Differences in bystander CPR by patient, caller, and telecommunicator sex: a retrospective analysis of emergency calls

Introduction

Individuals who receive prompt cardiopulmonary resuscitation (CPR) have a two-fold increase in chance of survival compared to those who do not receive CPR after out-of-hospital cardiac arrest (OHCA).¹ Female patients receive CPR in public less frequently.2, 3, 4, 5 This finding of sex disparities in the receipt of bystander CPR in public has been shown to occur broadly internationally.2, 3, 4, 5

Additionally, both sex and age of a hypothetical OHCA patient impacted the reported likelihood of initiating CPR for participants.⁶ Concerns expressed about CPR on female patients included causing injury to the patient, exposing patient breasts, concerns about pregnancy, and being accused of violating the patient.⁶ In a retrospective cohort study from Japan, younger women received both AED application and CPR at lower rates than males of the same age.⁵

Multiple studies have found individuals expressing fears of hurting the patient, fears of inappropriate touching, and fears of being accused of assault as barriers to bystander CPR for females.6, 7, 8 Additionally, barriers to performing CPR on female patients may vary based on the bystander sex, with male bystanders more likely to fear being accused of assault or touching the patient inappropriately with female bystanders more likely to fear hurting the patient.⁸

Telecommunicator CPR (T-CPR) is a promising intervention for improving rates of CPR. A 2024 scoping review of T-CPR studies showed that T-CPR was associated with trends toward improvement of CPR rates.⁹ A study using data from the nationwide OHCA registry of the Republic of Korea found that T- CPR improved rates of CPR for both females and males with OHCA in public, reducing but not eliminating the disparity in bystander CPR for females.¹⁰ A 2025 scoping review of interventions to improve T-CPR highlighted the important role the interaction between callers and telecommunicators plays in overall OHCA response.¹¹

However, it is not clear how the sex of the person calling and the telecommunicator influence receipt of CPR for female patients. In the current analyses, we compared the frequency of bystander CPR receipt for patients by the sex of the patient, caller, and telecommunicator.

Methods

We conducted a retrospective cohort study analyzing OHCA calls from Durham County, North Carolina, United States. We partnered with the Durham Emergency Communications Center (a primary 9-1-1 public safety answering point). A Communications Officer (CO) at 9-1-1 identified potentially eligible calls based on the nature code associated with the call. After cross-referencing with Emergency Medical System (EMS) run sheets, calls less likely to be associated with cardiac arrest were removed. From the final call list, the CO at 9-1-1 provided us with call recordings and Computer Assisted Dispatch (CAD) reports for 327 calls for suspected adult, non-traumatic OHCA. Durham is a mixed urban and rural county with 1133.7 people per square mile and an annual EMS 9-1-1 call volume of over 57,000 (2023).¹²

The study team reviewed transcripts and corresponding CAD reports to remove any calls not meeting our inclusion criteria. We excluded all calls for patients under 18 and those that did not appear to be OHCA calls (e.g., trauma). As a way of limiting calls to layperson responders, we excluded calls coming from healthcare sites (e.g., rehab facilities). We did not exclude calls where the caller or another bystander had started CPR in advance of the call. We also excluded calls where a “do not resuscitate” (DNR) order was in place. We finally excluded calls in languages other than English as these are placed through a call center and therefore include significant barriers beyond those present in the English-language calls.

For all calls included in the analysis, we identified patients (observer-reported age) and callers (relationship to the patient), as well as presence/absence of bystanders, location, duration, time of day, and receipt of CPR. In instances with multiple calls per patient, the primary call was used for patient-level analyses. For OHCA events where more than one person called about the event, one of these calls was determined to be primary. All calls, including non-primary calls, were used for caller and telecommunicator analyses.

Transcribers identified sex of caller and telecommunicator based on audio recordings and the study team assigned patient sex based on the transcript. We analyzed data descriptively. We present frequency of receipt of bystander CPR by the sex of the caller, patient, and telecommunicator.

Results

Patient characteristics

After excluding administrative calls,² calls with a “DNR” order in place,⁵ calls for problems other than OHCA,¹² calls at a healthcare site (24), patients under 18 (25), Non-English language calls,² those with no transcript (30), our final analysis sample included 197 calls, representing 188 patients. Of the cohort (n = 197 calls, 188 patients), Mean age was 56 ± 20 years and 37% of patients were female. Overall, 49% received CPR: 49% of these were female patients and 51% male patients (p = 0.66).

Call and caller characteristics

The mean call length was 10 ± 3 min and 60% of the callers were female. Overall, 60% of calls occurred during daytime hours (8 am–8 pm). Among calls placed by female callers, 63% were placed during the day, compared to 56% of calls placed by male callers (p = 0.37). Eighty percent of all calls were made in private locations. Among calls placed by female callers 83% were placed in private locations compared to 74% of calls placed by male callers (p = 0.15).

Twenty percent of calls were made by immediate family, 31% were other family (e.g., parent, uncle), 22% by significant others (e.g., spouse, girlfriend), 5% were friends, 18% had no relationship to the patient, while 24% the relationship was unidentified. Bystanders in addition to the caller were present in 49% of calls. Forty-eight percent of female callers and 51% of male callers performed CPR (p = 0.77).

Telecommunicator sex

In 71% of the calls, the telecommunicator was female. When the telecommunicator was female, 47% of callers performed CPR. When the telecommunicator was male, 57% of callers performed CPR (p = 0.21).

Association of sex of the caller and telecommunicator with receipt of CPR for females

When the caller was female, 44% of female and 51% of male patients received CPR (p = 0.54). When the caller was male, 51% of female and 53% of male patients received CPR (p = 1.0) (Fig. 1A).

Fig. 1.

Differences in receipt of bystander CPR by telecommunicator and caller sex (N = 197 calls, 188 patients).

When the telecommunicator was female, 40% of the females and 51% of males received CPR (p = 0.21). When the telecommunicator was male, 65% of females and 53% of males received CPR (p = 0.42) (Fig. 1B). Fig. 1C illustrates receipt of CPR for male and female patients for each caller and telecommunicator sex pairing.

Association of telecommunicator-caller sex pairings with CPR for female patients

The highest rate of CPR performance on female patients occurred when the caller and telecommunicator were both male. The lowest rate of CPR performance for female patients occurred when the telecommunicator and caller were both females.

Discussion

Females with suspected OHCA received CPR less frequently than males. Rates of CPR for females differed descriptively based on the sex of the caller and telecommunicator. The highest rates of CPR for females were observed when both the caller and telecommunicator were male. While none of the comparisons was statistically significant, the associations are worth exploring.

Mutual trust and effective communication are key to facilitating collaboration between callers and telecommunicators.¹³ Understanding the role that the sex of both the telecommunicator and the caller play in establishing or hindering both this trust and communication may be an important step in developing interventions to improve the effectiveness of telecommunicator CPR. Existing literature suggests several possible explanations for differences in CPR for different caller-telecommunicator sex pairings.

Benefits of telecommunicators in promoting CPR

While telecommunicator CPR improves CPR for patients experiencing OHCA in public,6, 9, 11 interventions to improve telecommunicator CPR have had mixed results.9, 11 In a review of the literature around barriers and facilitators of telecommunicator recognition of OHCA, sixteen studies examined factors related to the caller, telecommunicator, or the interaction between the two that promoted or hindered telecommunicator recognition of OHCA.¹⁴ Caller factors included caller emotional status and caller description of the situation.¹⁴ Telecommunicator factors included telecommunicator approach and specific telecommunicator behavior (e.g., insufficient questioning, inaccurate use of medical language).¹⁴ The way the caller responded to telecommunicator questions also impacted the effectiveness of the telecommunicator response.¹⁴ However, in this review, no studies assessed how the characteristics of the caller impacted OHCA recognition (21 studies investigated patient characteristics),¹⁴ meaning that the impact of caller sex on telecommunicator-caller communication and patient outcomes in OHCA have not been examined in previous literature.

Patient-provider communication

While sex and its role in communication have not been studied around OHCA response, sex concordance between patient and provider has been identified as a possible factor influencing patient outcomes, usually in a positive direction.15, 16, 17 Evidence from this literature may help guide future study of telecommunicator-caller interactions. Higher patient adherence has been observed in sex concordant patient-provider relationships.¹⁶ A recent study on both sex and racial/ethnic concordance between patients and providers did not show sex concordance as impacting patient trust of the physician, although racial/ethnic concordance did increase trust. This suggests that perhaps telecommunicator and caller concordance impacts receipt of CPR by the patient however, research in this area is preliminary and warrants further investigation. It is also of interest to investigate whether sex concordance matters more (or less) for females or males.

Implicit bias

Implicit bias may also impact telecommunicator-caller dyads differentially depending on the sex of both parties. Implicit bias contributes to observed disparities in healthcare. Implicit bias in patient-provider interactions impacts many aspects of care including communication, treatment adherence, provider decision making, and provider’s perspective of patient’s likelihood to adhere to treatment.¹⁸ Mechanisms for this include differential care and trust between provider and patient.¹⁸ Implicit bias has not been explored well in the prehospital setting specific to OHCA to date.

Barriers exist for initiating CPR on females related to discomfort with exposing or touching the patient’s breasts, fear of being accused of assault or inappropriate behavior, and perceiving the female’s body as vulnerable.6, 8 The specific barriers that impact a patient in a specific OHCA event may also vary based on the sex of the bystander responding.⁴ Sex of the caller and telecommunicator may also influence CPR on the patient. Potential explanations for the differences in CPR for different telecommunicator-caller dyads found in this study include (1) the influence of sex concordance on a caller’s comfort in following telephone-assisted CPR directions; (2) implicit bias on the part of the telecommunicator (e.g., perceptions of women callers as less capable by the telecommunicator; or (3) implicit bias on the part of the caller (e.g., perceiving women telecommunicators as less capable or authoritative). Further research is needed to understand the mechanisms underlying these differences. Evidence from interventions aimed at improving patient-provider interactions may also provide important insight into potential interventions to improve outcomes and address these known disparities.

Limitations

This is a small observational study that reviewed a limited number of calls, especially given the number of variables of interest in this dataset. As such, we used descriptive statistics alone and all observed trends or associations are hypothesis generating only. All calls from this study were from a single site that uses the Medical Priority Dispatch System (MPDS). This may limit generalizability to agencies that use other protocols or do not use protocols for directing caller-administered CPR.

Additionally, data collection from audio/transcripts of 9-1-1 calls is challenging and includes a fair amount of inference. Most notably, since we could not determine the biological sex of the caller, telecommunicator, or patient, when we refer to sex, we mean perceived sex.

Conclusion

We found differences in bystander CPR for women based on the sex of the caller and telecommunicator. The highest rates of CPR for women were observed when both the caller and telecommunicator were men. Future studies should examine sex-related barriers to telecommunicator-caller communication.

Statement of ethics review

The Duke University Health System Institutional Review Board deemed this protocol exempt from IRB review.

CRediT authorship contribution statement

Lauren Hart: Writing – original draft, Visualization, Formal analysis, Data curation. Anjni P. Joiner: Writing – review & editing, Methodology. Monique A. Starks: Writing – review & editing. Heather A. King: Writing – review & editing. Lee Van Vleet: Writing – review & editing, Methodology. Fahad J. Siddiqui: Writing – review & editing, Methodology, Conceptualization. Charlotte Patterson: Writing – review & editing, Investigation. Angel Gonzalez: Writing – review & editing, Investigation. Pedro Gomez Altamirano: Writing – review & editing, Investigation. Truls Østbye: Writing – review & editing. Janet P. Bettger: Writing – review & editing. Audrey L. Blewer: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Dr. Blewer is funded through an NIH NIAMS scholar award (K12AR084231), NIH NHLBI (5UH3-HL1463505 and 1R21HL175712-01), NIH NIMHD (1-R21-MD019370-01A1) the American Heart Association research grant (23HERNPRNPRH1150361), and a Laerdal Foundation grant. Dr. Blewer has also received in-kind support from the American Heart Association and World Point to support this work. Dr. King has funding awarded to her institutions from the Veterans Affairs (VA); National Institutes of Health (NIH); and Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. Dr. Joiner is funded through a Laerdal Foundation grant.