The New England Journal of Medicine
e-mail icon  FREE NEJM E-TOC    HOME   |   SUBSCRIBE   |   CURRENT ISSUE   |   PAST ISSUES   |   COLLECTIONS   |    Advanced Search
Sign in | Get NEJM's E-Mail Table of Contents — Free | Subscribe
 
Original Article
PreviousPrevious
Volume 330:1710-1716 June 16, 1994 Number 24
NextNext

Transmission of Tuberculosis in New York City -- An Analysis by DNA Fingerprinting and Conventional Epidemiologic Methods
David Alland, Gary E. Kalkut, Andrew R. Moss, Ruth A. McAdam, Judith A. Hahn, William Bosworth, Ernest Drucker, and Barry R. Bloom

 

This Article
-Abstract

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation
ABSTRACT

Background The incidence of tuberculosis and drug resistance is increasing in the United States, but it is not clear how much of the increase is due to reactivation of latent infection and how much to recent transmission.

Methods We performed DNA fingerprinting using restriction-fragment-length polymorphism (RFLP) analysis of at least one isolate from every patient with confirmed tuberculosis at a major hospital in the Bronx, New York, from December 1, 1989, through December 31, 1992. Medical records and census-tract data were reviewed for relevant clinical, social, and demographic data.

Results Of 130 patients with tuberculosis, 104 adults (80 percent) had complete medical records and isolates whose DNA fingerprints could be evaluated. Isolates from 65 patients (62.5 percent) had unique RFLP patterns, whereas isolates from 39 patients (37.5 percent) had RFLP patterns that were identical to those of an isolate from at least 1 other study patient; the isolates in the latter group were classified into 12 clusters. Patients whose isolates were included in one of the clusters were inferred to have recently transmitted disease. Independent risk factors for having a clustered isolate included seropositivity for the human immunodeficiency virus (HIV) (odds ratio for Hispanic patients, 4.31; P = 0.02; for non-Hispanic patients, 3.12; P = 0.07), Hispanic ethnicity combined with HIV seronegativity (odds ratio, 5.13; P = 0.05), infection with drug-resistant tuberculosis (odds ratio, 4.52; P = 0.005), and younger age (odds ratio, 1.59; P = 0.02). Residence in sections of the Bronx with a median household income below $20,000 was also associated with having a clustered isolate (odds ratio, 3.22; P = 0.04).

Conclusions In the inner-city community we studied, recently transmitted tuberculosis accounts for approximately 40 percent of the incident cases and almost two thirds of drug-resistant cases. Recent transmission of tuberculosis, and not only reactivation of latent disease, contributes substantially to the increase in tuberculosis. .


After decades of decline, the incidence of tuberculosis in the United States began to increase in 1986, resulting in 52,000 excess cases by 19921. New York City accounted for 14 percent of all cases of tuberculosis in the United States in 1992; the number of cases reported in the city has increased by over 150 percent since 19792. This increase has been especially dramatic among minorities and in specific areas.

The increase in tuberculosis has been attributed to coinfection with the human immunodeficiency virus (HIV),3,4 deterioration of the public health infrastructure,5 social disruption including homelessness and drug abuse,5 and immigration6. A high reactivation rate of latent tuberculous infection in persons coinfected with HIV is thought to be the principal mechanism underlying this phenomenon7. It is estimated that 90 percent of tuberculosis cases nationwide are due to the reactivation of latent, remote infection8. To reduce the risk of reactivation, tuberculosis-control strategies have emphasized preventive therapy in populations at high risk for latent infection8.

Outbreaks of tuberculosis in a residence for patients with the acquired immunodeficiency syndrome (AIDS),9 in shelters,10,11 and in hospitals12 show that transmission and rapid progression to disease can occur in institutional settings. Increases in tuberculosis among children13 indicate that transmission is occurring in the community. There has been a growing recognition that social conditions in poor urban areas where tuberculosis remains prevalent, combined with high rates of HIV infection, may facilitate transmission of tuberculosis14. However, the relative contribution of recent transmission to the overall incidence of disease and the risk factors involved have not been established. In this study, we used molecular and epidemiologic methods to assess these factors in one medical center in New York City.

