[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Next Article in Journal
Ecological Zoning Based on Suitability Evaluation of Ecological Product Development from the Value-Risk-Cost-Demand Perspective
Previous Article in Journal
The Effects of Employment Center Characteristics on Commuting Time: A Case Study of the Seoul Metropolitan Area
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJGI
Volume 14
Issue 3
10.3390/ijgi14030117
ijgi-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

How Did the Fever Visit Management Policy During the COVID-19 Epidemic Impact Fever Medical Care Accessibility?

by
Zhiyuan Zhao
1,2,3,*,
Youjun Tu
1,2 and
Yicheng Ding
1,2
1
Academy of Digital China (Fujian), Fuzhou University, Fuzhou 350108, China
2
Key Laboratory of Spatial Data Mining and Information Sharing, Ministry of Education, Fuzhou 350108, China
3
The Science and Technology Innovation Team of the Digital Economy Alliance of Fujian Province, Fuzhou 350000, China
*
Author to whom correspondence should be addressed.
ISPRS Int. J. Geo-Inf. 2025, 14(3), 117; https://doi.org/10.3390/ijgi14030117
Submission received: 15 December 2024 / Revised: 15 February 2025 / Accepted: 3 March 2025 / Published: 6 March 2025
(This article belongs to the Topic The Use of Big Data in Public Health Research and Practice)
Browse Figures
Figure 1
<p>Study area and distribution of medical facilities.</p> "> Figure 2
<p>Technical flow chart.</p> "> Figure 3
<p>The dynamic average accessibility of fever clinics in the main urban area of Xining city under driving and public transport modes.</p> "> Figure 4
<p>The hourly accessibility of fever clinics in the main urban area of Xining city for under 15 min of driving.</p> "> Figure 5
<p>Changes in the accessibility of fever clinics in the daytime compared with the nighttime for driving conditions of under 15 min.</p> "> Figure 6
<p>Spatial distribution of relative changes in the accessibility of fever clinics relative to general hospitals based on the dynamic population distribution in the main urban area of Xining city for the driving mode.</p> "> Figure 7
<p>Spatial distribution of relative changes in the accessibility of fever clinics relative to general hospitals based on the dynamic population distribution in the main urban area of Xining city for the public transport mode.</p> "> Figure 8
<p>Relative changes in accessibility based on the NTA versus accessibility based on the dynamic population distribution for the driving mode.</p> "> Figure 9
<p>Relative changes in accessibility based on DTA versus accessibility based on the dynamic population distribution for the driving mode.</p> "> Figure 10
<p>Spatial distribution of population displacement in the main urban area of Xining city.</p> "> Figure 11
<p>The changing dynamics of population flow in the study area.</p> ">
Versions Notes

Abstract

:
Fever visit management (FVM) played a critical role in reducing the risk of local outbreaks caused by positive cases during the coronavirus disease 2019 (COVID-19) pandemic under the dynamic zero-COVID-19 policy. Fever clinics were established to satisfy the healthcare needs of citizens with fever symptoms, including those with and without COVID-19. Learning how FVM affects fever medical care accessibility for citizens in different places can support decision making in establishing fever clinics more equitably. However, the dynamic nature of the population at different times has rarely been considered in evaluating healthcare facility accessibility. To fill this gap, we adjusted the Gaussian-based two-step floating catchment area method (G2SFCA) by considering the hourly dynamics of the population distribution derived from mobile phone location data. The results generated from Xining city, China, showed that (1) the accessibility of fever clinics explicitly exhibited spatial distribution patterns, being high in the center and low in surrounding areas; (2) the accessibility reduction in suburban areas caused by FVM was approximately 2.8 times greater than that in the central city for the 15 min drive conditions; and (3) the accessibility of fever clinics based on the nighttime anchor point was overestimated in central areas, but underestimated in suburban areas.

