Attitudes Towards Cardiopulmonary Resuscitation Situations and the Associations of Possible Affecting Factors – a Survey Among 3,602 in-hospital Healthcare Professionals

Background Attitudes among in-hospital healthcare professionals (HCPs) towards cardiopulmonary resuscitation (CPR) are poorly understood. The aim of this study was therefore to evaluate attitudes among in-hospital HCPs towards CPR and furthermore to assess the associations of possible affecting factors. Methods A questionnaire was distributed to n=6,055 HCPs. The associations of affecting factors were analysed using binary logistic regression.


Introduction
The competence and quality of skills of healthcare professionals (HCPs) regarding cardiopulmonary resuscitation (CPR) has previously been reported to be overestimated by HCPs and inadequate by observation [1][2][3]. Attitude is, together with knowledge and skills, seen as an essential component of competence [4]. The attitude among HCPs toward performing CPR is poorly understood, and it is furthermore unclear if there are factors possibly affecting attitudes.
The behaviour in a speci c situation is affected by intention and perceived behavior control. Intention is determined by behavior beliefs, subjective norm (the perceived social pressure to perform), and by the degree of perceived behavioral control [5]. Attitude consists of positive or negative feelings, beliefs and behavioural information about a speci c object [6]. Attitude have been found to be the strongest predictor regarding the intention to perform CPR among college students [7]. A cardiac arrest situation can be associated with stress [8,9], even in simulated situations [10]. Nurses have reported feelings of hesitation [11,12], anxiety [11,13] and fear of harming the patient [11]. Junior physicians have reported feeling stressed due to concerns that CPR felt inappropriate, a lack of CPR training, or to the situation itself [9]. CPR training have been proven to give a feeling of preparedness, decrease stress [8], increase positive attitude [14] and increase con dence [12]. After completion of a CPR training course, Swedish nurses and assistant nurses have shown decreased levels of anxiety and reduced fear of contracting an illness [15].
Advanced CPR training, attended by physicians and nurses on monitored wards, have shown to increase con dence [16]. Previous CPR experience in real life have been associated with positive attitudes towards CPR [14]. The attitude among HCPs toward CPR might affect their behavior in a cardiac arrest situation and attitudes are therefore important to evaluate. Performing CPR in a cardiac arrest situation is a team effort, and the attitude among all professions involved in the situation is important. Previous research has often derived from smaller samples and have often focused on speci c professions. There is a lack of a large sample study among all in-hospital professions regarding the attitude towards CPR including an assessment regarding possible associated factors.

Aim
The aim of this study was to evaluate attitudes among in-hospital healthcare professionals towards cardiopulmonary resuscitation and furthermore to assess the associations of possible affecting factors.

Design and data collection
This study was based on cross-sectional questionnaire surveys in two regions in Sweden [15,17] at ve secondary-care hospitals during the year 2008 and 2015-16. A total of n = 6,055 active-duty HCPs involved in patient contact were invited to participate (Additional le 1, owchart of participants). The HCPs consisted of physicians, nurses (including midwifes and radiology nurses), nursing assistants (including caretakers and dental nurses) and other university-educated staff. Participation was voluntary, anonymous in one region and con dential in the other. HCPs were informed of the study at staff meetings or by e-mail. The questionnaires were distributed either at staff meetings or to mailboxes on the wards in one region and directly to participating individual in the other region. Reminders were sent by email.

The questionnaire
Twenty questions were selected from a validated questionnaire [15,18]. Participants in both surveys in 2008 and 2015-16 responded to six questions concerning attitude towards being required to perform CPR, (Table 1), together with three questions concerning professional occupation, number of years in profession, and time since last CPR training course. In addition, participants responding to the survey in 2008 also responded to additional three questions regarding previous real-life CPR experience: experience of performing CPR in a cardiac arrest situation [yes/no], number of previous real-life CPR performances [one time to more than 30 times] and time since the last real-life CPR performance (last month to 24 months ago or more). Participants who reported having previous real-life CPR experience responded to eight additional questions concerning attitude in connection to their latest cardiac arrest situation ( Table 2). Participant's workplaces were noted during collection of the questionnaires. All questions were constructed as multiple-choice questions.