The presence of repetitive genetic insertional elements in Mycobacterium tuberculosis permits the identification of individual strains by DNA fingerprinting with restriction-fragment-length polymorphism (RFLP) analysis and can be used to demonstrate the transmission of particular M. tuberculosis strains in outbreaks9,15. This technique, combined with medical-record review and analysis of census data, allowed us to investigate the microbiologic, clinical, social, and demographic factors associated with recently transmitted disease.

Methods

Patient Population

The study was performed at a 765-bed hospital that is the largest provider of primary care to the 1.2 million residents of the Bronx, a borough of New York City. The population served is largely poor, Hispanic, and black, but also includes middle-class whites and minorities. The hospital also serves as a referral center for nearby counties. There have been no recognized outbreaks of nosocomially transmitted tuberculosis.

Accrual of Patients and Mycobacterial Samples

All patients who had at least one positive culture for M. tuberculosis from December 1, 1989, through December 31, 1992, were enrolled in the study. Each patient's medical records were reviewed with a standardized form, and one initial M. tuberculosis isolate was selected for RFLP analysis and antimycobacterial-susceptibility testing. Thirteen isolates could not be subcultured: 11 were nonviable, 1 was contaminated, and 1 was lost. Patients were considered to be infected with HIV only if they met the 1987 definition of AIDS devised by the Centers for Disease Control and Prevention (CDC) or were documented to be seropositive, and they were considered to have AIDS only if they met the 1987 CDC case definition. A patient's country of origin was not always recorded; consequently, patients were classified as either foreign-born or not foreign-born (U.S.-born or of unknown origin). The hospital's medical data base was used to provide a complete listing of each patient's previous hospitalizations at the institution from mid-1984 until the end of the study period. The investigators were unaware of the individual RFLP patterns of the study patients when they reviewed the medical records. This study was approved by the institutional review board of the hospital.

Analysis and Mapping of Census Blocks

Adults with tuberculosis who had confirmed Bronx addresses were located in specific census blocks and block groups (the smallest areas of census enumeration) with TIGER files, the comprehensive geographic locator system developed by the U.S. Census and U.S. Geological Survey. Demographic, social, and household characteristics were taken from the STF1 and STF3 reports, the major files of the 1990 U.S. Census. Mapping was done by the ATLAS/GIS mapping program. Fifteen patients were excluded from this analysis because their medical records listed an address outside the Bronx, and one was excluded because the address given could not be confirmed as an existing street address.

Susceptibility Testing

Susceptibility testing was done with the CDC version of the proportion method16. Susceptibility data were not available for three isolates; this was reflected in the denominator used in calculations involving drug resistance. A radiometric culture system (Bactec) was not used at the study hospital.

RFLP Analysis

RFLP analysis was performed according to previously described methods17. In brief, M. tuberculosis DNA was extracted, digested with PvuII, subjected to electrophoresis, and hybridized with Southern blot techniques with a fragment of the insertion element IS6110 measuring 245 base pairs and generated by the polymerase chain reaction. Images were generated by enhanced chemiluminescence (Amersham, Arlington Heights, Ill.). A subgroup of M. tuberculosis strains was subjected to secondary RFLP analysis. The DNA was digested with AluI and probed in a similar manner with a 36-base oligonucleotide homologous to the direct-repeat region of the M. tuberculosis genome18. Two investigators visually examined the RFLP fingerprints. A cluster was defined as a group of two or more isolates from different patients whose RFLP fingerprints were identical with respect to both the number and molecular size of all bands. Isolates that had unique fingerprints were deemed nonclustered. The isolates were divided into two groups: group 1 contained all nonclustered isolates, and group 2 contained all the clustered isolates.

