Impact of Emotions on Communication and Outcomes in OHCA-Related Emergency Medical Services Calls: A Scoping Review

Introduction

Rationale

Out-of-hospital cardiac arrest (OHCA) constitutes an emergency in which every second matters [1]. Therefore, communication during Emergency Medical Services (EMS) calls must function efficiently to minimize delays. Essential information—such as the exact location of the incident, the patient’s consciousness level, and breathing status—must be obtained rapidly to ensure immediate ambulance dispatch and prompt recognition of OHCA, enabling early bystander cardiopulmonary resuscitation (CPR). Each step performed without delay can improve survival chances [2]. Evidence indicates that a caller’s emotional intensity and degree of cooperation may affect whether bystander CPR is initiated and how quickly it begins [3-5]. Moreover, the call-taker’s interaction style can alter the caller’s emotional state—sometimes positively, sometimes negatively—across both OHCA-specific situations and other emergency contexts [6, 7]. A systematic review published in 2021 [8] summarized features that help or hinder call-taker recognition of OHCA and identified “emotional distress” as a major contributing factor. However, no prior review has specifically focused on emotions within emergency OHCA calls.

Objectives

This scoping review aims to map the literature concerning emotional content in OHCA emergency calls made to EMS. The specific objectives are to gather, evaluate, and synthesize studies that:

• (a) explored emotional aspects during emergency calls made by a bystander for a patient experiencing OHCA—this may encompass caller emotions, call-taker emotions, and how either party perceives or responds to the other’s emotional cues; AND/OR

• (b) presented an instrument or methodological approach for identifying or evaluating emotions in these calls; AND/OR

• (c) investigated how emotional expressions during the call relate to subsequent call outcomes (both process-related and/or patient-related outcomes).

Materials and Methods

Protocol and registration

The review was carried out using the methodological framework set out in the Joanna Briggs Institute (JBI) Manual for Evidence Synthesis for Scoping Reviews [9]. The review protocol was preregistered on the Open Science Framework (OSF) under DOI 10.17605/OSF.IO/3CJH8, with restricted-access viewing available at https://osf.io/ewhj3/?view_only=322a9bb5ad4a4b6e84fee312ddc52c38.

Eligibility criteria

Inclusion criteria

Participant/Population

Eligible material involved emergency calls to EMS in which the patient was undergoing an OHCA. The term “OHCA” encompasses both cases ultimately affirmed by EMS teams and incidents interpreted or managed as OHCA during the call itself. Only situations with second-party callers—individuals close enough to potentially deliver CPR or guide another bystander to do so—were included.

Concepts

Studies were required to address emotional aspects within EMS calls linked to OHCA. Inclusion depended on the presence of all three elements—emotion, EMS call, and OHCA—in the same study. Research comparing emotional features between OHCA and non-OHCA calls was also acceptable.

Context

Only emergency calls placed to EMS on a patient’s behalf in the setting of an OHCA event were considered.

Types and sources of evidence

The review included peer-reviewed original studies and review papers that examined how emotions were displayed, influenced the call process, or were measured during EMS interactions involving OHCA.

Exclusion criteria

Any record that did not fit the inclusion framework was excluded. Additional exclusions covered: work involving animals, simulation-only research, commentary-based publications, papers lacking an English abstract, and grey literature.

Information sources

Five databases—Medline, PsycInfo, CINAHL, Embase, and the Cochrane Review Database—formed the core search sources. No limits were placed on the time of publication.

Search strategies

A preliminary scan using web search tools and manual checking confirmed the absence of previous reviews on the topic, supporting the need for the current scoping review.

The authors then applied the JBI’s three-step search procedure [9].

• Step 1: The first author performed an exploratory search in Medline(R) via OVID and CINAHL via EBSCO. This stage used multiple combinations of keywords, including: “emotions,” “emotion assessment,” “cardiac arrest,” “heart arrest,” “OHCA,” “cardiopulmonary resuscitation,” “CPR,” “emergency,” “ambulance,” “emergency medical services,” “emergency medical dispatch,” “EMS,” and “dispatcher” [10]. Terms identified in titles, abstracts, and keyword lists were used to refine the vocabulary for the next step.