1. Introduction

Coronavirus disease 2019 (COVID-19) has evolved into a major global public health crisis, posing severe threats to both human health and societal functioning. Each country has adopted different epidemic prevention policies and measures. Since the control of COVID-19 began in 2020, China has gradually developed a dynamic zero-COVID-19 policy [1]. Fever is a typical symptom of COVID-19. To mitigate the risk of local outbreaks caused by positive individuals seeking medical care for fever, China implemented fever visit management (FVM) under the dynamic zero-COVID-19 policy. Specifically, residents were required to seek medical care exclusively at fever clinics when experiencing symptoms of fever, regardless of whether there was a local epidemic. Due to strict and effective epidemic control, only a few cities had limited local COVID-19 cases during a certain period. In other words, most cities experienced a normal situation during the three-year COVID-19 pandemic in China. However, the FVM policy decreased the convenience of direct fever medical care for all people with a fever. Evaluating the impact of the FVM policy on fever medical care accessibility can reveal what we have paid for effective epidemic control. Moreover, by understanding and enhancing the availability and distribution of essential healthcare services, we can better prepare for and respond to public health emergencies effectively [2].
During the COVID-19 pandemic, researchers from various disciplines, including epidemiology, nursing, public emergency management, architecture, and others, turned their focus toward exploring various aspects related to fever clinics. These included the exploration of process management and optimization strategies [3,4,5], the standardization and innovative construction of internal building space [6], and the allocation standards and emergency capacity of primary medical and health institutions [7]. However, despite the effective role of fever clinics in mitigating the risk of nosocomial infections, these facilities are mostly concentrated in urban centers, mainly due to their dependence on well-established medical institutions. Urban centers are typically characterized by a higher population density, greater access to medical resources, and a concentration of economic, cultural, and administrative functions, which lead to better accessibility to healthcare services. Consequently, residents in remote suburban areas, which surround the urban core and generally have a lower population density, fewer medical facilities, and greater challenges in terms of transportation, face considerable limitations in accessing fever clinic services. This geographical disparity between urban centers and suburban areas raises concerns about equitable healthcare access, especially for those residing in underserved regions [8,9]. Therefore, the rationality of the configuration and spatial accessibility of fever clinics has been discussed. Zhang et al. (2021) used the susceptible–exposed–infectious–recovered (SEIR) model to predict the number of daily infections in different control areas of Beijing and adopted the improved two-step floating catchment area (2SFCA) method to analyze the spatial accessibility of fever clinic services in Beijing [10]. Wang et al. (2021) analyzed the accessibility of fever clinics and designated hospitals by using diverse methods and found that it took 60 min to cover the whole area for most residents [11]. In addition, Xu et al. (2023) explored the evaluation and optimization of fever clinic allocation during public health emergencies in Harbin [12].
Spatial accessibility refers to the quantitative expression of residents’ desire and ability to overcome obstacles, such as distance, travel time, and cost, to reach a public service facility [13,14,15]. This metric reflects the opportunities for residents to access public service resources through transportation, serving as a common index to gauge the spatial equity and service level of public service facilities [16]. Currently, various methods are used to evaluate accessibility, including the nearest neighbor method from a temporal or spatial perspective [17], the network analysis method [18], and spatial interaction models such as the gravity model [19], Huff model [20], 2SFCA method [21], M2SFCA method [22], and E2SFCA method [23]. The 2SFCA method has become one of the most widely used spatial accessibility measurement methods in public service access research [22,24]. Although this method has been extensively applied by Wang et al. [25,26], 2SFCA assumes equal accessibility within the catchment area, while locations outside the threshold are considered completely inaccessible. This assumption ignores the impact of distance on service accessibility in reality, which may lead to the overestimation of accessibility in central areas and the underestimation of it in peripheral regions [27], thus affecting the accuracy of equity assessments. To address the limitations of 2SFCA, Luo (2009) and Delamater (2013) proposed the enhanced 2SFCA (E2SFCA) and modified 2SFCA (M2SFCA) methods [22,23], respectively. E2SFCA introduces stepwise distance decay by dividing the catchment area into multiple zones and assigning different weights to each, partially considering the effect of distance on accessibility. However, since it adopts a discrete decay approach, it still suffers from artificial boundary effects, meaning that the accessibility changes abruptly between different zones. M2SFCA further refines the decay function by introducing a flexible distance decay mechanism, allowing adjustments based on different research needs and improving the adaptability of the model. However, M2SFCA still relies on discrete decay settings, which cannot completely eliminate boundary discontinuities. In contrast, the Gaussian-based two-step floating catchment area (G2SFCA) method adopts a continuous Gaussian distance decay function [28], ensuring that the accessibility gradually decreases with increasing distance without sudden changes. Compared to 2SFCA, E2SFCA, and M2SFCA, G2SFCA eliminates the artificial boundary effects caused by predefined stepwise decay, providing a smoother and more realistic representation of service accessibility changes. This enhancement has been widely used in evaluating the accessibility of various public service facilities, such as medical facilities [29,30], parks and green spaces [31], and elderly care facilities [32].
The majority of related studies have estimated the demand for public service resources based on fixed population distributions derived from residential locations. Since human trajectories are dynamic, it is crucial to consider these temporal variations when evaluating people’s access to public services [33]. To address this limitation, some scholars have started incorporating temporal changes into spatial accessibility assessments, applying them to essential urban infrastructures such as medical facilities [34,35], park green space [36,37], and public transit [38,39,40]. For instance, Park et al. (2023) investigated daily changes in ICU bed accessibility using the enhanced two-step floating catchment area (E2SFCA) method in Texas [41]. Their research focused on the population with COVID-19, but ignored the accessibility of hospitals for people with an ordinary fever. This study utilized mobile phone location data to capture the dynamic population distribution of all groups. Compared to traditional methods based on census and statistical data, mobile phone location data offer considerable advantages, including shorter update cycles and greater spatiotemporal precision [42,43,44,45]. In addition, mobile phone data can continuously and passively reveal the location of a large population, which can provide dynamic population distribution information [46,47].
This paper used the population dynamic distribution as the demand information based on mobile phone location data and explored the dynamic changes in residents’ medical accessibility to fever clinics during the COVID-19 pandemic using the Gaussian-based two-step floating catchment area method. This study aimed to address the following three research questions: (1) How did the COVID-19 pandemic affect residents’ access to medical resources? (2) What were the differences in accessibility results under different transportation modes and time threshold conditions? (3) Did the traditional method of calculating accessibility based on a static population distribution produce biased results? Based on these research questions, we proposed the following hypotheses: (1) The COVID-19 pandemic reduced access to medical resources more in suburban areas than in urban centers. (2) Accessibility varies by transportation mode (private car vs. public transport) and time threshold (e.g., 15 vs. 30 min), with a greater impact in suburban areas. (3) Static population-based accessibility calculations overestimate the access in urban centers and underestimate it in the suburbs, while dynamic methods (e.g., mobile data) provide more accurate results. The findings of this study can offer valuable theoretical support for planning and establishing fever clinics in response to major public health emergencies in the future. By leveraging real-time population data and dynamic accessibility assessments, policymakers can develop more responsive and effective strategies to ensure equitable access to critical healthcare services during times of crisis.

2. Study Area and Datasets

2.1. Study Area

The study area encompassed the main urban area of Xining city (Figure 1), which is situated in the eastern part of Qinghai Province within the middle Huangshui River valley basin in Northwest China. Xining serves as the capital of Qinghai Province and is one of the region’s key core cities. The main urban area comprises four districts—the East District, West District, North District, and Central District—covering an approximate area of 360 square kilometers. According to the 2018 Statistical Bulletin of Xining’s National Economic and Social Development, the urban population of Xining was approximately 1.71 million, accounting for 72.1% of the total permanent population.

2.2. Datasets

This study utilized mobile phone data obtained in collaboration with a telecommunications operator. The data covered four workdays: 1–3 August and 6 August 2018, with an average of approximately 43.14 million daily records. Each record contained attributes such as anonymized user IDs, location coordinates, and time stamps. From these data, we identified approximately 2600 base stations. To ensure the protection of user privacy, the telecommunications company implemented strict encryption protocols prior to data access. All personally identifiable information (PII) was anonymized, and the data were processed to maintain user confidentiality. Ethical and legal standards were adhered to in order to safeguard individual and community privacy throughout the study.
Second, the list of fever clinics was collected from the official website of the Qinghai Province Health Commission https://wsjkw.qinghai.gov.cn/ (accessed on 7 July 2022). In the main urban area of Xining, 22 hospitals have established fever clinics. According to the Qinghai Judicial Administration website https://sft.qinghai.gov.cn/ (accessed on 12 July 2022), the fever clinics are open 24 h a day. We selected community health centers and grade II or above hospitals as normal fever treatment hospitals (referred to as “general hospitals” in this paper). Hospitals with emergency departments at level II and above are generally accessible 24 h a day. Community health centers are usually open from 8:00 to 17:00. A total of 176 general hospitals were identified. Additionally, we acquired basic information about fever clinics and general hospitals, including their grades, number of beds, and addresses, from their official websites. Subsequently, we obtained the geographical locations of the fever clinics and general hospitals from Baidu Map.
To enhance the reliability of the analysis results and mitigate the impact of invalid and abnormal data, data-cleaning and noise-reduction processing were performed on the mobile phone data. The cleaning process included the following steps: (1) the removal of duplicate values, (2) the deletion of null values, and (3) the elimination of records with typical error values, such as an abnormal latitude and longitude, anonymous user IDs, and erroneous timestamps. Furthermore, noise-reduction processing was applied to address the issue of location fluctuations caused by mobile phone data inconsistencies. This phenomenon, referred to as the ping-pong effect, occurs when a user’s location appears to rapidly switch between different points, often due to network or signal fluctuations. To detect and mitigate this, we employed a velocity threshold method to identify the discrepancies, and the nearest neighbor method was used to smooth the data and correct these fluctuations effectively.

3. Methodology

The technical flow (Figure 2) of this paper is as follows: first, mobile phone data were utilized to obtain dynamic and static population distributions. Second, the time cost for residents to reach medical facilities was calculated. Finally, the Gaussian-based two-step floating catchment area method was applied to assess residents’ accessibility to medical facilities for fever, and the results were analyzed.