Confirmatory Analysis

We have found that RFLP analysis based solely on hybridization with IS6110 may not distinguish between isolates with only two apparently identical bands (data not shown). In order to confirm strain identity among these isolates, we subjected most such strains to secondary RFLP analysis using a second enzyme and a different probe (see above). On this basis, we were able to sort strains with only two initially identical bands into three smaller clusters. One such strain had a unique secondary fingerprint and was reclassified into group 1. Two other strains could not be recultured for secondary RFLP analysis. One was placed in an established cluster of two-banded strains on the basis of a shared unusual pattern of resistance, and the other strain was excluded from the study. Two strains with only one apparently identical band detected on primary RFLP fingerprinting were also excluded to avoid a bias toward clustering.

Statistical Analysis

Chi-square tests (or Fisher's exact tests, when expected cell sizes were less than five) were performed to test the association of clustering with categorical predictor variables. Wilcoxon nonparametric tests and t-tests were performed to determine whether the distributions of the continuous variables differed between subjects with clustered strains and those with nonclustered strains. Predictor variables that were significantly associated with clustering (P<0.05) were included in a logistic-regression model. Statistical interaction with HIV status was also investigated. Logistic-regression analysis was also performed on the data on subjects who were Bronx residents to determine the independent effects of household income on clustering.

Results

Study Population

Of 130 patients who had culture-proved tuberculosis during the study period, 117 had M. tuberculosis cultures available for RFLP fingerprinting. The medical records were reviewed for 114 of the 117 patients (97 percent) with known RFLP fingerprints. Three patients with negative smears, a single positive culture, and no clinical or radiologic evidence of tuberculosis on chart review were excluded because their culture results were thought to represent possible laboratory contamination or mislabeling. We also excluded three patients whose isolates had two or fewer bands on primary RFLP analysis and for which secondary RFLP analysis was not done. For classification purposes, the remaining 108 patients were assigned to group 1 or group 2 on the basis of their DNA fingerprints. For all other analyses, the four children under 15 years of age were excluded, because many of the social variables evaluated in this study are not relevant to children. The final study population consisted of the 104 adult patients for whom complete data were available. The demographic and clinical characteristics of the study population were compared with those of two series of patients with tuberculosis in New York City who were studied in 199119,20. The study population was similar to those in the two earlier studies in most respects, including sex, age, proportion of foreign-born subjects, and proportion seropositive for HIV, but there was a greater proportion of Hispanic subjects, reflecting the ethnic makeup of the Bronx, and somewhat fewer patients with AIDS and homeless patients.

Characterization of Patients According to Tuberculosis Strain

Of the 104 study patients, 65 (62.5 percent) had strains of M. tuberculosis with unique RFLP fingerprints (group 1) and 39 (37.5 percent) had strains whose RFLP patterns were identical to those of 1 to 11 isolates from other patients, which we classified into 12 clusters (group 2). A total of 77 RFLP patterns were seen in 104 patients.

The clinical and epidemiologic characteristics of the groups are shown in Table 1. The patients in group 2 were significantly more likely than the patients in group 1 to be HIV-infected (67 percent vs. 31 percent, P<0.001) and to have drug-resistant isolates (49 percent vs. 16 percent, P<0.001), including resistance to multiple drugs (24 percent vs. 5 percent, P = 0.003). There was a significant association with Hispanic ethnicity (49 percent vs. 26 percent, P = 0.02); however, birth outside the United States and Puerto Rico was less common (8 percent vs. 28 percent, P = 0.01). The patients in group 2 were also significantly younger than those in group 1 (mean age, 36 vs. 46 years, P = 0.001). Other clinical and radiologic features of the two groups were similar. The patients in group 2 were not more likely than the patients in group 1 to have been hospitalized previously at the study hospital; before their disease developed, only three patients could possibly have been exposed to other patients infected with strains with identical RFLP fingerprints.

View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of the Two Groups of Patients with Tuberculosis.

