Abstract
Objective:
This prospective cohort study aimed to evaluate the clinical efficacy of a 5G-integrated emergency full-process system in telemedicine, specifically assessing its impact on prehospital-to-in-hospital information transfer, remote triage efficiency, and teleconsultation outcomes.
Methods:
A prospective single-center cohort study was conducted from January to December 2024, involving 300 patients in the 5G system group and 300 matched patients in the traditional care group. The primary endpoints included prehospital information warning time, remote rescue success rate, and remote triage accuracy.
Results:
The 5G system significantly reduced the prehospital warning time to 136 ± 22 sec (vs. 235 ± 41 sec in the control group, p < 0.001), increased the remote rescue success rate to 98.9% (p < 0.001), and improved remote triage accuracy to 96.3% (p < 0.001). Most notably, the 30-day mortality rate was significantly lower in the 5G group (4.0% vs. 9.3% in the control group, p = 0.007). Additionally, data reporting completeness for telemedicine cases improved by 32.2%, accompanied by a 45% reduction in reporting time.
Conclusion:
The 5G emergency full-process system demonstrates substantial advantages in facilitating real-time tele-emergency services, providing strong evidence to support its integration into evidence-based telemedicine protocols.
Keywords
Introduction
Research background
Telemedicine has become a pivotal strategy for mitigating health care disparities. However, traditional systems face significant challenges in real-time data transmission during emergency care. The inherent limitations of 4G technology—such as latency exceeding 500 milliseconds and bandwidth below 100 megabits per second—substantially constrain its application in time-sensitive emergency settings. In such scenarios, delays in remote consultation or prehospital data transfer are directly correlated with adverse patient outcomes. 1 The advent of fifth-generation (5G) mobile networks, characterized by ultra-low latency (<50 ms) and high bandwidth (up to 1 Gbps), presents unprecedented opportunities for redefining emergency telemedicine services. 2
Literature gap
Recent research has highlighted the potential of 5G in remote patient monitoring. Lloret et al. (2017) emphasized that 5G networks are essential for continuous remote patient surveillance due to their low latency and guaranteed bandwidth. 3 Ge et al. (2019) demonstrated that a 5G-based prehospital-to-in-hospital telemedicine platform could enable real-time video consultations and vital sign monitoring, effectively reducing response delays. 4 Despite these advancements, empirical evidence regarding the comprehensive application of 5G across end-to-end emergency telemedicine workflows—from pre-hospital remote triage to in-hospital teleconsultation—remains notably scarce.
Previous studies have established the feasibility and benefits of telemedicine in prehospital settings using earlier generation networks. For instance, tele- electrocardiograms (ECG) transmission for STEMI patients and video consultation for trauma victims have been shown to reduce door-to-balloon time and improve diagnostic accuracy.5,6 However, these systems often operated under the constraints of 4G technology, facing challenges such as video freezing, audio desynchronization, and unreliable data streams during high-speed transport or in crowded urban areas. Our study builds upon this existing body of work but fundamentally differs by leveraging the unique capabilities of 5G—ultra-low latency (<50 ms) and gigabit-level bandwidth—to enable a seamless, real-time, and immersive “telepresence” throughout the entire emergency process. This is not merely an incremental improvement but a qualitative shift that allows for complex remote-guided interventions (e.g., intubation, ultrasound) and AI-powered, real-time decision support that was previously impractical with conventional telemedicine systems.
Materials and Methods
Study design
A prospective, single-center cohort study was conducted at the Affiliated Second Hospital of Fujian Medical University between January and December 2024. The study protocol was approved by the hospital's Ethics Committee (reference number: 2023QH1263). All participants or their legal guardians provided written informed consent for the use of their medical data and telemedicine records, in accordance with the Declaration of Helsinki. Consent forms included specific clauses authorizing the transmission of 5G-collected biological signals (e.g., ECG, vital signs) and clinical images, with explicit statements on data anonymization procedures.
Participant enrollment
5G System Group: Comprised 300 patients transported via 5G-equipped ambulances who received emergency care integrated with telemedicine functionalities.
Control Group: Comprised 300 patients transported via conventional ambulances who received standard emergency care without 5G telemedicine integration.
Inclusion Criteria: Patients aged ≥18 years requiring prehospital emergency transport and subsequent in-hospital triage for chest pain, stroke, or trauma.
Matching Criteria: Patient groups were matched based on age (±5 years), gender, primary diagnosis, and prehospital severity score (Modified Early Warning Score, MEWS).