3.1. Dynamic Population Distribution Data Acquisition

There are various methods for obtaining dynamic population distributions, such as using Tencent YiChuXing data or mobile phone data. In this study, we used mobile phone data to acquire a 24 h dynamic population distribution. Specifically, we divided each user’s daily trajectory into 24 one-hour time windows and used the nearest neighbor method to interpolate the user’s trajectory, ensuring that each user had trajectory points in each time window. Then, we identified the most frequent location for each user in each time window as the anchor point for that period, resulting in the spatial distribution of all users within the 24 time windows. Drawing from the research of Ni et al., we divided the main urban area of Xining city into 500 m × 500 m grids as the analysis units [27]. The 500 × 500 m scale was chosen because it provided an optimal balance between spatial resolution and computational efficiency. This grid size allowed for a detailed analysis of accessibility patterns at a local level while maintaining practicality in the data processing and analysis. Additionally, it aligned with the spatial resolution of the available urban data, such as the population density and land use, which are often represented at similar scales. This approach ensured that the study was both comparable to similar urban accessibility research and efficient for computational modeling.

3.2. Typical Daytime/Nighttime Population Distribution Estimation

An activity anchor point refers to a location or place that individuals frequently visit during a specific period, which is identified and marked for mobile behavior research. The process of identifying and marking these activity anchor points is crucial for comprehending people’s activity patterns and social behaviors, as well as inferring their interests and behaviors. Moreover, activity anchor points play an important role in representing population distribution data for specific periods. In this study, the nighttime anchor point (NTA) was defined as the representative cellphone tower that indicated where a user spent at least four hours between 00:00 and 07:00, while the daytime anchor point (DTA) was defined as the tower that represented where a user spent at least six hours between 09:00 and 18:00 [48], based on the established routines of people.
We analyzed mobile phone data collected on 1–3 August and 6 August 2018, to identify users’ NTAs and DTAs over these four days. If the same location was identified for the same user on all four days, we considered it their NTA or DTA. However, if a user’s location varied over the four days, we selected the point closest to the center of gravity of these positions as their NTA or DTA. As a result, we identified 99,347 users who had both an NTA and a DTA. To assess the representativeness of mobile phone users, we compared the number of mobile phone users with the data from the sixth population census of China. After eliminating the outliers through box plots, the Pearson correlation coefficient between the two populations was 0.77 (p < 0.001) for the remaining streets, except for the populations in Linjiaya, Lejiawan, Yunjiakou, Zongzhai, and Pengjiazhai, which were significantly abnormal. This indicated that the number of mobile phone users was representative.

3.3. Time Cost Calculation While Considering Road Network Conditions

Considering the urban road connectivity, the traffic capacity, and the efficiency of different traffic modes, we obtained the travel time from demand points to supply points under the driving and public transport modes based on navigation apps (e.g., Baidu Map, Google Maps) commonly used by the public. Specifically, we obtained the time spent by residents visiting healthcare facilities on five consecutive weekdays, from 5–9 September 2022, between 8:00 and 17:00. The average travel time was then calculated as the time cost. Compared with traditional calculation results based on road network data, this method comprehensively considers factors such as the road grade, turn time, congestion, traffic lights, and other actual traffic conditions, providing more accurate evaluation results.

3.4. Accessibility Calculation While Integrating Population Dynamics

Instead of calculating a uniform supply–demand ratio for each supply point, this study considered population dynamics and calculated the supply–demand ratio for each supply point over M time windows, where M refers to the number of time windows. A Gaussian-based two-step floating catchment area method was used to calculate the accessibility of fever clinics and general medical care.
First, using the locations of hospitals as supply points ( j ) and the locations of grid centers as demand points ( i ), all the demand points that were within a threshold time ( d 0 ) from the catchment for each supply point were used to compute the supply–demand ratio ( R j ) as follows:
R j z = S j ∑ k ∈ { d k j ≤ d 0 } G ( d k j , d 0 ) D k z
where R j z denotes the supply-to-demand ratio per hour, S j represents the number of beds at a hospital ( j ), D k z indicates the population per hour at demand location ( k ) with its centroid in the catchment area ( j ) ( d k j ≤ d 0 ), and d k j is the time cost between demand point k and supply point j . The friction of distance G ( d k j , d 0 ) was calculated as follows:
G ( d k j , d 0 ) = e − ( 1 2 ) × ( d k j d 0 ) 2 − e − ( 1 2 ) 1 − e − ( 1 2 ) ,   i f   d k j ≤ d 0 0 ,   i f   d k j > d 0
Next, the overall supply points within d 0 were searched for each demand point i , and the supply–demand ratio ( R j z ) was summed to calculate the spatial accessibility per hour ( A i z ) at demand point i . To compare with the accessibility values calculated based on the static population distribution, the dynamic average accessibility ( A i ) was obtained by the average of the accessibility over M time windows.
A i z = ∑ j ∈ { d i j ≤ d 0 } G ( d i j , d 0 ) R j z
A i = ∑ A i z / M
where A i represents the accessibility of the population at demand point i to hospitals.

4. Results

4.1. Accessibility of Fever Clinics Based on the Dynamic Population Distribution

To facilitate a comparative analysis of the accessibility across different traffic modes and time thresholds, this study categorized accessibility values into six grades: no supply (0), low (0–0.03), relatively low (0.03–0.1), medium (0.1–0.2), relatively high (0.2–0.3) and high (>0.3). Based on this, other driving thresholds and bus thresholds were classified and compared.
Overall, the accessibility grades of fever clinics based on the dynamic population distribution in the main urban area of Xining showed high spatial distribution characteristics in the center and low spatial distribution characteristics in peripheral areas for both the driving and public transport modes (Figure 3).
For driving modes, when the time threshold was 15 min, the accessibility grades exhibited significant disparities across the study region, with the central area characterized by higher accessibility and some regions in the north, west, and southeast lacking access to fever clinic services. However, as the time threshold increased, the accessibility of fever clinics gradually improved in suburban areas, and the accessibility of fever clinics in suburban areas gradually converged with that in the central city. Only by driving for 60 min did fever clinic services cover the whole study area, which is consistent with the findings of Wang et al. [11].
Regarding the public transport mode, the fever clinic accessibility varied substantially with increasing time thresholds. At the 15 min threshold, only isolated areas had access to fever clinics due to the additional time required for walking to public transport stations and waiting for vehicles. Under the 30 min threshold, the accessibility grades in the central area improved notably, but some areas in the north, west, and southwest still lacked access to fever clinic services. However, even at the 60 min threshold, the entire study region remained without full coverage of fever clinic services. Note that the time window number in Equation (4) was set to 17, considering that the bus service operation time is mainly between 6:00 and 22:00.
To further investigate the dynamic changes in fever clinic accessibility over a 24 h period in the study area, this paper used the 15 min driving scenario as an example to showcase the accessibility pattern (Figure 4). The overall analysis revealed a consistent spatial distribution, with higher accessibility concentrated in the central city and lower accessibility in the periphery. In detail, in the central city, the grids with high accessibility levels almost did not change from 0:00 to 7:00. However, after 7:00, the number of grids with high accessibility gradually declined, reaching a steady state at approximately 10:00. Subsequently, after 18:00, the central city experienced a gradual increase in grids with high accessibility.
By calculating the mean accessibility values during the day (9:00–18:00) and night (0:00–7:00), we examined the variation in fever clinic accessibility between these two periods (Figure 5). The results showed that the daytime accessibility was generally lower than the nighttime accessibility in the central city, whereas the opposite trend was observed in suburban areas. In specific regions of the central city, such as Lirang, Yinma, Renmin, Shenglilu, and Xinghailu, the accessibility of fever clinics decreased by more than 10% during the day compared to that during the night. Conversely, in parts of Lejiawan and Nanchuandonglu, the accessibility of fever clinics increased by more than 15% during the daytime compared to the nighttime values.