 
Analysis of the Bronx residents according to census blocks and block groups permitted us to assess the local economic and demographic environment in which each patient lived. Patients in group 2 resided in block groups with lower median household incomes than those in group 1 ($17,676 vs. $22,338, P = 0.02) and lived in more crowded conditions (percentage of households on the block with more than one person per room, 22 percent vs. 17.8 percent; P = 0.04). The difference in median incomes was significant only among non-Hispanic patients ($17,713 vs. $24,480, P = 0.01) and HIV-seronegative patients ($14,977 vs. $23,476, P = 0.003). Hispanic patients as a whole lived in block groups with lower median incomes than non-Hispanic patients (mean income, $17,577 vs. $22,224; P = 0.02), but there was no significant difference in this variable between group 1 and group 2.

Age Differences

Younger age and HIV seropositivity were strongly associated with having a clustered isolate (group 2). However, group 1 also had young foreign-born and HIV-seropositive patients (Figure 1). The higher mean age of the patients in group 1 was due to the large number of HIV-seronegative patients over the age of 50 (21 of 45), as compared with group 2 (2 of 13, P = 0.04). Therefore, younger patients were at risk for both clustered and nonclustered isolates, whereas older patients had predominantly nonclustered isolates.


View larger version (24K):
[in this window]
[in a new window]
 
Figure 1. Age Distribution of Patients in Group 1 and Group 2 According to HIV Status and Place of Birth.

Group 1 comprises at least three populations: an older population of HIV-seronegative patients and two younger populations, one HIV-seropositive and the other HIV-seronegative and foreign-born. Patients in group 2 are more homogeneous, with similar age distributions of HIV-seropositive and HIV-seronegative patients.

 
Multivariate Analysis of Differences between Groups

In the multivariate model (Table 2), resistance to one or more drugs was strongly associated with clustering (adjusted odds ratio, 4.52; P = 0.005), as was younger age (adjusted odds ratio for a difference of 10 years as a continuous variable, 1.59; P = 0.02). Conversely, subjects known to be foreign-born were less likely to be in group 2 (adjusted odds ratio, 0.27; P = 0.08). There was an interaction between HIV status and ethnicity. The odds of clustering were significantly increased among patients who were Hispanic, HIV-seropositive, or both as compared with patients who were HIV-seronegative and not Hispanic. However, HIV seropositivity did not increase the risk of having a clustered isolate among the Hispanic patients, and conversely, being Hispanic did not increase this risk among HIV-seropositive patients.

View this table:
[in this window]
[in a new window]
 
Table 2. Multivariate Analysis of the Association of Study Variables with Clustering of Tuberculosis.

 
Multivariate analysis was performed on data on Bronx residents, including block-group data on median household income. Crowded living conditions (more than 20 percent of households in the census block with more than one person per room) was not included in the final model because it was correlated with median household income (R2 = 0.71, P<0.001). Previously noted associations with clustering remained the same, although their statistical significance was decreased as a result of the smaller number of observations and increased number of variables in the model. A median household income of less than $20,000 per year was associated with clustering (adjusted odds ratio, 3.22; P = 0.04). The geographic concentration of patients in group 2 in predominantly lower-income block groups and the relation of this variable to HIV status and ethnic group are shown in Figure 2.


View larger version (54K):
[in this window]
[in a new window]
 
Figure 2. Map of the Bronx, Showing the Ethnicity, HIV Status, and Geographic Location of Patients in Group 1 and Group 2 in Relation to Lower-Income Block Groups.

Patients in group 2 lived more often in poor block groups (median household income in 1989, below $20,000). This association was independent of ethnicity or HIV status. The lower-income block groups represented identifiable geographic areas where cases of recently transmitted tuberculosis were common.