The study was conducted within the urban and suburban areas of Quanzhou, Fujian Province, China. Ambulances from both groups operated from the same central emergency medical service stations, ensuring that the geographic regions covered, including transport distances from the ambulance base to the patient’s location and from the scene to the hospital, were comparable between the two groups. Analysis confirmed no significant differences in these distances (p > 0.05). To assess the general similarity of the patient populations, we compared key demographic and socioeconomic factors. The two groups were well-balanced in terms of age, gender, and clinical severity at enrollment (as per the matching criteria). Furthermore, analysis of residential zip codes and hospital billing data (as a proxy for socioeconomic status and health care access) revealed no systematic differences between the groups.
The allocation of 5G-enabled ambulances was not randomized but was based on a phased implementation plan by the hospital administration. Specific ambulance units were outfitted with 5G technology based on availability and operational readiness. During the study period, when a 5G-equipped ambulance was available, it was dispatched according to standard EMS protocols, regardless of the patient’s condition. The control group consisted of patients transported by conventional ambulances from the same pool of stations during overlapping time periods. This pragmatic approach was chosen to evaluate the system's effectiveness in a real-world setting.
5G Emergency Telemedicine System Architecture
Prehospital remote monitoring subsystem
5G Mobile Tele-ICU: Ambulances were outfitted with 5G-enabled multi-parameter monitors (Mindray ePM12M) capable of transmitting 12-lead ECGs, invasive blood pressure readings, and oxygen saturation (SpO2) data at 100 Mbps to the hospital’s telemedicine center, with end-to-end latency of <50 ms. 7
VR Telepresence System: A 360°VR camera (Insta360 Pro 2) and augmented reality (AR) glasses (Hololens 2) facilitated real-time visualization of the prehospital scene by remote experts, enabling guided interventions (e.g., intubation, trauma management) through 5G low-latency transmission.
AI-Powered Remote Triage: An embedded artificial intelligence algorithm, trained on a dataset of 10,000 emergency cases, automatically generated triage levels from patient vital signs, demonstrating 98% agreement with manual triage assessments by emergency physicians.
Remote consultation subsystem
5G Teleconsultation Platform: A Health Insurance Portability and Accountability Act (HIPAA)-compliant platform supported high-definition 4K video consultations (1080p at 60 frames per second) between prehospital teams and in-hospital specialists. The platform featured end-to-end encryption and dynamic bandwidth allocation to ensure uninterrupted communication. 8
Tele-EMSBot: An AI chatbot integrated with natural language processing (NLP) capabilities standardized teleconsultation workflows. It automatically recorded time stamps from call initiation to specialist response and generated structured consultation summaries, streamlining documentation processes.
Telemedicine data management subsystem
Distributed Telemetry Database: A blockchain-based database ensured tamper-proof storage of telemedicine data. Real-time synchronization across 5G nodes enabled remote quality control and data integrity verification. 9
Automated Tele-Reporting: NLP algorithms converted teleconsultation audio into Health Level Seven (HL7)-compliant structured reports, reducing documentation time by 60% compared to manual entry methods.
Data Security and Privacy Protection: All patient data were de-identified using a blockchain-based distributed database (Hyperledger Fabric v2.2). Biological signals were transmitted via a dedicated 5G medical network with end-to-end encryption (AES-256 algorithm). Data transmission complied with HIPAA security rules, and all telemedicine records were stored with irreversible anonymization to protect patient privacy.
Outcome measures
Telemedicine-specific metrics:
Prehospital information warning time (time from scene arrival to telemedicine center activation).
Remote rescue success rate (proportion of patients surviving to hospital admission with telemedicine guidance).
Remote triage accuracy (agreement between remote triage level and in-hospital final diagnosis).
Teleconsultation response time (time from consultation request to specialist engagement).
Data management metrics:
Completeness of telemedicine case data reporting.
Average time for telemedicine report generation.
Primary clinical endpoint:
30-day mortality rate (all-cause mortality within 30 days following hospital admission).
Statistical analysis
Continuous variables are presented as mean ± standard deviation and were compared using independent samples t-tests. Categorical variables are reported as counts and percentages and were analyzed using chi-square tests. Statistical significance was defined as a two-sided p-value <0.05. All analyses were performed using IBM SPSS Statistics 26.0 (IBM Corp., Armonk, NY, USA).
Results
Baseline characteristics
The two study groups exhibited comparable demographic, clinical, and operational characteristics (Table 1, Figure 1), confirming the adequacy of the matching process.