4.2. Impact of the COVID-19 Pandemic on the Population’s Access to Healthcare for a Fever

To gain insights into the impact of fever clinic location on residents seeking medical treatment for a fever during the epidemic, we conducted a comparative study of the accessibility of fever clinics and general hospitals for driving and public transport modes. Our analysis revealed noteworthy differences in the accessibility of fever clinics and general hospitals in the main urban area of Xining city, with the trends varying according to the time threshold (Figure 6 and Figure 7).
Regarding the driving mode, the accessibility of fever clinics demonstrated a downward trend compared to that of general hospitals. As the travel time cost acceptable to the user increases, the decline in the accessibility of medical care to a resident with a fever decreases. This suggests that, when the temporal threshold is high, it is easy to mask the spatial inequality of accessibility. Specifically, when driving for 15 min, the relative change in fever clinic accessibility compared to general hospitals decreased by more than 80%, with prominent changes observed in Nianlipu and Pengjiazhai (Figure 6a). As the time threshold increased, the decrease in accessibility to medical care for a fever gradually stabilized across the study area. The reductions in accessibility within the study area were all less than 15% when the time threshold was increased to 45 min at 0:00 (Figure 6c). At 10:00, when the time threshold was increased to 30 min, most of the area within the study area experienced less than a 50% reduction in accessibility.
In the public transport mode, the accessibility of fever clinics compared to general hospitals decreased. This suggests that, after the epidemic, the opportunity for residents to access fever medical services via public transportation decreased. When the time threshold was 30 min, the accessibility of residents seeking medical treatment for a fever decreased in central urban areas and decreased even more in suburban areas. Specifically, areas such as Nianlipu, Biotechnology Industrial Park, Pengjiazhai, and Nanchuandonglu Street experienced a reduction in accessibility of more than 80%. Furthermore, as the time threshold increased, the number of areas with significant declines in accessibility decreased. However, even under the 60 min time threshold, the accessibility in Nianlipu still decreased by more than 80%.
Table 1 and Table 2 show the changes in the accessibility of fever clinics compared to that of general hospitals based on the dynamic population distribution. The accessibility of fever clinics decreased compared with that of general hospitals, indicating that, after the occurrence of major public health events, it was more difficult for residents to seek medical treatment for a fever. In the driving mode, different time thresholds yielded higher decreases in accessibility at 10:00 than at 0:00. In addition, as the travel time cost acceptable to the user increases, the decrease in accessibility decreases, regardless of whether travel is by car or public transportation.
Interestingly, variations were observed between the suburban and central urban areas. Except for the 15 min public transport threshold, the decrease in accessibility was greater in suburban areas than in central urban areas. Specifically, at 10:00 a.m., for 15 min by car and 30 min by public transport, the accessibility to the suburbs decreased by more than 2.8 times as much as that in the central city. Under the condition of 15 min by public transport, fewer areas had access to fever medical services, which resulted in a smaller decline value. These results indicate that the establishment of fever clinics in Xining city will decrease the accessibility of medical care for residents with fever symptoms, and the accessibility of healthcare facilities is sensitive to time thresholds.

4.3. Comparative Accessibility of Fever Clinics Based on Static and Dynamic Population Distributions

To compare the differences in accessibility obtained from static and dynamic population distributions, we examined the relative changes in the accessibility of fever clinics based on the NTA and DTA relative to the dynamic population distribution, using the driving mode as an example (Figure 8 and Figure 9).
When comparing the NTA with the dynamic population distribution, we observed that the accessibility of the central city and inner suburban areas was overestimated, while the accessibility of the outer suburban areas was underestimated for driving conditions of under 15 min (Figure 8a). Specifically, the accessibility of fever clinics in the central city and Nanchuandonglu was overestimated by more than 3%, while the accessibility of fever clinics in Lejiawan was underestimated by more than 8%. Moreover, as the time threshold increased to 45 min, the accessibility of the entire study area decreased (Figure 8c).
When comparing the DTA with the dynamic population distribution, this study showed contrasting results in terms of accessibility to fever clinics. For driving conditions of under 15 min, the accessibility in the central city and inner suburban areas was underestimated, while that in the outer suburban areas was overestimated (Figure 9a). These findings presented opposite characteristics compared to the accessibility results based on the DTA. Notably, with an increased time threshold of 30 min, the accessibility of fever clinics in the outer suburban areas decreased (Figure 9b). Similarly, at the 45 min and 60 min time thresholds, the accessibility of fever clinics decreased throughout the entire study area to a similar extent (Figure 9c,d).

4.4. Typical Area Analysis

In this paper, we selected Shenglilu, Nanchuandonglu, Lejiawan, and Nianlipu as case studies to investigate the relationship between the accessibility of fever clinics and population displacement in the main urban area of Xining city. To investigate the spatial displacement of the population from nighttime to daytime, we analyzed the movement of the population at two time points from 3:00 to 15:00. Additionally, we examined the population movement from 0:00 to 1:00.
The population distribution remained relatively stable from 0:00 to 1:00, with minimal changes (Figure 10a,c). However, from 3:00 to 15:00, there was a considerable shift in the population distribution, especially with a notable increase in population in the central city (Figure 10b). This population displacement might explain the decrease in accessibility in the central city during the daytime relative to the nighttime. Notably, some areas of Lejiawan experienced an increase in population, but their accessibility improved. This can be attributed to the fact that accessibility is an average value within a given range. When the population increases in a small local area, it may not necessarily lead to decreased accessibility. This finding suggests that the relationship between population movement and accessibility should be analyzed on a relatively large scale. Furthermore, we observed that the population seemed to flow to adjacent areas from 3:00 to 15:00 (Figure 10d). For instance, the percentage of the population flowing from Shenglilu to Guchengtai and Hutai was greater than 16% (Table 3). Similarly, the percentage of the population flowing from Lejiawan to Yunjiakou was 19%, and the percentage of the population flowing from Nianlipu to Shengwukejichanyeyuan, Chaoyang, and Xiaoqiaodajie was more than 13%.
By further exploring the variation in the net mobile traffic for the population in the selected case areas over time, we found that the total number of trips made by the crowd in Shenglilu was significantly greater than that in other case areas, while Nianlipu had the lowest travel intensity. Between 6:00 and 10:00, a large influx of people into Shenglilu was observed, and after 17:00, there was a greater outflow than inflow of people. Many people left Nanchuandonglu between 6:00 and 10:00, but there was an influx after 17:00, indicating a close relationship with residents’ daily commuting patterns (Figure 11).

