Optimal Allocation of COVID-19 Test Kits Among Accredited Testing Centers in the Philippines

Christian Alvin H Buhat; Jessa Camille C Duero; Edd Francis O Felix; Jomar Fajardo Rabajante; Jonathan B Mamplata

We determine the GPS coordinates of all the cities and municipalities in the Philippines, and of the 11 accredited testing centers, using Geocode by Awesome Table app in google spreadsheets. This app returns the latitude and longitude coordinates of the testing centers' locations and the geographic centers of the cities and municipalities. We use these coordinates in the computations except for two: Datu Hoffer Ampatuan and Shariff Sardona Mustapha, Maguindanao. Geocode app cannot locate Shariff Sardona Mustapha, and it returned incorrect coordinates for Datu Hoffer Ampatuan. We use the coordinates of one barangay from each municipality instead. In addition, we compute for various dispersal length scale parameters k shown in (3) We utilize the 11 accredited testing centers to determine the optimal distribution of testing kits. Shown in Table 1 is the proportion of testing kits to be distributed among the 11 testing centers using different values for k. We can observe that UP-NIH will receive the biggest proportion, ranging from 16.25% to 44.80%, while Southern Philippines Medical Center will receive the smallest proportion which is 0.01%. From the results, we notice that San Lazaro Hospital, UP-NIH and Lung Center of the Philippines got the majority of the proportion, across all scenarios. If mass testing will push through, these three hospitals will be overworked and will not be able to deliver timely results on the tests. To address this, we propose to add more testing centers. We identify 10 communities with the highest number of infected individuals and no existing accredited testing center. We will treat these 10 communities as the location of the suggested additional testing centers. We then compute for the optimal distribution of testing kits to the 21 testing centers. The proportion of test kits allocated to each of UP-NIH, San Lazaro Hospital, Lung Center of the Philippines and Makati City is greater than 10%. We can say that this distribution poses concerns similar to that of only having 11 testing centers. To reduce the proportion of test kits allocated to each of these testing centers, we consider all communities with more than 5,000 infected individuals. We then identify 19 communities as locations of new testing centers in addition to the original 11 and compute for the optimal distribution of testing kits to the 30 testing centers.