“Emerging infectious diseases” typically denote highly contagious and lethal outbreaks that occur within a short timeframe in specific regions (1, 2). The widespread transmission of such diseases poses significant challenges to epidemic control, strains healthcare systems, and triggers public panic, leading to social unrest and crises that threaten national and regional stability and security (3). The COVID-19 outbreak that emerged at the end of 2019 led to numerous infections and hospitalisations (4, 5), severely impacting healthcare systems worldwide. For instance, Bhaskaran et al. (4) discovered that COVID-19 hospital admissions carried significantly higher risks of rehospitalization and death compared to the general population. Additionally, Timothy et al. (5) predicted a peak in COVID-19 hospitalizations from November to April of the following year, which poses challenges for the healthcare system. Therefore, identifying effective prevention and control measures and enhancing the carrying capacity of medical resources (CCMR) during an outbreak are crucial for controlling the spread of COVID-19 (6).

The CCMR is a key indicator for preventive implementation and epidemic response readiness, encompassing two main aspects (7): First, the production and allocation capacity of medical resources, which includes the ability to produce and distribute beds, medications, medical equipment, and related supplies required for patient treatment, reflecting a region’s ability to promptly supplement medical resources and its efficiency during an outbreak. Second, the maximum number of available medical resources, which is based on the production and allocation capacity of medical resources and their limited nature indicates the maximum amount of resources that medical institutions can provide promptly. For instance, following the COVID-19 outbreak in Taiyuan City, hospital admissions surged, and departments other than respiratory medicine were repurposed to accommodate COVID-19 patients, maxing out bed capacity across all departments. During the pandemic, as the number of COVID-19 patients increased, the shortage of medical resources and resulting congestion posed significant public health issues in various locations. In a study from Germany, early in the epidemic, to increase the capacity for COVID-19 patients, a reduction of bed and operating room occupancy of 50.8 ± 19.3% and 54.2 ± 19.1% was reported (8). The World Health Organization also highlighted that the severe shortage of nursing staff globally could threaten the safety of COVID-19 patients (9). Therefore, during an outbreak, assessing the carrying capacity of the medical resources, preventing the occurrence of medical congestion, shortening the duration of congestion, and developing corresponding prevention strategies are essential.

Vaccination is a pivotal strategy in controlling the spread of infectious diseases. Amidst the COVID-19 pandemic, nations worldwide have prioritised vaccine development and administration (10, 11). In China, governmental and professional institutions such as the Centres for Disease Control and Prevention, have recommended that the public expedite their vaccination process, particularly the administration of the booster doses (12). Studies indicate that completing a three-dose regimen, significantly reduces infection rates (13–17), thereby decreasing the demand for medical resources. Although the efficacy of vaccines in reducing infections has been widely confirmed (18), research on the relationship between vaccination and the CCMR remains insufficient. During the pandemic, quantifying the impact of vaccination on CCMR is paramount for improving local epidemic prevention and control measures.

Infectious disease dynamics modelling is an essential tool for analysing the mechanisms of disease transmission, predicting trends, and identifying influencing factors. Existing research encompasses three main areas. First, the use of infectious dynamic models to analyse the spread of COVID-19 within populations. For instance, Mandal et al. (19) established the SPFEIDR model to predict short-term trends in several severely affected regions in India, identifying that reducing interpersonal contact is key to controlling disease spread. Zhao et al. (20) developed a SUQC model to characterise the COVID-19 dynamics. By fitting it with actual daily incidence data, analysed the outbreak in Wuhan and four first-tier cities in China, predicting the pandemic’s end time and the final scale of infection numbers. Second, the assessment of vaccination effects, such as Ali et al.’s (21) SEIIAVR model which mathematically demonstrated that increasing the vaccination rate to over 50% can effectively reduce the infection rate. Paulo and colleagues (22) suggest that to maximise the number of individuals with partial COVID-19 protection, an optimisation model based on SEIR dynamics recommends the optimal delay for the second vaccine dose, considering crucial factors such as the efficacy of a single dose, the anticipated vaccine supply pipeline, and the potential emergence of more virulent COVID-19 variants, which could significantly reduce ICU admissions. Third, the evaluation of the CCMR in different regions. Wang et al. (7) assessed the CCMR of multiple countries post-outbreak using the SEIARSqTH₁H₂ model, finding that enhancing medical capabilities and testing in tandem can alleviate the epidemic and prevent excessive pressure on CCMR. Overall, current studies focus on the dynamics of transmission, vaccination rates, and the factors affecting CCMR, while there is a lack of discussion on the impact of different vaccination strategies on epidemic trends and CCMR.

As the capital city of Shanxi Province, Taiyuan boasts convenient transportation and abundant medical resources. Data from sentinel hospitals between December 6, 2022, and January 13, 2023, revealed a COVID-19 mortality rate in Taiyuan below 1/10,000, signifying notable success in pandemic control efforts. This study, taking Taiyuan as a case example, collected vaccination status and incidence data from residents through surveys. Utilising medical resources and patient information provided by sentinel hospitals, an SV1V2V3EIQHDR model simulated vaccination scenarios and epidemic trends during the COVID-19 period. The model assessed changes in hospital bed occupancy, reflecting the impact of different vaccination levels on infection rates, hospitalisations, severe cases, and trends, along with CCMR. This provides a scientific basis for optimising vaccination strategies and medical resource allocation in Taiyuan.

This study categorised the vaccinated population into four groups: those susceptible without vaccination (S), those who have received one dose (V1), those who have received two doses (V2), and those who have received three doses (V3). We use the number of available hospital beds as a quantitative indicator of CCMR. When the number of beds hits zero, the CCMR is deemed to have reached its limit, leading to a medical congestion. The onset, duration, and end time of this congestion serve as metrics for assessing the CCMR. The primary assumptions of the study include:

(H₁) Excludes transmission from asymptomatic infections and those in the incubation period in the model. It is because following the easing of control measures and the discontinuation of nucleic acid testing, it is widely accepted that the propagation of the novel coronavirus is primarily due to individuals exhibiting symptoms of the virus.

(H₂) Individuals exhibiting symptoms of COVID-19 will either be quarantined at home or receive treatment in a hospital.

(H₃) Individuals who have been infected will not contract the infection again within a short span of time.

(H₄) No contagiousness during home quarantine period: Patients are considered not to be infectious during the period of home isolation, given that they are unlikely to trigger community transmission while in home isolation.

(H₅) Provided that there are enough beds, there will also be ample healthcare staff and medical equipment. Even in times of healthcare stress, although there may be difficulties in housing all patients, those admitted to the hospital receive appropriate care.

(H₆) The capacity, contact rate, and severity conversion rate of medical resources will change with the development of the epidemic. Following the initial stage of an epidemic in a region, the rate of infection gradually peaks. A rapid increase in demand for hospital care corresponds to an increase in the demand for medical resources. This signifies the point at which the epidemic reaches its highest level. During this time, social interactions typically decrease as individuals minimise their contact with others. Simultaneously, hospitals see a marked increase in the number of severe and critical cases brought on by the novel coronavirus, compared with earlier levels. Therefore, we propose that the methods used to calculate medical resource capacity, contact rates, and severity conversion rates all change simultaneously.

(H₇) Given that most individuals did not undergo nucleic acid testing following the lifting of restrictions, we consider individuals presenting symptoms similar to those of COVID-19, such as fever or cough, as new cases.

During widespread outbreaks of acute infectious disease, the strain on medical resources is a common phenomenon and has garnered extensive attention and research from numerous scholars (3, 7, 19, 20, 32–36, 39). Taking COVID-19 as a case study, we investigate the impact of vaccination and resource allocation on the epidemic trends and CCMR during such outbreaks. Based on survey data on COVID-19 transmission in Taiyuan, we constructed a dynamic model to fit the daily reported new cases, hospitalised patients, and severe cases (as shown in Figures 2, 3), and estimated parameters (as shown in Table 2), predicting the ultimate scale of the epidemic (indicated by the solid green line in Figure 5). By calculating the trend of remaining bed numbers, we can determine the occurrence and duration of medical congestion. As shown in Figure 2, the peak times for new cases, hospitalizations, and severe cases typically lag in succession. Some existing research (37, 38) can fully support our conclusion. For instance, research by Wang et al. (37) reported that during the COVID-19 outbreak in Wuhan, the median time from the onset of initial symptoms to hospital admission for 260 patients was approximately 8 days; C Dananché and colleagues (38) noted that younger COVID-19 patients tended to have a longer interval between the onset of symptoms and hospitalization. Figure 4 indicates that until January 16, 2023, the demand for hospital beds remained high. To address the sharp increase in hospital admissions, most hospitals in Taiyuan City reallocated beds from other departments to accommodate COVID-19 patients, alleviating the bed shortage, but the continuous high demand for hospitalisation by COVID-19 patients temporarily reduced the number of available beds to zero.

In the sensitivity analysis, we first examined the impact of different vaccination scenarios on the numbers of infections, hospitalisations, and severe cases, as well as the change in the number of availible beds in an ideal vaccination scenario, reflecting the influence of vaccination on medical congestion. We further analysed the specific impact of vaccination on the CCMR. The data from Figures 5 and Table 3 suggests that vaccination can diminish the scope of infections and hospitalisations and reduce severe cases. Figure 6 further illustrates that a vaccination regimen of increasing three doses of vaccine can notably shorten the duration of medical congestion, and enhance the capacity of medical resources. Subsequently, we explored the influence of medical resource-related factors on the duration of medical congestion, as detailed in Figure 7. Figure 7A demonstrates that increasing the maximum number of available beds can end medical congestion earlier. Figure 7B reveals that increasing the number of maximum capacity of hospital beds at the onset of the epidemic can delay medical resource shortages but does not end the shortage earlier. Figure 7C indicates that under a fixed total amount of medical resources, higher efficiency in resource allocation can end medical congestion earlier. In other words, before the outbreak of an epidemic, preparation should begin in advance by increasing the number of beds, equipping medical professionals, and increasing medical supplies. During the outbreak, efforts should be made to accelerate the production and distribution of medical supplies and the deployment of medical professionals.

The capability, quantity, and timing of resource allocation significantly influence the occurrence, development, and closing time of CCMR overload. Thus, a thorough consideration of vaccination and medical resource allocation becomes imperative. In contrast to Wang et al. (7), our research delineates diverse vaccination scenarios and analyses their impact on the CCMR. In contrast to Mandal (19) and Zhao (20), we not only predicted the trend of COVID-19 transmission in Taiyuan but also anticipated potential medical congestion and its duration. Compared to Ali et al. (21), we used the model to validate the effectiveness of vaccination and explore the impact of different dosing regimens on medical congestion. Our study used the number of beds as a quantitative indicator of medical resource capacity. However, the CCMR is affected by multiple factors such as medical staff, ventilators, and patient’s hospital preference, which will be addressed in future research. Additionally, our study only considered symptomatic individuals as transmission sources. Future investigations should delve into transmission from asymptomatic and incubation period infections. Therefore, future research may need to consider these factors to improve the accuracy and applicability of the model. Current research focuses primarily on prevention and control strategies for COVID-19 (3, 7, 32–36). Given the constraint of limited medical resources, future research should concentrate on how to combine vaccination and the enhancement of CCMR to alleviate medical congestion and improve local medical resource capacity.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

JG: Conceptualization, Investigation, Methodology, Software, Visualization, Writing – original draft. YL: Data curation, Formal analysis, Investigation, Project administration, Validation, Writing – review & editing. YM: Writing – review & editing. SX: Validation, Writing – review & editing. JL: Validation, Writing – review & editing. TW: Writing – review & editing. LL: Writing – review & editing. LH: Writing – review & editing. HY: Formal analysis, Project administration, Writing – review & editing. JX: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Writing – review & editing.