Social network analysis methods to characterize tuberculosis transmission patterns

Guidelines for TB contact investigation are based upon expert opinion and follow the standard concentric circle approach.[1] Extensive amounts of information are collected as a result of these activities. Recent studies have documented limitations in contact investigation including problems with contact identification, screening completion, and therapy initiation.[2] The concerns with standard contact investigation are amplified in certain high-risk groups, such as substance users and the homeless.[3] 

Controlling TB transmission requires an understanding of the psychosocial and cultural differences within communities, yet standard contact investigation collects data on individual cases without assessing these cases within their social construct. This approach provides a static view of the existing disease pattern. In contrast, social network analysis is dynamic and predictive. Visualization and mathematical models are used to understand how people are connected within varying social frameworks by measuring the size, shape, and structure of the social network over time as well as group cohesion and individual prominence.[4] This process allows for the study of group interaction from data collected on individuals, the relationships between social entities, and the implications of those relationships.

The relevance of social network structure to transmission dynamics in sexually transmitted diseases has been well illustrated in that transmission includes but is not limited to sexual activity.[5] Locally, social networking methods have been implemented by the Division of STD/AIDS, BCCDC to control a syphilis epidemic. In the field of tuberculosis, network methods have been used predominantly to examine outbreaks.[6]

In February 2004, data collection began for the TB Epidemiology Studies Consortium (TBESC) Task Order #7: “The use of network analysis methods to characterize M. tuberculosis transmission patterns among women and other high-risk populations.” This year-long project involves a collaborative group including the Division of TB Control, BCCDC and the US CDC in Atlanta. The purpose of this project is to examine the utility and practicality of social network methods as a tool for analyzing contact investigation data and to improve the overall success of TB control efforts. In addition to routine contact investigation activities, each TB case and their named contacts will be interviewed to elicit places of social aggregation. Network analysis software will then be used to link data on cases, contacts, and multiply named locations. It is anticipated that social network methods will uncover previously unrecognized connections among cases, contacts, and places, as well as identify key persons and places involved in TB transmission.

The Downtown East Side (DTES) of Vancouver was identified as one of the locations for this study. The DTES has one of the highest rates of TB in British Columbia and struggles with the complex issues of inner-city life including drug addiction, sexually transmitted infections, and unstable housing.[7] Currently, only a small proportion of active cases in the DTES are identified by routine contact investigation, which highlights the need for alternative approaches. It seems possible that this community could benefit from improved contact investigation through social network methods. To date, 10 repeatedly named locations have been identified on the DTES and prioritized for TB screening.

It is anticipated that a network-informed approach to contact investigation may lead to more efficient and effective TB control activities. This novel approach may also expand our knowledge of TB transmission in high-risk communities and provide the infrastructure to implement and assess its influence in other settings within BC.

—Victoria J. Cook, MD, FRCPC
Director, TB Services to Aboriginals
TB Control, BCCDC

References

1. Long R. Canadian Tuberculosis Standards. 5th edition. Canadian Thoracic Society, Standards Committee (Tuberculosis). Ottawa, ON: Canadian Lung Association, 2000:175-186.
2. Reichler MR, Reves R, Bur S, et al. Evaluation of investigations conducted to detect and prevent transmission of tuberculosis. JAMA 2002;287:991-995. PubMed Abstract Full Text
3. Weis SE, Pogoda JM, Yang Z, et al. Transmission dynamics of tuberculosis in Tarrant County, Texas. Am J Respir Crit Care Med 2002;166:36-42. PubMed Abstract Full Text
4. Wasserman S, Faust K. Social Network Analysis: Methods and Implications. Cambridge, UK: Cambridge University Press, 1994:4-20.
5. Rothenberg RB, Long DM, Sterk CE, et al. The Atlanta Urban Networks Study: A blueprint for endemic transmission. AIDS 2000;14:2191-2200. PubMed Abstract Full Text
6. Klovdahl AS, Graviss EA, Yaganehdoost A, et al. Networks and tuberculosis: An undetected community outbreak involving public places. Soc Sci Med 2001;52:681-694. PubMed Abstract Full Text
7. BCCDC Tuberculosis Control. Tuberculosis Control Annual Report 2001. Vancouver: BC Centre for Disease Control, 2001. www.bccdc.org/downloads/pdf/tb/reports/TB%20annual%20vs%204.pdf (accessed 1 June 2004).