Objectives  At present, China has invested a large amount of resources in improving primary medical and health services. Aiming at 15 min medical circle, a variety of primary medical security facilities such as clinics and medical insurance stations have been built. Evaluation of their accessibility is of great significance for rational planning and optimization of facilities.

  Methods  We made full use of pan-map resources such as online map, population thermal map, path planning service and isochronous circle range service, and improved the Huff-based two-step floating catchment area method (Huff 2SFCA) from the perspective of travel capacity difference and day and night population distribution difference, and formed a 15 min nucleic acid sampling circle evaluation framework.

  Results  Taking Shanghai as an example to analyze the accessibility of medical facilities in 15 min living circle, the results show that: (1) In all administrative districts except Baoshan and Pudong in the study area, the area coverage and population coverage of the 15 min sampling circle are above 90%, among which Hongkou and Jing?an have the highest accessibility. (2) Affected by the separation between work and housing of commuters, the areas with large permanent residents tend to have high accessibility during the day, and the accessibility decreases at night. On the contrary, the areas with industrial clusters have significantly higher accessibility at night. (3) The surrounding development zones such as Pudong and Baoshan show the characteristics of polarization of accessibility. The accessibility of well-developed residential areas is high, while the accessibility of developing new industrial parks is low.The accessibility distribution of the old city is relatively more balanced, and can better meet the needs of the elderly and the working population.

  Conclusions  The results show that the current spatial layout of sampling stations considers the travel behavior of residents, and can meet the needs of most residents for testing near their residence and work place. For areas with high or low accessibility, the resource allocation of sampling points can be optimized according to the diurnal variation of population distribution.