 
Analysis of Clustered Cases According to Date of Presentation and Pattern of Drug Resistance

Cases of tuberculosis due to clustered strains tended to appear together (Figure 3A). The first positive cultures from the patients in each cluster were obtained over a limited period, a circumstance consistent with an episodic pattern of transmission. Strains in cluster J, the largest cluster, were isolated throughout the study, but they were found predominantly in 1990 and 1992; only one case was seen in 1991. This same strain has been implicated in an outbreak of tuberculosis at a men's shelter in northern Manhattan21.


View larger version (13K):
[in this window]
[in a new window]
 
Figure 3. Temporal Distribution and Patterns of Drug Resistance of Strains of Tuberculosis Isolated from Patients in Group 2.

Panel A shows that most cases caused by clustered strains presented within a limited period. The date of the initial positive M. tuberculosis culture is shown for each patient (39 adults and 2 children, the latter indicated by asterisks). The initial isolates in clusters A through F, H, I, and L were all cultured within 11 months of each other. New isolates with identical RFLP fingerprints were not otherwise seen during the 37-month study period. Isolates from clusters G and K were identified 12 and 14 months apart, respectively. Cases caused by strains from cluster J, the largest cluster, occurred throughout the study but were found predominantly in 1990 and 1992; only one case was seen in 1991.

Panel B shows that the patterns of drug resistance were similar within each cluster. Each entry describes the resistance profile of one M. tuberculosis isolate. Resistance profiles are listed according to cluster and are shown in the order in which they presented (from left to right) during the study period. In four of eight clusters with drug resistance, all isolates with known susceptibility profiles were resistant to the same drugs. In two clusters, one isolate was drug-sensitive whereas the others were resistant to a single drug, and in one cluster all isolates were multidrug-resistant but one was resistant to an additional drug. In only one cluster did all three isolates have different susceptibility profiles. I denotes isoniazid, R rifampin, E ethambutol, S streptomycin, a dash no resistance, and a question mark an unknown pattern of resistance. There were a total of 12 drug-sensitive isolates in cluster J (marked with a dagger).

 
Nearly 50 percent of group 2 isolates were drug-resistant. The isolates within an individual cluster had similar patterns of resistance (Figure 3B); most clusters were identical, and in three of four with different susceptibilities, the susceptibilities differed by only one drug. Five patients with drug-resistant strains had a history of tuberculosis, but the susceptibility of their previous isolate was unknown.

Discussion

We used DNA fingerprinting in conjunction with traditional epidemiologic methods to investigate the pathogenesis of resurgent tuberculosis in a New York City community. Traditional teaching has held that the majority of the cases of tuberculosis in developed countries result from reactivation during adulthood of an infection contracted decades before22,23. It is estimated that reactivation is responsible for up to 90 percent of the incident cases in the United States8. Excess cases have largely been attributed to increases in tuberculosis contracted outside the United States1 and to the high reactivation rate among persons infected with both tuberculosis and HIV3,4. Recently transmitted tuberculosis (usually defined as disease occurring within two years of infection) was generally thought to have a minor role24. However, investigations of institutional outbreaks caused by a single strain have clearly demonstrated that transmission and rapid progression to disease can occur, particularly in persons with AIDS9,12. Some experts suggest that recent transmission may have a substantial role in the current spread of tuberculosis14. However, few data are available on the relative contributions of recent transmission and reactivation to incident cases in the community or on risk factors associated with transmission.

DNA fingerprinting provides a new tool for distinguishing recently transmitted from reactivated tuberculosis. Investigations of numerous outbreaks have demonstrated that epidemiologically linked strains of M. tuberculosis have identical RFLP patterns, whereas unrelated strains have differing patterns9,15. In the Netherlands, where the incidence of tuberculosis is declining, all epidemiologically unrelated M. tuberculosis isolates have unique RFLP fingerprints25.

The substantial diversity of RFLP patterns among members of the study population suggests that the chance occurrence of identical RFLP fingerprints among unrelated cases would be unusual. We therefore infer that cases of tuberculosis caused by strains with identical RFLP fingerprints (group 2) are due to recently transmitted disease and that cases caused by strains with unique RFLP fingerprints (group 1) are primarily due to the reactivation of infection. Several findings support this conclusion. Cases within each cluster occurred over a limited period, as would be expected with new focal outbreaks of disease. Furthermore, almost half the clustered cases were drug-resistant, with similar patterns of resistance within each cluster. Until recently, drug resistance was uncommon in New York City20; therefore, clustering of RFLP-identical, drug-resistant isolates must be due to recently transmitted organisms. Conversely, many patients with nonclustered strains had demographic characteristics consistent with a finding of reactivated tuberculosis. Eighteen of 21 foreign-born patients were in group 1, as were 22 of 25 patients who were 50 years of age or older.

Our analysis probably underestimates the true extent of recent transmission. Some patients in group 1 might have been recently infected by persons outside the study population. Conversely, if each cluster includes one reactivated index case, it could overestimate the incidence of recent transmission. The latter possibility seems unlikely, since RFLP fingerprinting of strains from other area hospitals has revealed isolates with patterns identical to several of the clusters in this study,21,26,27 demonstrating that the chain of transmission extends beyond our study population. Although it is possible that a bias toward clustering exists in any study performed at a single institution, our population was comparable in most respects to that of all patients with tuberculosis in New York City in 1991.

Our study suggests that recently transmitted tuberculosis accounts for almost 40 percent of the incident cases in an inner-city community. The independent risk factors for recently transmitted disease include younger age, Hispanic ethnicity in HIV-seronegative patients, and infection with drug-resistant organisms. Living in a lower-income block group was an additional risk factor for recent transmission in some demographic groups. Forty-three percent of the cases of tuberculosis in the HIV-seropositive patients were in group 1 and can be attributed to the increased reactivation rate with HIV coinfection. We believe that the remaining 57 percent are due to recent transmission. There was no association between clustering and previous admissions, and only three cases of possible nosocomial transmission at the study hospital; hence, nosocomial transmission is an unlikely explanation for our findings. We also found that most foreign-born patients did not have clustered strains, which implies that imported cases are not a major cause of recent transmission of tuberculosis in this area. The association between recent transmission and HIV infection reflects both biologic and social factors. HIV-seropositive patients can have rapid disease progression after infection with tuberculosis12. HIV may modify or overwhelm other risk factors for transmission. For example, in this study, the impact of lower income was found to be greatest in the HIV-seronegative population.

This study has several implications for the future control of tuberculosis. In this urban population, Hispanic ethnicity, HIV infection, and residence in lower-income block groups were risk factors for recent transmission. The recently transmitted cases can be mapped to lower-income neighborhoods, thus identifying an environment where there are other potential risk factors for transmission, such as crowding. In populations in which most cases of tuberculosis are due to reactivation, screening with the tuberculin test and selective use of preventive therapy offer effective strategies for controlling disease. However, in communities with high rates of recent transmission, identification of groups at high risk for transmission, early identification of cases, reduction of institutional spread, and treatment until the disease is cured require more emphasis. In this study, almost half the isolates from patients with recently transmitted infections were drug-resistant, and one quarter were resistant to multiple drugs. Recently transmitted cases accounted for almost two thirds of drug-resistant M. tuberculosis. In 1991, many of the drug-resistant cases in New York City were due to primary drug resistance20. The difficulty of treating drug-resistant tuberculosis until it is cured may lead to a prolonged infectious state, increasing the risk of selection for and transmission of drug-resistant strains. Our study suggests that efforts to control the increase in drug-resistant tuberculosis must include a strategy to reduce the transmission of the disease. In addition to intensive inpatient and outpatient therapeutic programs, unconventional approaches, including active case finding in the areas of high transmission and the option of immunization with bacille Calmette-Guerin for persons at higher risk, merit serious consideration to protect the community, health care workers, and hospitalized patients from further transmission of drug-resistant disease.