5. Discussion

When comparing the traditional approach of calculating accessibility based on home or work locations with that based on the dynamic population distribution, we observed certain biases. The method based on home location tended to overestimate the accessibility in the central city, but underestimate the accessibility in distant suburban areas. Conversely, the method based on work location overestimated the accessibility in suburban areas and underestimated the accessibility in central urban areas. As a result, the traditional approach using a static population distribution led to biased results compared to the more accurate dynamic population distribution method. Yun et al. (2020) found that using census data solely for an accessibility analysis could lead to certain errors, and adopting mobile-based population data better represented real-world situations for solving problems of social inequity in primary medical care [49]. Scholars have focused on this issue not only in the context of accessibility studies. Notably, Tan et al. (2017) studied the impact of an uncertain geographic context in a space–time behavior analysis and found that the geographic context based on residential neighborhoods may lead to inaccurate results [50], such as overestimating the contextual effects of neighborhood areas and underestimating those of other activity places.
Furthermore, the accessibility analysis of fever clinics, considering the dynamic population distribution, revealed greater accessibility in urban centers than in the suburbs. This finding aligns with the typical spatial structure of rapidly urbanizing cities in China, which exhibit a compact pattern of outward expansion. The historical development of older urban centers has led to a concentration of high-quality medical facilities, resulting in medical resources being more abundant in the city center and scarce in the periphery. Since fever clinics often rely on existing high-quality medical facilities, their spatial distribution also displays a monocentric pattern, leading to lower accessibility for suburban residents. This geographical disparity between urban centers and suburban areas raises concerns about equitable healthcare access, especially for those living in underserved suburban regions. To address this inequity, it is essential to focus on improving medical resource access for suburban residents. Hashtarkhani et al. (2020) developed an age-included approach to measure the potential accessibility of emergency medical services (EMSs) across the urban and suburban areas of Mashhad city in Iran [51]. They also found that suburban areas had less potential accessibility than central urban areas due to the high density of EMS stations in the city center areas.
Additionally, the accessibility of fever clinics varied substantially based on different traffic patterns and time thresholds. For instance, within a 15 min drive, the population covered by the relatively low, low, and no-supply accessibility levels was three times larger than that within a 30 min drive. In addition, the population covered by relatively high and above accessibility was 1.6 times larger within a 30 min drive than within a 30 min commute by transit. The measurement of the accessibility of healthcare facilities is sensitive to time thresholds, as we have shown. Pereira (2019) studied the future accessibility impacts of transport policy scenarios. Their study showed that equity assessments of transport projects are influenced by the travel time threshold chosen for the cumulative opportunity accessibility analysis [52]. Consequently, decision-makers must consider these variations to make well-informed decisions and have a clear reference basis.
Based on the findings of this study, policymakers should take a series of measures to address spatial inequalities in healthcare accessibility. First, it is recommended to expand healthcare facilities, especially fever clinics, in suburban and underdeveloped areas, or extend the operating hours of existing facilities, particularly during peak demand periods. At the same time, improving public transportation infrastructure, expanding bus routes, increasing the frequency, and providing dedicated medical shuttle services will help mitigate the negative impact of transportation on healthcare accessibility. In addition, it is suggested to integrate real-time population movement data into healthcare planning to dynamically adjust the distribution of medical resources based on population shifts, ensuring that resources can flexibly respond to changing demands. When assessing the healthcare capacity, more metrics should be included, such as the number of doctors, the number of hospital beds, and the amount of medical equipment, to provide a comprehensive evaluation of healthcare service availability. Lastly, considering the impact of time thresholds on accessibility in different regions, policymakers should adjust time thresholds according to local transportation characteristics, ensuring the accuracy and fairness of accessibility assessments. By implementing these measures, the healthcare accessibility gap between urban and suburban areas can be effectively reduced, ensuring that all residents have equal access to essential healthcare services.
This study used Xining city as a case to analyze the impact of the fever visit management (FVM) policy on healthcare accessibility, with the results influenced by the local geography, transportation, and healthcare resource distribution. However, the Gaussian two-step floating catchment area method (G2SFCA) and the dynamic population distribution assessment method used in this study are highly transferable and have been validated in healthcare accessibility research across several cities [23,26], especially in rapidly urbanizing areas or cities with uneven healthcare resource distribution. The findings show that the FVM policy has had a significant spatially uneven impact on healthcare accessibility under different transportation modes and time thresholds, particularly with a clear disparity between urban centers and suburban areas. This phenomenon aligns with the global trend of healthcare resource distribution, where urban centers typically have denser healthcare resources, while the suburbs and remote areas face greater challenges in healthcare accessibility. Additionally, when studying healthcare equity in different urban contexts, Yun et al. (2020) suggested that dynamic population assessments based on mobile phone data provide a more accurate accessibility analysis than traditional static population distribution methods [49]. This study supports that view and indicates that static population distribution methods may overestimate the accessibility in urban centers while underestimating healthcare accessibility in suburban areas.
Despite these important findings, our study has certain limitations. First, we utilized mobile phone data from only four working days to identify users’ DTAs and NTAs. In future research, the use of data from more extended periods could enhance the reliability of such identifications. Second, the calculation of accessibility relied solely on the number of hospital beds as the supply. For more comprehensive assessments, future studies could incorporate factors such as the number of doctors and hospital floor space to better gauge the supply of hospital facilities. Additionally, this study focused on a single city (Xining), and while the findings are valuable for understanding local healthcare accessibility patterns, the generalizability of the results to other countries or healthcare systems is limited. This limitation should be emphasized, and future research could explore how the methods and findings may apply to other urban contexts with different healthcare infrastructures, transportation systems, and population characteristics. Finally, due to data limitations, particularly the lack of real-time traffic flow and detailed transportation mode data, we were unable to assess how variations in these assumptions might impact the results. Future studies with access to more comprehensive data could incorporate such analyses to improve the robustness and accuracy of the conclusions.

6. Conclusions

In this study, we investigated the impact of the COVID-19 pandemic on fever medical care accessibility utilizing the Gaussian two-step floating catchment area method, which incorporates a dynamic population distribution. Our findings led to the following conclusions. First, for both driving and public transport, the accessibility of fever clinics exhibited a spatial distribution pattern, with higher accessibility in central areas and lower accessibility in the outskirts. As the time threshold increased, the accessibility tended to stabilize, with less variation across the study area. Second, the impact of FVM on fever medical care accessibility varied in space. Notably, the decrease in accessibility was more pronounced in suburban areas than in the central city. Specifically, at 10:00 a.m., for travel conditions of 15 min by car and 30 min by public transport, the accessibility in the suburbs decreased by more than 2.8 times that in the central city. Moreover, the decrease in accessibility was more substantial for public transportation than for driving for 30 min, 45 min, and 60 min time thresholds. Third, for a 15 min drive, the accessibility of fever clinics based on the NTA was overestimated in central areas, but underestimated in suburban areas. Conversely, the accessibility based on the DTA showed the opposite trend. Additionally, when the time threshold was increased to 45 min, the accessibility of the entire study area was underestimated.
These conclusions support the hypotheses proposed earlier and underscore the importance of integrating dynamic population data in assessing healthcare accessibility, particularly during crises. By incorporating a real-time population distribution, this study provides valuable insights into optimizing healthcare resource allocation and ensuring equitable access to medical services across all areas, especially in future public health emergencies.

