Successful pandemic management through computer science: a case study of a financial corporation with workers on premises

Abstract

Background: In November 2019, an infectious agent that caused a severe acute respiratory illness was first detected in China. Its rapid spread resulted in a global lockdown with negative economic impacts. In this regard, we expose the solutions proposed by a multinational financial institution that maintained their workers on premises, so this methodology can be applied to possible future health crisis.

Objectives: To ensure a secure workplace for the personnel on premises employing biomedical prevention measures and computational tools.

Methods: Professionals were subjected to recurrent COVID-19 diagnostic tests during the pandemic. The sanitary team implemented an individual following to all personnel and introduced the information in databases. The data collected were used for clustering algorithms, decision trees, and networking diagrams to predict outbreaks in the workplace. Individualized control panels assisted the decision-making process to increase, maintain, or relax restrictive measures.

Results: 55,789 diagnostic tests were performed. A positive correlation was observed between the cumulative incidence reported by Madrid's Ministry of Health and the headcount. No correlation was observed for occupational infections, representing 1.9% of the total positives. An overall 1.7% of the cases continued testing positive for COVID-19 after 14 days of quarantine.

Conclusion: Based on a combined approach of medical and computational science tools, we propose a management model that can be extended to other industries that can be applied to possible future health crises. This work shows that this model resulted in a safe workplace with a low probability of infection among workers during the pandemic.

Keywords: COVID-19; epidemiology; health informatics; information management; occupational and industrial medicine.

Figure 1

Return to the office flow chart. Depending on the initial state (COVID-positive, contact with positive, and presence of compatible symptoms), the return to the office strategy was adapted to minimize the risk of outbreaks. Ag, Antigen test; Ct, qRT-PCR cycles; IgM, IgM antibody serologic test; IgG, IgG antibody serologic test.

Figure 2

Monthly COVID-19 tests performed in headquarters depending on their type. qRT-PCR (PCR): gray, antigens (Ag): black, quantitative IgM/IgG antibodies serologies (q-Ab): dashed lines, and qualitative IgM/IgG antibodies rapid tests (r-Ab): dotted.

Figure 3

Networking diagram of confirmed cases of COVID-19 infections and occupational contacts. Current contacts and infections are shown in bright colors, resolved quarantines and infections are shown in dimmed colors, and occupational outbreaks (three or more related infections) are rounded. Each node represents an employee that had contact with an infected coworker and are connected through lines, being central nodes the first case of infection. Contacts who developed the disease are shown in orange, yellow, red, and pink. Infections exclusively attributed to occupational contacts are shown in red and pink.

Figure 4

14-day cumulative incidence related with the population of the perimeter (habitants or employees). Community of Madrid incidence (dotted line), HQ (solid black line), and occupational infections (solid gray line). Holiday periods (summer, Christmas, and Holy week) are shaded, showing a relationship with outbreak waves. The first wave (between March and June 2020) is not represented due to the absence of validated data, as the diagnostic tests were only available for severe cases.