Skip Navigation

Transmission Network Analysis in Tuberculosis Contact Investigations

  1. Dr. Victoria J. Cook1,2,
  2. Sumi J. Sun3,
  3. Jane Tapia4,
  4. Stephen Q. Muth7,
  5. D. Fermín Argüello5,a,
  6. Bryan L. Lewis3,a,
  7. Richard B. Rothenberg4 and
  8. Peter D. McElroy6,a
  1. 1 Department of Medicine, University of British Columbia, Vancouver, Canada
  2. 2 Division of TB Control, British Columbia Centre for Disease Control, Vancouver, Canada
  3. 3 Tuberculosis Control Branch, California Department of Health Services, Richmond, California
  4. 4 Institute of Public Health, Georgia State University, Atlanta, Georgia
  5. 5 Rollins School of Public Health, Emory University, Atlanta, Georgia
  6. 6 Division of Tuberculosis Elimination, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia
  7. 7 Quintus-ential Solutions, Colorado Springs, Colorado
  1. Reprints or correspondence: Dr. Victoria J. Cook TB Control British Columbia Centre for Disease Control 655 W. 12th Ave. Vancouver Canada V5Z 4R4 (mailvictoria.cook{at}bccdc.ca).

  • ↵Present affiliations: Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia (P.D.M.); Division of Parasitic Disease, Centers for Disease Control and Prevention, Atlanta, Georgia (D.F.A.); Virginia Bioinformatics Institute, Virginia Polytechnic Institute and State University, Blacksburg, Virginia (B.L.L.).

 
Next Section

Abstract

Background. Social network analysis (SNA) is an innovative approach to the collection and analysis of infectious disease transmission data. We studied whether this approach can detect patterns of Mycobacterium tuberculosis transmission and play a helpful role in the complex process of prioritizing tuberculosis (TB) contact investigations.

Methods. We abstracted routine demographic and clinical variables from patient medical records and contact interview forms. We also administered a structured questionnaire about places of social aggregation to TB patients and their contacts. All case-contact, contact-contact, case-place, and contact-place dyads (pairs and links) were considered in order to analyze the structure of a social network of TB transmission. Molecular genotyping was used to confirm SNA-detected clusters of TB.

Results. TB patients not linked through conventional contact-investigation data were connected through mutual contacts or places of social aggregation, using SNA methods. In some instances, SNA detected connected groups prior to the availability of genotyping results. A positive correlation between positive results of contacts' tuberculin skin test (TST) and location in denser portions of the person-place network was observed (P < .01).

Conclusions. Correlation between TST-positive status and dense subgroup occurrence supports the value of collecting place data to help prioritize TB contact investigations. TB controllers should consider developing social network analysis capacity to facilitate the systematic collection, analysis, and interpretation of contact-investigation data.

As the number of tuberculosis (TB) cases decreases in the United States and Canada, it becomes more important and challenging to detect persons recently exposed to Mycobacterium tuberculosis and to treat those with latent M. tuberculosis infection [1–3]. Contact investigation is the principal method used to detect additional TB patients and recently exposed persons with latent M. tuberculosis infection at risk for progression to TB [4–7]. TB-detection programs have varying rates of success in eliciting, locating, and evaluating contacts of TB patients and initiating treatment for latent M. tuberculosis infection [9–12]. These programs often have limited success when they involve high-risk or vulnerable groups [13,14]. Although molecular genotyping of M. tuberculosis can help identify case clusters and inform some contact investigations, additional novel approaches are needed.

The term “social network” is used to describe a set of persons (nodes) and the connections (ties) among them. Social network analysis (SNA) measures the nature of these ties (e.g., sexual activity, needle sharing, cohabitation, work- or school-associated activity, kinship, and leisure activity) and the effect on activities within a network [15] and can help explore the relevance of social structure to transmission of human disease. The basicunit of analysis in SNA is the tie linking 2 nodes (e.g., person-person dyads and person-place dyads). The relevance of social network structure to the transmission of infectious diseases has been an area of intense investigation [16–20].

During standard TB contact investigations, considerable data are collected on individual patients and their contacts. Few TB-control programs assemble these dyad data to systematically analyze case-contact or case-place characteristics and dynamics. In contrast, a network-informed perspective uses quantitative analyses and visual diagrams of these dyads to explore the emerging patterns and structure of interdependent case-contact-place nodes linked in a transmission network [21–23].

SNA has been used retrospectively to characterize M. tuberculosis outbreaks and highlight the importance of places of social aggregation in sustaining transmission [24–28]. However, the use of SNA to complement routine contact investigation and the subsequent impact of such a program on TB control have not been tested [5]. In this study, we implemented a strategy involving SNA methods to collect and interpret contact tracing data in order to determine whether important transmission patterns not otherwise detected by routine contact investigation would emerge. Traditional case-contact dyad data obtained by routine contact investigation was supplemented with nontraditional case-place and contact-place dyad data. We assessed whether more-densely connected contacts of persons infected with M. tuberculosis were more likely to have positive tuberculin skin test (TST) results and compared the occurrence of disease clusters detected through network connections with clusters detected by molecular genotyping.

