Background

Out-of-hospital cardiac arrest is associated with poor survival rates [1]. The initial professional treatment of out-of-hospital cardiac arrest patients requires the prehospital resuscitation providers (providers) to decide whether to withhold, initiate, continue, or terminate the prehospital cardiopulmonary resuscitation. This medical decision-making is a deliberative process that combines conscious and unconscious influences derived from medical evidence, personal and medical beliefs, and, where possible, knowledge of patient preferences [2]. Clinical guidelines based on medical factors, such as comorbidities or initial cardiac rhythm assist decision-making in prehospital resuscitation in some countries, while other countries rely on individual judgement [3]. Even though resuscitation decision-making is based on medical evidence, non-medical factors can influence decision-making, even where guidelines and clinical rules are applied, which may in turn influence patient survival [2, 4].

Anderson et al. described the factors influencing the treatment of out-of-hospital cardiac arrest and identified five themes: the arrest event, the patient characteristics, the resuscitation scene, the perspective of the resuscitation provider, and medico-legal concerns [2]. The study explored both medical and non-medical factors and found that non-medical factors may change the course of action especially if the medical factors point in diverging directions [2]. As the field of resuscitation evolves, the context in which healthcare professionals make decisions changes. Recent studies describe new aspects, including non-medical factors, in decision-making in prehospital resuscitation [5,6,7,8,9]. We aimed to review original studies on non-medical factors that pre-hospital care providers describe as important for decision-making in adult resuscitation..

Methods

We conducted a mixed-methods systematic review with a narrative synthesis and reported this according to the Cochrane Qualitative Research Methods Group guidelines [10] and the PRISMA guidelines for reporting of systematic reviews [11]. As PRISMA is not designed for mixed-methods systematic reviews, we applied the fields relevant to our study (See Additional file 1). A systematic mixed-method review allows for the inclusion of multiple methodologies to combine the strengths of quantitative and qualitative methods but also makes it possible to compensate for the limitations in each method [12]. Our protocol was published on PROSPERO on March 17, 2021 (PROSPERO registration number CRD42021237078).

Selection criteria and search strategy

We searched for peer-reviewed quantitative, qualitative, and mixed-methods studies containing empirical data on non-medical factors in prehospital resuscitation of out-of-hospital cardiac arrest and their influence on the providers’ decision-making. Non-medical factors are medically extraneous factors, not necessarily based on symptoms or tests, and therefore can be difficult to define precisely [13, 14]. We used the existing literature on non-clinical and non-medical factors to outline the definition of non-medical factors [15,16,17]. This literature does not necessarily originate from the emergency medical setting and the non-medical factors had to be adapted to the prehospital resuscitation decision-making. See Table 1 for examples. We included studies concerning decision-making in prehospital resuscitation of adults > 18 years and non-medical factors as described by the resuscitation providers employed in EMS systems. We only included mixed populations of in-hospital healthcare personnel and prehospital providers if we could determine that the majority of the study population were EMS workers. Mixed populations of paediatric patients and adult patients were included if the majority of the patients were adults. We contacted the authors of two studies with a mixed population for additional information on the proportion of prehospital participants. We excluded studies concerning in-hospital decision-making, paediatric resuscitation, post-resuscitation care, and register-based studies where decision-making was not described by the providers.

Table 1 Examples of non-medical factors
Full size table

We conducted our search guided by the PRISMA Search Report Extension [18]. A research librarian assisted the search in the following electronic databases: PubMed, MEDLINE, EMBASE, CINDAHL, Cochrane Library, and PsycINFO (See Additional files 2 and 3 for the full search strategy and the list of search terms). We performed the final search on June 11, 2021. To ensure data completion, we used Scopus to identify articles that cited the included papers and used the snowballing method to identify additional eligible studies from the reference lists of the included studies [19] and screened the reference lists of systematic reviews meeting our inclusion criteria for eligible studies. We used Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at www.covidence.org) for data management. Authors LM and JK conducted the title and abstract screening independently. Conflicts were resolved in consultation with author SM.

Data extraction

Authors LM and JK independently extracted data on study title, journal title, methodology, and participants, setting characteristics (e.g., country of origin, type of emergency medical system, legislation concerning termination of resuscitation), phenomena of interest, and outcomes of relevance to the review questions. Qualitative data consisted of themes and subthemes including relevant illustrations. Quantitative data consisted of outcomes based on descriptive or inferential statistical tests. Disagreements were resolved through discussion with author DN.