Author Contributions

Zhiyuan Zhao was responsible for conceptualizing the study and securing funding. Youjun Tu contributed to the manuscript drafting and revisions. Yicheng Ding handled data processing and algorithm development. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (No. 42201500).

Data Availability Statement

The data that support the findings of this study are available from the mobile operator, but restrictions apply to the availability of these data, which were used under license for the current study, and so, they are not publicly available. The data are, however, available from the authors upon reasonable request and with the permission of the mobile operator.

Conflicts of Interest

The authors declare that they have no competing interests.

References

  1. Wang, Q.; Huang, L. China’s “Dynamic Clearing” Epidemic Prevention Policy: Achievements, Challenges, and Prospects. Front. Public Health 2022, 10, 978944. [Google Scholar] [CrossRef] [PubMed]
  2. Azar, K.M.J.; Shen, Z.; Romanelli, R.J.; Lockhart, S.H.; Smits, K.; Robinson, S.; Brown, S.; Pressman, A.R. Disparities in Outcomes Among COVID-19 Patients in A Large Health Care System in California. Health Aff. 2020, 39, 1253–1262. [Google Scholar] [CrossRef]
  3. Gingras, L.F.; Pelzman, F.N.; Marquez, A.; Arias, D.; Tung, J. The Cough Cold and Fever Clinic: A Model for the Safe Management of Ambulatory Patients with Suspected COVID-19. Qual. Manag. Healthc. 2022, 31, 38. [Google Scholar] [CrossRef]
  4. Nayan, N.; Kumar, M.K.; Nair, R.K.; Manral, I.; Ghosh, S.; Bhalla, S.; Singh, J.; Monga, A.; Afzal, M.; Kapoor, R. Clinical Triaging in Cough Clinic Alleviates COVID-19 Overload in Emergency Department in India. SN Compr. Clin. Med. 2021, 3, 22–27. [Google Scholar] [CrossRef] [PubMed]
  5. Shen, J.; He, Q.; Shen, T.; Wu, Z.-Q.; Tan, M.-M.; Chen, Y.-L.; Weng, Q.; Nie, L.-M.; Zhang, H.-F.; Zheng, B.; et al. “An Integrated System, Three Separated Responsibilities”, a New Fever Clinic Management Model, in Prevention and Control of Novel Coronavirus Pneumonia. World J. Clin. Cases 2021, 9, 9050. [Google Scholar] [CrossRef] [PubMed]
  6. Cheng, A.; Chen, Y.; Gao, Y.; Sun, P.; Chang, R.; Zhou, B.; Qiu, H.; Cheng, F. Mobile Isolation Wards in a Fever Clinic: A Novel Operation Model during the COVID-19 Pandemic. Epidemiol. Infect. 2021, 149, e61. [Google Scholar] [CrossRef]
  7. Wolfson, J.A.; Leung, C.W. An Opportunity to Emphasize Equity, Social Determinants, and Prevention in Primary Care. Ann. Fam. Med. 2020, 18, 290–291. [Google Scholar] [CrossRef] [PubMed]
  8. Schnake-Mahl, A.S.; Sommers, B.D. Health Care in The Suburbs: An Analysis of Suburban Poverty and Health Care Access. Health Aff. 2017, 36, 1777–1785. [Google Scholar] [CrossRef]
  9. Chen, X.; Orom, H.; Hay, J.L.; Waters, E.A.; Schofield, E.; Li, Y.; Kiviniemi, M.T. Differences in Rural and Urban Health Information Access and Use. J. Rural Health 2019, 35, 405–417. [Google Scholar] [CrossRef]
  10. Zhang, J.; Han, P.; Yang, L. Spatial accessibility of fever clinics for multi-tiered prevention and control on COVID-19 in Beijing. J. Peking Univ. Health Sci. 2021, 53, 543–548. [Google Scholar] [CrossRef]
  11. Wang, D.; Tian, Y.; Qian, Z.; Tang, Y. Analysis on spatial accessibility of fever clinics and designated hospitals in response to epidemic events of infectious diseases: A case study of Hunan province. Chin. J. Health Policy 2021, 14, 74–79. [Google Scholar]
  12. Xu, D.; Wu, Q.; Feng, Y.; Wu, S. Evaluation and Optimization of Fever Clinic Allocation Under Public Health Emergencies: A Case Study of Harbin. J. Geomat. 2023, 48, 157–163. [Google Scholar] [CrossRef]
  13. Hansen, W.G. How Accessibility Shapes Land Use. J. Am. Plan. Assoc. 1959, 25, 73–76. [Google Scholar] [CrossRef]
  14. Marwal, A.; Silva, E. Literature Review of Accessibility Measures and Models Used in Land Use and Transportation Planning in Last 5 Years. J. Geogr. Sci. 2022, 32, 560–584. [Google Scholar] [CrossRef]
  15. Wang, M.; Mu, L. Spatial Disparities of Uber Accessibility: An Exploratory Analysis in Atlanta, USA. Comput. Environ. Urban Syst. 2018, 67, 169–175. [Google Scholar] [CrossRef]
  16. Taleai, M.; Sliuzas, R.; Flacke, J. An Integrated Framework to Evaluate the Equity of Urban Public Facilities Using Spatial Multi-Criteria Analysis. Cities 2014, 40, 56–69. [Google Scholar] [CrossRef]
  17. Rosenthal, M.B.; Zaslavsky, A.; Newhouse, J.P. The Geographic Distribution of Physicians Revisited. Health Serv. Res. 2005, 40, 1931. [Google Scholar] [CrossRef]
  18. Nicholls, S. Measuring the Accessibility and Equity of Public Parks: A Case Study Using GIS. Manag. Leis. 2001, 6, 201–219. [Google Scholar] [CrossRef]
  19. Talen, E. Neighborhoods as Service Providers: A Methodology for Evaluating Pedestrian Access. Environ. Plan. B Plan. Des. 2003, 30, 181–200. [Google Scholar] [CrossRef]
  20. Ogryzek, M.; Podawca, K.; Cienciała, A. Geospatial Tools in the Analyses of Land Use in the Perspective of the Accessibility of Selected Educational Services in Poland. Land Use Policy 2022, 122, 106373. [Google Scholar] [CrossRef]
  21. Radke, J.; Mu, L. Spatial Decompositions, Modeling and Mapping Service Regions to Predict Access to Social Programs. Ann. GIS 2000, 6, 105–112. [Google Scholar] [CrossRef]
  22. Delamater, P.L. Spatial Accessibility in Suboptimally Configured Health Care Systems: A Modified Two-Step Floating Catchment Area (M2SFCA) Metric. Health Place 2013, 24, 30–43. [Google Scholar] [CrossRef]
  23. Luo, W.; Qi, Y. An Enhanced Two-Step Floating Catchment Area (E2SFCA) Method for Measuring Spatial Accessibility to Primary Care Physicians. Health Place 2009, 15, 1100–1107. [Google Scholar] [CrossRef]
  24. Kocatepe, A.; Ulak, M.B.; Ozguven, E.E.; Horner, M.W.; Arghandeh, R. Socioeconomic Characteristics and Crash Injury Exposure: A Case Study in Florida Using Two-Step Floating Catchment Area Method. Appl. Geogr. 2017, 87, 207–221. [Google Scholar] [CrossRef]
