| 摘要: |
| 发热门诊作为应对突发公共卫
生事件的重要防线，是保障城乡居民生
命健康、完善疫情防控体系的前沿关口。
其空间布局的科学合理性直接关系到城
乡疫情防控的响应效率。研究以哈尔滨
市为例，以“高效可达与全覆盖的发热
门诊医疗资源布局”为目标，提出了发
热门诊“可达性评价—选址模型优化—
适宜性评估辅助决策”的规划评价与实
施框架。通过高斯两步移动搜索法对哈
尔滨市人口需求数据和发热门诊供给数
据的可达性特征评价，识别发现，部分
城市外围地区和偏远乡镇地区成为发热
门诊的薄弱区和空白区。通过粒子群算
法在发热门诊匮乏区进行设施点选址。
最后，结合发热门诊作为传染性监测型
医疗设施的特性，构建基于人口需求与
医疗基底、交通条件、风险要素、危险
要素等4 类14 项因子指标的风险要素指
标体系，对新增发热门诊的规划适宜性
进行评价，提出具有可实施性的分类优
化策略。研究为我国发热门诊等应急医
疗设施优化布局提供参考借鉴。 |
| 关键词: 发热门诊 多源数据 粒子群
算法 适宜性评价 哈尔滨 |
| DOI：10.13791/j.cnki.hsfwest.20241112008 |
| 分类号: |
| 基金项目:国家自然科学基金面上项目（52478047）；国家自然科学基金青年项目（41001096） |
|
| Evaluation and configuration optimization of fever clinic based on multi-source data: Acase study of Harbin |
|
XU Daming,WU Qian,FENG Yingkun
|
| Abstract: |
| As a crucial frontier for responding to public health emergencies, fever clinics play an
irreplaceable role in safeguarding the life and health of urban and rural residents and improving the
epidemic prevention and control system. The scientific rationality of their spatial layout directly
determines the response efficiency of urban and rural epidemic prevention and control, as well as the
effectiveness of early monitoring and intervention of infectious diseases. Taking 128 streets (townships)
in seven municipal districts of Harbin as the research area, this study aims to achieve “efficient
accessibility and full coverage of fever clinic medical resource layout” and proposes a comprehensive
planning evaluation and implementation framework for fever clinics, which consists of three core
links: “accessibility evaluation-site selection model optimization-suitability assessment aided decisionmaking”.
To accurately assess the current layout characteristics of fever clinics, this study adopts a 1 km
grid as the minimum research unit, which does not only capture the spatial differences of population and
POI with high precision but also aligns with the 15-minute community living circle planning. Based on
multi-source data including administrative division data, population distribution raster data, road traffic
network data, fever clinic information, and POI data, the Gaussian two-step floating catchment area
method (2SFCA) is used to evaluate the accessibility of population demand and fever clinic supply. This
method is improved in three aspects: demand point data precision, actual travel time calculation based on
road network speed using OD cost matrix, and spatial attenuation rule establishment with Gaussian
function, ensuring the accuracy of accessibility evaluation results. The evaluation results reveal
significant inequities in the accessibility of fever clinic services in Harbin. Fever clinics show an obvious
single-center agglomeration feature in old urban areas such as Nangang and Xiangfang, while peripheral
urban areas and remote townships have become weak areas and blank areas with poor accessibility.
Specifically, Songbei District has no fever clinics at all, and 8 townships in Hulan District have zero
accessibility due to long transportation distances. To address this problem, the K-means clustering
algorithm is first used to determine the minimum number of new fever clinics required under the 30-
minute travel threshold, and the result shows that 10 new facilities need to be added. Subsequently, the
Particle Swarm Optimization (PSO) algorithm, which is advantageous for global optimization with fast
convergence speed and few parameters, is applied to select the location coordinates of the 10 new fever
clinics, with the goal of minimizing the total travel distance weighted by population demand. Considering
the characteristics of fever clinics as infectious disease surveillance medical facilities, a risk factor index
system consisting of four primary categories and 14 secondary indicators is constructed to evaluate the
planning suitability of newly added fever clinics. The primary categories include population demand and
medical base (population density, elderly density, medical density), traffic conditions (traffic convenience,
road density), risk factors (distance to airports/stations, subway entrances, comprehensive markets,
commercial facilities, catering establishments, entertainment venues, and schools), and hazard
factors (distance to chemical enterprises and gas stations). The combined weighting method, integratingthe analytic hierarchy process (AHP) for subjective weighting and the entropy weight method for objective weighting (with equal weight of 50% each), is used
to determine the weight of each indicator, ensuring the comprehensiveness and rationality of the evaluation. Based on the suitability evaluation results, the
initial site selection results of the PSO algorithm are adjusted to enhance the feasibility and operability of the planning. The adjustment principle prioritizes
existing medical facilities with high suitability evaluation levels within a 4 877 m buffer zone (half of the average travel distance between townships). Finally,
the optimized site selection plan includes two secondary medical institutions, one community health service center, six township health centers, and one village
clinic, among which one village clinic needs to be upgraded and 1 township health center requires relocation and merger to meet the standardized configuration
requirements of fever clinics.This study comprehensively addresses the practical problems of uneven distribution, insufficient coverage, and poor operability of
planning in the current layout of fever clinics. The proposed technical framework and optimization strategies do not only provide a feasible solution for the
spatial layout optimization of fever clinics in Harbin but also offer valuable references for the planning and construction of emergency medical facilities such as
fever clinics in other cities in China. It helps to improve the overall capacity of epidemic prevention and control, strengthen social system resilience, and
promote the construction of a healthy and reasonable urban medical and health epidemic prevention facility system. |
| Key words: fever clinic multi-source data Partide Swarm Optimizaticm (PSO) suitability evaluation Harbin |