Previous SectionNext Section

Methods

Collaborators. A team from the US Centers for Disease Control and Prevention (CDC) and the Tuberculosis Epidemiologic Studies Consortium [29] supplemented routine TB contact-investigation procedures with an interview to record places of social aggregation [30]. This prospective observational study was performed in the following 3 demographicallyand epidemiologically distinct locations: Contra Costa County, California (2000 population, 948,816; area, 1159 km2); DeKalb County, Georgia (2000 population, 665,865; area, 431 km2); and the Downtown Eastside area of Vancouver, Canada (2000 population, 50,873; area, 1.6 km2).

Definitions. A confirmed TB patient was one that met the CDC surveillance criteria for TB [31]. A contact was any person named by a patient with pulmonary TB during the course of the routine contact investigation [32]. High-risk contacts were persons with characteristics known to increase the risk of latent M. tuberculosis infection (e.g., being a contact of a sputum smear-positive TB patient) or of progression from latent M. tuberculosis infection to TB (e.g., having HIV infection) [6]. A contact with latent M. tuberculosis infection was defined as a person with a positive TST result, either at the time of the current contact investigation (induration, ⩾5 mm) or through documentation of a previous positive TST result (induration, ⩾10mm) [6].Recent infection was confirmed if the TST caused an induration ⩾5 mm in diameter 2–3 months after the result of an initial TST was negative [5]. All other contact-investigation procedures were defined according to current practices within each TB control jurisdiction. A multiply named contact included any person named by at least 2 TB patients, by a TB patient and a contact, or by 2 contacts. Places of social aggregation refer to locations where a TB patient and their contacts regularly shared air space during the 6-month period preceding TB diagnosis or evaluation as a contact. A multiply named place included any place of social aggregation named by at least 2 persons.

Study design. During a 6-month study period (1 February through 31 August 2004), we abstracted routinely collected demographic and clinical information from TB investigation data forms in each jurisdiction for persons with confirmed TB and their contacts. A standardized, pretested, staff-administered, in-person, open-ended interview instrument [30] was used to collect information on places of social aggregation and was used for all TB patients <70 years of age and for contacts who, on the basis of jurisdiction-specific criteria, were deemed to be at high risk for latent M. tuberculosis infection. Multiply named contacts were also interviewed to elicit the names of their close contacts (i.e., secondary contacts); this procedure is not usually performed as part of standard TB-control practice. Local TB-control staff directly involved in routine contact investigation interviewed TB patients at the time they were initially reported to the local TB-control programs; contacts were interviewed after they were identified by TB patients.

Variables. Study variables abstracted for TB patients included sociodemographic characteristics (name, address, age, sex, race or ethnicity, and country of origin), clinical features (sputum smear and culture results, anatomic site of disease, chest radiograph findings, and HIV infection status), and genotyping results for M. tuberculosis isolates obtained from patients with culture-confirmed TB. Variables abstracted for contacts included similar sociodemographic features, as well TST results and history of treatment for latent M. tuberculosis infection. Traditionally collected data for case-contact pairs included contact environment, categorized as home, work, or leisure; and strength of relationship, categorized as close (i.e., prolonged, frequent, or intense contact with a person with infectious TB) or casual (anything other than a close relationship) [32]. The interview to determine places of social aggregation permitted identification of specific physical structures, categorized and defined on the data collection forms as daytime (6 AM–5 PM), evening (5AM–10 PM), or night (10 PM–6 AM) for the 6-month period preceding TB diagnosis or evaluation as a contact.

Genotyping. Molecular genotyping was completed for all available M. tuberculosis isolates recovered from patients. Iso lates from the 2 US sites were processed through the genotyping services provided by the CDC, including spoligotyping and multiple interspersed repetitive units variable number tandem repeats (MIRU-VNTR) analyses [33–36]. Isolates with a matching spoligotype and MIRU-VNTR pattern were then typed using standard IS6110-based restriction fragment length polymorphism (RFLP) analysis to increase the discriminatory power to distinguish between isolates [36–37]. M. tuberculosis isolates from Vancouver were processed and analyzed using standard RFLP methodology at the British Columbia reference laboratory. At the 2 US sites, a TB cluster was defined as ⩾2 patients from whom M. tuberculosis isolates with matching spoligotype, MIRU-VNTR, and RFLP patterns were recovered. A TB cluster at the Vancouver site was defined as ⩾2 patients from whom M. tuberculosis isolates with the same RFLP pattern were recovered (isolates with a pattern that deviated by ⩾1 band were not included).

Statistical and network analyses. Statistical analyses were performed using Epi Info, version 3.3.2 (CDC), and SAS, version 9.0 (SAS Institute). Social network metrics and diagrams were generated using the software programs UCINET, version 6.101 (Analytic Technologies) [38], and Pajek, version 1.09 [39]. TB transmission depends on close contact between individuals; the more closely connected infectious and uninfected persons are to each other (e.g., by sharing air spaces), the higher the probability of M. tuberculosis transmission. K-plex modeling, a network metric for characterizing subgroup formation on the basis of a group's degree of cohesion relative to individuals outside the group, was used to rank the level of cohesiveness of contact between TB patients and their contacts [40]. Rank scores were created to measure the extent of cohesion through Seidman 2-plexes [40], groups of nodes with a higher extent of interconnection. The UCINet output was then processed with SAS. Association between this group-cohesiveness rank score with the TST positivity of contacts was examined using a 1-sided Wilcoxon (i.e.,Ridit) test [41,42]. To determine the degree to which genotyping results corresponded with how cases were linked in groups, we considered all combinations of cases, taken 2 at a time, and evaluated whether both were in the same network-connected group and had the same genotype; any differences in proportions were examined with the McNemar test [43].

Study approval. This study protocol was approved by the local ethics review board at each study site and by the CDC. Written informed consent was obtained from multiply named contacts prior to eliciting names of their contacts.

Previous SectionNext Section

Results

TB patients. We enrolled 87 TB patients at the 3 study sites; 42 patients were in Contra Costa, 36 were in DeKalb, and 9 were in Vancouver (table 1). No significant differences in age or sex were detected across the 3 sites. The race and ethnicity distribution varied widely, with Asian, black, and Aboriginal persons making up the largest proportion of TB patients in Contra Costa, De-Kalb, and Vancouver, respectively. Foreign-born persons made up the majority of TB patients in both Contra Costa and DeKalb. The majority (70 patients [80%]) received a diagnosis of pulmonary TB. The rate of M. tuberculosis culture positivity ranged from 67% (in Contra Costa) to 100% (in Vancouver), and the percentage of TB patients with sputum smears positive for acid-fast bacilli (AFB) ranged from 36% (in DeKalb) to 89% (in Vancouver). Two-thirds of the TB patients in Vancouver were coinfected with HIV. Of 64 patients with culture-confirmed TB, all but 1 had an isolate genotyped. Genotyping confirmed 5 TB clusters at the 3 sites; 1 was detected in Contra Costa (a 2-case cluster [2 copies of IS6110]), 2 in DeKalb (two 2-case clusters [6 and 12 copies of IS6110]), and 2 in Vancouver (one 4-case cluster [8 copies of IS6110] and one 2-case cluster [11 copies of IS6110]).

Figure 1.
View larger version:
Figure 1.

Networks of relationships between patients with tuberculosis (diamonds), their contacts (circles), and places of social aggregation (houses) in Contra Costa County, California (A), DeKalb County, Georgia (B), and Vancouver, Canada (C), February through August 2004.

Figure 2.
View larger version:
Figure 2.

Network of relationships between patients with tuberculosis (TB), their contacts, and places of social aggregation, Vancouver, Canada, February through August 2004. A, Networks without places; B, networks with places named at least 3 times (the k-plex analysis upon which this diagram is based included all places of social aggregation). Case, contact, and place nodes linked together through social network analysis are referred to as "connected groups." TB patients are denoted by diamonds that include a plus sign if HIV positive and a minus sign if HIV negative. Contacts are denoted by circles that are black if TST positive, grey if TST negative, and white if the TST status is unknown. Places are denoted by houses, which include a plus sign if frequented by HIV-infected TB patients. Node sizes are directly proportional to the extent of their involvement in more-densely connected regions of the network. More-densely connected nodes are visualized closer to the center and are larger, whereas less densely connected nodes are smaller and located closer to the periphery. Connections between TB patients are denoted by solid lines, connections between multiply named contacts and their secondary contacts are denoted by dotted lines, and connections between TB patients who named each other as contacts are denoted by double arrows.

Table 1.
View larger version:
Table 1.

Demographic and clinical characteristics of patients with tuberculosis (TB), by study site, February through August 2004.

Contacts. Contact investigations for the TB patients with pulmonary disease generated the names of 440 distinct persons during the study period; 179 contacts were in Contra Costa, 205 were in DeKalb, and 56 were in Vancouver (table 2). The median number of recorded contacts per TB patient was 5, with no significant differences across the sites. The mean age of the contacts was 29 years, and 51% were male. Although many data were missing, the race and ethnicity distribution of the contacts mirrored those of the patients at each respective site. Only 1 site (Vancouver) accrued multiply named contacts (n= 6). Additionally, 2 multiply named Vancouver contacts shared a mutually identified contact.

Table 2.
View larger version:
Table 2.

Demographic characteristics and latent Mycobacterium tuberculosis infection status for contacts of patients with tuberculosis, by study site, February through August 2004.

Places of social aggregation. We performed 211 interviews for TB patients and contacts, which elicited 1056 places of social aggregation (table 3). Vancouver participants identified a greater median number of places per interview (n = 8) and the highest number of multiply named places (n = 31). The mean duration of the interviews to determine places of social aggregation are also presented in table 3. Demographic characteristics of TB patients interviewed about places of social aggregation were not significantly different from those for uninterviewed TB patients, both at individual sites and overall (data not shown). The same was true for contacts, with the following exceptions: interviewed contacts from DeKalb were less likely than those from the other 2 study sites to be foreign born (48% vs. 87%; P<.001), and all interviewed contacts were more likely than uninterviewed contacts to be TST positive (45% vs. 23%; P<.001).

Table 3.
View larger version:
Table 3.

Characteristics of places of social aggregation named by patients with tuberculosis (TB) and their contacts in a supplemental interview during routine contact investigation, by study site, February through August 2004.

Case-contact, contact-contact, case-place, and contact-place dyads. A total of 453 case-contact dyads were recorded at the 3 sites (table 4). The contact environment (i.e., household) and relationship strength (i.e., close) for these dyads were similar for Contra Costa and DeKalb, whereas in Vancouver the dyad ties were more likely to be nonhousehold and casual. In Vancouver, 13 case-contact dyads (19%) involved 5 multiply named contacts linked in various combinations. Three of the 5 multiply named contacts from Vancouver consented to an interview to elicit names of their contacts, resulting in 15 contact-contact dyads. One secondary contact was named by 2 multiply named contacts. Case-place versus contact-place dyads are presented in table 4 and include only dyads comprised of multiply named places (i.e., places named ⩾2 times). A greater number of daytime places were elicited at all 3 sites from both patients and contacts.

Table 4.
View larger version:
Table 4.

Characteristics of dyad combinations, by study site, February through August 2004.

Network visualizations and metrics. Few interconnections were found among TB patients in either Contra Costa (figure 1A) or DeKalb (figure 1B). The Contra Costa network showed that several multiply named places were reported, but common places did not link any 2 patients together. The patients from the single genotype cluster in Contra Costa were not linked by SNA or conventional epidemiologic analysis. Although the DeKalb network showed a series of TB patients and contacts linked through multiply named places (arranged at the top of figure 1B), these connected groups were confirmed by genotyping not to be the same strain. Also, the patients from the 2-case genotype clusters in DeKalb were not linked by SNA or conventional epidemiologic analysis.

The Vancouver network represented a single connected group of 9 TB patients, 53 distinct contacts, and 31 multiply named places of social aggregation (figure 1C). In the absence of the multiply named places (as would be the case in a traditional contact investigation), no association was demonstrated between TST positivity and more-densely connected contacts (figure 2A) (P = .4). Inclusion of multiply-named places of social aggregation reveals a single connected network (figure 2B). Two main groups emerge as connected by 2 multiply named places and 1 contact, located approximately in the center of the diagram; 1 group includes HIV-negative TB patients, the other includes HIV-positive TB patients. A positive association (P < .01) between contacts' TST positivity and their location in denser portions of the network can be seen by noting the frequency and size of the nodes nearer the center of the diagram, which represent TST-positive persons. For all 3 sites, the contact-investigation data and the network visualizations they generated provided evidence of the presence or absence of case clustering before genotyping results were available.

Previous SectionNext Section

Discussion

Achieving TB elimination in the United States and Canada will require health departments to adopt innovative methods that augment the commonly used “concentric circle” approach for TB contact investigations [1, 8]. Yet limited resources have prevented many TB-control programs from benefiting from even the most basic information-management technologies. In this study, we implemented a contact-investigation strategy in 3 TB-control jurisdictions, using readily accessible tools to document, store, analyze, and interpret contact-tracing data. With the exception of place data and the identification of secondary contacts, this strategy used data typically collected by TB-control programs to monitor completeness of contact evaluations or to track the number of TST-positive contacts eligible for treatment for latent M. tuberculosis infection. In this study, local TB-control programs were able to use network analysis and visualization tools to help understand the patterns of M. tuberculosis transmission in their jurisdictions.

The small social-geographic area and the high population density of the Vancouver site permitted more-extensive network analysis and visualization despite the short 6-month period of data collection. For example, we were able to better visualize how HIV-infected persons and their associated social venues were situated with respect to the larger picture ofall detected relationships in this TB network (figure 2B). Through this approach, we were also able to identify the presence or absence of M. tuberculosis transmission clusters before genotyping results were available at the local level. In 3 instances where TB “clusters” were first identified through rapid PCR-based methods but subsequently characterized by RFLP as false matches, SNA instead found no connections among the patients, simultaneously confirming the results of traditional contact tracing methods and clarifying the sometimes erroneous and confusing genotyping data. This observation suggests some benefit in using network analysis to enhance the specificity of PCR-based methods in detecting TB clusters. The converse was also true. In DeKalb, 2 groups of interconnected patients detected by SNA were not confirmed as transmission clusters by genotyping methods. Although these patients had unique isolates and thus were not considered to be part of true disease clusters stemming from recent transmission, we believe the detection of common places of so cial aggregation that link multiple TB patients is important for TB-control efforts, because knowledge of these locations may lead to the detection of a higher than average number of new cases.

Detection of TB clusters with the help of SNA is also relevant because a high proportion of diagnosed cases are not confirmed by culture (22% in the United States during 2004) [44] and thus lack M. tuberculosis isolate for genotyping. In the absence of SNA or another method, no clear approach currently exists to accommodate culture-negative TB cases during investigations of suspected transmission clusters. Additionally, molecular clustering does not always indicate recent transmission, and circulating endemic stains can result in “clusters” of epidemiologically unrelated cases, generating confusion for outbreak investigators [36, 44, 45]. Combining the results of SNA with the results of genotyping should help improve the specificity of molecular genotyping methods and help avoid unnecessary public health actions.

Though electronic management of routine contact-investigation data is recommended [5], most TB-control programs' efforts to store and creatively analyze contact-tracing data need strengthening [9]. This is particularly important for higher-population jurisdictions that commonly collect data from thousands of contacts each year. Many of those contacts are repeatedly identified over time but are not always immediately recognized as important links to ongoing transmission. Contact investigations augmented by SNA offer several advantages. Monitoring of the evolution of interconnected cases, contacts, and places may facilitate earlier detection of ongoing transmission and help prioritize which contacts or locations to investigate. Updated diagrams could be compared with historical diagrams to highlight critical nodes responsible for expanded transmission. These “spot reviews” or summary views of contact investigations could identify specific places for on-site or location-based screening for TB and latent M. tuberculosis infection. The network approach could be a resource to outreach workers planning contact investigations and targeted TST activities or to TB-control program managers making decisions about resource allocation [5, 46].

Few sites in North America systematically collect and electronically store place data, despite current recommendations [5]. Even where collected, such data are not periodically analyzed to spot increased reporting of specific locations named by TB patients and contacts. Our results confirm the importance of named places in TB transmission, because more TST-positive contacts in densely connected networks were identified after places of social aggregation were taken into account. Although similar observations have been reported, and the cost of recording place data is unlikely to be prohibitive for most TB-control programs, few practice this strategy. A simple spreadsheet could be used to cumulatively store place data, with multiply named places considered for location-based screening activities. Several SNA software applications, including free noncommercial and commercial packages, are available. One rate-limiting step for implementing this strategy is the training of key staff to use the software and interpret the output. The training strategy successfully used to implement TB genotyping at the national level could serve as a model for educating TB controllers about other new technologies, including SNA [47].

Despite training, some stressed TB-control programs maybe at their limit for expansion of investigation activities. Nonetheless, an expansion of contact lists and ascertainment of place names may be of value even in a setting where the data are not computerized or analyzed. Investigators' memory is a powerful tool, and simple enumeration of contacts and places, without further analysis, may provide clues to investigators of possible clustering of people within places. In addition, the collection of such data provides the potential for subsequent analysis that would be otherwise impossible.

Several factors impacted the effectiveness of SNA during this pilot study. Given the large geographic size of Contra Costa and DeKalb counties, the 6-month data collection period was likely too short for elucidation of transmission dynamics in these locales. The short study duration was chosen to allow for assessment of the burden of implementing the SNA paradigm. A study period more commensurate with the speed of M. tuberculosis propagation would permit the accrual of more multiply named places of importance and could make a greater contribution in larger, less dense areas such as Contra Costa and DeKalb counties. The varying yield of multiply named contacts and places may also be explained by the characteristics of the populations at the 3 sites and the reduced likelihood of recent transmission. For example, if the majority of TB cases in a community are due to reactivation of remotely acquired infection (e.g., as is often the case for older, foreign-born persons with TB), the usefulness of SNA as an adjunct to genotyping for the assessment of recent transmission may not be as convincing. The discrepancy in interview times across the 3 sites may also have influenced the number of multiply named contacts and multiply named places. Finally, complete data were missing for many contacts. Our results, although incomplete, are consistent with findings on the incompleteness of data in many routine contact investigations [9].

Overall, we found that contact investigations were improved when data on places of social aggregation were collected and SNA was implemented. SNA provides the analytic framework for placing TB patients, their contacts, and the places they frequent into a social-geographic context. Qualitative SNA measures can be used as an adjunct to investigation of more well-established individual-level TB risk factors (e.g., recent TST conversion, presence of HIV infection, and high infectivity of the source case) and may offer TB control programs a systematic approach to prioritizing persons, places, neighborhoods, or entire communities. If the voluminous data from routine contact tracing can be successfully harnessed, analyzed, and interpreted using SNA, the pace of TB elimination in North America may be accelerated.

Previous SectionNext Section

Network Analysis Project Team

Henry M. Blumberg, Institute of Public Health, Georgia State University, and Epidemiology Department, Grady Memorial Hospital, Atlanta, Georgia; J. Mark FitzGerald, Department of Medicine, University of British Columbia, and Centre for Clinical Epidemiology and Evaluation, Vancouver General Hospital, Vancouver, Canada; Jennifer M. Flood, Tuberculosis Control Branch, California Department of Health Services, Richmond, California; Sheila P. McCarthy, Department of Medicine, University of British Columbia, Vancouver; Alawode Oladele, Tuberculosis, Refugee Health, and Richardson Laboratory, DeKalb County Board of Health, Georgia; Ameisha R. Sampson and Maureen Wilce, Division of Tuberculosis Elimination, National Center for Human Immunodeficiency Virus, Viral Hepatitis, Sexually Transmitted Disease, and Tuberculosis Prevention, Centers for Disease Control and Prevention, Atlanta; and Francie Wise, Tuberculosis Control Program, Contra Costa County Department of Health, California.

Previous SectionNext Section

Acknowledgments

We thank the staff members from the tuberculosis-control programs in DeKalb County, Georgia; Contra Costa County, California; and the Downtown Eastside Tuberculosis Program in Vancouver, Canada. Specifically, we acknowledge Beverly DeVoe-Payton, Rose Sales, and the Georgia DHR/Division of Public Health/TB Control Program, as well as the staff at the De-Kalb County Board of Health; Mark Condit at the Contra Costa County Tuberculosis Program; and Dr. Kevin Elwood, Nash Dhalla, Jerry Cyr, and Shelley Dean at the Division of TB Control, British Columbia Centre for Disease Control, Vancouver. We are also grateful to the Division of Tuberculosis Elimination at the US Centers for Disease Control and Prevention (CDC) and the Tuberculosis Epidemiologic Studies Consortium (TBESC) for funding this project. We thank Dr. Rachel Albalak (CDC), Dr. Earl Hershfield (University of Manitoba), and Dr. Tom Navin (CDC) for their leadership in guiding the TBESC during this study. Dr. Patrick Moonan (CDC) provided specific genotyping data at the state and national level. Genotyping for DeKalb County and Contra Costa County isolates was performed by one of two contracting reference laboratories: the Division of Infectious Disease, Bureau of Laboratories, Department of Community Health (Lansing, MI), or the California Department of Health Services, Microbial Disease Laboratory (Richmond, CA). Vancouver isolates were genotyped by the National Reference Centre for Mycobacteriology, National Microbiology Laboratory (Winnipeg, Canada).

Previous SectionNext Section

Footnotes

  • ↵Study group members are listed after the text.

  • Potential conflicts of interest: none reported.

  • Financial support: Tuberculosis Epidemiologic Studies Consortium, which is supported by the US Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention.

  • Received February 16, 2007.
  • Accepted June 8, 2007.
Previous Section
 

References

  1. 1.↵
    Institute of Medicine. Ending Neglect: The Elimination of Tuberculosis in the United States. Washington, DC: National Academy Press; 2000.
  2. 2.
    Health Canada. Annual Reports Tuberculosis in Canada (1996–2002). (Available at: http://www.phac-aspc.gc.ca/publications_4_e.html+).
  3. 3.↵
    1. Taylor Z,
    2. Nolan CM,
    3. Blumberg HM
    . American Thoracic Society Centers for Disease Control and Prevention Infectious Diseases Society of America: Controlling Tuberculosis in the United States. Am J Respir Crit Care Med 2005;172:1169-227.
    FREE Full Text
  4. 4.↵
    Centers for Disease Control and Prevention. Screening for tuberculosis and tuberculosis infection in high-risk populations: recommendations of the Advisory Committee for Elimination of Tuberculosis. MMWR Recomm Rep 1990;39:1-7.
    Medline
  5. 5.↵
    National Tuberculosis Controllers Association and Centers for Disease Control and Prevention (CDC). Guidelines for the investigation of contacts of persons with infectious tuberculosis: recommendations from the National Tuberculosis Controllers Association and CDC. Morb Mortal Wkly Rep Recomm Rep 2005;54:1-47.
  6. 6.↵
    American Thoracic Society and Centers for Disease Control and Prevention. Targeted tuberculin testing and treatment of latent tuberculosis infection. Am J Respir Crit Care Med 2000;161:S221-47.
    FREE Full Text
  7. 7.↵
    1. Long R
    1. Tannenbaum TN,
    2. Fanning A
    . Contact follow-up and outbreak management in tuberculosis control. In: Long R, editor. Canadian tuberculosis standards. 5th ed. Ottawa: Canadian Lung Association, Canadian Thoracic Society and Tuberculosis Prevention and Control Centre for Infectious Disease Prevention and Control, Health Canada; 2000. p. 175-86.
  8. 8.↵
    1. Reichman LB,
    2. Hershfield ES
    1. Etkind S
    . Contact tracing. In: Reichman LB, Hershfield ES, editors. TB: a comprehensive international approach. New York: Marcel Dekkar; 1993. p. 275-89.
  9. 9.↵
    1. Reichler MR,
    2. Reves R,
    3. Bur S,
    4. et al
    . Evaluation of investigations conducted to detect and prevent transmission of tuberculosis. JAMA 2002;287:991-5.
    Abstract/FREE Full Text
  10. 10.
    1. Marks SM,
    2. Taylor Z,
    3. Qualls NL,
    4. Shrestha-Kuwahara RJ,
    5. Wilce MA,
    6. Nguyen CH
    . Outcomes of contact investigations of infectious tuberculosis patients. Am J Respir Crit Care Med 2000;162:2033-8.
    Abstract/FREE Full Text
  11. 11.
    1. Hopewell PC
    . Targeting tuberculosis prevention. Am J Respir Crit Care Med 2000;162:2017-8.
    FREE Full Text
  12. 12.↵
    1. Mohle-Boetani JC,
    2. Flood J
    . Contact investigations and the continued commitment to control tuberculosis. JAMA 2002;287:1040-2.
    FREE Full Text
  13. 13.↵
    1. Weis SE,
    2. Pogoda JM,
    3. Yang Z,
    4. et al
    . Transmission dynamics of tuberculosis in Tarrant County, Texas. Am J Respir Crit Care Med 2002;166:36-42.
    Abstract/FREE Full Text
  14. 14.↵
    1. Hernandez-Garduno E,
    2. Kunimoto D,
    3. Wang L,
    4. et al
    . Predictors of clustering of tuberculosis in Greater Vancouver: a molecular epidemiology study. CMAJ 2002;167:349-52.
    Abstract/FREE Full Text
  15. 15.↵
    1. Wasserman S,
    2. Faust K
    . Social network analysis: methods and applications. Cambridge: Cambridge University Press; 1994.
  16. 16.↵
    1. Rothenberg RB,
    2. Narramore J
    . Commentary: the relevance of social network concepts to sexually transmitted disease control. Sex Transm Dis 1996;23:24-9.
    MedlineWeb of Science
  17. 17.
    1. Rothenberg RB,
    2. Long DM,
    3. Sterk CE,
    4. et al
    . The Atlanta Urban Networks Study: ablueprint for endemic transmission. AIDS 2000;14:2191-200.
    CrossRefMedlineWeb of Science
  18. 18.
    1. Galavotti C,
    2. Saltzman LE,
    3. Sauter SL,
    4. Sumartojo E
    . Behavioral science activities at the Centers for Disease Control and Prevention: a selected overview of exemplary programs. Am Psychol 1997;52:154-66.
    CrossRefMedline
  19. 19.
    1. Potterat JJ,
    2. Rothenberg RB,
    3. Muth SQ
    . Network structural dynamics and infectious disease propagation. Int J STD AIDS 1999;10:182-5.
    Abstract/FREE Full Text
  20. 20.↵
    1. Wylie JL,
    2. Cabral T,
    3. Jolly AM
    . Identification of networks of sexually transmitted infection: a molecular, geographic, and social network analysis. J Infect Dis 2005;191:899-906.
    Abstract/FREE Full Text
  21. 21.↵
    1. Rothenberg RB,
    2. McElroy PD,
    3. Wilce MA,
    4. Muth SQ
    . Contact tracing: comparing the approaches for sexually transmitted diseases and tuberculosis. Int J Tuberc Lung Dis 2003;7:S342-8.
    MedlineWeb of Science
  22. 22.
    1. Potterat JJ,
    2. Rothenberg RB,
    3. Woodhouse DE,
    4. Muth JB,
    5. Pratts CI,
    6. Fogel JS
    . Gonorrhea as a social disease. Sex Transm Dis 1985;12:25-32.
    MedlineWeb of Science
  23. 23.↵
    1. Figueroa JP,
    2. Brewer CB,
    3. Dale D,
    4. Hileman SB,
    5. Weir S
    . An assessment of sites where persons go to meet sexual partners in St. James, Jamaica, using the PLACE method. Sex Transm Dis 2007;34:410-5.
    MedlineWeb of Science
  24. 24.↵
    1. Klovdahl AS,
    2. Graviss EA,
    3. Yaganehdoost A,
    4. et al
