INTRODUCTION

Background

Timely and effective management of trauma patients is a core component of modern emergency medicine, as delays in initial care are strongly linked to increased morbidity and mortality [1–4]. Although there is ongoing debate regarding what constitutes a critical delay, one study reported that every 10-minute increase in prehospital time was associated with a 9% rise in in-hospital mortality [4].

In recent years, Korea’s emergency medical services (EMS) has faced increasing strain. A major challenge was the COVID-19 pandemic, which disrupted emergency services through stringent infection control protocols and diversion of resources. As a result, delays within EMS were reported, with scene-to-door times lengthening during this period [5,6].

Although the system gradually began recovering after the pandemic, a new crisis emerged. In early 2024, a substantial disruption occurred when a large proportion of resident physicians joined a nationwide walkout in opposition to government policies on medical workforce expansion [7–9]. This sudden walkout created a significant shortage of physicians in emergency departments (EDs) and trauma centers, raising serious concerns about whether patients could continue to receive timely and reliable care [9,10].

While the value of rapid and effective trauma care is well established, little research in Korea has examined how large-scale workforce disruptions, such as the resident walkout, affect both prehospital and in-hospital trauma care. Understanding the impact of such disruptions is essential for shaping effective healthcare policy and ensuring that the emergency system can maintain function during crises.

Objectives

This study investigated changes in scene-to-door time for trauma patients before and after the resident walkout at our center. By doing so, we aimed to provide timely evidence on how medical workforce stability influences the performance of an urban trauma system and to inform future policy decisions.

METHODS

Ethics statement

This study was approved by the Institutional Review Board of Seoul National University (No. 2506-028-1646). The requirement for informed consent was waived due to the use of deidentified data and the retrospective nature of the study. All methods were carried out in accordance with relevant guidelines and regulations, including the Declaration of Helsinki.

Study design and setting

We conducted a retrospective cohort study at Seoul National University Hospital (Seoul, Korea). Trauma patients who visited the ED between June 1, 2022, and March 31, 2025, were analyzed.

Patients were eligible if they met all of the following criteria: (1) presented with trauma-related complaints; (2) transported via the EMS system; and (3) classified as major trauma requiring trauma team activation upon ED arrival. Exclusion criteria were as follows: (1) patients dead on arrival who did not receive ED care; (2) missing data on scene arrival time or ED arrival time; (3) interhospital transfers; and (4) trauma cases that did not require trauma team activation.

Patients were grouped according to the timing of the nationwide resident walkout in Korea. Group 1 included patients who visited the ED between June 1, 2022, and February 20, 2024, representing the post–COVID-19 normalization period when emergency services were operating routinely. Group 2 included patients who visited after February 21, 2024, when the walkout began and medical professional shortages occurred.

Most data were obtained from the Korean Trauma Data Bank, a national trauma registry managed by the National Emergency Medical Center of Korea [11,12]. For additional variables not available in the registry, we reviewed electronic medical records at Seoul National University Hospital.

Variables

Scene-to-door time was defined as the interval (in minutes) from EMS arrival at the scene to the patient’s arrival at the ED. Time points were collected from EMS run sheets and the electronic medical records.

Seoul National University Hospital is designated by the Seoul Metropolitan Government as a definitive care trauma center, responsible for providing time-critical and comprehensive interventions such as emergency surgery, interventional procedures, and intensive care for major trauma patients. For this study, the primary service area was defined as six districts in northwestern Seoul, for which Seoul National University Hospital is officially responsible for definitive trauma care: Jongno-gu, Jung-gu, Yongsan-gu, Seodaemun-gu, Mapo-gu, and Eunpyeong-gu (Fig. S1). Patients injured within these districts were classified as being from the primary service area, whereas those injured in other districts of Seoul or outside the city were categorized as outside the primary service area.

The Korean Triage and Acuity Scale (KTAS) is a five-level triage tool implemented nationwide in 2012, adapted from the Canadian Triage and Acuity Scale, and used in Korean EDs to classify patients by clinical urgency and resource needs [13,14]. In this study, trauma team activation generally occurred for patients triaged as KTAS levels I–III upon ED arrival. Triage was performed by certified nurses or physicians. KTAS levels are defined as follows: level I (resuscitation), requiring immediate life-saving intervention; level II (emergent), involving potentially life- or limb-threatening conditions needing rapid care; level III (urgent), conditions that may worsen if not promptly treated; level IV (less urgent), stable cases requiring evaluation or treatment; and level V (nonurgent), minor complaints that can be safely delayed.