  25. Li, M.; Wang, F.; Kwan, M.-P.; Chen, J.; Wang, J. Equalizing the Spatial Accessibility of Emergency Medical Services in Shanghai: A Trade-Off Perspective. Comput. Environ. Urban Syst. 2022, 92, 101745. [Google Scholar] [CrossRef]
  26. Wang, F. Measurement, Optimization, and Impact of Health Care Accessibility: A Methodological Review. Ann. Assoc. Am. Geogr. 2012, 102, 1104–1112. [Google Scholar] [CrossRef] [PubMed]
  27. Langford, M.; Higgs, G.; Fry, R. Multi-Modal Two-Step Floating Catchment Area Analysis of Primary Health Care Accessibility. Health Place 2016, 38, 70–81. [Google Scholar] [CrossRef]
  28. Dai, D. Racial/Ethnic and Socioeconomic Disparities in Urban Green Space Accessibility: Where to Intervene? Landsc. Urban Plan. 2011, 102, 234–244. [Google Scholar] [CrossRef]
  29. Conte, D.; Locascio, A.; Amoruso, J.; McNamara, M. Modeling Multimodal Access to Primary Care in an Urban Environment. Transp. Res. Interdiscip. Perspect. 2022, 13, 100550. [Google Scholar] [CrossRef]
  30. Kim, K.; Kwon, K. Time-Varying Spatial Accessibility of Primary Healthcare Services Based on Spatiotemporal Variations in Demand, Supply, and Traffic Conditions: A Case Study of Seoul, South Korea. J. Transp. Health 2022, 27, 101531. [Google Scholar] [CrossRef]
  31. Li, Z.; Fan, Z.; Song, Y.; Chai, Y. Assessing Equity in Park Accessibility Using a Travel Behavior-Based G2SFCA Method in Nanjing, China. J. Transp. Geogr. 2021, 96, 103179. [Google Scholar] [CrossRef]
  32. Han, F.; Luo, R. Matching of Supply and Demand for Community Service-Oriented Home Care Facilities Based on Accessibility Measurement: A Case Study of Nanjing. Econ. Geogr. 2020, 40, 91–101. [Google Scholar] [CrossRef]
  33. Halas, M. Temporality in the Delimitation of Functional Regions: The Use of Mobile Phone Location Data. Reg. Stud. 2024, 58, 2175–2187. [Google Scholar] [CrossRef]
  34. Chang, X.; Ma, L.; Cui, X.; Tao, T.; Zhao, S. Spatial Equity Evaluation of Rural Eldercare Service Resources Based on Accessibility: A Case Study of Huanxian County of Gansu Province, China. Chin. Geogr. Sci. 2024, 34, 869–885. [Google Scholar] [CrossRef]
  35. Xia, T.; Song, X.; Zhang, H.; Song, X.; Kanasugi, H.; Shibasaki, R. Measuring Spatio-Temporal Accessibility to Emergency Medical Services Through Big GPS Data. Health Place 2019, 56, 53–62. [Google Scholar] [CrossRef]
  36. Colbert, J.; Chuang, I.-T.; Sila-Nowicka, K. Measuring Spatial Inequality of Urban Park Accessibility and Utilisation: A Case Study of Public Housing Developments in Auckland, New Zealand. Landsc. Urban Plan. 2024, 247, 105070. [Google Scholar] [CrossRef]
  37. Wu, W.; Zheng, T. Establishing a “Dynamic Two-Step Floating Catchment Area Method” to Assess the Accessibility of Urban Green Space in Shenyang Based on Dynamic Population Data and Multiple Modes of Transportation. Urban For. Urban Green. 2023, 82, 127893. [Google Scholar] [CrossRef]
  38. Chen, W.; Zeng, Y.; Zeng, J. Impacts of Traffic Accessibility on Ecosystem Services: An Integrated Spatial Approach. J. Geogr. Sci. 2021, 31, 1816–1836. [Google Scholar] [CrossRef]
  39. Lee, W.K.; Sohn, S.Y.; Heo, J. Utilizing Mobile Phone-Based Floating Population Data to Measure the Spatial Accessibility to Public Transit. Appl. Geogr. 2018, 92, 123–130. [Google Scholar] [CrossRef]
  40. Liu, L.; Miller, H.J. Measuring the Impacts of Dockless Micro-Mobility Services on Public Transit Accessibility. Comput. Environ. Urban Syst. 2022, 98, 101885. [Google Scholar] [CrossRef]
  41. Park, J.; Michels, A.; Lyu, F.; Han, S.Y.; Wang, S. Daily Changes in Spatial Accessibility to ICU Beds and Their Relationship with the Case-Fatality Ratio of COVID-19 in the State of Texas, USA. Appl. Geogr. 2023, 154, 102929. [Google Scholar] [CrossRef] [PubMed]
  42. Blondel, V.D.; Decuyper, A.; Krings, G. A Survey of Results on Mobile Phone Datasets Analysis. EPJ Data Sci. 2015, 4, 10. [Google Scholar] [CrossRef]
  43. Liu, J.; Lai, Z.; Meng, B.; Guo, Z.; Liu, X. Assessing Spatial Configuration of Barrier-Free Facilities from the Perspective of Age-Friendliness: A Case Study of Beijing, China. Appl. Geogr. 2024, 172, 103426. [Google Scholar] [CrossRef]
  44. Zhao, Z.; Shaw, S.-L.; Yin, L.; Fang, Z.; Yang, X.; Zhang, F.; Wu, S. The Effect of Temporal Sampling Intervals on Typical Human Mobility Indicators Obtained from Mobile Phone Location Data. Int. J. Geogr. Inf. Sci. 2019, 33, 1471–1495. [Google Scholar] [CrossRef]
  45. Zhu, L.; Shi, F. Spatial and Social Inequalities of Job Accessibility in Kunshan City, China: Application of the Amap API and Mobile Phone Signaling Data. J. Transp. Geogr. 2022, 104, 103451. [Google Scholar] [CrossRef]
  46. Li, J.; Zhao, P.; Zhang, M.; Deng, Y.; Liu, Q.; Cui, Y.; Gong, Z.; Liu, J.; Tan, W. Exploring Collective Activity Space and Its Spatial Heterogeneity Using Mobile Phone Signaling Data: A Case of Shenzhen, China. Travel. Behav. Soc. 2025, 38, 100920. [Google Scholar] [CrossRef]
  47. He, L.; Li, W.; Li, J.; Sun, J. Urban Mobility Analytics amid COVID-19 Pandemic: A Framework for Promoting Work Resumption Based on Mobile Phone Data. J. Transp. Geogr. 2024, 117, 103887. [Google Scholar] [CrossRef]
  48. Xu, Y.; Shaw, S.-L.; Fang, Z.; Yin, L. Estimating Potential Demand of Bicycle Trips from Mobile Phone Data—An Anchor-Point Based Approach. ISPRS Int. J. Geo Inf. 2016, 5, 131. [Google Scholar] [CrossRef]
  49. Yun, S.B.; Kim, S.; Ju, S.; Noh, J.; Kim, C.; Wong, M.S.; Heo, J. Analysis of Accessibility to Emergency Rooms by Dynamic Population from Mobile Phone Data: Geography of Social Inequity in South Korea. PLoS ONE 2020, 15, e0231079. [Google Scholar] [CrossRef]
  50. Tan, Y.; Chai, Y.; Guan, M. The impact of the uncertain geographic context on the space-time behavior analysis: A case study of Xining, China. Acta Geogr. Sin. 2017, 72, 657–670. [Google Scholar]
  51. Hashtarkhani, S.; Kiani, B.; Bergquist, R.; Bagheri, N.; VafaeiNejad, R.; Tara, M. An Age-Integrated Approach to Improve Measurement of Potential Spatial Accessibility to Emergency Medical Services for Urban Areas. Int. J. Health Plan. Manag. 2020, 35, 788–798. [Google Scholar] [CrossRef] [PubMed]