Supported by training grants from the National Cancer Institute (2T32CA09173-16) and the National Institute of Allergy and Infectious Diseases (T32AI07183-13 and T32AI07183-14), by the Henry J. Kaiser Foundation, and by the Howard Hughes Medical Research Institute.

We are indebted to Mary Motyl, Ph.D., John McKitrick, Ph.D., and William R. Jacobs, Jr., Ph.D., for their expertise and support; to Charles Lawson and Barry Shapiro of the Audio Visual Center of the Albert Einstein College of Medicine for assistance with illustrations; and to Peter Alpert, M.D., for help with data collection and analysis.


Source Information

From the Division of Infectious Diseases, Department of Medicine, Montefiore Medical Center-North Central Bronx Hospital (D.A., G.E.K.), the Department of Epidemiology and Social Medicine, Montefiore Medical Center (E.D.), and the Department of Microbiology and Immunology and the Howard Hughes Medical Institute (R.A.M., B.R.B.), Albert Einstein College of Medicine, Bronx, N.Y.; the Department of Epidemiology and Biostatistics, University of California, San Francisco, and the Division of Epidemiology and Medicine, San Francisco General Hospital, San Francisco (A.R.M., J.A.H.); and the Department of Political Science, Lehman College, City University of New York, Bronx, N.Y. (W.B.).

Address reprint requests to Dr. Alland at the Division of Infectious Diseases, Department of Medicine, Montefiore Medical Center, 111 East 210 St., Bronx, NY 10467.

