Evaluating the Effectiveness of Contact Tracing on Tuberculosis Outcomes in Saskatchewan Using Individual-Based Modeling
Authored by Yuan Tian, Nathaniel D. Osgood, Assaad Al-Azem, Vernon H. Hoeppner
Date Published: 2013-10
DOI: 10.1177/1090198113493910
Sponsors:
Canadian Institutes for Health Research
Platforms:
Java
Model Documentation:
Other Narrative
Flow charts
Model Code URLs:
Model code not found
Abstract
Tuberculosis (TB) is a potentially fatal disease spread by an airborne pathogen infecting approximately one third of the globe. For decades, contact tracing (CT) has served a key role in the control of TB and many other notifiable communicable diseases. Unfortunately, CT is a labor-intensive and time-consuming process and is often conducted by a small and overworked nursing staff. To help improve the effectiveness of CT, we introduce a detailed, individual-based model of CT for the Canadian province of Saskatchewan. The model captures the detailed operation of TB CT, including loss to follow-up, and prophylactic and case treatment. This representation is used to assess the impact on active TB cases and TB infection prevalence of differential scoping, speed, prioritization of the CT process, and reduced loss to follow-up. Scenario results are broadly consistent withbut provide many additional insights beyondour previously reported findings using an aggregate model. In the context of a stylized northern community, findings suggest that age- and ethnicity-prioritized schemes could improve CT effectiveness compared to unprioritized schemes by dramatically reducing TB infection and preventing on average roughly 11% (p < .0001) of active TB cases over a period of 20 years. Reducing loss to follow-up to 10% could yield 5.4% (p = .02) TB cases prevented on average with lower prevalence of TB infection, but improving the CT speed does not yield significant improvement in TB outcomes. Finally, although the work emphasized the value of social network analysis, we found that caution should be exercised in directly translating social network analysis-observed associations into prioritization recommendations.
Tags
Agent-based modeling
scale-free network
Infection
networks
Individual-based modeling
Contact tracing
infection control
tuberculosis
Impact
Risks