Assessment of methodological quality

Authors LM and JK critically appraised the included studies independently using the mixed methods appraisal tool (MMAT) [20]. The MMAT assesses both qualitative, quantitative, and mixed methods studies using different templates for each method. Every template contains five different criteria to be assessed, thus allowing one robust score to be used for multiple study types [20]. This enables appraisal scores ranging from 0% (no criteria met) to 100% (all five criteria met).

Data synthesis

We undertook a sequential exploratory synthesis in which qualitative data are synthesised first followed by a synthesis of the quantitative findings to generalise and test the qualitative findings [12]. The findings are then incorporated into an overall synthesis and interpretation. Our qualitative synthesis consisted of a thematic content analysis inspired by Malterud [21]. The subsequent quantitative synthesis including a meta-analysis was not possible due to heterogeneity of interventions and outcome measures, and instead, we used a narrative approach. In the final synthesis, we incorporated the quantitative findings with the themes and subthemes identified during the thematic content analysis. See Additional file 4.

Ethics

Due to the nature of the study, ethical approval was not required.

Results

Study selection and study characteristics

We identified 19,943 papers through database searches and included 27 papers after screening. Of the 143 papers excluded after full-text review, 59 had study outcomes irrelevant to our study such as decision-making factors based on registry data and not data described by providers, 51 were opinion or discussion papers without data, 16 were from an in-hospital setting, seven contained in-hospital providers’ opinions on OHCA. In eight papers, a full-text article could not be obtained, and two papers were abstracts without full-texts available. See Fig. 1 for a flowchart of the study process. The included papers contained data from 25 unique studies. Of these papers, 12 were qualitative studies [5, 8, 22,23,24,25,26,27,28,29,30,31], two mixed-method studies [32, 33], and 13 quantitative studies [34,35,36,37,38,39,40,41,42,43,44,45,46]. The papers were published between 1993 and 2021. See Table 2 for the study characteristics. Of the included papers, 24 concerned prehospital providers only, while three papers [34, 35, 42] included both prehospital and in-hospital healthcare professionals (with a majority of prehospital providers). See Table 3 for additional study characteristics.

Fig. 1
figure 1

PRISMA flowchart of the inclusion process

Full size image
Table 2 Study characteristics and critical appraisal score
Full size table
Table 3 Additional study characteristics
Full size table

Quality assessment

For quality appraisal scores, see Table 2. The MMAT quality score varied among the studies. The qualitative papers had a median score of 100% (range 0–100%), the quantitative papers had a median score of 80% (range 60–100%), and the mixed methods papers had a median score of 60% (range 20–100%).

Descriptions of non-medical factors influencing prehospital providers’ decision-making

Our analysis revealed five main themes and 13 subthemes. See Fig. 2. Both qualitative and quantitative studies covered all themes, but we identified differences between qualitative and quantitative findings regarding emotions, values, and personal beliefs of providers and their influence on decision-making. These were found mainly in qualitative studies. Quantitative studies mostly concerned the patients’ age, the providers’ characteristics and experience, and advance directives. All five main themes were covered by studies from various geographical origin. See Table 4 for an overview of factors and how they influence decision-making.

Fig. 2
figure 2

A visual presentation of the identified themes and their relations

Full size image
Table 4 Non-medical factors and their influence on decision-making
Full size table

Applying patient-related factors and perspectives

Patient characteristics

Patient age formed a part of the decision-making [41]. In younger patients, some providers would almost always perform resuscitation [42] and continue resuscitation for longer [8, 27, 29, 31, 32]. Old age in itself was not described as a reason to cease treatment, but young age was reported to be a reason to initiate resuscitation [29] and treat the patient more aggressively [31]. The providers felt a greater weight of responsibility and emotional burden when terminating resuscitation in younger patients [8]. In line with this finding, patients’ age above 79 years increased the perception of inappropriate resuscitation attempts [35].

Social status was considered in decision-making [29, 31]. A patient’s social status was thus explicitly reported not to influence resuscitation [29] while another study noted the patient’s social value did influence the decision-making [31]. Social value was described as certain personal patient attributes that are highly valued by an ensemble of members of a society [31]. Patient gender was not mentioned in any studies.


