Modeling the Airborne Infection Risk of Tuberculosis for a Research Facility in eMalahleni, South Africa

• Published in: Risk analysis Vol. 39; no. 3; pp. 630 - 646
• Main Authors: Küsel, Ralf R., Craig, Ian K., Stoltz, Anton C.
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 01.03.2019
• Subjects: Airborne infection; Airborne infectious disease transmission; biological system modeling; Computer simulation; Control rooms; Disease transmission; Efficacy; Electric power; Environmental control; Environmental testing; Experiments; Health risks; Infections; Infectious diseases; Irradiation; Mathematical models; Measurement; Mycobacterium tuberculosis; parameter extraction; Power; Reliance; Risk; Risk assessment; Sensitivity analysis; Simulation; Tuberculosis; Ultraviolet radiation; Uncertainty; South Africa; biological system modeling; Mycobacterium tuberculosis; Airborne infectious disease transmission; parameter extraction; sensitivity analysis
• ISSN: 0272-4332; 1539-6924
• DOI: 10.1111/risa.13180

A detailed mathematical modeling framework for the risk of airborne infectious disease transmission in indoor spaces was developed to enable mathematical analysis of experiments conducted at the Airborne Infections Research (AIR) facility, eMalahleni, South Africa. A model was built using this framework to explore possible causes of why an experiment at the AIR facility did not produce expected results. The experiment was conducted at the AIR facility from August 31, 2015 to December 4, 2015, in which the efficacy of upper room germicidal ultraviolet (GUV) irradiation as an environmental control was tested. However, the experiment did not produce the expected outcome of having fewer infections in the test animal room than the control room. The simulation results indicate that dynamic effects, caused by switching the GUV lights, power outages, or introduction of new patients, did not result in the unexpected outcomes. However, a sensitivity analysis highlights that significant uncertainty exists with risk of transmission predictions based on current measurement practices, due to the reliance on large viable literature ranges for parameters.