Impact of Digital Contact Tracing and Other Nonpharmaceutical Interventions on Pandemic Control: Microlevel, Behavior-Driven Agent-Based Model Analysis
Journal of Medical Internet Research ·
Background: Nonpharmaceutical interventions (NPIs), including digital contact tracing (DCT), are central to control the spread of airborne pathogens. Nevertheless, the effectiveness of individual interventions and the role of personal behavior remain insufficiently understood. Objective: This study aimed to understand how NPI combination and the individual’s behavior contribute to reducing airborne pathogen transmission. Methods: We disentangle the efficacy of individual NPIs, including DCT, with a novel microlevel, behavior-driven agent-based model (ABM) that simulates individual human behavior to analyze the effectiveness of nonpharmaceutical interventions on airborne pathogen propagation. Our model’s Zeitgeber architecture delineates contextual characteristics, including daytime, daily routines, locations, and activities. Our method determines each agent’s current location and behavior in a realistic environment under NPI restrictions. We model viral load transfer between agents from contact duration, distance, and the infected agent’s infectiousness level. We examine the effects of DCT-related behavior parameters, including adoption, adherence, and compliance, with a default intervention, and with further restricting and relaxing NPIs, on key pandemic indicators. Results: The effect analysis of personal choices regarding DCT indicates that a high DCT adoption rate (activate DCT) should be the first goal of a DCT implementation campaign, followed by adherence (notify others), and compliance (follow recommendations, if notified). There is no monotonic path in the behavior parameter space to improve pandemic characteristics. Assuming realistic behavior with regard to DCT, total infections were reduced by 43% in the default intervention, while DCT combined with other NPIs reduced total infections by up to 52%. Surprisingly, however, some restricting NPI combinations do not improve pandemic characteristics. Conclusions: DCT implementations will face challenges, as pandemic characteristics do not consistently improve when behavior parameters ( adoption, adherence, and compliance) are increased. When considering realistic behavior, more is not always better; NPI combinations can interfere with each other to the detriment of pandemic control. Our approach offers fine-grained insight on the effectiveness of NPI combinations that cannot be obtained in human studies due to confounding effects. Thus, our approach can guide future pandemic control efforts and prioritization for pandemic preparedness.
Background: Nonpharmaceutical interventions (NPIs), including digital contact tracing (DCT), are central to control the spread of airborne pathogens. Nevertheless, the effectiveness of individual interventions and the role of personal behavior remain insufficiently understood. Objective: This study aimed to understand how NPI combination and the individual’s behavior contribute to reducing airborne pathogen transmission. Methods: We disentangle the efficacy of individual NPIs, including DCT, with a novel microlevel, behavior-driven agent-based model (ABM) that simulates individual human behavior to analyze the effectiveness of nonpharmaceutical interventions on airborne pathogen propagation. Our model’s Zeitgeber architecture delineates contextual characteristics, including daytime, daily routines, locations, and activities. Our method determines each agent’s current location and behavior in a realistic environment under NPI restrictions. We model viral load transfer between agents from contact duration, distance, and the infected agent’s infectiousness level. We examine the effects of DCT-related behavior parameters, including adoption, adherence, and compliance, with a default intervention, and with further restricting and relaxing NPIs, on key pandemic indicators. Results: The effect analysis of personal choices regarding DCT indicates that a high DCT adoption rate (activate DCT) should be the first goal of a DCT implementation campaign, followed by adherence (notify others), and compliance (follow recommendations, if notified). There is no monotonic path in the behavior parameter space to improve pandemic characteristics. Assuming realistic behavior with regard to DCT, total infections were reduced by 43% in the default intervention, while DCT combined with other NPIs reduced total infections by up to 52%. Surprisingly, however, some restricting NPI combinations do not improve pandemic characteristics. Conclusions: DCT implementations will face challenges, as pandemic characteristics do not consistently improve when behavior parameters ( adoption, adherence, and compliance) are increased. When considering realistic behavior, more is not always better; NPI combinations can interfere with each other to the detriment of pandemic control. Our approach offers fine-grained insight on the effectiveness of NPI combinations that cannot be obtained in human studies due to confounding effects. Thus, our approach can guide future pandemic control efforts and prioritization for pandemic preparedness.