in patients with OHCA who presented with ventricular fibrillation [9].

Although it is well documented that defibrillation by a member of the public improves outcomes following an OHCA, access to, and knowledge of AEDs is an issue and public access defibrillation is only used in 0.2–4.3% of all OHCA [10]. Reported bystander defibrillation use has seen little change with time, with some studies showing only 1.1% increase in bystander defibrillation in a 9 year period [11].

Previous research investigating public knowledge of bystander intervention has been conducted in a number of countries and has shown general low understanding. In an Australian telephone survey, around half (50.5%) of participants stated they would give chest compressions during an cardiac arrest (CA) scenario [12], and a UK based survey study recorded lower levels, with 41.4% of participants spontaneously stating they would perform CPR during an OHCA [13], while an Austrian questionnaire study found 33% of participants would be willing to perform CPR [14]. In terms of AED intervention, even lower public confidence has been found with only 5.3% of participants in a Taiwanese telephone interview study stated they had the confidence to use an AED [15]. In a US study, 2.2% of respondents of a convenience survey spontaneously mentioned defibrillation when questioned on their attitudes to a CA [16], and similar results were found in the UK with only 2.1% of participants stating they would apply and use a PAD [13]. Additionally, common misconceptions of CPR and AED knowledge were seen in Saudi Arabia and 92.3% of participants were unsure on how to use an AED correctly [17].

Barriers to bystander AED use have also been investigated, the initial fear of using the AED incorrectly and legal liability were stated by 57% and 28% of respondents respectively in a survey in a US shopping mall [18]. Furthermore, a UK study has shown more than 60% of community PADs lack signage [19], likely exacerbating low public knowledge.

Public knowledge can be increased by education, and studies have found training in AED use [18, 20] and Basic Life Support (BLS) [20, 21] increases the likelihood of intervention by bystanders. Furthermore, a meta-analysis showed community interventions, such as training the lay public in CPR, were associated with greater bystander CPR rate (OR, 1.28 and 95% CI, 1.06, 1.54) and increased survival to discharge or 30-day survival (OR 1.34 and 95% CI 1.14, 1.57) [22].

In 2013 the UK government’s Cardiovascular Disease Outcomes Strategy for England set a target of increasing survival from OHCA by 50% giving 1000 additional lives saved each year [23], and improvements in patient outcomes following an OHCA is part of the National Health Service (NHS) long term plan published in 2019 [24], as well as there being international efforts to improve OHCA outcomes [25]. For this to be supported, increasing the public knowledge and confidence of public access AED (PAD) is essential.

Public knowledge and confidence of PAD has been previously explored in Southampton, in the South of England [13], however, it is unknown what the beliefs are of the members of public in the North of England. This is of interest as there are documented differences, such as areas of deprivation [26] and population demographics [27, 28] between the north and south of England, and bystander resuscitation varies by area in England [29]. The overall aim of this study was to ascertain the public’s level of knowledge and awareness of CA and PAD, the perceived barriers to helping during an OHCA, and the experience of AED use in the north of England. We also aimed to provide educational information on OHCA and AED use to help improve understanding and attitudes to the use of an AED to address the barriers to responding to an OHCA and improve the rate of bystander AED use.

Methods

Setting

This study used a face-to-face cross-sectional survey method to collect data. The study was performed on Gosforth High Street in Newcastle-upon-Tyne, in the North East of England. Data collection took place from January and August 2018 between 11:00 and 18:00 on both weekdays and weekends. Gosforth High Street is approximately 1.2km long [30], with a variety of amenities including shops, restaurants, pubs, a gym, a church and businesses on either side of a road. In 2018, Gosforth High Street had two PADs within approximately 200 m of each other. Ethical approval was granted by the University of Sunderland Research Ethics Committee.

Participants and sampling

In 2011, the resident population of Gosforth was 20,136 [31]. From a sample size calculation considering the total population size of the area (20,136), the required confidence interval (95%), the margin of error (5%) and z-score (1.96), to support an accurate reflection of the total population [32], this study aimed to recruit a minimum convenience sample of 377 members of the public. A lower age limit was not set since there are endorsements by the World Health Organisation in training school children in CPR [33, 34], and children under 16 were questioned if they volunteered to participate and there was consent from an adult guardian.

The research team stood approximately in the middle of Gosforth High Street. Interviewers either wore full paramedic outfits or wore a jacket with the North Ambulance Service (NEAS) logo on and were accompanied by a parked NEAS Cardiac Arrest Response Unit (CARU) car to distinguish the research team. Members of the public were approached by researchers and invited to take part in a questionnaire. The interviewers were non-discriminatory in their choice of participants and the closest member of the public was invited to participate. No participant exclusion criteria were used, and all members of the public on Gosforth High Street at the time of the study session were considered potential participants.

Questionnaire

Members of public that agreed to participate were presented with a questionnaire consisting of nine questions, and was based on the semi-structured questionnaire used by Brooks et al. [13] which included questions regarding BLS, bystander CPR, AED knowledge and first aid training. In our study we separated question 2 from Brooks et al.’s questionnaire into 2 different questions, and participants were asked separately if they knew what an AED was, and where they might find one. Additionally, our questionnaire also investigated whether a participant knew the difference between a CA and a heart attack, whether there was anything that would prevent them from helping in a situation when someone had collapsed following a presumed CA, if they knew where the closet AED was, and how often they visit Gosforth High Street. A copy of our questionnaire is available (see additional file 1).

Participants were required to spontaneously suggest answers to the survey questions and were not prompted. Paper based questionnaires were used and participant answers were marked in pen on the questionnaire by the researchers. Since this study sought to help improve knowledge, the researchers provided informative answers to participants if they do not know the answer to the survey questions with the aim to help improve the public’s knowledge.

Data analysis

Questionnaire answers were collated and entered into a Microsoft Excel spreadsheet. Results were analysed using SPSS Statistics for Windows®, Version 24 (IBM, Armonk, NY: IBM Corp). Questionnaire answers are reported as count and percentage, and continuous data are presented as median and range. Chi-Square analysis was used to test the association of the independent variables sex (male/female) and first aid trained (yes/no) on the participants’ responses. Answers to question 3 “Is there anything that might prevent you from the above/helping?” were coded and grouped into categories and overarching themes. Content analyses were used to calculate frequencies of themes [35].

Results

Participants

A total of 421 participants were recruited to our study across 10 study sessions; the average number of participants sampled per session was 34. The median participant age was 47 years (range 10–92 years) and 52.3% (n = 220/421) of the participants were female (Table 1). The majority of participants visited Gosforth High Street weekly or more often (72.9%, n = 307/421). Information on the number of members of public declining participation or withdrawing the questioning was not recorded.

Table 1 Participant characteristics
Full size table

First aid training

Just over half of the respondents claimed to have first aid training (50.8%, n = 214/421), with significantly more females than males being first aiders (57.3% n = 126/220 females and 43.9% N = 87/198 males, p = 0.01 χ2 = 7.41, Table 2). The location or source of first aid training and the length of time since training are shown in Table 3. Of the participants who could recall the location of their first aid training, training received from a charity and through work were the most frequently mentioned by participants (14.0%, n = 30/214 and 11.7%, n = 25/214 respectively). Nearly half of participants reported they received first aid training more than 3 years ago (48.1% n = 103/214), while 20.6% (n = 44/214) had received first aid training within the past year.

Table 2 Cardiac Arrest Knowledge, N (%)
Full size table
Table 3 Location or Source of First Aid Training and Length of Time since First Aid Training
Full size table

Cardiac arrest knowledge

Out of all the respondents, 30.9% (n = 130/421) identified the difference between a heart attack and a CA (Table 2), and little difference between sex was seen (29.8%, n = 59/198 of males and 32.4%, n = 71/220 of females, p = 0.56 χ2 = 0.33). Significantly more first aiders were aware of the difference between a heart attack and a CA than non-first aiders by a relative percentage difference of 70.2% and an absolute percentage difference of 16.0% (38.8% n = 83/214 versus 22.8% n = 47/207 respectively, p < 0.001 χ2 = 12.53, Table 2).

Response to a collapse following an out of hospital cardiac arrest

Most respondents (82.9%, n = 349/421) reported that they would know what to do if someone were to collapse in front of them with a presumed CA (Table 2). Moreover, 91.6% (n = 196/214) of first aiders compared to 73.9% (n = 153/207) of non-first aiders stated they would know what to do (p < 0.001 χ2 = 23.19) and little difference was seen between sex (80.3%, n = 159/198 of males versus 85.5%, n = 188/220 of females, p = 0.16 χ2 = 1.96).

When asked what they would specifically do in response to a collapse following a CA, most participants spontaneously reported that they would call 999 (64.1% n = 270/421) and commence CPR (58.2%, n = 245/421), while 24.0% (n = 101/421), 17.6% (n = 74/421) and 14.3% (n = 60/421) reported they would check breathing, shout for help and locate a defibrillator respectively (Table 2). Only 9.7% (n = 41/421) participants reported they would check for a response, while 4.5% (n = 19/421) claimed they would apply the defibrillator, and 3.3% (n = 14/421) reported they would continue CPR until the ambulance arrives. Compression ratio or rate was rarely spontaneously mentioned by participants, and only 3.1% (n = 13/421) and 1.4% (n = 6/421) claimed they would use a compression/breath ratio of 30:2 or a compression rate of 100–120/min respectively.

Out of the actions for cardiac arrest mentioned by participants, little difference was seen between sex, except an absolute difference of 10% more females than males reported they would call 999 (59.1% n = 117/198 of males versus 69.1% n = 152/220 of females, p = 0.03 χ2=4.54, Table 2).

A greater proportion of first aiders mentioned actions for cardiac arrest than non-first aiders for all actions (Table 2). Additionally, statistically significant differences were seen between first aiders and non-first aiders for shout for help (22.0%, n = 47/214 versus 13.0% N = 27/207, p = 0.02 χ2 = 5.78), check for response (15.4%, n = 33/214 versus 3.9% n = 8/207, p < 0.001 χ2 = 15.99), check breathing (31.8%, n = 68/214 versus 15.9%, n = 33/207, p < 0.001 χ2 = 14.47), commence CPR (68.2%, N = 146/214 versus 47.8%, n = 99/207, p < 0.001 χ2 = 18.00), locate defibrillator (20.6%, n = 44/214 versus 7.7%, n = 16/207, p < 0.001 χ2 = 14.18), and apply defibrillator to patient and use if appropriate (7.0%, n = 15/214 versus 1.9%, n = 4/207, p = 0.01 χ2 = 6.29).

Factors preventing intervention during an out of hospital cardiac arrest

The barriers to helping during a CA stated by participants are shown in Table 4. Almost a third of participants (29.9%, n = 126/421) remarked that there were barriers preventing them from helping, and little difference was seen between males and females (27.3%, n = 54/198 of males and 32.7%, n = 72/220 of females, p = 0.23 χ2 = 1.47 Table 2). More first aiders (40.7%, n = 87/214) commented that there were barriers preventing them from helping than non-first aiders (18.8%, n = 39/207) by an absolute percentage difference of 21.9% (p < 0.001 χ2 = 23.87) (relative percentage difference of 114.9%) (Table 2). The most common barrier stated by 42.1% (n = 53/126) of our participants was lack of knowledge (Table 4).

Table 4 Perceived Barriers to Helping during an OHCA
Full size table

AED and AED location

The majority of participants knew what a defibrillator was (80.3%, n = 338/421, Table 5), with significantly more first aiders than non-first aiders accurately describing an AED (91.1%, n = 195/214 versus 69.1%, n = 143/207 p < 0.001 χ2 = 32.29). Similar findings were found amongst females and males for AED knowledge (Table 5).

Table 5 Self-reported AED knowledge
Full size table

At 74.3% (n = 313/421), most respondents knew where defibrillators were generally located (Table 5), with more first aiders (81.8%, n = 175/214) than non-first aiders (66.7%, n = 138/207) describing AED locations (p < 0.001 χ2 = 12.59). Statistically more females than males were aware of AED locations by an absolute percentage difference of 10.3% (79.5%, n = 175/220 of females and 69.2%, n = 137/198 of males, p = 0.02 χ2 = 5.90). The most common locations to find an AED suggested by participants were sport centres (22.7%, n = 71/313) and hospitals (21.4%, n = 67/313) (Table 6).

Table 6 Suggested AED locations and knowledge of AED location based on frequency of Gosforth High Street visit
Full size table

Of all respondents, only 20.9% (n = 88/421) recounted that they knew where the closest defibrillator was, with small differences observed between sex (17.7%, n = 35/421 of males and 24.1%, n = 53/421 of females). Significant difference, however, was seen amongst first aiders and non-first aiders with more first aiders recounting the location of the nearest AED (28.5%, n = 61/421 of first aiders versus 13.0%, n = 27/421 non-first aiders, relative percentage difference of 119.2%).

A higher proportion of respondents who reported they visited Gosforth High Street weekly or more often knew where the closest AED was compared to those who visited less than weekly (25.7%, n = 79/421 for ≥ weekly and 10.2% n = 9/421 for < weekly, p < 0.001 χ2 = 16.00, Table 6).

Knowledge of AED operation

Out of all the respondents, 34.7% (n = 146/421) stated they knew how to operate an AED, and no statistical difference was observed between sex for any AED action (Table 5).

Of all participants, 16.2% (n = 68/421) freely recounted they would open the AED and 25.2% (n = 106/421) of participants claimed they would listen and follow instructions. 19.7% (n = 83/421) of respondents recounted they would apply adhesive pads in the correct area on chest, and 5.7% (n = 24/421) stated they would await analysis of rhythm (Table 5). Additionally, 9.5% (n = 40/421) and 11.9% (n = 50/421) of respondents spontaneously claimed they were aware to be clear of the patient before shocking, and to shock the patient if advised. Only 3.8% (n = 16/421), 1.7% (n = 7/421) and 1.9% (n = 8/421) of participants stated they would continue CPR should shock not be advised/successful, call 999 and follow instructions from 999, respectively (Table 5).

All AED actions were statistically different between first aiders and non-first aiders, with a larger proportion of first aiders having a greater knowledge of AED use than non-first aiders (Table 5). This was most apparent for self-reported defibrillator knowledge (51.9%, n = 111/214 versus 16.9%, n = 35/207 of first aiders and non-first aider respectively, p < 0.001 χ2 = 56.77), listen and follow instructions by 26.5% (38.3%, n = 82/214 versus 11.8%, n = 24/207 of first aider and non-first aider respectively, p < 0.001 χ2 = 40.83) and apply adhesive pads in the correct area on chest by 20.6% (29.9%, n = 64/214 versus 9.3%, n = 19/207 of first aider and non-first aider respectively, p < 0.001 χ2 = 29.19).

More first aiders recounted that they would open the AED than non-first aiders (23.8%, n = 51/214 versus 8.3% n = 17/207 respectively, p < 0.001 χ2 = 19.38, Table 5), were aware people should be clear of patient before shocking (17.2%, n = 36 versus 2.0% n = 4 respectively, p < 0.001 χ2 = 27.50), and freely mentioned they would shock a patient (21.1%, n = 44/214 versus 2.9%, n = 6/207 respectively, p < 0.001 χ2 = 31.82). Moreover, the AED actions which were lesser mentioned, including await analysis of rhythm, continue CPR should shock not be advised/successful, and follow instructions from 999, were infrequently mentioned by non-first aiders (≤ 0.5%), with no non-first aiders mentioning that they would call 999. Although more first aiders claimed they would perform these AED actions, low counts were still observed for spontaneous mention of call 999 and follow instructions from 999 (both 3.3%, n = 7/214). However, a much higher relative difference was seen between first aiders and non-first aiders for await analysis of rhythm (11.0%, n = 23/214 of first aiders versus 0.5%, n = 11/207 of non-first aiders, p < 0.001 χ2 = 20.85).

Discussion

Summary of main findings

This study investigated the knowledge and confidence of the public in intervening during an OHCA. Although most participants stated they would know what to do during an OHCA and knew what an AED was, when questioned further, low knowledge was recorded, and generally low numbers of participants spontaneously mentioned specific bystander OHCA responses and AED actions. Moreover, over half of our participants stated that lack of knowledge was a barrier to helping during an OHCA and first aid training was found to statistically increase knowledge of OHCA responses and AED actions.

Comparison with other literature

In comparison to a similar survey study by Brooks et al., conducted in South West England, similarities in the percentage of participants reporting that they would know what to do if someone were to collapse in front of them with a presumed CA, the CA actions call 999 and 30:2 compression ratio were found [13]. Similar numbers have also been reported for those who stated they would commence CPR in an Australian survey study [12], however, this study reported that women were less likely than men to state that they would give chest compressions [12], moreover, females reported a lower willingness to initiate basic life support (BLS) attempts and to use an AED device in a Austrian survey study [14]. In our study, out of most questions asked, little difference was seen between sex, except for statistically more females than males reported they would call 999, and more females than males were aware of AED locations. This may reflect the statistically higher levels of first aid training in females in our study cohort, whilst the above studies did not compare first aid training between sexes. Although there was mostly little difference between our male and female responses, the higher proportion of first aid training within the female cohort suggests that initiatives to increase recruitment of males to first aid training may be needed in the North East of England.

Lower numbers of first aiders who spontaneously mentioned check for response and check breathing were recorded in our study compared to the study by Brooks et al. [13] (15.4% versus 50% and 31.8% versus 58% respectively). Moreover, only 5.2% of Brooks et al.’s participants spontaneously recounted that they would locate a defibrillator while 20.6% of our participants did. Overall, although both conducted in England, the discrepancies noted above between studies may have occurred due to differences in locations, populations and data collection year used in the studies. Moreover, variation in bystander responses has been reported in the literature, for example, a systematic review of surveys and qualitative interviews concluded that overall awareness of the purpose of an AED ranged between 15 and 89% [10], which accounts for the variation seen between our participants (80.3%) and those from Brooks et al.’s study (69%) who stated they knew what a defibrillator was.

Another discrepancy between our results and those published included an Austrian telephone survey study which reported more than two-fold more participants would check for breathing compared to our study (52% versus 24% respectively) [14], however, significantly higher levels of our participants self-reported that they would commence CPR (58% versus 33%). Again, these differences could be attributable to the variances between survey study stated above, and specifically the questionnaire used with different phrases and 4 different answer options provided to the participant by the researchers.

Similar to our findings, research has also found more first aiders self-reported actions for cardiac arrest [12, 13] and AED actions [13] than non-first aiders. Significantly more females than males were first aiders in our study, with similar findings in other studies from the UK [20] and Australia [36, 37], but not in Singapore [38], additionally, more females than males were aware of AED locations.

Of all our respondents, 74.3% could name potential AED locations, however, only 20.9% recounted that they knew where the closest defibrillator was, which is at the top end of the range of 5–22% of people who collectively reported they were able to locate their nearest PAD in a systematic review [10]. This is considerably higher than the 5.1% of people who knew where or how to find their nearest PAD in the study by Brooks et al. [13].

More CA actions and all AED actions were reported statistically more by first aiders than non-first aiders in our study, and more first aiders recounted the location of the nearest AED than first aiders. This aligns with the literature where previous training in CPR and/or AED use is associated with improved CPR psychomotor skills [39], increased likelihood of performing CA actions [40], increased AED and PAD location knowledge [12] and confidence in AED use [12, 16].

A recent systematic review and meta-analysis of patients with OHCA found rates of bystander CPR and survival to discharge were lower for patients from more deprived communities than those from higher socioeconomic status (SES) communities [41]. Since Gosforth is an affluent area of Newcastle, a relatively high proportion of bystander intervention could be assumed. Moreover, in 2011 it was found that patients in the least deprived areas were more likely to receive bystander CPR than those from the most deprived areas (23.3% versus 14.5%) in the North East of England [42]. These proportions are lower than the findings from our study (58.2% freely stated they would commence CPR) and may be due to the discrepancies between participant responses to a survey compared to actual bystander intervention, but may also indicate an increase in knowledge and awareness over time within the region.

Almost a third of our participants remarked that there were barriers preventing them from helping during a collapse following a CA with the most common barrier stated by these participants being lack of knowledge. Fear of not having the skills and/or causing harm are also the most common barriers previously mentioned for CA actions [12, 15, 38] and PAD use [10, 15, 38] in the literature, which supports the need of training to enhance knowledge and increase bystander confidence.

Limitations and future implications

A main limitation of this study is self-selection of participants, reducing the representativeness of our findings. Furthermore, answers to the survey questionnaire may not represent what an individual would do under pressure at a time of a OHCA collapse, and we did not record the participant response rate. Gosforth is an affluent area, and similar research in deprived areas of Newcastle and the UK would provide deeper insights. However, 10.0% (N = 42) of participants stated they only visit Gosforth once a year, and our findings still show low knowledge within a potentially affluent participant sample.

Since disparities in intervention, healthcare and outcomes have been seen in deprived communities and ethnic minorities [43], further exploration of the demographic factors such as ethnicity, education level and employment [10] in the UK and their role on AED coverage, knowledge, training and confidence in their use would provide deeper insights. Moreover, data were collected in 2018 and so may be somewhat dated, and investigation into the affect the COVID-19 pandemic [44] and the novel developments in OHCA intervention such as smartphone-based use [45], or drone delivery of AEDs [46,47,48] on bystander response, warrants further research. Notwithstanding the above limitations, this survey study used a suitable sample size and provides further understanding into public knowledge and confidence in the use of AEDs, the barriers to helping during a CA and acquiring an AED on a public high street in the North East of England.

Conclusion

Although most participants reported they would know what to do during an OHCA and what an AED was, low numbers of participants spontaneously mentioned OHCA responses and AED actions, suggesting low knowledge. The most frequent OHCA actions mentioned were call 999 and commence CPR. Just over half of participants were first aid trained, with more females being first aiders. Only around a fifth of participants knew where the closest AED was.

Knowledge of CA and AED actions, and location of the nearest AED are barriers to public OHCA intervention. Our results show that first aid training improves knowledge in helping during an OHCA collapse. Increasing knowledge and confidence through campaigns such as the annual Restart a Heart initiative [49] will help address the low public involvement and under use of PAD during OHCAs, and improve OHCA outcomes.