Statistics
The results from the attitude questions are presented descriptively. Differences regarding attitudes between groups of profession were analysed using Pearson's chi squared test, and Fischer's exact test were used when expected counts in one cell were ≤ 5. The associations of possible affecting factors on attitudes were analysed using binary logistic regression. Each of the eight questions concerning attitude in connection to the latest cardiac arrest situation (yes/no) served as a dependent variable, where odds ratio for the response "yes" were estimated. Independent variables were number of previous real-life CPR performances (1-3 times, 4-10 times and more than 10 times), time since last real-life CPR performance (24 months ago or more, 4-23 months ago and 1-3 months ago), workplace (monitored versus nonmonitored ward), profession (physicians, nurses, assistant nurses, and other university educated staff), time since last CPR training course (0-6 months ago, 7-11 months ago and 12 months ago or more), and number of years in profession (from 0-46 years). Each independent variable was regressed with the dependent variables in a crude model, one at a time, and only independent variables which were found associated (at p < 0.1) were included in a multiple model. Again, in the multiple models, independent variables that were considered not to be associated (p > 0.1) were excluded. In the nal multiple model, independent variables that were considered to be associated with the dependent variable at a level of p < 0.05 were considered statistically signi cant. Data were analysed using IBM Statistical Package for the Social Sciences 27.

Results
The response rate was 59% (n = 3,602). Mean work experience was 18.5 (SD 11.2) years and 39% (n = 1,392) of HCPs had attended CPR training during the last year. Twenty-two percent (n = 777) worked on monitored wards. A total of 45% (n = 967/2154) of the participants in one region had previous real-life CPR experience, with a higher proportion found on monitored wards (44% vs. 22%). Forty-two percent (n = 401/967) of those with previous experience had performed CPR up to three times and 47% (n = 457/967) had performed CPR more than two years ago (Table 3).  4 out of a total n = 2154 in one region, missing n = 2, 5 missing n = 37, 6 missing n = 35.
Attitudes towards cardiopulmonary resuscitation  A worry to contract an illness during CPR, feelings of discomfort in initiating CPR and feelings of failure after CPR performance were not possible to analyse in multiple models due to few observations after exclusion of outliers. However, in a crude model, there was an association regarding time since last CPR training course and a worry to contract an illness, where CPR training 7-11 months ago was associated with higher odds compared to 0-6 months ago. Also, in a crude model, ≥ 4 months since last real-life CPR performance and ≥ 4 previous real-life CPR performances were associated with lower odds of feeling discomfort in initiating CPR and working on a non-monitored ward meant higher odds. Furthermore, crude models showed higher odds regarding feelings of failure among HCPs working in nonmonitored wards, and lower odds regarding feelings of failure with increased real-life CPR experience.

Discussion
The main ndings in this study showed that a majority of HCPs felt con dent in their CPR knowledge, would know what to do and would be able to take command if necessary in a cardiac arrest situation. A third of HCPs reported being worried about making mistakes or to cause complications. Stress and anxiety were common in connection to the latest cardiac arrest situation. A short time since the last reallife CPR performance and a high number of previous real-life CPR performances were associated with lower odds of a worry about making mistakes/cause complications, lower odds of feeling stressed or anxious, and higher odds of feeling calm. Despite previous real-life CPR experience, there were differences in attitudes between groups of profession, and between monitored and non-monitored wards. Time since last CPR training course was associated with anxiety.
The associations regarding previous real-life CPR experience and attitudes remained signi cant in all applicable multiple models. A positive effect of real-life CPR experience on attitudes [14] and con dence [16] have been seen previously. Exposure to a certain object can create a more positive attitude toward the object. This means that the more familiar one is with an object the more likely it is to have a positive attitude towards it [6]. Also, previous experience with a behavior affects behavioral control and perceived behavioral control can mediate planned behavior [5]. Considering this, it seems that increased experience of real-life CPR performance in cardiac arrest situations would increase positive attitudes towards CPR and increase perceived behavioral control of performing CPR.
The results in the present study showed differences regarding attitude between professions and between HCPs on monitored and non-monitored wards, despite of previous real-life CPR experience. This could be due to prede ned roles and responsibilities in the cardiac arrest situation and due to different levels of CPR training. Physicians and nurses on monitored wards attend advanced CPR training and working in high-risk areas of IHCA have been shown to be associated with increased con dence in performing CPR skills and in being a part of a resuscitation team [16]. In this study, differences in attitude were seen between HCPs on monitored wards and non-monitored wards regardless of previous real-life CPR experience which suggests that the level of CPR training might affect attitude.
A majority of HCPs thought they would be able to take command if necessary in a cardiac arrest situation. Previous results have shown that only 7% of Swedish HCPs agreed with the ability to lead a cardiac arrest team [17]. The importance of perceived behavior control increases in situations where deliberate choice to perform a speci c behavior decreases [5]. During teamwork, team members tend to anticipate task ability based on outer factors such as profession or sex, not necessarily associated with competence [19]. Considering this, in the absence of a more competent leader, a majority of HCPs in the present study would try to take command of the situation if necessary although the self-assessed ability to lead among HCPs have previously been shown to be low.
In this study, anxiety was reported among 26% of HCPs when being required to perform CPR, and among 27% after the latest real-life CPR performance. It has been reported that only a third of rst responders manage to treat a cardiac arrest correctly in simulated situations and that many teams initially fail in building a team structure [20]. Perceived stress has been proven to delay initiation of CPR in simulated situations [21]. Simulated cardiac arrest situations have been shown to increase negative emotions [10] and anxiety is the most common emotion during stress overload [22]. An overload of stress often has a negative impact on the individual and happens when the resources are insu cient to the demands of the situation. Stress overload could affect attention, memory and decision-making and therefore have a negative effect on managing a sudden cardiac arrest [22].
The present study showed that the compliance to recommended CPR training intervals, every six months or at least once a year, was low among all groups of profession. Furthermore, many HCPs work in low-risk areas of cardiac arrest, which means a lack of real-life CPR experience. It has previously been shown that CPR training increase positive attitudes towards CPR [14,15]. The results in the present study indicates that 12 months or more since CPR training is associated with increased odds of anxiety despite of previous real-life experience. This shows the importance of refresher-training to maintain the knowledge and skills necessary to manage the situation, which seems to reduce stress overload and anxiety. CPR training need to be prioritized among all in-hospital HCPs.

Limitations
There are limitations in this study. Higher response rates were found in one of the two regions, probably due to differences in how the questionnaires were distributed. Equal response rates among professions, but lower on monitored wards compared to non-monitored wards, were observed in one region. In the other region, the response rate was higher on monitored wards than non-monitored wards, but the profession of the non-responders were not known and therefore cannot be analysed. Experience in reallife cardiac arrest situations was only requested in one region, due to differences in focus of the surveys. There was a time difference regarding data collection in the two regions. However, the attitude among HCPs toward performing CPR is not considered to be affected by this time difference.

Conclusion
Despite overall positive attitudes towards performing CPR, feelings of stress and anxiety were common which could re ect possible stress overload. Many HCPs work in low-risk areas of cardiac arrest, therefore regular CPR training is important to maintain competence and reduce anxiety. The possible effect of attitudes on performing CPR needs to be studied further.
Abbreviations CPR Cardiopulmonary resuscitation HCPs Healthcare professionals Declarations Ethics approval and consent to participate This study was given an additional amendment by the ethical review authority in Uppsala, (Dnr 2006/201/2) and was registered at ClinicalTrials.gov (ID nr NCT04321213). The study was performed according to the principles of the declaration of Helsinki [23]. HCPs gave consent to participation by responding to the questionnaire.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during this study are available from the corresponding author on reasonable request.