Ethical reflections concerning the patient’s wishes and quality of life Perceptions of the patient’s prognosis and expected quality of life were described as influencing decisions [5, 22, 27, 30, 34, 41, 44]. The subjective assessment of morbidity or the patient’s clinical presentation tended to dominate decision-making over chronological age, explained by the fact that exact age and known comorbidities are often not known prehospitally [22]. The perception that resuscitation was inappropriate was significantly higher in patients where the providers found the first physical impression of the patient to be “poor” or “bad”. Providers mentioned the quality of life [27, 30], where an expected low quality of life in case of successful resuscitation would make providers question the treatment. When providers were legally obliged to initiate resuscitation, they described having concerns about the patient’s expected quality of life after resuscitation [44].

In some studies, patient wishes and advance directives were important factors in decision-making [23, 28, 33], while others reported a lack of consideration for patient wishes and formal documentation [22]. In some studies, the majority of participants would honour an advance directive [33, 39, 45]. In other studies, half of the participants would not honour an advance directive [42, 43]. Honouring an advance directive could be influenced by the age of the patient and the gender and experience of the provider [36, 39, 43]. Providers would most likely seek documentation of an advance directive if they were male or more experienced or if the patient was 60 years or older [39, 43].

Some studies highlighted concern for the patient’s best interests [22, 30, 47], either by discontinuing resuscitation in patients with a low expected quality of life in case of survival [30] or by giving the patient the benefit of the doubt continuing resuscitation for a longer period [23]. The patient’s best interests were not evaluated exclusively through advance directives expressing the patient’s wishes, but rather by conferring with the family and crewmembers [22].

Some providers believed that a seriously ill and old patient had the right to die a “natural death” or die with dignity [29]. Dignity was not only described as upholding the patient’s wishes but also the act of avoiding apparently futile procedures [25]. Providers highlighted the right to a dignified death [25] or to die without interference [29] and were concerned with the patients’ dignity in cases where they could not legally terminate resuscitation [44].

Involving and involvement of bystanders and family members

Family wishes and emotions

The patient’s family and their wishes, emotions, and/or expectations could pressure providers to the initiation, continuation [24, 29, 30, 32], and termination [29] of resuscitation. Some providers mentioned that the family’s religion could lead to pressure to continue obviously futile resuscitation when according to the religious beliefs, everything possible had to be done including transportation to the hospital [32]. In cases where the family wished for termination, the providers described the family as expressing fear of suffering or a permanent vegetative state for the patient [29]. Some providers would not follow family wishes [42]. Others believed family members’ verbal wishes should be honoured [45]. Family wishes were mentioned as a reason for not complying with the guidelines [37, 46]. The family’s despair led them to beg or plead for the continuation of resuscitation—a request that the providers often complied with [24, 29], but the emotions of the families also increased the providers’ uncertainty if their decision went against the expressed wishes of the family [24, 30]. Some providers voluntarily involved the relatives in the decision-making process to obtain information or to aid the families in their mourning [29, 41].

Some providers continued resuscitation to accommodate the perceived needs of the family members [5, 24, 29]. This included continuing resuscitation to ensure that everything possible had been done to save the patient [24], but also to give the family time to realise the patient’s imminent death [5]. The family’s acceptance of a severely ill or older patient’s death could make providers more likely to withhold resuscitation [29]. On the other hand, providers feared increasing unrealistic expectations of survival if they continued resuscitation to allow the family to say goodbye [24, 33, 44]. Continuation of resuscitation and transportation to the hospital was the providers' way to deal with their inadequacy in meeting the family’s needs [24] and avoid facing the relatives with information that resuscitation was futile [29]. The in-hospital environment was perceived by providers as making it easier for the family to accept termination of resuscitation [32]. Close identification with the family made some providers more subjective in their decision-making and lead to the continuation of futile resuscitation [24]. Cultural barriers and an inability to identify with the emotional response of the family could lead to transportation of the patient to avoid confrontation [37].

The presence of bystanders

The presence of other bystanders than the family could influence the providers to continue resuscitation [36] to show that something had been done to attempt to save the patient’s life [24] or for providers to look their best in front of spectators [27]. The providers in one study displayed frustration with the bystanders’ unrealistic expectations [8]. These high expectations could influence the providers to continue resuscitation and transport the patient [29, 32]. When bystanders had initiated resuscitation, providers felt an obligation to continue resuscitation to respect the resuscitation attempt [5], to assure bystanders that their effort was important for a good outcome [29], and encourage them to do resuscitation again in a similar situation [42].

The personal conditions of providers

Provider’s characteristics and experience

Pre-existing conditions among others, length of service, type of daily work, or training were described as influencing decision-making [23]. Surgeons could be more likely to initiate resuscitation compared to anaesthesiologists and general practitioners [40]. One other study, however, found no association between profession and the perception of the appropriateness of resuscitation [34]. The level of training was associated with doubts concerning withholding resuscitation with paramedics being more likely to be troubled than EMTs [38]. Providers who had received specific training were more inclined to find resuscitation appropriate [34] and providers with palliative training found it more reasonable to resuscitate dying and terminal patients [39].

The provider’s gender and age was described as influencing decisions [31, 36, 42]. One study did not find an association between age and withholding resuscitation [45]. Personal experience and years of service made the providers more confident in their decision-making [8, 22, 23, 31, 32, 38, 39, 41,42,43]. Two papers did not find an association between experience and honouring a do not attempt resuscitation order (DNACPR) [44, 45]. Some províders described that specific experiences had influenced their decision-making [5, 29, 30], e.g. receiving flowers or letters from elderly patients who survived a cardiac arrest, and made them more prone to initiate resuscitation despite old age [29].

Emotions and personal values

The providers described an uncertainty in decision-making [8, 22, 24, 28, 29, 31, 46] in unexpected situations [8]. This was often described in situations with a lack of information [22, 31]. Uncertainty could lead to prolonged resuscitation [8, 46]. providers feared overlooking information that was important for the treatment and not having done everything possible to save the patient [22, 29]. Others feared legal consequences, criticism, or disciplinary procedures if they did not initiate or continue resuscitation [23, 27,28,29, 32, 44].

The providers acknowledged that their emotions, beliefs, and values provided a lens through which they viewed the patient’s and family’s emotions and reactions and that this might influence the decision-making [8]. Some providers noted that their own religious beliefs influenced their decision-making [33] while two studies did not support this association between religion and decision-making [34, 39].

Team interaction

Agreement within the team and the respect of other team members’ opinions were important [5, 23, 26, 32]. Team members who wanted to continue resuscitation prevailed over those who would not [5, 22, 23, 26, 27, 30, 32]. Several studies reported that non-physician providers contacted a physician to ensure everything had been done before terminating resuscitation [22, 23, 27, 30, 32]. Some providers mentioned that the composition of the prehospital crew (e.g., type of work, experience, etc.) influenced the decision-making [23].

Being influenced by external factors

Emergency medical service system and work environment

The providers described being influenced by the level of support from their employer [23] and the emergency healthcare system’s reputation [29,30,31]. The providers wished to contribute to a positive EMS reputation and sometimes adjusted their actions according to the perceived wishes of bystanders [29].

Some providers continued resuscitation for teaching or training purposes [29, 31, 42]. One study found provider fatigue as an influence referring to the long and odd working hours in the EMS [23], which led to deciding faster to transport or continuing longer at the scene to avoid an accusation of not having done enough [23]. Threats from the family could lead providers to commence or continue resuscitation and transport the patient [8, 29, 32, 46]. Scene safety was not cited as an issue in one study [38]. Rural versus urban area of service was a factor in one study with providers in rural areas more likely to terminate resuscitation compared to providers in urban areas [36].

Legislation and official guidelines

Providers found that legislation and official guidelines influenced decisions [35, 39]. Some found them helpful [25], while others felt uncertain about legal obligations [23, 33]. In one study, only 9.8% of providers felt competent to handle advance directives [39], while in another study, 73.7% felt confident terminating resuscitation when a DNACPR was present [33].

The cardiac arrest setting

The providers reported that the prehospital setting influenced decision-making [8, 22, 32, 37]. Specific areas with high mortality and morbidity, such as nursing homes and low socioeconomic areas, influenced treatment [22] as did the location of arrest e.g. public place [37]. Weather conditions and other environmental conditions e.g., cramped, dark places could influence resuscitation [8, 32, 37]. Logistical limitations e.g. difficulties of doing CPR during transport were mentioned in one study [8], while another study did not find the distance to the hospital to influence the decision-making [32].

Navigating conflicts


Ethical conflicts between guidelines, legislation, and beliefs The providers experienced conflicts between the law, the guidelines, and the patient’s wishes [23, 26] [44]. Specifically, they mentioned the lack of formal documentation in situations where family members stated that the patient did not wish to be resuscitated [23] and believed that ethically correct decisions sometimes resulted in deviations from guidelines or legislation [26, 28].

Conflicting values

The providers described various values and beliefs that created conflicts during decision-making [33]. The family could express different opinions on resuscitation than those noted by the patient in advance directives [23, 25, 26, 29, 32]. To avoid conflicts, providers would continue or transport the patient during resuscitation [32]. Providers experienced conflicts between their own values and the expectations from the legislative system [24, 30]. They feared the negative consequences if they did not transport the patient [30] and felt a conflict between the perceived legal obligation to save lives contrasting with the best interest of the patient [24].

The providers' values could conflict within the resuscitation team [5, 8, 23, 31]. In all studies, disagreements arose over the appropriate level of aggressiveness of treatment. The providers balanced futile care against a well-founded decision and described conflicts and challenges in choosing whether to initiate or continue resuscitation in cases where information was lacking [5, 8]. For example, almost all participants in one study would initiate resuscitation despite the feeling of providing futile care [5].

Discussion

In this systematic review concerning non-medical considerations in decision-making during prehospital resuscitation, we identified various themes influencing prehospital providers' decision-making, including patient-, family-, and provider-related factors as well as external factors such as legislation. Furthermore, we identified conflicts occurring between the influences coming from various actors, and differences between findings in qualitative and quantitative studies. Our findings underline the importance of contributions from both study methodologies to gain a better understanding of the decision-making process and the various influences. The study aims and the topics in the included studies varied from end-of-life to sudden cardiac death, which may explain variation in study results. However, this is a reflection of the wide spectrum of cardiac arrest scenarios that providers attend.

Non-medical factors are diverse. We identified several areas of potential improvement in out-of-hospital cardiac arrest decision-making: First, several factors, which should ideally be avoided, were described as part of decision-making. Examples of this include social status, the location of the cardiac arrest, and provider bias which risk challenging the bioethical principle of justice and hence should be avoided in decision-making [48]. However, complete avoidance of such factors may be difficult to attain. Instead, efforts to encourage providers to consciously reflect on non-medical factors in decision-making may be helpful.

Secondly, some non-medical factors, which should be included, were not always considered. One of the most researched non-medical factors were DNACPRs. Interestingly, studies on DNACPRs and their use in decision-making showed diverging results. The handling of advance directives differed and we found diversity in individual studies where participants were divided almost equally between their beliefs on DNACPR and their handling in practice [41,42,43]. Current guidelines underline that information about the patient’s wishes and values on resuscitation should be sought and included in resuscitation decision-making [3], but our findings suggest that providers find it difficult to do this in practice. Up to 10% of out-of-hospital cardiac arrests involve DNACPRs and in these cases, conflicts can occur between the patient’s family and providers [49]. Most of the included studies were conducted in a prehospital setting without a universal termination of resuscitation rule [50], but in the few studies where these rules were implemented, non-medical factors were described as influencing treatment and could also collide with the provider’s personal beliefs [32]. These conflicts of priorities, together with ethical conflicts, conflicts of expectations, fear of litigation and uncertainties, can influence decision making in unfortunate ways and furthermore cause moral distress in healthcare professionals [51]. Interventions to support providers’ prehospital resuscitation decision-making could target these situations, and future studies could shed further light on the conflicting factors in resuscitation and the dilemmas in providers’ coping.

Limitations

Our study has several limitations. The study aims and the topics in the included studies varied from end-of-life to sudden cardiac death. Thus, variation in study results may have occurred. The included studies are from a wide range of countries and have different study aims, EMS structure, and publication years. This precluded a meta-analysis. Furthermore, non-medical factors are difficult to define. We may thus have overlooked some studies. The resuscitation guidelines are updated every 5 years. This may induce changes in the resuscitation practices and thus contribute to limitations in our study. However, we found representations of all themes in both newer and older studies. As clinical decision-making including ethical considerations is a sensitive area, response bias may have occurred. Three qualitative studies rated as high quality with the critical appraisal tool [5, 30, 31], did not state their methodological orientation nor supported the analysis process with any references. As qualitative studies explore the content and meaning of empirical data, not precisely stating the methodological orientation and referencing the theoretical approach makes it difficult to decipher how the authors analysed their data. Thus, it should be noted that high reporting rates according to MMAT do not necessarily equate to high‐quality studies.

Conclusion

When deciding whether to initiate, continue, terminate, or refrain from resuscitation, prehospital resuscitation providers are influenced by a plethora of factors of which some are not strictly medically related. The providers report that patient- and family-related factors influence their decision-making process. They further report that they are influenced by personal conditions and external factors. Additionally, the providers may experience that conflicts between various factors complicate decision-making. Future research should consider non-medical factors and their role in decision-making.