[PYTHON] If the people of Tokyo become seriously ill with the new coronavirus, they may be taken to a hospital in Kagoshima prefecture.

table of contents

[Introduction](# Introduction) [Current status of medical collapse risk](#Current status of medical collapse risk) [Whether or not to secure a bed to accept the severely ill](# About whether or not to secure a bed to accept the severely ill) [Severely ill person transport model](#Severely ill person transport model) [About the destination of the severely ill](#About the destination of the severely ill) Conclusion

Introduction

-** Background ** Infectious disease designated medical institutions are tight, and it is extremely difficult to secure sufficient beds to accommodate severely ill patients. In various places, measures have been taken to recuperate mildly ill people at hotels and homes in order to secure beds for the severely ill.

• Task Now that the infection of the new coronavirus is spreading, it will be important how quickly the seriously ill people who occur at the same time can be transported to the designated medical institution for infectious diseases. In addition, it is desirable that the distance be as short as possible to avoid the risk of infection due to transportation and further aggravation.
• Purpose --Visualize the tightness of medical institutions designated for infectious diseases due to the spread of new coronavirus infection and understand the current state of medical collapse risk ――To judge by simulation whether it is possible to secure a bed to accept the severely ill when all the mildly ill are treated at a hotel or home. --Construct a mathematical model that minimizes the total transport distance of the severely ill and consider its behavior ――Based on the above analysis results, raise or lower your awareness of the crisis regarding the new coronavirus. (** From the conclusion, the result was that the people of Tokyo were transported to Kagoshima prefecture, and the sense of crisis exploded **)

Writer's sense of crisis (5-grade evaluation): ☆☆☆☆☆ (Initial value: Everyone Corona Corona is noisy)

Current status of medical collapse risk

We will visualize the relationship between the number of beds and the number of infected people at designated medical institutions for infectious diseases by prefecture, and consider the current risk of medical collapse. For the number of beds by prefecture and the number of infected people by prefecture, only New Coronavirus Patients Open Data was used. Obtained from New Coronavirus Countermeasure Dashboard as of 20:00 on April 23, 2020.

A heat map of bed usage rates by prefecture (left figure) and top 10 bed usage rates (center table) and Worst 10 (right table) when all infected people use designated medical institutions for infectious diseases.

--There are many prefectures with 100% or more, and ** there is no choice but to take measures to recuperate mildly ill people at hotels, homes, etc. ** --Toyama and Okinawa, which have few beds, are also in the Top 10, and ** the risk of medical collapse is not just a matter of population **

Writer's sense of crisis: ★ ☆☆☆☆ (I don't feel like it's going to happen for the time being)

Whether or not to secure a bed to accept severely ill people

From the current state of the risk of medical collapse, it turned out to be a dangerous atmosphere. Next, what is worrisome is that even if all mildly ill people are treated at a hotel or home, is it possible to secure enough beds to accommodate severely ill people in the first place?

According to a report by Novel Coronavirus Pneumonia Emergency Response Epidemiology Team (2020), infected persons in China (4) Of the 46,672 people), about 81% were mildly ill and about 19% were severely ill, including severely ill.

Based on this report, we will simulate the status of securing beds. Assuming that 19% of all infected people are severely ill, the bed usage rate will be calculated for each prefecture. We will consider the bed usage rate of each prefecture when the random number is changed and repeated 1000 times.

The heat map of the average bed usage rate by prefecture (left figure) and the top 10 (center table) and Worst10 (right table) of the average bed usage rate of the simulation results.

――The average bed utilization rate is 100% or more in 9 prefectures, and ** even severely ill people may not be admitted to designated medical institutions for infectious diseases in their area ** -** In the worst case, you may have to transport to a neighboring prefecture **

The author's sense of crisis: ★★★ ☆☆ (The average bed usage rate in Tokyo is over 100% ...)

Severely ill transport model

From the simulation results, it turned out that the atmosphere was just like that. The next thing to worry about is ** what should I do if I am a seriously ill person who could not be admitted to an infectious disease designated medical institution in my area **? This time, although it is a painstaking measure, we will consider transporting it to another prefecture. There may be many pros and cons, but [trolley problem](https://ja.wikipedia.org/wiki/%E3%83%88%E3%83%AD%E3%83%83%E3% 82% B3% E5% 95% 8F% E9% A1% 8C) I had no choice but to do this.

In order to avoid the risk of infection and further aggravation due to transportation, it is desirable that the distance be as short as possible, so we will consider a mathematical model that minimizes the total transportation distance of the severely ill.

Click here if you are interested in the details of the mathematical model

--The set of severely ill persons is $ I $, and the set of designated medical institutions for infectious diseases is $ J $. --Transport $ j (\ in J) $ to the infectious disease designated medical institution $ i (\ in I) $ for the severely ill. --The distance when a seriously ill person $ i $ moves to an infectious disease designated medical institution $ j $ is $ d_ {ij} $. --If you want to allocate $ i $ for a seriously ill person to $ j $, a medical institution designated for infectious diseases, $ x_ {ij} = 1 $, otherwise $ x_ {ij} = 0 $. --Find $ x_ {ij} $ that minimizes the total transport distance. --The upper limit of the number of beds $ C_j $ of the designated medical institution for infectious diseases $ j $ shall not be exceeded. --All severely ill persons will be transported to one of the designated medical institutions for infectious diseases $ j $.

As a numerical experiment, we will optimize the destination of 19% of all infected people who are severely ill. However, due to lack of data on location information, there is only one location information for severely ill people in each prefecture, the location of the prefectural office of residence, and the distance is [Vincenty method](https: //qiita.com/r-fuji/items/99ca549b963cedc106ab) will be derived.

Heat map of average bed utilization by prefecture before and after optimization (left / left center map), summary statistics of transport distance (right center table), distribution (right figure).

--The bed usage rate in all areas other than Kagoshima and Okinawa prefectures will be 100%, and ** the situation where medical care will completely collapse unless the number of infected people stops increasing ** ――The total transportation distance was 201940km, which was 187km per person. --While many people are transported to designated medical institutions for infectious diseases in their area of residence, some people are transported in units of several hundred kilometers **.

Writer's sense of crisis: ★★★★ ☆ (Isn't it bad that Japan is bright red ?!)

About the destination of severely ill people

From the simulation results, it was found that the atmosphere seems to be realistic. The next thing to worry about is ** how did the seriously ill person move **? Visualize the transition network diagram of the severely ill.

--Severely ill people in Tokyo are transported to Hokkaido in the north and Kagoshima prefecture in the south to designated medical institutions for infectious diseases. -** It was found that there are cases where acceptance of other prefectures is prioritized in order to minimize the total transportation distance ** -** The maximum transport distance of 962 km is when a seriously ill person in Tokyo is transported to Kagoshima prefecture, and there are 31 people **

Writer's sense of crisis: ★★★★★ (I'm tired of traveling 962km even when I'm fine)

in conclusion

――Thank you very much for reading this far. I think that there are many points that I have been arguing about, but I would appreciate it if you could comment if you have any impressions or advice. ――In SIGNATE's COVID-19 Challenge (Phase 2) for the purpose of data analysis that contributes to the decision making of social distance strategy (at least it will be the subject of discussion) We are also conducting a similar analysis called "Visualization of the tightness of designated medical institutions for infectious diseases and proposal of a transportation plan model for the severely ill". If you are interested, please read, like, and comment.