Prehospital systolic blood pressure and prehospital respiratory rate were defined as the first vital signs recorded by EMS personnel at the scene before ED arrival.

Statistical analysis

Baseline characteristics and clinical variables were summarized using descriptive statistics. Normality of continuous variables was assessed with the Shapiro-Wilk test. Depending on distribution, data were presented as mean with standard deviation or median with interquartile range. Comparisons of continuous variables were performed using either the independent t-test or the Mann-Whitney U-test, and categorical variables were compared using the chi-square or Fisher exact test, as appropriate. Scene-to-door time was evaluated using restricted cubic spline regression to examine temporal trends. Model fit was assessed with the F-test and adjusted R² values. To enhance visual interpretation and illustrate local variation, locally estimated scatterplot smoothing (LOESS) analysis was additionally applied as a model-independent method. To evaluate the effects of study period (group 1 vs. group 2) and service area (primary vs. outside the primary area) on scene-to-door time, we applied a generalized linear model after confirming assumptions of normality and homogeneity of variance. All statistical analyses were performed using R ver. 4.4.3 (R Foundation for Statistical Computing). A P-value of <0.05 was considered statistically significant.

RESULTS

Baseline characteristics

A total of 271 trauma patients were included in the study, with 117 in group 1 and 154 in group 2. Baseline characteristics are summarized in Table 1. No significant differences were observed between the two groups with respect to age, sex, or mechanism of injury. However, the proportion of patients from outside the primary service area was significantly higher in group 2 than in group 1 (16.7% vs. 31.3%, P=0.018). Initial physiological status, including prehospital systolic blood pressure and respiratory rate, did not differ significantly between groups. The median Glasgow Coma Scale (GCS) score at ED arrival was higher in group 2 (14 vs. 15, P<0.001). KTAS distribution also differed significantly, with a larger proportion of group 2 patients triaged as KTAS II or III (46.2% vs. 68.2%, P<0.001).

In-hospital outcomes and time intervals

In-hospital outcomes and time intervals are presented in Table 2. General ward admissions were more frequent in group 2 than in group 1 (17.9% vs. 53.2%), whereas intensive care unit (ICU) admissions were substantially more common in group 1 (65.0% vs. 22.1%, P<0.001). Interhospital transfers from the ED also increased significantly during the walkout period (7.7% vs. 18.8%, P=0.015).

Scene-to-door time was significantly longer in group 2 than in group 1, with the median increasing from 28 to 44.5 minutes (P<0.001). When stratified by service area, patients from the primary service area demonstrated an increase from 28 minutes (interquartile range [IQR], 22.5–33.5 minutes; group 1, n=75) to 36 minutes (IQR, 27.5–48 minutes; group 2, n=103; P<0.001). For patients from outside the primary service area, the delay was more pronounced, with the median rising from 27 minutes (IQR, 23–34.5 minutes; group 1, n=15) to 64 minutes (IQR, 52–85.5 minutes; group 2, n=47; P<0.001). These results indicate a marked prolongation across both subgroups, particularly among patients transferred from outside the primary service area.

Despite longer scene-to-door times, the overall ED length of stay remained similar between groups (189 minutes vs. 190.5 minutes, P=0.994).

Scene-to-door time trends by study period

Trends in scene-to-door time are illustrated in Fig. 1. In group 1, spline regression did not demonstrate a significant temporal trend. The model had poor explanatory power (adjusted R²=–0.030), and overall fit was not statistically significant (F(4,112)=0.153, P=0.961), suggesting that scene-to-door times remained stable after normalization of EMS following the pandemic. In group 2, no significant temporal trend was observed either. Although a visual inspection indicated fluctuations, spline regression again demonstrated low explanatory power (adjusted R²=–0.006, F(4,149)=0.760, P=0.556). This suggests substantial variability in scene-to-door times and the possible influence of unmeasured confounding factors. Overall, spline regression did not support the presence of a consistent temporal trend in either study period.