    . Networks and tuberculosis: an undetected community outbreak involving public places. Soc Sci Med 2001;52:681-94.
    CrossRefMedlineWeb of Science
  25. 25.
    1. McElroy PD,
    2. Rothenberg RB,
    3. Varghese R,
    4. et al
    . A network-informed approach to investigating a tuberculosis outbreak: implications for enhancing contact investigations. Int J Tuberc Lung Dis 2003;7:S486-348.
    MedlineWeb of Science
  26. 26.
    1. Fitzpatrick LK,
    2. Hardacker JA,
    3. Heirendt W,
    4. et al
    . A preventable outbreak of tuberculosis investigated through an intricate social network. Clin Infect Dis 2001;33:1801-6.
    Abstract/FREE Full Text
  27. 27.
    1. Sterling TR,
    2. Thompson D,
    3. Stanley RL,
    4. et al
    . A multi-state outbreak of tuberculosis among members of a highly mobile social network: implications for tuberculosis elimination. Int J Tuberc Lung Dis 2000;4:1066-73.
    MedlineWeb of Science
  28. 28.↵
    1. Barnes PF,
    2. Yang Z,
    3. Preston-Martin S,
    4. et al
    . Patterns of tuberculosis transmission in Central Los Angeles. JAMA 1997;278:1159-63.
    Abstract/FREE Full Text
  29. 29.↵
    1. Katz D,
    2. Albalak R,
    3. Wing JS,
    4. Combs V
    ; for the Tuberculosis Epidemiologic Studies Consortium. Setting the agenda: a new model for collaborative tuberculosis epidemiologic research. Tuberculosis 2007;87:1-6.
    CrossRefMedline
  30. 30.↵
    Supplemental interview for places of social aggregation. Centers for Disease Control and Prevention and Tuberculosis Epidemiologic Studies Consortium. [25 January 2007]. (Available at: http://www.cdc.gov/tb/Samples/network-forms.htm).
  31. 31.↵
    American Thoracic Society and Centers for Disease Control and Prevention. Diagnostic standards and classification of tuberculosis in adults and children. Am J Respir Crit Care Med 2000;161:1376-95.
    FREE Full Text
  32. 32.↵
    Centers for Disease Control and Prevention (CDC). Division of Tuberculosis Elimination and Public Health Practice Office, Division of Media and Training Services. Atlanta: CDC; 1999. Contact investigation for tuberculosis: self-study modules on tuberculosis.
  33. 33.↵
    Centers for Disease Control and Prevention. New CDC program for rapid genotyping of Mycobacterium tuberculosis isolates. MMWR Morb Mortal Wkly Rep 2005;54:47-47.
    Medline
  34. 34.
    1. Cowan LS,
    2. Diem L,
    3. Brake MC,
    4. Crawford JT
    . Transfer of a Mycobacterium tuberculosis genotyping method, spoligotyping, from a reverse line-blot hybridization, membrane-based assay to the Luminex multianalyte profiling system. J Clin Microbiol 2004;42:474-7.
    Abstract/FREE Full Text
  35. 35.
    1. Supply P,
    2. Lesjean S,
    3. Savine E,
    4. Kremer K,
    5. van Soolingen D,
    6. Locht C
    . Automated high-throughput genotyping for study of global epidemiology of Mycobacterium tuberculosis based on mycobacterial interspersed repetitive units. J Clin Microbiol 2001;39:3563-71.
    Abstract/FREE Full Text
  36. 36.↵
    1. Cowan LS,
    2. Diem L,
    3. Monson T,
    4. et al
    . Evaluation of a two-step approach for large-scale, prospective genotyping of Mycobacterium tuberculosis isolates in the United States. J Clin Microbiol 2005;43:688-95.
    Abstract/FREE Full Text
  37. 37.↵
    1. van Embden JD,
    2. Cave MD,
    3. Crawford JT,
    4. et al
    . Strain identification of Mycobacterium tuberculosis by DNA fingerprinting: recommendations for a standardized methodology. J Clin Microbiol 1993;31:406-9.
    Abstract/FREE Full Text
  38. 38.↵
    1. Borgatti SP,
    2. Everett MG,
    3. Freeman LC
    . UCINET 6 for Windows. Harvard: Analytic Technologies; 2002.
  39. 39.↵
    1. Batagelj V,
    2. Mrvar A
    . Pajek—program for large network analysis. Connections 1998;21:47-57.
  40. 40.↵
    1. Seidman S,
    2. Foster B
    . A graph theoretic generalization ofthe clique concept. J Math Soc 1978;6:139-54.
    CrossRef
  41. 41.↵
    1. Wilcoxon F
    . Individual comparisons by ranking methods. Biometrics 1945;1:80-3.
    CrossRefWeb of Science
  42. 42.↵
    1. Selvin S
    . Statistical analysis of epidemiologic data. 2nd ed. New York: Oxford University Press; 1996.
  43. 43.↵
    1. Fleiss JL
    . Statistical methods for rates and proportions. 2nd ed. New York: John Wiley & Sons; 1981.
  44. 44.↵
    Centers for Disease Control and Prevention (CDC). Reported tuberculosis in the United States, 2004. Atlanta: CDC; 2005 September.
  45. 45.↵
    1. Cowan LS,
    2. Crawford JT
    . Genotype analysis of Mycobacterium tuberculosis isolates from a sentinel surveillance population. Emerg Infect Dis 2002;8:1294-302.
    MedlineWeb of Science
  46. 46.↵
    1. Andre M,
    2. Ijaz K,
    3. Tillinghast JD,
    4. et al
    . Transmission network analysis to complement routine tuberculosis contact investigations. Am J Public Health 2007;97:470-7.
    Abstract/FREE Full Text
  47. 47.↵
    National Tuberculosis Controllers Association Centers for Disease Control and Prevention (CDC) Advisory Group on Tuberculosis Genotyping. Guide to the application of genotyping to tuberculosis prevention and control. Atlanta: CDC; 2004 [25 January 2007]. (Available at: http://www.cdc.gov/nchstp/tb/genotyping/toc.htm).
| Table of Contents

This Article

  1. J Infect Dis. (2007) 196 (10): 1517-1527. doi: 10.1086/523109
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most