• Step 2: The revised keywords were then applied systematically across all five databases. This second stage was completed on 18 November 2021. Table 1 outlines the PsycInfo strategy, noting the use of “exp” (explode) and “mp” (multiple purpose).

Table 1. Example of search strategy (step 2) using PsycInfo database

exp: explode

mp: multiple purpose

• Step 3: Reference lists from all studies found in step 2 were screened to capture any missing published work, supplemented by additional incidental manual searches.

A Curtin University librarian contributed to the development and adjustment of search terms throughout the process. All search outputs were transferred into Rayyan [11] for the screening phase.

Study selection

The initial screening was performed by the first author, who reviewed all search outputs in Rayyan by checking Titles and Abstracts and categorizing each record as Include, Exclude, or Maybe. Full-text assessment was then undertaken jointly by the first two authors for all studies not excluded during the first pass. At this stage, decisions favoured inclusion rather than exclusion, while still following the previously defined eligibility criteria. If the two reviewers had encountered unresolved disagreements, a third independent reviewer would have been consulted. Final selections were organized and archived in EndNote.

Data extraction (Charting)

A preliminary charting template was created to guide data extraction by the review authors (HN and TB). The template included the following fields:

(i) authors and study title;

(ii) publication year;

(iii) country or region of origin;

(iv) stated aims;

(v) sample details and population;

(vi) study design, methods, and comparison groups (if any);

(vii) measured outcomes;

(viii) primary findings; and

(ix) which review objectives were addressed.

Synthesis of results

Evidence meeting the inclusion standards was organized according to the three objectives presented in the Introduction. Descriptive summaries of the extracted data were first compiled into tables structured similarly to the charting tool. These summaries were then integrated into a narrative synthesis, with concepts grouped according to their relevance to the three review aims.

Results and Discussion

Selection of sources of evidence

The database search across all five platforms produced 183 records. After removing duplicates, 160 citations remained. Author HN screened these 160 items by Title and Abstract and retained 10 studies [3-5, 12-18] for full-text review. An additional four papers [6, 19-21] were located through manual or post-hoc searches and were also reviewed in full (Figure 1). These four studies did not surface in the formal search because each lacked at least one of the required concepts in their indexed keywords. For example, Clark et al. [19] did not include emotion as an indexed term, and the other three lacked cardiac arrest in their keyword lists.

Characteristics of sources of evidence

The two primary reviewers were able to resolve all minor discrepancies during the screening phase, so arbitration by another author was unnecessary. Table 2 outlines key characteristics of the 14 studies examined in full, of which 13 were ultimately included. One study [19] was excluded because, although it addressed OHCA and EMS calls, its reference to emotional content was minimal (“hysterical”) and did not satisfy any of the review objectives. Each included article was treated as a unique study.

Table 2. Description of 14 studies retained for full-text review

DA-CPR: Dispatcher-Assisted Cardiopulmonary Resuscitation. EMS: Emergency Medical Service. NA: Not Applicable. OHCA: Out-of-Hospital Cardiac Arrest.

† indicates studies found through manual searching; the remaining 10 were retrieved from database searches.

The 13 included publications appeared between 1986 [16] and 2021 [14]. Countries represented were the United States [5, 6, 16, 17], Taiwan [4, 14, 18], Sweden [13, 21], Australia [3], Denmark [12], Norway [20], and Switzerland [15].

Descriptions of emotional features in OHCA calls varied considerably. Several studies examined broader determinants that influence dispatcher recognition of OHCA or initiation of bystander CPR, with emotional expression functioning as one of multiple influencing elements [3, 5, 12, 17]. Only a few papers explained or detailed the tools used to evaluate emotions during calls [6, 14, 21].

All included research involved analysis of EMS call audio recordings. Most studies—except for Chin et al. [14] and Svensson and Pesämaa [21]—used consecutive or opportunistic sampling from a defined time interval. In contrast, Svensson & Pesämaa [21] prospectively sampled the first EMS call occurring in every 30-minute period. Chin et al. [14] used call recordings mainly to support the calibration or validation of newly developed artificial intelligence models.

Specific Objective 1: Examination of emotions in emergency calls involving OHCA

All 13 studies contributed data relevant to this objective (Table 3). Emotional states expressed during OHCA calls were described using terms such as calm [3, 12, 13], hysteria [3], emotional distress [15, 17], fear, sadness, anger [21], high emotional state [4], or simply emotional [16]. In a number of studies, emotional expression was intertwined with perceived attitudes or behaviours—e.g., cooperation, resistance, resignation, or aggression—shown by the caller from the perspective of the call-taker [3, 4, 6, 18, 20].

Table 3. Descriptions of the 13 included studies according to the three objectives

DA-CPR: Dispatcher-Assisted Cardiopulmonary Resuscitation. ECCS: Emotional Content and Cooperation Score. EMS: Emergency Medical Service. OHCA: Out-of-Hospital Cardiac Arrest.

Using the Emotional Content and Cooperation Score (ECCS), created by the U.S. National Academy of Emergency Medical Dispatch, Clawson and Sinclair [6] reported that callers involved in OHCA emergencies tended to be far more composed than expected when compared with earlier research by Eisenberg et al. [16]. They noted that the “demeanor, voice tone, empathy, and attitude of the dispatcher have an effect on the caller” [6], suggesting that enhanced dispatcher training over time might account for callers’ increased calmness. More information on the ECCS appears under Objective 2.

Specific Objective 2: Description of methods or approaches to measuring/assessing emotions in emergency calls involving OHCA

A total of six studies (46%) [4, 6, 13, 14, 18, 21] detailed the approaches used to evaluate or quantify emotions.

Bang et al. [13], Clawson and Sinclair [6], Ma et al. [18], and Chien et al. [4] all employed the ECCS instrument [6]. The ECCS measures the emotional intensity and cooperation level of the caller, not the dispatcher. It features five categories:

1 = normal conversational tone;

2 = anxious but cooperative;

3 = moderately distressed but cooperative;

4 = uncooperative, shouting, and not following instructions;

5 = highly uncontrolled or hysterical.

Because caller emotions often fluctuate across a call, Chien et al. [4] analysed the highest ECCS score reached in each recording. Ratings were made by a registered nurse and a paramedic, then reviewed by a physician, with inter-rater reliability reported as good (κ = 0.63) [4].

In Bang et al. [13], ECCS scoring was performed at both the beginning and end of the call by the physician overseeing medical responsibility, together with the dispatch centre’s operational leader. In Clawson and Sinclair [6], a certified “case review professional” assigned the ECCS score, capturing only the rating at call initiation. Ma et al. [18] did not state whether the rating applied to the entire call or specific portions; two investigators completed the ECCS scoring, achieving good reliability (κ = 0.65) [18].

Chin et al. [14] developed a two-step method incorporating artificial intelligence (AI) for emotion detection. First, human raters scored calls using ECCS; disagreements were re-evaluated until agreement was reached. Second, the AI component utilized mel-frequency cepstral coefficients and a support vector machine. The authors explained that mel-frequency coefficients align frequency scales with human auditory perception and are widely used in automated speech analysis, while support vector machines are robust classifiers used to detect “emotionally unstable callers” [14]. When compared with human ECCS ratings, the AI model demonstrated 39% sensitivity, 73% PPV, 98% specificity, and 93% NPV. Similar or slightly improved performance occurred when only the first 10 seconds of calls were analysed, leading the authors to propose the tool as a rapid detector of callers with high emotional instability [14].

Svensson and Pesämaa [21] categorized caller emotions into three basic groups—fear, sadness, and anger. Dispatchers rated the intensity of each on an 11-point scale (10 = very pronounced, 0 = none) and assessed perceived caller need for assistance on an 8-point scale (1 = no assistance, 8 = severe need). Only the first one minute of each call was used for these ratings. Additional information is presented under Objective 3.

Specific Objective 3: Links between emotional expression and call outcomes in OHCA-related emergency calls

Seven of the 13 studies (54%) 12., 17., 20., 4., 5. addressed this objective, although they explored different patient and call outcomes and reported inconsistent associations.

Chien et al. [4] examined the broadest collection of predetermined call-process and patient variables in relation to caller emotional expression. They reported that varying ECCS emotion levels (1–5; ‘calm’ to ‘yelling/hysteria’) did not correspond to differences in how accurately callers reported consciousness or breathing. Likewise, higher emotional arousal (higher ECCS) did not hinder OHCA identification. Instead, stronger emotional expression appeared to assist call-takers in recognising OHCA, possibly because callers’ distress signalled the urgency of the situation. In contrast, greater emotional distress was linked to lower agreement to perform DA-CPR. Interestingly, distressed callers initiated the first chest compressions sooner than calmer callers once instructions were given. [4]

Findings on whether emotions affected DA-CPR implementation were contradictory. Lerner et al. [5] noted that caller emotion accounted for 14% of failures to complete DA-CPR instructions. Conversely, Case et al.³ observed that both “hysteria/panic” and “calm resignation” appeared among 20% and 39% of calls, respectively, where OHCA was recognised but DA-CPR did not occur, implying that both heightened distress and disengaged calmness might obstruct DA-CPR. In contrast, Hauf et al. [17] concluded that emotional distress was not a frequent reason for DA-CPR omission. All three studies [3, 5, 17]. focused on calls where DA-CPR was not delivered, but without parallel analysis of a comparison group in which DA-CPR was successfully carried out, which limits the certainty of these findings.

Through qualitative thematic analysis of two small datasets (unrecognised OHCA, n = 13; recognised OHCA, n = 8), Alfsen et al. [12] identified “emotional distance” (i.e., maintaining composure) as a notable theme in the recognised-OHCA group. They suggested that calmer callers were better able to collaborate with call-takers, thereby supporting OHCA recognition. [12]

Svennevig [20] undertook an in-depth conversational analysis of three calls that ended in a fatal OHCA. Hostility developed rapidly between callers and call-takers, was not resolved, and ultimately resulted in no DA-CPR instructions, no attempts at DA-CPR, and delayed EMS dispatch. [20]

Svensson and Pesamaa [21] reported that fear and sadness were more likely than anger to signal that the caller required assistance. Fear was considered more intense than sadness and was perceived as indicating the highest need for help. [21] They proposed that such perceptions could influence call-taker decisions (e.g., dispatching an ambulance or initiating DA-CPR instructions). [21] Unlike most studies in the review, this investigation did not specifically target OHCA calls; rather, OHCA incidents were part of a broader analysis of emotional expression. Nevertheless, it was included due to its relevance to emotional content in OHCA-related calls.

Only one study [20] evaluated how emotions affected patient outcomes, examining fatality alone. Other outcomes—such as return of spontaneous circulation during CPR or survival to hospital discharge—were not assessed in relation to caller emotional expression.

Summary of evidence

This scoping review identified a limited set of studies examining emotions during OHCA-related emergency calls, with only a few detailing how emotional states were quantified. Many studies explored how emotions related to call-process metrics, although the extent and specificity of these analyses varied considerably. According to Chien et al. [4] and Hauff et al. [17], emotional distress did not obstruct OHCA recognition or DA-CPR. In contrast, Case et al. [3] reported that both high distress and subdued resignation were frequently noted in cases without DA-CPR, whereas Alfsen et al. [12] suggested that emotional neutrality (calmness) improved caller cooperation and thereby aided recognition. Only one study [20] linked intense emotions during OHCA calls with mortality, attributing the outcome to deteriorating caller–call-taker interaction, delayed assistance, and ultimately the patient’s death. No study evaluated other survival indicators, such as return of spontaneous circulation or discharge survival.

Marked heterogeneity was evident among the included studies, spanning methodological approaches (quantitative, qualitative, mixed methods), study design (mainly retrospective, often without control groups), sample sizes (from n = 3²⁰ to n = 6449⁶ calls), aims, and outcome variables.

Very few investigations described the method used to classify emotions in OHCA-related EMS calls. The Emotional Content and Cooperation Scale (ECCS) [6] was the most frequently adopted tool. Another study [14] applied artificial intelligence to build a new model capable of identifying the caller’s emotional state early in the call, proposing that this could help call-takers tailor their communication for callers experiencing heightened distress.

Given the complexity, variability, and intensity of emotional expression, emotions remain challenging to evaluate and predict in time-critical OHCA situations. This review highlights the potential impact of how caller emotions are interpreted by call-takers, which may influence OHCA recognition and the rapid initiation of DA-CPR.

Strengths and limitations of this scoping review

This work represents the first systematic scoping review specifically examining emotional expression in emergency calls involving OHCA. The three guiding objectives shaped the review process and enabled a consolidated overview of current evidence, while also drawing attention to gaps requiring further investigation.

The review has limitations

First, a manual supplementary search conducted after the main database search located three additional eligible studies that had not been retrieved earlier. Likewise, the paper by Lewis et al. [22], which noted that a small proportion of cases showed associations between caller “emotional states,” lack of DA-CPR instruction delivery, and/or delays in initiating initial chest compressions, may also have been inadvertently overlooked. Such occurrences are not uncommon and are generally considered acceptable within systematic and scoping review methodologies.

Second, this review did not undertake a critical evaluation of the methodological quality of the included studies. This omission aligns with typical scoping review practices, as formal appraisal is not usually viewed as a core requirement [23].

Where to from here

Drawing on the findings of this scoping review, three main recommendations emerge for future investigations. Studies exploring emotions during OHCA-related emergency calls—and their relationships with process or patient outcomes—should: clearly articulate the research question and predetermined outcome measures; include an appropriate comparison group; and, for quantitative work, ensure sufficient sample size.

Emotions are inherently multifaceted, with variable expressions, triggers, and subtleties. In several studies, emotional descriptors also overlapped with interpretations relating to attitudes, behaviours, or perceptions, further complicating analysis—especially within the high-pressure context of OHCA emergency communication. A robust measurement tool for this setting must account for these complexities. Although the ECCS [6] appeared most frequently in the reviewed literature, it does not reflect the evolving nature of emotional expression, nor does it incorporate contextual influences. Moreover, as previously indicated, confusion among similar emotional categories can occur, indicating the need for a more precise instrument for evaluating emotions in OHCA calls.

The work of Svennevig [20] and Svensson and Pesamaa [21] demonstrates the value of linguistic approaches in tracing how emotional dynamics develop and escalate throughout a call. A recent editorial authored by members of the present review team has also emphasized the importance of such research directions [24].

Building on insights from Chin et al. [14], Clawson and Sinclair [6], Svennevig [20], and Svensson and Pesamaa [21], two additional questions arise for future inquiry and practical application:

1. What forms of training should be provided to EMS call-handlers to better support effective management of caller emotional states?
2. Would it be beneficial to educate the public about the structure and expectations of the emergency-calling process, so that callers are more prepared for required questions and better able to contribute to life-saving actions?

Conclusion

This scoping review identified a limited body of literature addressing emotional expression during emergency OHCA calls, with only 13 studies meeting inclusion criteria. Emotional assessment varied widely, and only six studies fully explained how emotions were evaluated. Several studies examined links between emotional expression and call outcomes, though findings lacked consistency. Only one investigation considered the effect of caller emotions on OHCA patient survival. Overall, this review highlights a substantial research gap with important implications for advancing both future studies and clinical practice.

Acknowledgments: None.

Conflict of interest: None.

Financial support: All authors are employees of Curtin University of Technology, Perth, Australia. JF receives salary support and research funds from a NHMRC Investigator grant (#1174838). The authors gratefully acknowledge Ms Vanessar Varis, Faculty Librarian, Curtin University's Faculty of Health Sciences, for her help with the literature search strategy.

Ethics statement: None. Studies involving humans and animals must have been performed with the approval of an appropriate ethics committee and provide the reference number.