Temporal trends in scene-to-door time: LOESS analysis

To further investigate short-term variation, LOESS analysis was performed as a complementary model-independent method. As shown in Fig. 2, group 1 demonstrated a relatively flat trajectory with limited fluctuations, consistent with spline regression findings. In contrast, group 2 exhibited more variability, with transient increases followed by declines in scene-to-door time. The broader confidence intervals observed in group 2 highlight greater instability and fluctuation in prehospital transfer times during the walkout.

Interaction effects of group and area on scene-to-door time

A generalized linear model with log-transformed values was constructed to evaluate the combined effects of group and area (Table 3). Scene-to-door time was significantly longer in group 2 than in group 1 (Exp=1.42, P<0.001), representing a 42% increase during the walkout period. The effect of service area alone was not statistically significant (Exp=1.21, P=0.235). However, the interaction term between group and area suggested an additional 40% increase in scene-to-door time for patients from outside the primary service area in group 2, although this did not reach statistical significance (Exp=1.40, P=0.073).

DISCUSSION

This study demonstrated that scene-to-door time significantly increased during the 2024 resident walkout, reflecting delays in prehospital care. Additionally, shifts in in-hospital patient flow were observed, suggesting that the walkout influenced multiple stages of trauma care, from prehospital transfer to in-hospital management.

In group 2, the proportion of patients from outside the primary service area rose significantly (16.7% to 31.3%, P=0.018), suggesting that nearby hospitals had diminished capacity to manage trauma cases during the walkout. This finding aligns with previous studies indicating that transferred patients are more vulnerable to systemic limitations during periods of restricted capacity or after-hours care [15,16]. The higher ED GCS scores (14 to 15, P<0.001) and the increased proportion of patients triaged as KTAS II or III (P<0.001) further indicate that even less physiologically critical patients could not be admitted locally and were instead referred to higher-level centers. Although the absolute number of KTAS I patients was lower in group 2, the daily average was slightly higher than in group 1 (0.12 vs. 0.10), implying that the reduction reflected the shorter study period rather than a true decline in severely injured patients. Moreover, the overall number of trauma patients increased during the walkout, with the daily average nearly doubling in group 2 (0.186 vs. 0.380). Together, these findings suggest that moderately injured patients were more frequently transferred to higher-level centers, while the volume of critically ill patients remained steady. A similar situation was reported in the United States during the COVID-19 pandemic: Vohra et al. [17] found that more than half of hospital referral regions (53.1%) exhibited ICU capacity imbalances, with some hospitals overwhelmed while nearby facilities still had available beds. Larger hospitals and those treating more vulnerable populations were more likely to experience crowding, whereas smaller or rural hospitals often retained unused capacity.

The proportion of ICU admissions decreased markedly in group 2 (65.0% to 22.1%), likely reflecting reduced ICU capacity in our center during the walkout. Transfers to other hospitals also more than doubled (7.7% to 18.8%, P=0.015), suggesting that even tertiary centers faced challenges in accommodating trauma patients during this period. At our institution, ICU availability was constrained by staffing shortages, limited access to operating rooms, and reduced inpatient bed capacity. These factors hindered admission of trauma patients who otherwise would have required inpatient management.

These challenges may have been caused not only by the shortage of ICU beds but also by challenges in arranging appropriate inpatient care or specialty services after stabilization in the ED. During the COVID-19 pandemic, hospitals frequently encountered situations where ICU beds were technically available but unusable due to operational limitations such as staffing deficits or resource constraints [18,19]. Furthermore, strategies such as canceling elective operations, commonly employed to create ICU capacity, proved only partially effective, since most ICU demand originated from medical emergencies rather than scheduled procedures [20,21]. These observations highlight that ICU capacity depends not merely on bed numbers, but also on the availability of trained professionals and supporting infrastructure. This perspective helps explain both the decrease in ICU admissions and the rise in transfers observed during the resident walkout, even when physical ICU beds may still have existed.

Interestingly, ED length of stay did not differ significantly between groups (189 minutes vs. 190.5 minutes, P=0.994), suggesting that in-hospital trauma care processes remained largely stable despite external disruptions. This stability may reflect the continued presence of the trauma team, which enabled the ED to sustain routine care despite workforce shortages and systemic challenges [22,23].