Comparison of baseline characteristics between the 5G system group and the control group.
Baseline Characteristics and Comparability of the Study Participants
All p-values > 0.05 indicate no statistically significant differences between groups at baseline.
MEWS, Modified Early Warning Score.
Demographic and Clinical Characteristics:
Mean age: 52.3 ± 14.7 years in the 5G group versus 51.8 ± 15.2 years in the control group (p = 0.68). Male proportion: 58.7% in the 5G group versus 57.3% in the control group (p = 0.74). Mean MEWS score: 3.2 ± 1.1 in the 5G group versus 3.1 ± 1.2 in the control group (p = 0.59).
Primary diagnosis distribution (chest pain/stroke/trauma): 32.3%/29.4%/38.3% in the 5G group versus 31.7%/29.3%/39.0% in the control group (p = 0.92).
Geographic and Operational Comparability:
Mean distance from ambulance base to scene: 5.2 ± 2.1 km in the 5G group versus 5.4 ± 2.3 km in the control group (p = 0.45). Mean distance from scene to hospital: 8.7 ± 3.5 km in the 5G group versus 8.9 ± 3.8 km in the control group (p = 0.61). Urban versus suburban case distribution: 78% urban in the 5G group versus 76% urban in the control group (p = 0.65).
Socioeconomic Proxy Indicators:
Proportion of patients with government health insurance: 85% in the 5G group versus 83% in the control group (p = 0.55). Distribution across residential area socioeconomic tiers (based on zip code analysis) showed no significant differences between groups (p = 0.71).
Prehospital telemedicine outcomes
Tele-Information Warning Time: The 5G group demonstrated a significantly shorter warning time of 136 ± 22 sec compared to 235 ± 41 sec in the control group (p < 0.001), representing a 42% reduction in time to telemedicine team activation.
Remote Rescue Success Rate: The 5G group achieved a remote rescue success rate of 98.9% (297/300), significantly higher than the 91.2% (274/300) observed in the control group (p < 0.001), reflecting the effectiveness of remote resuscitation guidance via 5G telepresence (Figure 2).
Comparison of remote triage efficiency between the 5G system group and the control group.
Remote triage efficiency
Remote Triage Completion Time: Triage completion time was reduced by 38% in the 5G group (136 ± 18 sec) compared to the control group (220 ± 35 sec, p < 0.001), attributed to 5G-enabled remote data pre-fetching.
Remote Triage Accuracy: Remote triage accuracy reached 96.3% (289/300) in the 5G group, significantly surpassing the 82.5% (247/300) observed in the control group (p < 0.001), indicating reliable remote severity assessment through real-time data streaming (Figure 2).
Teleconsultation and data management
Teleconsultation Response Time: The 5G group experienced a 51% reduction in teleconsultation response time (10.2 ± 3.1 min) compared to the control group (20.8 ± 5.4 min, p < 0.001), facilitated by 5G’s low-latency communication capabilities.
Tele-Data Completeness: Data completeness for telemedicine cases was 95.6% (287/300) in the 5G group, significantly higher than 72.3% (217/300) in the control group (p < 0.001), with automated NLP capturing 92% of critical teleconsultation details.
Tele-Reporting Time: The 5G group achieved a 45% reduction in tele-reporting time (198 ± 25 sec) compared to the control group (356 ± 48 sec, p < 0.001), streamlining documentation workflows (Figure 3).
Comparison of teleconsultation response time and data management efficiency between the 5G system group and the control group.
Primary clinical outcome: 30-day mortality
The 5G system group demonstrated a significantly lower 30-day all-cause mortality rate compared to the control group (12/300, 4.0% vs. 28/300, 9.3%; p = 0.007). Subgroup analysis based on primary diagnosis revealed consistent trends toward reduced mortality across all emergency categories, with the most pronounced benefit observed in stroke patients (5G group: 3.4% vs. control: 10.2%) (Table 2, Figure 4).
Kaplan-Meier curve illustrating survival probability within 30 days post-admission for the 5G and control groups.
Comparison of 30-Day All-Cause Mortality between the 5G System Group and the Control Group, Stratified by Primary Diagnosis
Discussion
Key telemedicine insights
This study provides novel evidence on the efficacy of 5G-enabled tele-emergency services:
Real-time Telepresence Equivalence: With latency below 50 ms, the 5G system enabled remote experts to provide effective prehospital procedural guidance, aligning with the telemedicine principle of “presence equivalence.” 10
Scalable Remote Triage Capacity: The AI-powered triage system maintained high accuracy across diverse case presentations, suggesting potential for expanding remote triage services during mass casualty events or resource-constrained scenarios.
Tele-data Interoperability: The blockchain-based database ensured the integrity of telemedicine data, addressing long-standing challenges related to telehealth record-keeping and legal compliance.
Impact on patient survival
The most clinically significant finding of this study is the notable reduction in 30-day mortality associated with the 5G system. The absolute risk reduction of 5.3% represents a number needed to treat (NNT) of approximately 19, indicating that for every 19 patients treated using the 5G-enabled system, one death could be prevented within 30 days. This mortality benefit can be attributed to the synergistic effect of multiple system advantages: (1) earlier intervention through reduced prehospital time, (2) more accurate triage leading to appropriate resource allocation, and (3) real-time specialist guidance during critical resuscitation phases. These findings suggest that 5G-enabled telemedicine not only improves process efficiency but also translates to meaningful improvements in patient survival.
Technical advantages in telemedicine
The superior performance of the 5G system can be attributed to its telemedicine-optimized design:
Adaptive Latency Control: A dual-mode 5G/4G fallback mechanism maintained telemedicine connectivity in areas with intermittent 5G coverage, ensuring uninterrupted remote care with a connectivity rate of 99.7%.
Intelligent Bandwidth Allocation: Machine learning algorithms prioritized medical data streams (e.g., ECG data over video) during network congestion, safeguarding critical telemedicine signals from delay.
Tele-expert Workflow Optimization: The Tele-EMSBot reduced cognitive burden on specialists by 35%, enabling focused remote decision-making without administrative distractions.
Limitations and future directions
Single-center Limitation: This study was conducted in a tertiary referral center with mature 5G infrastructure, which may limit generalizability to rural areas with limited 5G coverage or hospitals with varying telemedicine capabilities. Future multi-center trials in resource-constrained settings are needed to validate these findings and explore context-specific optimization strategies.
Long-term Outcome Expansion: While this study evaluated both short-term metrics and 30-day mortality, longer-term functional outcomes such as 90-day neurological recovery in stroke patients or quality of life measures were not assessed. Extended follow-up studies are warranted to comprehensively evaluate the system's impact on long-term patient prognosis.
Cost-benefit Analysis: The economic implications of 5G system implementation are an important consideration for wider adoption. While a formal cost-effectiveness analysis was beyond the scope of this clinical efficacy study, we acknowledge that the initial investment for outfitting a single ambulance with the 5G telemedicine system (including hardware, software, and integration) is approximately 30,000-50,000 USD. Outside the research context, this cost would likely be borne by health care providers, hospital systems, or through public health grants. However, this must be weighed against the potential long-term savings from improved operational efficiency and, more importantly, the significant reduction in mortality and potential reduction in long-term disability as observed in our study. The absolute risk reduction of 5.3% for 30-day mortality translates to an NNT of 19, suggesting a substantial clinical benefit that may justify the investment. Future studies incorporating detailed cost-utility analyses are urgently needed to provide a stronger foundation for health care policy and reimbursement decisions.
Clinical implications for telemedicine
The 5G emergency full-process system has several immediate implications for telemedicine practice:
Tele-emergency Service Standardization: Its modular design supports phased implementation, enhancing adaptability across diverse health care settings.
Telemedicine Workforce Expansion: By minimizing the need for physical presence, 5G enables specialists to provide remote support across multiple locations concurrently.
Disaster Telemedicine Preparedness: The system’s resilience during network stress, such as natural disasters, warrants further exploration for emergency response telemedicine applications.
Conclusion
The 5G emergency full-process system represents a transformative advancement in tele-emergency medicine, enabling real-time remote monitoring, triage, and consultation with unprecedented reliability. More importantly, the system demonstrated a significant reduction in 30-day mortality, confirming its potential to improve not only operational efficiency but also critical patient outcomes. By overcoming the fundamental limitations of traditional telemedicine, this system significantly enhances emergency response capabilities in both urban and underserved areas.
Authors’ Contributions
X.H.: Conceptualization, methodology, investigation, writing—original draft, project administration. H.Z.: Conceptualization, methodology, investigation, writing original draft, data curation. S.X.: Formal analysis, software, visualization, writing—review and editing. G.W.: Resources, validation, writing—review and editing. W.L.: Supervision, funding acquisition, writing—review and editing, resources. All authors have read and agreed to the published version of the article.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
Funding Information
This study was supported by the Qihang Fund of Fujian Medical University (Grant No. 2023QH1263). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the article.