© 2002 American Thoracic Society
Contact InvestigationsHow Do They Need to Be Designed for the 21st Century?Department of Medicine University of North Texas Health Science Center at Fort Worth Fort Worth, Texas In countries with a low prevalence of tuberculosis, persons with newly reported tuberculosis are interviewed in a process referred to as contact investigation. Contact investigations identify and evaluate persons who share air with the index case, and are standard public health measures designed to identify and treat additional persons with tuberculosis to prevent further transmission (1). These investigations are an important component of the overall strategy for tuberculosis elimination, but for many reasons, they are not done well even in successful tuberculosis elimination programs (2). Although there are many reasons for this, commonly there are more investigations that need to be done than investigators. In this issue of AJRCCM (pp. 1122–1127), Gerald and coworkers (3) develop an algorithm to focus efforts on persons most likely to have tuberculosis. Contact investigations are currently recommended for persons suspected of having pulmonary tuberculosis (1). These investigations are focused on the highest priority contacts, defined as individuals most likely to be infected or most likely to develop disease (1). This study evaluates a new approach to defining the "highest priority contacts." Current recommendations define the initial evaluation of highest priority contacts as including a medical history and a tuberculosis skin test (1). On the basis of the methodology described by Gerald and colleagues, the initial evaluation would be a categorization of risk using an algorithm (3). Evaluation would then be limited to persons characterized as being at high risk for having tuberculosis. Contact investigations are multistep processes in which contacts are systematically evaluated on the basis of the amount of time spent with the infectious person, the environment in which the exposure took place, and the susceptibility of those in contact (1). Limiting evaluations to persons who are likely to be identified with tuberculosis, using algorithms like those proposed by Gerald and coworkers, can reduce investigator work at the potential cost of missing some persons with tuberculosis (3). It is conceivable, however, that better-focused investigative efforts would identify more persons with tuberculosis. Improving the methods of contact investigation, as described by Gerald and coworkers (3), will allow programs to focus contact investigations, although this should be regarded as an interim step. Further reductions in tuberculosis will come not from reducing the scope of contact investigations, but by increasing it; in other words, what is needed to more effectively prevent tuberculosis is not a way to make investigations smaller but, rather, practical ways to expand investigations so that persons who are infected and missed by current contact investigations can be identified. The current approach for contact investigation, developed at a time when most contacts occurred within the home and family, is insufficient to identify transmission pathways in multicultural urban areas of the United States. That the traditional contact investigation currently fails to completely recognize and interrupt tuberculosis transmission has been demonstrated in multiple settings. Population-based molecular studies in Europe and the United States are consistent with between 35 and 60% of tuberculosis cases resulting from recent infection (4–11). Several groups have compared results of genotyping investigations with results of conventional methods of contact investigation. Less than 30% of cases in these studies were linked both molecularly and epidemiologically (4–7). The concentric circle model used in conventional contact investigations has also been evaluated by molecular methods (11). When the relationships of persons with culture-positive tuberculosis were evaluated, 31% of contacts were identified in the first concentric circle, 29% were identified in the second, and 40% in the third (11). Studies analyzing molecularly defined clusters, conducted in Europe and in urban U.S. populations, demonstrated tuberculosis transmission occurring at times and places that conventional means of contact investigation fail to detect (4–6, 9, 12). Taken together, these data show both the importance of recently transmitted tuberculosis to overall morbidity, confirming the need for contact investigations, as well as their lack of success. It is also apparent that a major proportion of tuberculosis transmission, destined to result in tuberculosis, occurs in individuals who are loosely associated with, or not identified by, the source case. The contact investigation needs an updated design. The article by Gerald and coworkers (3) will promote discussions in health departments that have more investigations than investigators. Several cautions are in order. Only confirmed tuberculosis cases were used in this model; however, contact investigations are initiated before cultures are available. Currently it is recommended that contact investigation be performed on smear-positive cases and that interviews be initiated no more than three working days after a suspected case is reported to the health department (1). Waiting to perform contact investigations until there is culture confirmation limits the workable field utility of this algorithm. Including in the analysis the results of contact investigations of persons suspected of having tuberculosis may change the reported values of sensitivity and specificity. The "total contact hours" used in this algorithm are not easy to determine from a practical standpoint. The effectiveness of contact investigations and of this algorithm depends on the index patient providing an accurate social history. We all know examples of mundane, illicit, or illegal social connections minimized by persons with tuberculosis. Investigations routinely fail to reveal these social connections until additional tuberculosis transmission occurs (4–6). In addition, as the authors acknowledge, their population includes a low proportion of homeless, foreign-born, and human immunodeficiency virus–coinfected persons, and the findings may not be directly generalized to other areas. The contact investigation is an important public health measure that identifies many persons with tuberculosis. The basic investigational design, however, has remained unchanged despite an evolving understanding of tuberculosis transmission. This evolving understanding of transmission creates additional opportunities to interrupt the spread of tuberculosis that are currently not exploited. In the short term, algorithms such as that described by Gerald and colleagues (3) can potentially improve the process for contact investigation by allowing investigators to focus more on persons who are likely to be infected and less on those unlikely to be infected. In the long term, contact investigations need to be redesigned by incorporating new methods for identifying persons with tuberculosis consistent with the current understanding of tuberculosis transmission. To be able to fully exploit such a model, we will need sensitive, specific tests that can be rapidly performed on more identified contacts than are currently being evaluated. Decision trees may be important in these expanded investigations. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