The significant increase in scene-to-door time observed in this study highlights a serious challenge to the trauma care system. In group 2, the median time rose to 44.5 minutes, representing a 16.5-minute increase compared to group 1. This delay is concerning, as it may reduce the likelihood of timely treatment during the critical first hour after injury, often referred to as the “golden hour.” Delays in definitive trauma care, even within the first 2 hours, have been associated with higher mortality [23,24]. Similar findings have been reported in penetrating trauma, where each additional minute of prehospital delay, including EMS response and scene time, was linked to increased mortality [25]. These results underscore the urgent need to minimize prehospital delays and reinforce the resilience of emergency care systems.

The negative adjusted R² values in our spline regression analysis should not be interpreted as poor model performance but rather as an absence of clear temporal trends in scene-to-door time. The results in group 1 (adjusted R²=–0.030, P=0.961) suggest that trauma transfer systems remained relatively stable once healthcare services normalized after the COVID-19 pandemic. This stability may reflect the continued use of standardized protocols and the consistent availability of medical professionals. In contrast, group 2 (adjusted R²=–0.006, P=0.556) also showed no systematic temporal pattern, but the wider confidence intervals from the LOESS analysis indicated greater variability. This variability may reflect fluctuating influences of limited bed availability, shortages of medical professionals, and ED crowding during the resident walkout [26,27]. Clinically, the stability observed in group 1 suggests a well-functioning system, while the irregular fluctuations in group 2 point to potential risks to patient safety.

Our generalized linear model further demonstrated a 42% increase in scene-to-door time during the resident walkout period (group 2), indicating that the disruption directly influenced hospital selection and patient transport. Although service area alone was not a significant factor, the interaction between group and area suggested a trend toward longer delays for patients outside the primary service area (Exp=1.40, P=0.073). Although not statistically significant, this pattern raises concerns that during periods of workforce instability, patients from outside core service areas may face disproportionately longer transport times, jeopardizing timely trauma care.

This study examined trauma care under a unique condition in which delays arose not from increased patient demand but from a temporary shortage of medical professionals. By focusing on prehospital delays during the resident walkout, our findings provide important insights for health policy and resource planning. The longer and more variable scene-to-door times observed highlight the need for robust response plans to ensure stable trauma care during workforce disruptions. To prepare for similar crises, policies should prioritize strengthening system flexibility [28]. Strategies may include training additional personnel who can provide temporary support, establishing clear interhospital transfer protocols, and preparing capacity-expansion measures to be activated when needed [29,30].

Limitations

Several limitations should be considered when interpreting these findings. First, this was a single-center retrospective study, which may limit the generalizability of the results. Scene-to-door time could have been influenced by unmeasured confounders at both the patient and system levels. In particular, prehospital factors such as EMS resource availability or protocol changes could not be fully assessed. Additionally, our study did not include operational data from the EMS system, such as dispatch times, ambulance availability, or staffing levels, which limited our ability to determine whether changes in EMS capacity or procedures directly contributed to the observed delays during the walkout period. The relatively small sample size may also have reduced statistical power, especially for detecting significant interaction effects. Given the relatively low trauma volume at our center, these results may not be generalizable to other institutions. Further multicenter studies with larger case numbers are needed to validate and expand these findings. Finally, this study focused on process-related outcomes rather than direct clinical endpoints such as mortality or functional recovery.

Conclusions

Our findings indicate that disruptions in the medical workforce, such as the resident walkout, can result in substantial delays in trauma patient transport, as evidenced by prolonged scene-to-door times. These results highlight the importance of careful planning, flexible resource allocation, and robust regional cooperation to ensure that trauma patients receive timely and consistent care, even during periods of systemic strain.

ARTICLE INFORMATION

Author contributions

Conceptualization: all authors; Data curation: YSC, JYJ, JWW, CYP; Formal analysis: YSC; Methodology: YSC, CYP; Resources: all authors; Software: YSC; Supervision: CYP; Visualization: YSC; Writing–original draft: YSC; Writing–review & editing: JYJ, JWW, DJP, CYP. All authors read and approved the final manuscript.

Conflicts of interest

Chan Yong Park is an editorial board member of this journal, but was not involved in the peer reviewer selection, evaluation, or decision process of this article. The authors have no other conflicts of interest to declare.

Funding

The authors received no financial support for this study.

Data availability

Data analyzed in this study are available from the corresponding author upon reasonable request.

Supplementary materials

Fig. 1.

Spline regression of scene-to-door time in (A) group 1 (post–COVID-19 normalization period) and (B) group 2 (post–resident walkout period). Nonlinear changes in scene-to-door time were modeled using restricted cubic spline regression. In group 1, no significant temporal trend was observed (adjusted R²=–0.030, F(4,112)=0.153, P=0.961). Similarly, group 2 showed no statistically significant trend (adjusted R²=–0.006, F(4,149)=0.760, P=0.556). Scatterplots represent individual patients, while spline curves illustrate the fitted trends over time.

Fig. 2.

Locally estimated scatterplot smoothing analysis of scene-to-door time in trauma patients. Curves with 95% confidence intervals illustrate temporal trends in scene-to-door time for group 1 (post–COVID-19 normalization period) and group 2 (post–resident walkout period). Group 1 demonstrates a relatively stable trajectory, whereas group 2 shows greater variability with larger fluctuations across the observation period.

REFERENCES

• 1. Siripakarn Y, Triniti L, Srivilaithon W. Association of scene time with mortality in major traumatic injuries arrived by emergency medical service. J Emerg Trauma Shock 2023;16:156–60.ArticlePubMedPMC
• 2. Pham H, Puckett Y, Dissanaike S. Faster on-scene times associated with decreased mortality in Helicopter Emergency Medical Services (HEMS) transported trauma patients. Trauma Surg Acute Care Open 2017;2:e000122ArticlePubMedPMC
• 3. Chen CH, Shin SD, Sun JT, et al. Association between prehospital time and outcome of trauma patients in 4 Asian countries: a cross-national, multicenter cohort study. PLoS Med 2020;17:e1003360ArticlePubMedPMC
• 4. Gauss T, Ageron FX, Devaud ML, et al. Association of prehospital time to in-hospital trauma mortality in a physician-staffed emergency medicine system. JAMA Surg 2019;154:1117–24.ArticlePubMedPMC
• 5. Kwon J, Lee M, Moon J, et al. National follow-up survey of preventable trauma death rate in Korea. J Korean Med Sci 2022;37:e349ArticlePubMedPMCLink
• 6. Lim YJ, Park SY. Increased prehospital emergency medical service time interval and nontransport rate of patients with fever using emergency medical services before and after COVID-19 in Busan, Korea. J Korean Med Sci 2023;38:e69ArticlePubMedPMCLink
• 7. The Lancet Regional Health-Western Pacific. Junior doctor strikes in South Korea: more doctors are needed? Lancet Reg Health West Pac 2024;44:101056.ArticlePubMedPMC
• 8. You HS, Park KH, Roh H. Physicians' collective actions in response to government health policies: a scoping review. J Korean Med Sci 2025;40:e90ArticlePubMedPMCLink
• 9. Park J, Shin CH, Lee JY. Why did all the residents resign? Key takeaways from the junior physicians' mass walkout in South Korea. J Grad Med Educ 2024;16:402–6.ArticlePubMedPMCPDF
• 10. The Korea Times. Hospitals experience disruptions on extended doctors' walkout [Internet]. The Korea Times; 2024 [cited 2025 Dec 11]. Available from: https://www.koreatimes.co.kr/southkorea/health/20240224/hospitals-experience-disruptions-on-extended-doctors-walkout
• 11. Kim DH, Kim M, Lee DS, Hong TH, Park H, Cho H. Role of laparoscopic surgery in managing hemodynamically stable abdominal trauma patients: a single level I trauma center, propensity score matching study. Eur J Trauma Emerg Surg 2024;50:2517–25.ArticlePubMedPMCPDF
• 12. Ha M, Yu S, Kim BC, Lee JH, Choi HJ, Cho WH. Characteristics of patients with traumatic brain injury in a regional trauma center: a single-center study. Korean J Neurotrauma 2023;19:6–19.ArticlePubMedPMCLink
• 13. Ryu JH, Min MK, Lee DS, et al. Changes in relative importance of the 5-level triage system, Korean Triage and Acuity Scale, for the disposition of emergency patients induced by forced reduction in its level number: a multi-center registry-based retrospective cohort study. J Korean Med Sci 2019;34:e114ArticlePubMedPMCLink
• 14. Beveridge R, Ducharme J, Janes L, Beaulieu S, Walter S. Reliability of the Canadian emergency department triage and acuity scale: interrater agreement. Ann Emerg Med 1999;34:155–9.ArticlePubMed
• 15. Di Bartolomeo S, Marino M, Ventura C, Trombetti S, De Palma R. A population based study on the night-time effect in trauma care. Emerg Med J 2014;31:808–12.ArticlePubMed
• 16. Ono Y, Ishida T, Iwasaki Y, et al. The off-hour effect on trauma patients requiring subspecialty intervention at a community hospital in Japan: a retrospective cohort study. Scand J Trauma Resusc Emerg Med 2015;23:20.ArticlePubMedPMCPDF
• 17. Vohra AS, Khullar D, Kaushal R, Schpero WL. Many intensive care units were overloaded while nearby hospitals had excess capacity during the COVID-19 pandemic. Health Aff (Millwood) 2023;42:937–45.ArticlePubMed
• 18. Douin DJ, Ward MJ, Lindsell CJ, et al. ICU bed utilization during the coronavirus disease 2019 pandemic in a multistate analysis-March to June 2020. Crit Care Explor 2021;3:e0361ArticlePubMedPMC
• 19. Ohbe H, Hashimoto S, Ogura T, et al. Association between regional critical care capacity and the incidence of invasive mechanical ventilation for coronavirus disease 2019: a population-based cohort study. J Intensive Care 2024;12:6.ArticlePubMedPMCPDF
• 20. Tonna JE, Hanson HA, Cohan JN, et al. Balancing revenue generation with capacity generation: case distribution, financial impact and hospital capacity changes from cancelling or resuming elective surgeries in the US during COVID-19. BMC Health Serv Res 2020;20:1119.ArticlePubMedPMCPDF
• 21. Nurok M, Kahn JM. Intensive care unit capacity, cancellation of elective surgery, and the US pandemic response. Anesth Analg 2020;131:1334–6.ArticlePubMed
• 22. Otto R, Blaschke S, Schirrmeister W, Drynda S, Walcher F, Greiner F. Length of stay as quality indicator in emergency departments: analysis of determinants in the German Emergency Department Data Registry (AKTIN registry). Intern Emerg Med 2022;17:1199–209.ArticlePubMedPMCPDF
• 23. Moy E, Coffey RM, Moore BJ, Barrett ML, Hall KK. Length of stay in EDs: variation across classifications of clinical condition and patient discharge disposition. Am J Emerg Med 2016;34:83–7.ArticlePubMed
• 24. Hsieh SL, Hsiao CH, Chiang WC, et al. Association between the time to definitive care and trauma patient outcomes: every minute in the golden hour matters. Eur J Trauma Emerg Surg 2022;48:2709–16.ArticlePubMedPDF
• 25. Nasser AA, Nederpelt C, El Hechi M, et al. Every minute counts: the impact of pre-hospital response time and scene time on mortality of penetrating trauma patients. Am J Surg 2020;220:240–4.ArticlePubMed
• 26. Badkur M, Kharkongor M, Sharma N, et al. Factors influencing treatment delays in trauma patients: a three-delay model approach. J Emerg Trauma Shock 2024;17:172–7.ArticlePubMedPMC
• 27. Shakir M, Altaf A, Irshad HA, et al. Factors delaying the continuum of care for the management of traumatic brain injury in low- and middle-income countries: a systematic review. World Neurosurg 2023;180:169–93.e3.ArticlePubMed
• 28. Garcia-Castrillo L, Petrino R, Leach R, et al. European Society for Emergency Medicine position paper on emergency medical systems' response to COVID-19. Eur J Emerg Med 2020;27:174–7.ArticlePubMedPMC
• 29. May JS, McCafferty JA, Read DJ, Gumm K, Shakerian R. Defining futile and potentially avoidable interhospital trauma transfers. Injury 2024;55:111629.ArticlePubMed
• 30. Gauss T, de Jongh M, Maegele M, Cole E, Bouzat P. Trauma systems in high socioeconomic index countries in 2050. Crit Care 2024;28:84.ArticlePubMedPMCPDF

Figure & Data

References

Citations

Citations to this article as recorded by  

• Impact of the Korean Medical–Political Conflict on Septic Shock Care in the Emergency Department: A Retrospective Single-Center Clinical Analysis
  Seongmun Kang, Jae Hwan Kim, Chiwon Ahn, Young Taeck Oh, Sojune Hwang
  Medicina.2026; 62(8): 1604.     CrossRef