References

  1. Tuberculosis morbidity -- United States, 1992. MMWR Morb Mortal Wkly Rep 1993;42:696-7, 703. [Medline]
  2. Tuberculosis in New York City, 1992: information summary. New York: New York City Department of Health, 1993.
  3. Barnes PF, Bloch AB, Davidson PT, Snider DE Jr. Tuberculosis in patients with human immunodeficiency virus infection. N Engl J Med 1991;324:1644-1650. [Medline]
  4. Rieder HL, Cauthen GM, Comstock GW, Snider DE Jr. Epidemiology of tuberculosis in the United States. Epidemiol Rev 1989;11:79-98. [Free Full Text]
  5. Brudney K, Dobkin J. Resurgent tuberculosis in New York City: human immunodeficiency virus, homelessness, and the decline of tuberculosis control programs. Am Rev Respir Dis 1991;144:745-749. [Medline]
  6. Tuberculosis among foreign-born persons entering the United States: recommendations of the Advisory Committee for Elimination of Tuberculosis. MMWR Morb Mortal Wkly Rep 1990;39:1-5. [Medline]
  7. Selwyn PA, Hartel D, Lewis VA, et al. A prospective study of the risk of tuberculosis among intravenous drug users with human immunodeficiency virus infection. N Engl J Med 1989;320:545-550. [Abstract]
  8. The use of preventive therapy for tuberculous infection in the United States: recommendations of the Advisory Committee for Elimination of Tuberculosis. MMWR Morb Mortal Wkly Rep 1990;39:9-12. [Medline]
  9. Daley CL, Small PM, Schecter GF, et al. An outbreak of tuberculosis with accelerated progression among persons infected with the human immunodeficiency virus: an analysis using restriction-fragment-length polymorphisms. N Engl J Med 1992;326:231-235. [Abstract]
  10. Nardell E, McInnis B, Thomas B, Weidhaas S. Exogenous reinfection with tuberculosis in a shelter for the homeless. N Engl J Med 1986;315:1570-1575. [Abstract]
  11. Nolan CM, Elarth AM, Barr H, Saeed AM, Risser DR. An outbreak of tuberculosis in a shelter for homeless men: a description of its evolution and control. Am Rev Respir Dis 1991;143:257-261. [Medline]
  12. Nosocomial transmission of multidrug-resistant tuberculosis among HIV-infected persons -- Florida and New York, 1988-1991. MMWR Morb Mortal Wkly Rep 1991;40:585-591. [Medline]
  13. Starke JR, Jacobs RF, Jereb J. Resurgence of tuberculosis in children. J Pediatr 1992;120:839-855. [CrossRef][Medline]
  14. Nardell EA. Beyond four drugs: public health policy and the treatment of the individual patient with tuberculosis. Am Rev Respir Dis 1993;148:2-5. [Medline]
  15. Pearson ML, Jereb JA, Frieden TR, et al. Nosocomial transmission of multidrug-resistant Mycobacterium tuberculosis: a risk to patients and health care workers. Ann Intern Med 1992;117:191-196.
  16. Kent PT, Kubica GP. Public health mycobacteriology: a guide for the level III laboratory. Atlanta: Department of Health and Human Services, 1985.
  17. van Soolingen D, Hermans PWM, de Haas PEW, Soll DR, van Embden JDA. Occurrence and stability of insertion sequences in Mycobacterium tuberculosis complex strains: evaluation of an insertion sequence-dependent DNA polymorphism as a tool in the epidemiology of tuberculosis. J Clin Microbiol 1991;29:2578-2586. [Free Full Text]
  18. Hermans PWM, van Soolingen D, Bik EM, de Haas PEW, Dale JW, van Embden JDA. Insertion element IS987 for Mycobacterium bovis BCG is located in a hot-spot integration region for insertion elements in Mycobacterium tuberculosis complex strains. Infect Immun 1991;59:2695-2705. [Free Full Text]
  19. Tuberculosis in New York City, 1991: information summary. New York: New York City Department of Health, 1992.
  20. Frieden TR, Sterling T, Pablos-Mendez A, Kilburn JO, Cauthen GM, Dooley SW. The emergence of drug-resistant tuberculosis in New York City. N Engl J Med 1993;328:521-526. [Erratum, N Engl J Med 1993;329:148.] [Free Full Text]
  21. Dobkin J, Bangsberg B, Brudney K, Kalkut G, Bloom BR, Alland D. Detection by DNA fingerprinting (DNA FP) of a "hidden" tuberculosis outbreak among HIV+ homeless patients. In: Abstracts of the Ninth International Conference on AIDS/Fourth STD World Congress, Berlin, Germany, June 6-11, 1993. London: Wellcome Foundation, 1993:324. abstract.
  22. Styblo K. The epidemiology of tuberculosis. The Hague, the Netherlands: VEB Gustav Fischer Verlag Jena, 1984.
  23. American Thoracic Society. Control of tuberculosis in the United States. Am Rev Respir Dis 1992;146:1623-1633. [Medline]
  24. Screening for tuberculosis and tuberculous infection in high-risk populations: recommendations of the Advisory Committee for Elimination of Tuberculosis. MMWR Morb Mortal Wkly Rep 1990;39:1-7.
  25. Hermans PWM, van Soolingen D, Dale JW, et al. Insertion element IS986 from Mycobacterium tuberculosis: a useful tool for the diagnosis and epidemiology of tuberculosis. J Clin Microbiol 1990;28:2051-2058. [Free Full Text]
  26. Small PM, Shafer RW, Hopewell PC, et al. Exogenous reinfection with multidrug-resistant Mycobacterium tuberculosis in patients with advanced HIV infection. N Engl J Med 1993;328:1137-1144. [Free Full Text]
  27. Hewlett D, Franchini D, Horn D, et al. Outbreak of multidrug-resistant tuberculosis at a hospital -- New York City, 1991. MMWR Morb Mortal Wkly Rep 1993;42:427, 433-427, 434. 

 

This Article
-Abstract

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation

This article has been cited by other articles:



HOME  |  SUBSCRIBE  |  SEARCH  |  CURRENT ISSUE  |  PAST ISSUES  |  COLLECTIONS  |  PRIVACY  |  TERMS OF USE  |  HELP  |  beta.nejm.org

Comments and questions? Please contact us.

The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved.