Mobile applications enhance out-of-hospital cardiac arrest outcomes: a systematic review and meta-analysis

Abstract

Introduction

Mobile applications, as innovative tools for promoting bystander cardiopulmonary resuscitation (CPR), have demonstrated potential to improve outcomes for patients experiencing out-of-hospital cardiac arrest (OHCA). This meta-analysis sought to systematically review the technical features of existing mobile applications and evaluate their impact on OHCA patient outcomes under various emergency response strategies. The findings aimed to guide the development and optimization of prehospital public emergency response systems.

Methods

A systematic search was conducted in databases including China National Knowledge Infrastructure (CNKI), Wanfang Database, Chinese Scientific Journals Database (VIP), SinoMed, PubMed, Embase, Web of Science, and the Cochrane Library, from inception to August 2023. The included studies involved notifying citizens via text messages or smartphone applications to act as first responders or volunteers in OHCA cases. Using a random effects model and subgroup analysis, we synthesized the results to identify sources of heterogeneity and assess outcomes.

Results

Thirteen mobile applications were included, with an average activation rate of 35.3% among patients and a volunteer arrival rate of 53.3%. Compared to traditional emergency medical services, mobile applications significantly improved survival to discharge or 30-day survival rates (RR = 1.34, 95% CI: 1.24–1.44; P < 0.05), return of spontaneous circulation (ROSC) rates upon hospital admission (RR = 1.23, 95% CI: 1.09–1.40; P < 0.05), bystander CPR rates (RR = 1.25, 95% CI: 1.13–1.37; P < 0.05), and bystander defibrillation rates (RR = 1.23, 95% CI: 1.00–1.51; P = 0.05). Subgroup analyses revealed consistent results for bystander CPR rates and survival outcomes, while variations in defibrillation rates and ROSC at admission were observed, indicating potential influences of application design and operational parameters.

Conclusions

This study highlighted the significant potential of mobile applications in enhancing bystander interventions and improving patient outcomes. Addressing challenges such as improving access to automated external defibrillators and raising public awareness remained essential to maximizing their overall effectiveness.

PROSPERO registration number

CRD42023477676.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-025-12416-2.

Introduction

Out-of-hospital cardiac arrest (OHCA) poses a global public health challenge, characterized by poor treatment outcomes and a grim prognosis for patients. The annual incidence rates in Europe, North America, Australia, and China are reported as 86.4, 98.1, 112.9, and 97.4 per 100,000 individuals respectively [1, 2]. Despite considerable efforts, survival to discharge remains dismally low, ranging from 8.6% during 1976–1999 to 9.9% in 2010–2019 [3]. Effective cardiopulmonary resuscitation (CPR) has become a central focus of research, with studies indicating that increased bystander involvement and early defibrillation are associated with improved survival rates [4]. Consequently, attention has turned to enhancing cardiac arrest recognition, expanding public CPR education, and promoting the use of automated external defibrillators (AEDs).

Despite substantial investments in public training, bystander CPR rates remain disappointingly low. Many individuals who have undergone basic life support training report never having performed CPR or used an AED [5]. Since the concept was first documented in 2007 [6], mobile applications have been increasingly recognized for their potential to address these challenges. These technologies utilize smartphone capabilities, including mobile positioning and short messaging services, to promptly notify nearby volunteers to provide CPR or AED assistance before emergency medical services (EMS) arrive. This approach significantly reduces the "blank time" during emergency response [7]. More recently, the integration of 5G networks, mobile internet, and video technologies into smart emergency response and dispatch systems has further enhanced prehospital emergency care capabilities.

Mobile applications are now widely implemented across the globe [8–11]. However, most existing studies are either observational or based on randomized controlled trials (RCTs) with limited sample sizes, which may constrain the generalisability of their findings. Furthermore, considerable heterogeneity exists in terms of technological design, responder coverage, and integration with prehospital emergency services, raising questions about whether such variations influence patient outcomes [12]. A global systematic review conducted in 2020 [13] highlighted the potential of mobile applications to reduce intervention times and improve outcomes by mobilizing citizens as first responders for OHCA patients [7]. However, the precise impact of these variations remains unclear. Given the emergence of new evidence in recent years, this meta-analysis aimed to systematically review the technical features of mobile applications and assess their impact on OHCA patient outcomes under various emergency response strategies. The findings were expected to provide valuable insights for the development and optimization of prehospital public emergency response systems.

Methods

This systematic review and meta-analysis strictly adhered to the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) Statement 2020 [14], with the protocol preregistered on PROSPERO prior to data extraction (ID: CRD42023477676). As the study exclusively utilized data derived from previously published articles, ethical approval and informed consent were not required.

The research question was developed within the Population, Intervention, Comparison, and Outcome (PICO) framework, as follows: Does notifying nearby volunteers via a mobile application to perform CPR or defibrillation in cases of OHCA improve survival rates compared to standard EMS?

Search strategy

The search was conducted across multiple databases, including PubMed, Embase, Web of Science, the Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang Data, Chinese Scientific Journals Database (VIP), and SinoMed. The search spanned from the inception of the databases to August 1, 2023. Search terms comprised a combination of Medical Subject Headings (MeSH) and keywords related to OHCA and mobile applications. Only clinical trials and RCTs were included, while reviews, commentaries, and conference papers were excluded. Additionally, the references of included studies were manually reviewed to identify further relevant citations. The detailed search strategy is provided in Supplementary Fig. 1.

Study selection

The search results from each database were imported into EndNote X9, and duplicates were removed. Two researchers (QQT and XHL) independently screened titles and abstracts of potentially relevant studies. The full texts of shortlisted studies were then reviewed in detail to confirm eligibility. Discrepancies were resolved through discussion with a third researcher (RJH). Studies were included if they involved notifying citizens via text messages or smartphone applications to act as first responders or volunteers in OHCA scenarios. Exclusion criteria included data inaccuracies, inaccessible full texts, involvement of professional volunteers (e.g., police, firefighters), and studies published in languages other than Chinese or English.

Data extraction

Two researchers (QQT and JML) independently extracted data using a standardized form. Extracted data included basic study information (first author, country, journal, publication year, study type) and details of the mobile application (name, development year, country of use, application type, volunteer location, alert range, volunteer mobilization mode, and registration requirement). Key outcome indicators included the bystander CPR rate, application activation rate, volunteer arrival rate, AED retrieval rate, bystander defibrillation rate, return of spontaneous circulation (ROSC) rate, survival to discharge or 30-day survival rate, and favorable neurological recovery rate.

Quality assessment

The risk of bias and quality were independently assessed by two researchers (LL and XYP), with any disagreements resolved by consultation with a third researcher (RJH). For randomized controlled trials, the Cochrane Risk of Bias 2.0 (RoB 2) tool [15] was employed, evaluating five domains: the randomization process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selection of reported results. Bias was classified as "low risk", "some concerns" or "high risk" based on predefined criteria.

For observational studies, the Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) tool [16] was used, which includes additional domains such as confounding factors, participant selection, and intervention classification. Each domain was rated as "low risk", "moderate risk", "serious risk", "critical risk", or "no information". An overall risk of bias rating was assigned based on these individual assessments, and the results were visualized using the robvis tool [17].

The overall certainty of evidence was evaluated using GRADE (Grading of Recommendations Assessment, Development and Evaluation) criteria through the GRADEpro tool, classifying the evidence as very low, low, moderate, or high.

Statistical analysis

Statistical analyses were conducted using RevMan 5.4. All outcomes were presented as the median and interquartile range [M (Q1, Q3)]. The Mantel–Haenszel method was used to calculate pooled risk ratios (RR) with 95% confidence intervals (CI) for dichotomous outcomes. Heterogeneity was quantified by [tau]² and I² statistic. A fixed effects model was applied if P_interaction > 0.1 and I² < 50%, indicating low heterogeneity. If P_interaction ≤ 0.1 or I² ≥ 50%, suggesting significant heterogeneity, a random effects model was utilized. Subgroup analyses were performed, stratified by volunteer mobilization mode, application type, volunteer location, and activation radius. Sensitivity analyses were conducted by excluding individual studies systematically to identify potential sources of heterogeneity and ensure the robustness of the meta-analysis findings. Statistical significance was defined as P < 0.05.

Results

Study selection

An initial search identified 2220 citations, supplemented by four additional references through citation tracking. After removing duplicates, 1,500 citations underwent preliminary screening of titles and abstracts, with 93 proceeding to full-text review. Ultimately, 25 citations were included in the systematic review. Among these, one RCT [9] and 12 cohort studies [8, 11, 18–27] formed the basis of a meta-analysis examining the impact of mobile applications on outcomes in patients with OHCA. The study selection process are depicted in Fig. 1.

Fig. 1.

PRISMA flow diagram

Study characteristics

Between 2011 and 2023, 25 studies were published, including one in Chinese [28], one RCTs [9], and 15 cohort studies [8, 11, 18–27, 29–31]. These studies assessed 13 distinct text messaging or smartphone applications across 12 countries. In the meta-analysis, one study was rated as having a low risk of bias [9], seven as moderate risk [8, 11, 20, 21, 23, 25, 26], and five as high risk [18, 19, 22, 24, 27] (Supplementary Fig. 2). A summary of the studies' key characteristics is provided in Table 1.

Table 1.

Characteristics of included studies

First author	Country	Journal	Year	Study design	Overall risk of bias
Scholten [32]	Netherlands	Resuscitation	2011	Cross-sectional study	N/A
Zijlstra [8]	Netherlands	Resuscitation	2014	Prospective cohort study	Moderate
Ringh [9]	Sweden	N Engl J Med	2015	RCT	Low
Brooks [10]	America	Resuscitation	2016	Cross-sectional study	N/A
Pijls [18]	Netherlands	Resuscitation	2016	Prospective cohort study	Serious
Caputo [29]	Switzerland	Resuscitation	2017	Prospective cohort study	N/A
Dainty [33]	Canada	JMIR Mhealth Uhealth	2017	Cross-sectional study	N/A
Smith [34]	England	Resuscitation	2017	Descriptive study	N/A
Berglund [35]	Sweden	Resuscitation	2018	Cross-sectional study	N/A
Pijls [30]	Netherlands	Eur Heart J Acute Cardiovasc Care	2018	Prospective cohort study	N/A
Auricchio [36]	Switzerland	Resuscitation	2019	Descriptive study	N/A
Lee [19]	Korea	Resuscitation	2019	Prospective cohort study	Serious
Pips [37]	Netherlands	Neth Heart J	2019	Cross-sectional study	N/A
Andelius [11]	Denmark	J Am Coll Cardiol	2020	Prospective cohort study	Moderate
Blewer [20]	Singapore	Lancet Public Health	2020	Prospective cohort study	Moderate
Derkenne [21]	French	Acad Emerg Med	2020	Prospective cohort study	Moderate
Sarkisian [31]	Denmark	Resuscitation	2020	Retrospective cohort study	N/A
Stroop [22]	Germany	Resuscitation	2020	Prospective cohort study	Serious
Ng [38]	Singapore	Prehosp Emerg Care	2021	Descriptive study	N/A
Wong [23]	Singapore	Ann Acad Med Singap	2021	Retrospective cohort study	Moderate
Ran [28]	China	Chin J Emerg Med	2022	Descriptive study	N/A
Nielsen [24]	Denmark	Front Cardiovasc Med	2022	Prospective cohort study	Serious
Smith [25]	England	Eur Heart J Acute Cardiovasc Care	2022	Retrospective cohort study	Moderate
Jonsson [26]	Sweden	J Am Coll Cardiol	2023	Retrospective cohort study	Moderate
Oosterveer [27]	Netherlands	Neth Heart J	2023	Retrospective cohort study	Serious

N/A not applicable

Mobile application performance

From 2006 to 2021, a variety of mobile applications were developed across countries such as the Netherlands, Sweden, the United States, and China. Upon detecting an OHCA incident, these applications dispatched alerts to nearby volunteers via smartphones, with three using text messages, eight relying on smartphone applications, and two transitioning from text-based messaging to app-based notifications. Volunteer location was determined using GPS, registered addresses, community data, or electronic identities. Alert ranges varied between 300 and 5,000 m, with each incident triggering between two and 30 notifications.

The AED retrieval range extended up to 750 m and can display nearby or the nearest available AED locations. Volunteer mobilization included the following three approaches: (1) Priority CPR response: Volunteers were primarily dispatched to the OHCA scene to administer CPR [9, 10, 19, 20, 22, 23, 25, 26, 28, 29, 32–34, 36, 38], occasionally retrieving an AED when necessary, which significantly reduced response times and expedited the initiation of CPR. (2) Priority AED retrieval: The nearest volunteers were prioritized in retrieving an AED before proceeding to the scene, while additional responders focused on the patient [8, 18, 27, 30, 37], thereby ensuring swift access to defibrillation equipment; (3) Coordinated team response: Volunteer teams were organized to divide tasks between on-site resuscitation and AED retrieval, facilitating a coordinated and efficient response to the emergency [11, 21, 24, 31, 35]. Detailed information is summarized in Table 2.

Table 2.

Mobile application performance

Application name	Year	Country	Volunteer location	Alert range	AED retrieval method	Volunteer mobilization mode	Registration requirement
AED-Alert [32]	2008	Netherlandsa	Registered address	1000m	-	AED trained person retrieves AED	BLS
TM-Alert [8, 18, 27, 30, 37]	2010 2016	Netherlandsa	Registered address	1000m 30 alarms within 750 m	1000m 750m	1/3 for CPR, 2/3 for AED (within 500m)	BLS or AED
SMS-lifesavers [9, 35]	2010 2015	Swedena+b	GPS	500m 20 alarms within 240-1200m	2400m	10 for CPR, 10 for AED	CPR
Pulse Point Respond [10, 33]	2010	USA Canadab	GPS	400m	-	-	BLS or AED
First Responder [29, 36]	2006 2014	Switzerlanda+b	City or community GPS	Automatically exclude people later than EMS	-	Prioritize CPR	BLS or AED, 2/year
Good SAM [25, 34]	2015	UKb	GPS	3 alarms within 300m	Indicate nearby AED	Automatically alerts the next responder	CPR
Lee [19]	2015	South Koreaa	Registered address	Within the same area	SMS to send AED location	-	CPR
First AED [31]	2012	Denmarkb	GPS	9 alarms within 5000m	Indicate the nearest AED	Select the three nearest responders, one for AED and two for CPR	BLS/year
Heart Runner [11, 24]	2017	Denmarkb	Registered address	20 alarms within 1800m	Indicate the nearest AED	The first responder for CPR, and the remaining four for AED	Recommend to learn CPR or AED
My Responder [20, 23, 38]	2015	Singaporeb	Electronic identity	400m, taxi driver assists OHCA within 1500 m	Indicate nearby AED	-	CPR and AED
Staying Alive [21]	2017	Franceb	GPS	500m	Indicate the nearest AED	The second bystander retrieves the AED	BLS
Mobile Rescuers [22]	-	Germanyb	GPS	2 alarms nearby	-	-	BLS
Interconnect first aid [28]	2021	Chinab	GPS	1000-1500m	Indicate the nearest AED	-	BLS, 2/year

GPS global position system, SMS short messaging service

^atext message

^bsmartphone application

The median activation rate for OHCA was 35.3%, with 53.3% of volunteer arriving at the scene. Denmark [11, 24] showed exceptional performance, with over 85.1% of volunteers arriving before the EMS [31]. Pooled data revealed median rates of bystander CPR and defibrillation at 54.8% and 9.4%, respectively. ROSC rates upon hospital admission were 35.1%, while survival to discharge or 30-day survival rates reached 17.1%. Three studies [19, 22, 27] reported favourable neurological outcomes, as detailed in Table 3.

Table 3.

Mobile application-related outcomes (%)

Application name	Activation rate	Volunteer on-scene arrival rate	Bystander CPR rate	AED retrieval rate	Bystander defibrillation rate	ROSC rate upon hospital admission	Survival to discharge rate	Favourable neurological recovery rate
AED-Alert [32]	-	579/2047a	37/579a	5/579a	37/579a	-	-	-
TM-Alert [8, 18, 27, 30, 37]	50.7	68.9	49.8	-	64.9	41.6	27.1	12.8
SMS-lifesavers [9, 35]	35.6	58.9	27.3	8.6	2.0	29.4	11.2	-
Pulse Point Response [10, 33]	-	135/1199a	11/14a	-	-	-	-	-
First Responder [29, 36]	36.4	53.3	-	-	-	-	-	-
Good SAM [25, 34]	6.7	16.0	67.9	-	9.4	39.6	17.6	-
Lee [19]	35.3	-	59.8	-	-	13.1	12.7	8.3
First AED [31]	-	61.9	-	65.6	-	-	17.1	-
Heart Runner [11, 24]	59.7	70.8	126/184a	91/184a	19/184a	27.4	16.1	-
My Responde [20, 23, 38]	-	49.2	64.4	-	5.5	30.6	4.8	-
Staying Alive [21]	8.9	137/762a	16/137a	18/137a	18/137a	48.0	34.8	-
Mobile Rescuers [22]	13.0	46.0	33.0	-	25.0	45.0	18.0	11.0
Interconnect first aid [28]	33.8	6.6	33/169a	14/35a	3/14a	-	-	-
[M (Q1, Q3)]	35.3 (11.0, 43.6)	53.3 (31.0, 65.4)	54.8 (31.6, 65.3)	-	9.4 (3.8, 44.9)	35.1 (27.9, 44.1)	17.1 (11.9, 22.5)	-

The volunteer on-scene arrival rate is defined as the number of OHCA with volunteer participation divided by the total number of OHCA and is typically reported as the median. However, some studies chose to present outcome from the perspective of volunteer interventions, defining the on-scene arrival rate as the number of volunteers arriving at the scene divided by the number of volunteers who received the alert. To ensure comprehensive data representation, these instances were marked with an 'a', but were not included in the calculation of the median

Effect of mobile application on survival

Twelve studies demonstrated a significant improvement in survival to discharge or 30-day survival rates with the activation of mobile applications (9.6% vs. 8.0%; RR = 1.34, 95% CI: 1.24–1.44; P < 0.05), as depicted in Fig. 2. Subgroup analyses stratified by volunteer mobilization modes, application types, volunteer locations, and activation radii consistently reinforced these findings. Notably, there were markedly higher in coordinating team response (RR = 1.57, 95% CI: 1.18–2.10; P < 0.05). According to the GRADE criteria for quality, the randomized trials were rated as moderate, while observational studies were assessed as very low (Supplementary Table 2), due to variability in survival reported across studies. A funnel plot illustrating potential publication bias is presented in Supplementary Fig. 3.

Fig. 2.

Forest plot showing the association between volunteer mobilization modes and survival to discharge or 30-day survival rates

Nine studies encompassing 11,006 participants reported on ROSC rates upon hospital admission. Given the substantial heterogeneity, a random effects model was applied. The findings indicated a significant improvement in ROSC rates with mobile application activation (24.8% vs. 22.0%; RR = 1.23, 95% CI: 1.09–1.40; P < 0.05) (Fig. 3). Subgroup analyses similarly highlighted enhanced ROSC rates when bystander CPR was prioritized (RR = 1.24, 95% CI: 1.05–1.47; P < 0.05) and alert radius exceeded 500 m (RR = 1.30, 95% CI: 1.09–1.55; P < 0.05) (Supplementary Table 1). The quality was rated as moderate for randomized trials and very low for observational studies.

Fig. 3.

Forest plot showing the association between volunteer mobilization modes and ROSC rates upon hospital admission

Effect of mobile application on bystander interventions

Eleven studies encompassing 31,784 participants demonstrated a significant increase in bystander CPR rates following the activation of mobile applications (69.1% vs. 55.4%; RR = 1.25, 95% CI: 1.13–1.37; P < 0.05) (Fig. 4). Subgroup analyses consistently confirmed the robustness of these findings. Among the different volunteer mobilization strategies, prioritizing CPR response yielded the most favorable outcomes (RR = 1.32, 95% CI: 1.08–1.62; P < 0.05), followed by prioritizing AED retrieval and coordinated team response. The quality was rated as high for randomized trials and very low for observational studies (Supplementary Table 2).

Fig. 4.

Forest plot showing the association between volunteer mobilization modes and bystander CPR rates

Twelve studies encompassing 33,389 participants examined the impact of mobile applications on bystander defibrillation rates. The findings indicated a modest increase in bystander defibrillation, though the effect size was limited, and the possibility of chance could not be excluded (12.2% vs. 8.6%; RR = 1.23, 95% CI: 1.00–1.51; P = 0.05) (Fig. 5). Sensitivity analysis highlighted the notable influence of a single study [8], which demonstrated a stronger effect in EMS scenarios (RR = 1.35, 95% CI: 1.14–1.60; P < 0.05). Subgroup analyses further illuminated a higher bystander defibrillation rate in cases involving coordinated team responses (RR = 2.57, 95% CI: 1.62–4.06; P < 0.05) and in those utilizing smartphone applications (RR = 1.45, 95% CI: 1.11–1.88; P < 0.05) (Supplementary Table 1). The quality was rated as moderate for randomized trials and very low for observational studies.

Fig. 5.

Forest plot showing the association between volunteer mobilization and bystander defibrillation rates

Discussion

Summary of findings

This systematic review and meta-analysis comprehensively assessed the impact of mobile applications on patients with OHCA, confirming significant positive effects in improving survival to discharge or 30-day survival rates, ROSC rates upon hospital admission, bystander CPR rates and defibrillation rates. Subgroup analyses further explored the influence of volunteer mobilization modes, application types, volunteer locations, and activation radii on these outcomes, providing deeper insights into the factors shaping their effectiveness.

The stratified analysis of volunteer mobilization modes revealed more significant effects in improving survival rates, ROSC rates, bystander CPR rates and defibrillation rates when prioritizing CPR response or coordinating team-based interventions. These could be attributed to the rapid reduction in emergency response and CPR initiation time when volunteers were directly dispatched to the cardiac arrest scene to perform CPR without initially prioritizing AED retrieval [22]. This finding highlighted the synergistic role of early bystander resuscitation, aligning with previous research [39]. While prioritizing AED retrieval significantly shortened defibrillation time [8], its relative effectiveness may be limited as it can delay the initiation of early CPR, especially in residential or rural areas [36]. However, retrieving an AED before reaching the patient retains valuable when other responders have already initiated CPR or high-quality bystander CPR is in progress. Regional differences in public awareness, AED availability, and cultural attitudes toward bystander intervention remained significant barriers. Studies [37] have shown that practical training and simulated emergency situations were effective in building public confidence and improving response readiness. Innovative strategies, such as drone-assisted AED delivery [40, 41], have been implemented in some areas to enhance AED accessibility. Despite these advancements, the rate of bystander-initiated defibrillation remained relatively low, highlighting persistent challenges in public awareness and AED usage. To maximize the effectiveness of mobile applications in OHCA management, it should focus on optimizing volunteer mobilization, increasing AED availability, and implementing educational programme to raise public confidence and competence in performing CPR and using AEDs.

The application type-stratified analysis revealed that smartphone APP significantly improved survival to discharge rates, bystander CPR rates, and defibrillation rates compared to text messaging systems. This advantage likely stems from the ability to deliver real-time information, precise geographic positioning, and advanced communication functionalities. A study [29] investigating laypersons alerted via an APP to initiate earlier CPR recorded 593 OHCAs, of which 198 cases were notified by SMS and 134 cases by the APP. The median time for first responder or lay responder to arrival at the scene was significantly shorter with the APP (3.5 min vs. 5.6 min). Additionally, the proportion of lay responders arriving first at the scene increased significantly (70% vs. 15%). Similarly, an APP developed on Langland Island, Denmark [31], revealed that in more than four-fifths of OHCA cases, volunteer responders dispatched via the APP arrived at the scene earlier than EMS. Building on this foundation, a dynamic response mechanism incorporating voice guidance and video connectivity could be implemented in the future. This would facilitate real-time communication between callers, emergency command centers, and volunteers, improving coordination and rescue efficiency, particularly in regions with low public awareness of first aid.

Subgroup analysis revealed that GPS-enabled mobile applications had a significant positive impact on improving survival to discharge rates, while non-GPS systems demonstrated notable benefits in increasing bystander CPR rates and ROSC rates. However, neither group showed a significant difference in bystander defibrillation rates compared to EMS. This discrepancy may be attributed to the limited effectiveness of defibrillation in areas with relatively low AED utilization [42], despite mobile applications offering precise location information to facilitate rapid identification and timely emergency responses. Factors such as limited public awareness of AED use and insufficient AED availability may further hinder defibrillation outcomes [43]. Moreover, due to privacy restrictions, most systems are unable to track volunteers in real-time. Instead, they rely on preregistered residential or contact addresses to issue alerts, which undermines the accuracy and reliability of the results to some extent. Future research should prioritize raising public awareness of AED use while addressing challenges related to the accuracy of volunteer location tracking and improving AED availability. These efforts could encourage wider AED utilization and enhance defibrillation outcomes, particularly in regions with limited resources.

A stratified analysis based on volunteer activation radius revealed that mobile application with an alert radius of ≤ 500 m significantly improved survival rates and bystander CPR rates, while application with an alert radius of > 500 m were associated with a significant increase in ROSC. This may be attributed to the smaller radius enabling emergency responders to arrive at the scene more quickly, thereby enhancing the timeliness of interventions and improving patient outcomes, whereas a larger radius allows for broader volunteer engagement. However, no significant differences in bystander defibrillation rates were observed between the two groups compared with EMS, aligning with findings from the stratified analysis on volunteer positioning. A study [44] demonstrated that when one or more volunteers arrived before EMS, the proportion of bystander CPR and defibrillation increased, with a further rise in defibrillation as the number of volunteers increased. This suggests that the optimal activation radius and the number of volunteers depend on multiple factors, including volunteer density, AED density, public awareness of first aid, and EMS response times. Currently, there is no international consensus on the ideal number of activated volunteers or the most effective activation radius. Existing studies have reported a minimum activation radius of 300 m [25, 34] and a maximum radius of 5000 m [31], with alert ranges extending from the nearest accessible responders to as many as 30 registered volunteers. Moreover, an observational study in the Netherlands [45] indicated that having two AEDs per square kilometre and more than 10 volunteers was significantly associated with shorter defibrillation times. Looking ahead, it is essential to further optimize the volunteer activation radius and numbers, increase the density of AED distribution, and enhance public awareness of first aid to improve the efficiency of emergency responses and outcomes for OHCA patients.

Implications for clinical practice and research

Compared to previous studies [13], this research observed higher rates of bystander CPR and defibrillation, which can be attributed to the expanded scope of the literature search that included nine newly published studies. One study [8] noted that, compared to traditional EMS responses, the proportion of bystanders connecting an AED at the scene was significantly lower when the mobile application was activated (7.3% vs. 15.4%; P < 0.001). This disparity was primarily due to 41.8% of OHCA incidents not triggering the volunteer alert system. The reasons cited included the anticipated short arrival time of EMS and the dispatchers’ failure to identify cardiac arrest cases. Although this finding significantly influenced the overall results, it was considered to have clinical relevance and thus included in the analysis. Additionally, some of the studies included in this review were conducted during the COVID-19 pandemic. During this period, physical contact restrictions may have reduced both the willingness and ability of bystanders and volunteers to provide emergency assistance. For example, due to concerns about infection risk, volunteers may have been reluctant to perform chest compression or other first aid procedures requiring close contact, which could have directly affected the efficiency and effectiveness of the interventions.

Limitations

The limitations of this study are as follows: Firstly, the included studies varied significantly in design, population characteristics, and methodologies, contributing to heterogeneity in the meta-analysis. Although subgroup analyses were conducted, the differences between studies may have impacted the stability of the results. Secondly, despite employing rigorous bias risk assessments, some studies exhibited moderate or high risks, potentially affecting the overall quality of the evidence. While funnel plots were used to identify potential publication bias, the risk of bias could not be entirely eliminated. Future research should focus on conducting standardized, high-quality studies to enhance the reliability of findings.

Conclusion

This study evaluated the effectiveness of mobile applications under different emergency response strategies through subgroup analyses. The results highlighted the significant potential of mobile applications in enhancing bystander interventions and improving patient outcomes. They provided valuable guidance for addressing challenges such as improving access to AED and raising public awareness, thus offering insights into optimizing mobile applications design and refining emergency response practices.

Abbreviations

OHCA

Out-of-hospital Cardiac Arrest

CPR

Cardiopulmonary Resuscitation

CNKI

China National Knowledge Infrastructure

VIP

Chinese Scientific Journals Database

ROSC

Return of Spontaneous Circulation

AED

Automated External Defibrillator

EMS

Emergency Medical Services

GPS

Global Position System

RCT

Randomized Controlled Trial

Authors’ contributions

QQT, MHZ, and RJH conceived the study. XHL and JML searched and selected the studies and extracted and analyzed the data. LL, XYP and HMG reviewed and reanalyzed the extracted data. All authors provided advice on methodology and data analysis. QQT, MHZ, and RJH drafted the manuscript. All authors contributed substantially to the revision.

Funding

The study was supported by the research and talent development fund of Kweichow Moutai Hospital (MTyk2022-18), the "Thousand" level talents of high-level innovative talents in Guizhou Province (xmrc/20240205) and the 2023 graduate research fund project, Zunyi Medical University (ZYK266).

Data availability

The datasets supporting the conclusions of this article are included within the article and its supplement.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.