Open-access geodata as a resource against infectious diseases

Abstract

In the era of big data, open-access information online is ever increasing. Harnessing its potential can find useful application in many scientific sectors, including public health. One pressing concern within this realm is the well-being of food animals, whose importance extends beyond economics to encompass human health. Infectious diseases, antimicrobial resistance (AMR), and a limited access to healthcare professionals currently pose a threat to food animals and the 1.3 billion people worldwide who rely on them for their subsistence. This doctoral thesis explores the potential of using open-access data, in conjunction with geospatial models and user-friendly platforms, to provide new resources that can aid decision-makers in improving the health of animals raised for food. Chapter 2 introduces the open-access platform resistancebank.org, an online repository that centralizes 2,045 point prevalence surveys (PPS) reporting AMR prevalence estimates in food animals. Launched in 2019 as a shiny application (R programming language), the platform consists of 42,891 resistance prevalence estimates of foodborne pathogens sampled from terrestrial and aquatic species in low- and middle-income countries. Besides individual PPS, the platform provides access to AMR maps at the 10x10 km2 resolution, country-level reports of AMR, and allow users to upload the results of their PPS through an online form or an Excel template. Chapter 3 illustrates how open-access addresses of veterinarians were used to identify areas where veterinarians are farther than 1 hour of travel time from food animals (i.e., “veterinary coldspots”). First, the assembling of a global address book of 303,745 veterinarians’ addresses using web-scraping technique is presented. Log-Gaussian Poisson Regression models and spatial covariates were then used to predict the global distribution of veterinarians at the 10x10 km2 resolution. The resulting map showed that 43% of veterinarians are in high-income countries, which, however, account for just 21.2% of the global production of food animals. As a consequence, the highest percentages of all food animals in coldspots were identified in Asia (44.1%), Latin America (27.7%), and Africa (18.7%). In Chapter 4, AMR and veterinary capacity were used as case studies to present analyses about the optimal allocation of health facilities in underserved areas. Two geographic approaches were used to identify the locations of an additional 5% of the national number of i) health facilities (hospitals and clinical laboratories) to equip with laboratories for antimicrobial susceptibility tests (ASTs) in five African countries and ii) veterinarians predicted in Chapter 3, with the aim to reduce the number of food animals in coldspots. The allocation of laboratories testing for AMR resulted in a coverage of ~21 million people that can reach a laboratory within 1 hour. The allocation of veterinarians in countries with coldspots ensured veterinary care for 32.6% of the global number of cattle, chickens, and pigs living in coldspots. In Chapter 5, the focus shifts on the global health crisis sparked by COVID-19. As for many students worldwide, the pandemic affected the trajectory of my Ph.D. However, thanks to the skillset acquired while working on the project presented in Chapter 2, I was actively involved in research efforts aimed in monitoring the spread of the disease and the availability of healthcare resources in Switzerland. I set-up the open-access platform icumonitoring.ch used to summarize bi-weekly forecasts of COVID-19 cases, hospitalizations, intensive care units’ occupancy, and ventilators’ availability at the regional-, cantonal-, and hospital-level in Switzerland. Chapter 5 also includes my contribution in providing the Swiss Federal Office of Public Health with daily geocodings of COVID-19 cases. The output of this process was used to produce daily hotspots’ maps of COVID-19 cases and to highlight how hospitalizations, deaths, and the prevalence of positive COVID-19 cases were higher among the population living in the poorest neighbors of Switzerland.