  52. Pereira, R.H.M. Future Accessibility Impacts of Transport Policy Scenarios: Equity and Sensitivity to Travel Time Thresholds for Bus Rapid Transit Expansion in Rio de Janeiro. J. Transp. Geogr. 2019, 74, 321–332. [Google Scholar] [CrossRef]
Ijgi 14 00117 g001
Figure 1. Study area and distribution of medical facilities.
Figure 1. Study area and distribution of medical facilities.
Ijgi 14 00117 g001
Ijgi 14 00117 g002
Figure 2. Technical flow chart.
Figure 2. Technical flow chart.
Ijgi 14 00117 g002
Ijgi 14 00117 g003
Figure 3. The dynamic average accessibility of fever clinics in the main urban area of Xining city under driving and public transport modes.
Figure 3. The dynamic average accessibility of fever clinics in the main urban area of Xining city under driving and public transport modes.
Ijgi 14 00117 g003
Ijgi 14 00117 g004
Figure 4. The hourly accessibility of fever clinics in the main urban area of Xining city for under 15 min of driving.
Figure 4. The hourly accessibility of fever clinics in the main urban area of Xining city for under 15 min of driving.
Ijgi 14 00117 g004
Ijgi 14 00117 g005
Figure 5. Changes in the accessibility of fever clinics in the daytime compared with the nighttime for driving conditions of under 15 min.
Figure 5. Changes in the accessibility of fever clinics in the daytime compared with the nighttime for driving conditions of under 15 min.
Ijgi 14 00117 g005
Ijgi 14 00117 g006
Figure 6. Spatial distribution of relative changes in the accessibility of fever clinics relative to general hospitals based on the dynamic population distribution in the main urban area of Xining city for the driving mode.
Figure 6. Spatial distribution of relative changes in the accessibility of fever clinics relative to general hospitals based on the dynamic population distribution in the main urban area of Xining city for the driving mode.
Ijgi 14 00117 g006
Ijgi 14 00117 g007
Figure 7. Spatial distribution of relative changes in the accessibility of fever clinics relative to general hospitals based on the dynamic population distribution in the main urban area of Xining city for the public transport mode.
Figure 7. Spatial distribution of relative changes in the accessibility of fever clinics relative to general hospitals based on the dynamic population distribution in the main urban area of Xining city for the public transport mode.
Ijgi 14 00117 g007
Ijgi 14 00117 g008
Figure 8. Relative changes in accessibility based on the NTA versus accessibility based on the dynamic population distribution for the driving mode.
Figure 8. Relative changes in accessibility based on the NTA versus accessibility based on the dynamic population distribution for the driving mode.
Ijgi 14 00117 g008
Ijgi 14 00117 g009
Figure 9. Relative changes in accessibility based on DTA versus accessibility based on the dynamic population distribution for the driving mode.
Figure 9. Relative changes in accessibility based on DTA versus accessibility based on the dynamic population distribution for the driving mode.
Ijgi 14 00117 g009
Ijgi 14 00117 g010
Figure 10. Spatial distribution of population displacement in the main urban area of Xining city.
Figure 10. Spatial distribution of population displacement in the main urban area of Xining city.
Ijgi 14 00117 g010
Ijgi 14 00117 g011
Figure 11. The changing dynamics of population flow in the study area.
Figure 11. The changing dynamics of population flow in the study area.
Ijgi 14 00117 g011
Table 1. Changes in the accessibility of fever clinics relative to general hospitals in the main urban area of Xining city by driving mode (unit: %).
Table 1. Changes in the accessibility of fever clinics relative to general hospitals in the main urban area of Xining city by driving mode (unit: %).
d15d30d45d60
Time0:0010:000:0010:000:0010:000:0010:00
Entire study area−16.80−25.63−11.31−18.24−10.94−17.03−10.96−17.01
Central city−9.95−16.67−10.92−16.89−10.92−16.91−10.95−16.95
Suburbs−31.52−46.55−12.15−21.36−10.98−17.33−11.00−17.15
Table 2. Changes in the accessibility of fever clinics relative to general hospitals in the main urban area of Xining city by transit mode at 10:00 (unit: %).
Table 2. Changes in the accessibility of fever clinics relative to general hospitals in the main urban area of Xining city by transit mode at 10:00 (unit: %).
t15t30t45t60
Entire study area−16.15−33.50−24.65−20.27
Central city−19.68−21.57−17.36−16.96
Suburbs−7.91−61.37−41.67−27.99
Table 3. The percentages of population flow in the study area from 3:00 to 15:00.
Table 3. The percentages of population flow in the study area from 3:00 to 15:00.
3:00 Position15:00 PositionPercentage of People
ShengliluGuchengtai16.17
Hutai15.79
Pengjiazhai9.12
NanchuandongluPengjiazhai10.91
Nanchuanxilu9.82
Guchengtai8.06
LejiawanYunjiakou19.12
Bayilu8.35
Pengjiazhai6.84
NianlipuShengwukejichanyeyuan14.89
Chaoyang14.35
Xiaoqiaodajie13.27
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zhao, Z.; Tu, Y.; Ding, Y. How Did the Fever Visit Management Policy During the COVID-19 Epidemic Impact Fever Medical Care Accessibility? ISPRS Int. J. Geo-Inf. 2025, 14, 117. https://doi.org/10.3390/ijgi14030117

AMA Style

Zhao Z, Tu Y, Ding Y. How Did the Fever Visit Management Policy During the COVID-19 Epidemic Impact Fever Medical Care Accessibility? ISPRS International Journal of Geo-Information. 2025; 14(3):117. https://doi.org/10.3390/ijgi14030117

Chicago/Turabian Style

Zhao, Zhiyuan, Youjun Tu, and Yicheng Ding. 2025. "How Did the Fever Visit Management Policy During the COVID-19 Epidemic Impact Fever Medical Care Accessibility?" ISPRS International Journal of Geo-Information 14, no. 3: 117. https://doi.org/10.3390/ijgi14030117

APA Style

Zhao, Z., Tu, Y., & Ding, Y. (2025). How Did the Fever Visit Management Policy During the COVID-19 Epidemic Impact Fever Medical Care Accessibility? ISPRS International Journal of Geo-Information, 14(3), 117. https://doi.org/10.3390/ijgi14030117

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop