SUMMARY
OBJECTIVE :
To determine risk factors for multidrug-resistant tuberculosis (MDR-TB) and describe MDR-TB according to three characteristics: previous TB disease, recent transmission of MDR-TB, and reactivation of latent MDR-TB infection.
SETTING AND DESIGN :
We used 2011–2016 surveillance data from the US National Tuberculosis Surveillance System and National Tuberculosis Genotyping Service and used logistic regression models to estimate risk factors associated with MDR-TB.
RESULTS :
A total of 615/45 209 (1.4%) cases were confirmed as MDR-TB; 111/615 (18%) reported previous TB disease; 41/615 (6.7%) were attributed to recent MDR-TB transmission; and 449/615 (73%) to reactivation. Only 12/41 (29%) patients with TB attributed to recent transmission were known to be contacts of someone with MDR-TB. For non-US-born patients, the adjusted odds ratios of having MDR-TB were 32.6 (95%CI 14.6–72.6) among those who were known to be contacts of someone with MDR-TB and 6.5 (95%CI 5.1–8.3) among those who had had previous TB disease.
CONCLUSION :
The majority of MDR-TB cases in the United States were associated with previous TB disease or reactivation of latent MDR-TB infection; only a small proportion of MDR-TB cases were associated with recent transmission.
Keywords: recent transmission, previous TB, MDR-TB contacts, reactivation
RÉSUMÉ
OBJECTIF :
Estimer les facteurs de risque de tuberculose multirésistante (MDR-TB) et décrire la MDR-TB en fonction de trois caractéristiques : TB maladie préalable, transmission récente de la MDR-TB et réactivation d’une infection latente de MDR-TB.
CONTEXTE ET SCHÉMA :
Nous avons utilisé les données de surveillance 2011–2016 du Système national US de surveillance de la Tuberculose et du service national de génotypage de la TB. Nous avons eu recours à des modèles de régression logistique afin d’estimer les facteurs de risque associés à la MDR-TB.
RÉSULTATS :
Un total de 615/45209 (1,4%) cas ont été confirmés comme étant des MDR-TB; 111/615 (18%) ont rapporté des antécédents de TB maladie; 41/615 (6,7%) ont été considérés comme attribués à une transmission récente de MDR-TB; et 449/615 (73%), à une réactivation. Seulement 12/41 (29%) patients atteints de TB attribuée à une transmission récente etaient des contacts connus d’un patient MDR-TB. En cé qui concerne les patients nés hors des Etats-Unis, les odds ratios ajustés de MDR-TB et leurs intervalles de confiance à 95% (IC95%) ont été de 32,6 (IC95% 14,6–72,6) parmi ceux qui étaient des contacts connus de patients MDR-TB et de 6,5 (IC95% 5,1–8,3) parmi ceux qui avaient déjà eu une TB maladie.
CONCLUSION :
La majorité des cas de MDR-TB aux Etats-Unis ont été associés à une TB préalable ou à une réactivation d’infection latente de MDR-TB; une petite proportion de cas de MDR-TB seulement a été associée à une transmission récente.
RESUMEN
OBJETIVO:
Estimar los factores de riesgo de contraer la tuberculosis multirresistente (MDR-TB) y describirla con respecto a tres características, a saber: el antecedente de enfermedad tuberculosa, la transmisión reciente de MDR-TB y la reactivación de una infección latente de MDR-TB.
MARCO DE REFERENCIA Y MÉ TODO:
Se utilizaron los datos de vigilancia del 2011 al 2016 del sistema nacional de vigilancia y del servicio nacional de genotipificación de los Estados Unidos y se aplicaron modelos de regresión logística con el fin de evaluar los factores de riesgo asociados con la presencia de MDR-TB.
RESULTADOS:
De los 45209 casos, se confirmó la MDR-TB en 615 (1,4%); de los cuales 111 (18%) refirieron un antecedente de enfermedad tuberculosa; se atribuyeron 41 casos (6,7%) a una transmisión reciente de MDR-TB y 449 (73%) a una reactivación. Solo en 12 de los 41 pacientes (29%) con TB atribuida a una transmisión reciente se conocía el antecedente de contacto con un caso de MDR-TB. El cociente de posibilidades ajustado de contraer la TB en los pacientes nacidos fuera de los Estados Unidos fue 32,6 (IC95% 14,6–72,6) en quienes tenían contacto conocido con un caso de MDR-TB y 6,5 (IC95% 5,1–8,3) en los pacientes con antecedente de enfermedad tuberculosa.
CONCLUSIÓN:
La mayoría de los casos de MDR-TB en los Estados Unidos se asocia con un antecedente de enfermedad tuberculosa o con la reactivación de una infección latente de MDR-TB; solo una proporción pequena de casos de MDR-TB se asoció con unã transmisión reciente.
MULTIDRUG-RESISTANT TUBERCULOSIS (MDR-TB) is a form of TB disease caused by strains of Mycobacterium tuberculosis that are resistant to at least isoniazid (INH) and rifampin (RMP), the two most efficacious TB drugs. MDR-TB often requires complex treatment regimens that are associated with more adverse events, longer treatment time, higher cost, and higher treatment failure and mortality.1–7 In 2016, 3.3% of new TB cases worldwide and 1.2% of the reported US TB cases were MDR-TB.1,8 In the United States, the majority of TB incident cases (69%) and the vast majority of MDR-TB incident cases (92%) reported during 2016 were among non-US-born persons.8
New TB cases can result from recent M. tuberculosis transmission or reactivation of longstanding, untreated latent infection. Cases resulting from recent transmission, especially MDR-TB cases, are of major public health concern because they might represent uncontrolled transmission or an outbreak. Availability of genotyping data in the United States for culture-confirmed TB cases allowed for estimates of recent transmission because clustered cases sharing the same genotype are more likely attributable to recent transmission.9–12 The US Centers for Disease Control and Prevention (CDC) has developed a plausible source-case method for refining estimates of recent TB transmission in the United States by using nationally available genotyping and surveillance data; using the algorithm, it was estimated that 14% of genotyped TB cases were attributable to recent transmission during 2011–2014.13,14 The method can help identify populations and geographic areas with ongoing transmission or an outbreak, which can prompt additional control efforts for minimizing transmission.
MDR-TB cases result either from a person being infected through contact with a person who already had MDR-TB (i.e., MDR-TB transmission), or it can occur in a patient who originally had non-MDR-TB but then acquired MDR-TB as a result of inadequate or inappropriate TB treatment (i.e., acquired MDR-TB).15 Acquired MDR-TB is often assumed to be the cause of drug resistance among persons who have been treated for TB previously, or as defined in this study, among persons who reported a previous diagnosis of TB. During 2016, an estimated 19% of TB cases worldwide among persons who had been treated previously for TB disease were diagnosed as MDR-TB or at least RMP-resistant.1,15–20 With M. tuberculosis genotyping, however, we can now examine MDR-TB transmission. Furthermore, using genotyping methods, we can estimate which MDR-TB cases are associated with recent TB transmission and which are associated with the reactivation of latent TB infection. Analyzing data regarding MDR-TB cases by these three groups—previous TB disease, recent transmission, and reactivation from latent MDR-TB infection—can provide important insights into how to prevent MDR-TB. Previous studies have been based on relatively limited sample sizes, and the majority of studies were set in high MDR-TB incidence countries.19–26 Our study describes the demographic and epidemiologic characteristics of MDR-TB patients in the United States, a low MDR-TB incidence country. We were also able to estimate risk factors for MDR-TB among non-US-born and US-born persons, two population groups with different TB risk characteristics.8
STUDY POPULATION AND METHODS
We used 2011–2016 US surveillance data from the National Tuberculosis Surveillance System (NTSS) and the National Tuberculosis Genotyping Service (NTGS). NTSS collects clinical, demographic, and risk factor data for all reported US TB cases; NTGS collects genotyping results for culture-positive TB cases. These two data systems are linked through unique case numbers. All data collected and analyzed were part of routine public health surveillance. This project was determined not to be human subject research by the US CDC and therefore did not require ethical approval by an institutional review board.
We combined 6 years of NTSS data (n = 45 209 cases) to obtain better statistical power and used NTGS data and an established plausible source-case method (described elsewhere).13 We defined a case of MDR-TB as TB disease that is resistant to at least INH and RMP on the basis of initial drug susceptibility testing (DST) results. We analyzed the following demographic variables and potential risk factors for MDR-TB: US/non-US-born, age group, sex, race/ethnicity, years of residence in the United States in case of non-US-born persons, incarceration at diagnosis, homelessness any time during the previous year, excess alcohol use, injection drug use any time during the 12 months before diagnosis, human immunodeficiency virus (HIV) infection, history of previous contact with an MDR-TB patient during the past 2 years (known contact of an MDR-TB patient, only values of ‘Yes’ are coded), previous history of TB disease, and whether the case was attributed to recent transmission.
We classified each MDR-TB case into one of three distinct, non-overlapping categories according to the following hierarchical order: 1) previous TB (PreTB); 2) recent transmission only (RT; cases were attributable to recent transmission [i.e., a plausible source case within 2 years before diagnosis could be identified] but the patient had no previous TB); or 3) reactivation (RA; persons had no previous TB and infections were not attributable to recent transmission). An MDR-TB case attributed to recent transmission had at least one plausible source case identified by the plausible source-case method that had either MDR- or INH- or RMP-monoresistant TB (this definition for recent transmission of MDR-TB is slightly different from the definition of recent transmission of TB13,14). We included plausible monoresistant sources because the surveillance system only collects a single initial DST result, which sometimes leads to incomplete levels of reported drug resistance. If an MDR-TB case was not in one of these three groups, we classified the case as ‘Other’ (cases with missing values of previous TB or recent transmission).
We calculated the proportions of MDR-TB cases by each demographic variable and potential risk factor. We also computed the proportions of MDR-TB cases in each of the three source categories (PreTB, RT, RA) stratified by demographic variable and potential risk factor. We then used logistic regression models to identify risk factors associated with MDR-TB among all culture-positive MDR-TB and non-MDR-TB cases with genotyping data and initial DST results. All variables with a P < 0.3 were included in the bivariate analyses. Both manual forward and backward eliminations were used to estimate adjusted odds ratios (aORs) and 95% confidence intervals (95%CIs). The results from both selection methods were identical. For multivariable analyses, we used separate logistic models for non-US-born and US-born persons because they had different risk factors. Variables with P < 0.05 were retained in the models. We focused on main effects for each of these models. The interaction terms of recent transmission with previous TB disease (P > 0.8) and age group (P > 0.25) were also examined, but these were not significant. Sparse data bias can be worrying in case of small case counts and wide 95%CIs,27 but in this study, after the bivariate analyses, the case counts of the remaining variables used in multivariable models were found to be adequate. Records with missing values for variables removed from earlier models were nevertheless included in the final model. Statistical analyses were performed using SAS® v9.4 (SAS Institute, Cary, NC, USA).
RESULTS
During 2011–2016, 30 211 (66.8%) of 45 209 TB cases with both initial INH and RMP DST results were diagnosed among non-US-born persons. Proportions for MDR-TB cases were 1.8% (n = 537) among non-US-born and 0.5% (n = 77) among US-born persons (Tables 1 and 2). Among non-US-born persons, the MDR-TB rate was highest (3.0%) in those who had resided in the United States for ≤4 years. MDR-TB rates were not higher in persons reporting incarceration, homelessness, excess alcohol use, injecting drug use, or HIV infection than in TB patients who did not have these characteristics (Table 1). MDR-TB rates were high in persons who had known contact with an MDR-TB patient, compared with those who had no known contacts (38.0% vs. 1.7% in non-US-born and 25.0% vs. 0.5% in US-born, respectively). The MDR-TB rate was much higher in non-US-born persons with previous TB disease than in persons without previous TB disease (7.7% vs. 1.5%). Among cases attributed to recent transmission, MDR-TB rates were respectively 1.3% and 0.3% in non-US-born and US-born persons (Table 2).
Table 1.
Characteristics of MDR-TB patients, United States, 2011–2016
| Characteristic and category | All MDR- and non-MDR-TB n | MDR-TB n (%) |
|---|---|---|
| All | 45 209 | 615 (1.4) |
| Origin | ||
| Non-US-born | 30 211 | 537 (1.8) |
| US-born | 14 947 | 77 (0.5) |
| Missing | 51 | 1 (2.0) |
| Age, years | ||
| ≤14 | 885 | 9 (1.0) |
| 15–44 | 18 831 | 384 (2.0) |
| 45–64 | 14 297 | 147 (1.0) |
| ≥65 | 11 187 | 74 (0.7) |
| Missing | 9 | 1 (11.1) |
| Sex | ||
| Female | 17 203 | 294 (1.7) |
| Male | 28 003 | 321 (1.1) |
| Missing | 3 | 0 |
| Race/ethnicity | ||
| White | 6 279 | 68 (1.1) |
| Black | 9 208 | 80 (0.9) |
| Hispanic | 12 678 | 124 (1.0) |
| Asian | 14 889 | 313 (2.1) |
| Other | 2 022 | 30 (1.5) |
| Missing | 133 | 0 |
| Years in the United States (non-US-born) | ||
| ≤4 | 8 944 | 264 (3.0) |
| 5–19 | 10 299 | 183 (1.8) |
| ≥20 | 8 372 | 58 (0.7) |
| Missing/US-born | 17 594 | 110 (0.6) |
| Incarcerated at diagnosis | ||
| Yes | 1 627 | 18 (1.1) |
| No | 42 712 | 576 (1.3) |
| Missing | 870 | 21 (2.4) |
| Homeless any time during the previous year | ||
| Yes | 2 540 | 11 (0.4) |
| No | 42 382 | 602 (1.4) |
| Missing | 287 | 2 (0.7) |
| Excess alcohol use | ||
| Yes | 5 402 | 34 (0.6) |
| No | 39 229 | 577 (1.5) |
| Missing | 578 | 4 (0.7) |
| Injection drug use | ||
| Yes | 691 | 8 (1.2) |
| No | 43 979 | 603 (1.4) |
| Missing | 539 | 4 (0.7) |
| HIV infection | ||
| Yes | 2 562 | 30 (1.2) |
| No | 37 249 | 547 (1.5) |
| Missing | 5 398 | 38 (0.7) |
| Known contact of an MDR-TB patient | ||
| Yes | 74 | 25 (33.8) |
| Missing | 45 135 | 590 (1.3) |
| Previous TB disease | ||
| Yes | 2 003 | 111 (5.5) |
| No | 42 963 | 501 (1.2) |
| Missing | 243 | 3 (1.2) |
| Recent transmission | ||
| Yes | 6 382 | 47 (0.7) |
| No | 35 892 | 553 (1.5) |
| Missing | 2 935 | 15 (0.5) |
MDR-TB = multidrug-resistant tuberculosis; HIV = human immunodeficiency virus.
Table 2.
Characteristics of MDR-TB patients by origin of birth, United States, 2011–2016
| Characteristic and category | Non-US-born | US-born | ||
|---|---|---|---|---|
| All MDR- and non-MDR-TB n | MDR-TB n (%) | All MDR- and non-MDR-TB n | MDR-TB n (%) | |
| All | 30 211 | 537 (1.8) | 14 947 | 77 (0.5) |
| Age, years | ||||
| ≤14 | 243 | 2 (0.8) | 641 | 7 (1.1) |
| 15–44 | 14 000 | 352 (2.5) | 4816 | 32 (0.7) |
| 45–64 | 8 558 | 116 (1.4) | 5719 | 31 (0.5) |
| ≥65 | 7 404 | 66 (0.9) | 3 768 | 7 (0.2) |
| Missing | 6 | 1 (16.7) | 3 | 0 |
| Sex | ||||
| Female | 12 241 | 259 (2.1) | 4 947 | 34 (0.7) |
| Male | 17 968 | 278 (1.5) | 9 999 | 43 (0.4) |
| Missing | 2 | 0 | 1 | 0 |
| Race/ethnicity | ||||
| White | 1 323 | 47 (3.6) | 4953 | 21 (0.4) |
| Black | 3 633 | 59 (1.6) | 5 572 | 21 (0.4) |
| Hispanic | 9 746 | 106 (1.1) | 2 908 | 18 (0.6) |
| Asian | 14 350 | 301 (2.1) | 530 | 11 (2.1) |
| Other | 1 069 | 24 (2.2) | 943 | 6 (0.6) |
| Missing | 90 | 0 | 41 | 0 |
| Years in the United States (non-US-born) | ||||
| ≤4 | 8 944 | 264 (3.0) | — | — |
| 5–19 | 10 299 | 183 (1.8) | — | — |
| ≥20 | 8 372 | 58 (0.7) | — | — |
| Missing/US-born | 2 596 | 32 (1.2) | — | — |
| Incarcerated at diagnosis | ||||
| Yes | 944 | 13 (1.4) | 680 | 5 (0.7) |
| No | 28 755 | 509 (1.8) | 13 915 | 66 (0.5) |
| Missing | 512 | 15 (2.9) | 352 | 6 (1.7) |
| Homeless any time during the previous year | ||||
| Yes | 825 | 8 (1.0) | 1 706 | 3 (0.2) |
| No | 29 188 | 527 (1.8) | 13 161 | 74 (0.6) |
| Missing | 198 | 2 (1.0) | 80 | 0 |
| Excess alcohol use | ||||
| Yes | 2135 | 22 (1.0) | 3 260 | 12 (0.4) |
| No | 27 717 | 512 (1.8) | 11 485 | 64 (0.6) |
| Missing | 359 | 3 (0.8) | 202 | 1 (0.5) |
| Injection drug use | ||||
| Yes | 172 | 4 (2.3) | 518 | 4 (0.8) |
| No | 29 716 | 530 (1.8) | 14 233 | 72 (0.5) |
| Missing | 323 | 3 (0.9) | 196 | 1 (0.5) |
| HIV infection | ||||
| Yes | 1 427 | 25 (1.8) | 1 134 | 5 (0.4) |
| No | 25 349 | 478 (1.9) | 11 873 | 68 (0.6) |
| Missing | 3 435 | 34 (1.0) | 1 940 | 4 (0.2) |
| Known contact of an MDR-TB patient | ||||
| Yes | 50 | 19 (38.0) | 24 | 6 (25.0) |
| Missing | 30 161 | 518 (1.7) | 14 923 | 71 (0.5) |
| Previous TB disease | ||||
| Yes | 1 349 | 104 (7.7) | 654 | 7 (1.1) |
| No | 28 689 | 431 (1.5) | 14231 | 69 (0.5) |
| Missing | 173 | 2 (1.2) | 62 | 1 (1.6) |
| Recent transmission | ||||
| Yes | 2 651 | 34 (1.3) | 3721 | 13 (0.3) |
| No | 25 674 | 489 (1.9) | 10 183 | 63 (0.6) |
| Missing | 1 886 | 14 (0.7) | 1 043 | 1 (0.1) |
MDR-TB = multidrug-resistant tuberculosis; HIV = human immunodeficiency virus.
Of the 615 MDR-TB cases reported during 2011–2016, 6 of 111 PreTB cases were also attributed to recent transmission; 14 were classified as Other. In non-US-born persons, the proportions of PreTB, RT only, RA, and Other cases were respectively 19.4%, 5.2%,73.2% and 2.2%; those residing in the United States for ≤4 years had the highest rates of PreTB (24.6%) and low rate of RT only (1.9%). The proportions of PreTB, RT only, and RA cases in persons who had known contact with an MDR-TB patient were respectively 4.0%, 48.0%, and 48.0%. Only 12 of 41 patients from the RT group had documented contact with an MDR-TB patient (Table 3).
Table 3.
Characteristics of MDR-TB patients by previous TB disease, recent transmission only, reactivation, United States, 2011–2016
| Characteristic and category | MDR-TB n | Previous TB n (%) | Recent transmission only* n (%) | Reactivation n (%) | Other† n (%) |
|---|---|---|---|---|---|
| All | 615 | 111 (18.0) | 41 (6.7) | 449 (73.0) | 14 (2.3) |
| Origin | |||||
| Non-US-born | 537 | 104 (19.4) | 28 (5.2) | 393 (73.2) | 12 (2.2) |
| US-born | 77 | 7 (9.1) | 13 (16.9) | 55 (71.4) | 2 (2.6) |
| Missing | 1 | 0 | 0 | 1 (100.0) | 0 |
| Age, years | |||||
| ≤14 | 9 | 0 | 2 (22.2) | 7 (77.8) | 0 |
| 15–44 | 384 | 63 (16.4) | 24 (6.3) | 289 (75.3) | 8 (2.1) |
| 45–64 | 147 | 36 (24.5) | 10 (6.8) | 97 (66.0) | 4 (2.7) |
| ≥65 | 74 | 12 (16.2) | 5 (6.8) | 55 (74.3) | 2 (2.7) |
| Missing | 1 | 0 | 0 | 1 (100.0) | 0 |
| Sex | |||||
| Female | 294 | 55 (18.7) | 25 (8.5) | 207 (70.4) | 7 (2.4) |
| Male | 321 | 56 (17.4) | 16 (5.0) | 242 (75.4) | 7 (2.2) |
| Race/ethnicity | |||||
| White | 68 | 12 (17.6) | 5 (7.4) | 50 (73.5) | 1 (1.5) |
| Black | 80 | 13 (16.3) | 4 (5.0) | 63 (78.8) | 0 |
| Hispanic | 124 | 17 (13.7) | 8 (6.5) | 95 (76.6) | 4 (3.2) |
| Asian | 313 | 65 (20.8) | 17 (5.4) | 223 (71.2) | 8 (2.6) |
| Other | 30 | 4 (13.3) | 7 (23.3) | 18 (60.0) | 1 (3.3) |
| Years in the United States (non-US-born) | |||||
| ≤4 | 264 | 65 (24.6) | 5 (1.9) | 193 (73.1) | 1 (0.4) |
| 5–19 | 183 | 29 (15.8) | 13 (7.1) | 136 (74.3) | 5 (2.7) |
| ≥20 | 58 | 9 (15.5) | 1 (1.7) | 46 (79.3) | 2 (3.4) |
| Missing/US-born | 110 | 8 (7.3) | 22 (20.0) | 74 (67.3) | 6 (5.5) |
| Incarcerated at diagnosis | |||||
| Yes | 18 | 4 (22.2) | 2 (11.1) | 12 (66.7) | 0 (0) |
| No | 576 | 105 (18.2) | 35 (6.1) | 423 (73.4) | 13 (2.3) |
| Missing | 21 | 2 (9.5) | 4 (19.0) | 14 (66.7) | 1 (4.8) |
| Homeless any time during the previous year | |||||
| Yes | 11 | 5 (45.5) | 1 (9.1) | 5 (45.5) | 0 |
| No | 602 | 105 (17.4) | 40 (6.6) | 444 (73.8) | 13 (2.2) |
| Missing | 2 | 1 (50.0) | 0 | 0 | 1 (50.0) |
| Excess alcohol use | |||||
| Yes | 34 | 5 (14.7) | 5 (14.7) | 24 (70.6) | 0 |
| No | 577 | 106 (18.4) | 36 (6.2) | 423 (73.3) | 12 (2.1) |
| Missing | 4 | 0 | 0 | 2 (50.0) | 2 (50.0) |
| Injection drug use | |||||
| Yes | 8 | 1 (12.5) | 1 (12.5) | 6 (75.0) | 0 |
| No | 603 | 110 (18.2) | 40 (6.6) | 440 (73.0) | 13 (2.2) |
| Missing | 4 | 0 | 0 | 3 (75.0) | 1 (25.0) |
| HIV infection | |||||
| Yes | 30 | 7 (23.3) | 0 | 22 (73.3) | 1 (3.3) |
| No | 547 | 99 (18.1) | 36 (6.6) | 400 (73.1) | 12 (2.2) |
| Missing | 38 | 5 (13.2) | 5 (13.2) | 27 (71.1) | 1 (2.6) |
| Known contact of an MDR-TB patient | |||||
| Yes | 25 | 1 (4.0) | 12 (48.0) | 12 (48.0) | 0 |
| Missing | 590 | 110 (18.6) | 29 (4.9) | 437 (74.1) | 14 (2.4) |
Did not include six cases with both previous TB and recent transmission; these were included in the previous TB category.
Cases with missing values of previous TB or recent transmission.
MDR-TB = multidrug-resistant tuberculosis; HIV = human immunodeficiency virus.
The results of variables with P < 0.3 in bivariate analyses are presented in Table 4. These variables were included in multivariable analyses. For non-US-born persons, risk factors that remained statistically significant in the multivariable logistic regression model were age, sex, race/ethnicity, years residing in the United States, being a known contact of an MDR-TB patient, previous TB disease, and recent transmission. The aORs were 32.6 (95%CI 14.6–72.6) for being an MDR-TB patient contact and 6.5 (95%CI 5.1–8.3) for persons with previous TB disease. For US-born persons, the significant risk factors were age, race/ethnicity, being an MDR-TB patient contact, previous TB disease, and recent transmission. The aORs were 46.2 (95%CI 16.3–130.7) for MDR-TB contacts and 2.8 (95%CI 1.2–6.1) for persons with previous TB disease (Table 4).
Table 4.
Risk factors from bivariate and multivariable models for MDR-TB among MDR- and non-MDR-TB patients by place of birth, United States, 2011–2016
| Bivariate models* | Multivariable models† | ||||
|---|---|---|---|---|---|
| Risk factor and category | Non-US-born OR (95%CI) | US-born OR (95%CI) | Non-US-born aOR (95%CI) | US-born aOR (95%CI) | |
| Age, years | |||||
| ≤14 | 0.9 (0.2–3.8) | 5.9 (2.1–17.0) | 1.0 (0.3–4.4) | 4.9 (1.5–15.5) | |
| 15–44 | 2.9 (2.2–3.7) | 3.6 (1.6–8.2) | 2.4 (1.7–3.2) | 3.2 (1.4–7.5) | |
| 45–64 | 1.5 (1.1–2.1) | 2.9 (1.3–6.7) | 1.6 (1.2–2.2) | 3.1 (1.4–7.2) | |
| ≥65 | Reference | Reference | Reference | Reference | |
| Sex | |||||
| Female | 1.4 (1.2–1.6) | 1.6 (1.0–2.5) | 1.3 (1.0–1.5) | ||
| Male | Reference | Reference | Reference | ||
| Race/ethnicity | |||||
| White | Reference | Reference | Reference | Reference | |
| Black | 0.4 (0.3–0.7) | 0.9 (0.5–1.6) | 0.3 (0.2–0.4) | 0.9 (0.5–1.6) | |
| Hispanic | 0.3 (0.2–0.4) | 1.5 (0.8–2.8) | 0.3 (0.2–0.4) | 1.3 (0.7–2.6) | |
| Asian | 0.6 (0.4–0.8) | 5.0 (2.4–10.4) | 0.5 (0.4–0.7) | 2.7 (1.1–6.4) | |
| Other | 0.6 (0.4–1.0) | (0.6–3.7) | 0.7 (0.3–1.3) | 1.2 (0.5–3.3) | |
| Years in the United States (non-US-born) | |||||
| ≤4 | 4.4 (3.3–5.8) | 3.1 (2.3–4.4) | |||
| 5–19 | 2.6 (1.9–3.5) | 2.0 (1.4–2.7) | |||
| ≥20 | Reference | Reference | |||
| Homeless any time during the previous year | |||||
| Yes | 0.5 (0.3–1.1) | 0.3 (0.1–1.0) | |||
| No | Reference | Reference | |||
| Excess alcohol use | |||||
| Yes | 0.6 (0.4–0.9) | 0.7 (0.4–1.2) | |||
| No | Reference | Reference | |||
| Known contact of an MDR-TB patient | |||||
| Yes | 35.1 (19.7–62.5) | 69.8 (26.9–180.9) | 32.6 (14.6–72.6) | 46.2 (16.3–130.7) | |
| Missing | Reference | Reference | Reference | ||
| Previous TB disease | |||||
| Yes | 5.5 (4.4–6.8) | 2.2 (1.0–4.9) | 6.5 (5.1–8.3) | 2.8 (1.2–6.1) | |
| No | Reference | Reference | Reference | Reference | |
| Recent transmission | |||||
| Yes | 0.7 (0.5–0.9) | 0.6 (0.3–1.0) | 0.6 (0.4–0.9) | 0.4 (0.2–0.8) | |
| No | Reference | Reference | Reference | Reference | |
P < 0.3.
P < 0.05.
MDR-TB = multidrug-resistant tuberculosis; OR = odds ratio; CI = confidence interval; aOR = adjusted OR.
DISCUSSION
US national TB genotyping surveillance coverage, defined as the percentage of culture-positive TB cases having at least one isolate genotyped, increased from 52.6% in 2004 to 96.4% in 2016.8 During 2011–2016, the genotyping coverage for MDR-TB cases was 97.6%, which allowed for the classification of the majority of MDR-TB cases by recent transmission status. Among the seven MDR-TB cases classified as reactivation among children aged ≤14 years, two were under 5 years. These two children are likely to have disease that is a result of recent transmission, but as they did not have a culture result, they did not have genotyping results available. Because only six of 111 persons with previous TB disease in this study were attributed to recent transmission, the current hierarchical classification (without further classification of these six persons) should be adequate.
Among persons with MDR-TB, we estimated that recent transmission was low, at 6.7%, in 2011–2016. The low rate might be the result of prompt public health action or lower fitness of MDR-TB stains,28 but is more likely due to program effectiveness. Our data revealed that non-US-born persons had a higher proportion of previous TB than US-born persons (19.4% and 9.1%, respectively). Furthermore, among persons with previous TB disease, the rate of MDR-TB among the non-US-born was higher than among the US-born (7.7% vs. 1.1%). These differences likely indicate that domestic TB treatment practices are effective in terms of preventing acquired drug resistance.
The characteristics and risk factors for MDR-TB differ between non-US-born and US-born persons.8,29,30 On comparing the results between the two groups (Table 4), we determined that age groups 15–44 and 45–64 years, being a known contact of an MDR-TB patient, previous TB disease, and recent transmission were statistically significant risk factors for both non-US-born and US-born persons. Residing in the United States for ≤4 and 5–19 years and being White were additional risk factors for non-US-born persons; age group ≤ 14 years and being Asian were additional risk factors for US-born persons.
During 2000–2005, the average number of annual US MDR-TB cases decreased from 145 to 121; HIV co-infection prevalence among these patients ranged from 8% to 22%.7 In contrast, the average number of annual MDR-TB cases during 2011–2016 was 103; HIV co-infection prevalence was 5%, decreasing markedly between 2000 and 2005. In previous studies, homelessness was identified as a risk factor for TB disease.10,13,22,30 However, patients reporting homelessness within the previous year did not have higher risk for having MDR-TB than patients not reporting homelessness. This might be because of low case counts of MDR-TB among homeless populations. The low case counts may lead to wide confidence intervals and non-significant results. Although previous studies reported that incarceration, excess alcohol use, and HIV co-infection were risk factors for TB or MDR-TB,4,10,31,32 we did not identify statistically significant associations with MDR-TB for these characteristics. Previous studies that reported risk factors of excess alcohol use and HIV infection for MDR-TB were conducted in Malaysia and Ethiopia,4,31,32 countries with higher TB, MDR- and RMP-resistant TB incidence rates. This indicates that associations of excess alcohol use and HIV co-infection with MDR-TB might differ between high- and low-incidence countries.
We determined that a known contact with an MDR-TB patient and previous TB disease were the two risk factors most strongly associated with MDR-TB. In small-scale MDR-TB risk factor studies conducted in Ethiopia,32 previous contact with TB and MDR-TB patients was reported to be a risk factor for acquiring MDR-TB. In other studies, a history of previous TB disease was identified as a risk factor for MDR-TB.18,21,24,25,33–35 Although the majority of these studies had limited sample sizes, they included different settings, from low- to high-incidence countries (e.g., France, Western and Eastern European countries, Malaysia, China, Brazil, and Ethiopia). Therefore, unlike excess alcohol use and HIV co-infection, which vary in their associations with MDR-TB in different countries, previous TB disease remained a major risk factor universally; however, the magnitude might vary in different settings. MDR-TB intervention strategies in low-incidence countries will differ from those in higher TB incidence countries.
Our analysis was subject to certain limitations. First, we assumed that persons who reported previous TB disease had acquired drug resistance during the previous TB episode, without direct evidence that the persons had received previous treatment. Second, the current episode of TB that we studied might be associated with new transmission that occurred after the previous TB diagnosis. Third, although the method for attributing cases to recent transmission was validated by using field epidemiologic data and sensitivity and specificity analyses, misclassification errors might have occurred. Fourth, case counts were low for certain variables, contributing to non-significant results or wider confidence intervals for some variables. Fifth, we do not know whether treatment for previous TB disease of non-US-born persons occurred on shore or offshore. Sixth, incarceration at diagnosis may underrepresent infection acquired in the recent past; homelessness in the previous 12 months may not capture all MDR-TB acquired in such a setting.
In conclusion, we determined that among MDR-TB patients, the proportion of previous TB disease was higher among non-US-born persons than among US-born persons, and the proportion of recent transmission was low. Our data revealed that being a previous contact of an MDR-TB patient was the dominant risk factor for MDR-TB, followed by previous TB disease for non-US-born persons. The low proportions of recent transmission of MDR-TB might reflect the successful efforts of US TB programs in limiting transmission of MDR-TB in recent years.
Acknowledgements
The authors thank A Hill, T Navin, C Winston, A Langer, and B Silk for their suggestions; and K Smith for her editorial assistance.
Disclaimer:
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Footnotes
Conflicts of interest: none declared.
References
- 1.World Health Organization. Global tuberculosis report, 2017. WHO/HTM/TB/2017.23. Geneva, Switzerland: WHO, 2017. [Google Scholar]
- 2.Hannah HA, Miramontes R, Gandhi NR. Sociodemographic and clinical risk factors associated with tuberculosis mortality in the United States, 2009–2013. Public Health Rep 2017; 132(3): 366–375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Salinas JL, Armstrong LR, Silk BJ, Haddad MB, Cegielski JP. Factors associated with all-cause mortality among patients with multidrug-resistant tuberculosis—United States, 1993–2013. Clin Infect Dis 2017; 65(11): 1924–1926. [DOI] [PubMed] [Google Scholar]
- 4.Elmi OS, Hasan H, Abdullah S, Mat Jeab MZ, Ba Z, Naing NN. Treatment outcomes of patients with multidrug-resistant tuberculosis (MDR-TB) compared with non-MDR-TB infections in peninsular Malaysia. Malays J Med Sci 2016; 23(4): 17–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Marks SM, Mase SR, Morris SB. Systematic review, meta-analysis, and cost-effectiveness of treatment of latent tuberculosis to reduce progression to multidrug-resistant tuberculosis. Clin Infect Dis 2017; 64(12): 1670–1677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Shah NS, Pratt R, Armstrong L, Robison V, Castro KG, Cegielski JP. Extensively drug-resistant tuberculosis in the United States, 1993–2007. JAMA 2008; 300(18): 2153–2160. [DOI] [PubMed] [Google Scholar]
- 7.Wells CD, Cegielski JP, Nelson LJ, et al. HIV infection and multidrug-resistant tuberculosis: the perfect storm. J Infect Dis 2007; 196 (Suppl 1): S86–S107. [DOI] [PubMed] [Google Scholar]
- 8.Centers for Disease Control and Prevention. Reported Tuberculosis in the United States, 2016. Atlanta, GA, USA: US Department of Health and Human Services, CDC, 2017. [Google Scholar]
- 9.Long R, Langlois-Klassen D. Increase in multidrug-resistant tuberculosis (MDR-TB) in Alberta among foreign-born persons: implications for tuberculosis management. Can J Public Health 2013; 104(1): e22–e27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Althomsons SP, Kammerer JS, Shang N, Navin TR. Using routinely reported tuberculosis genotyping and surveillance data to predict tuberculosis outbreaks. PLoS One 2012; 7(11): e48754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Moonan PK, Ghosh S, Oeltmann JE, Kammerer JS, Cowan LS, Navin TR. Using genotyping and geospatial scanning to estimate recent mycobacterium tuberculosis transmission, United States. Emerg Infect Dis 2012; 18(3): 458–465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nguyen LN, Gilbert GL, Marks GB. Molecular epidemiology of tuberculosis and recent developments in understanding the epidemiology of tuberculosis. Respirology 2004; 9(3): 313–319. [DOI] [PubMed] [Google Scholar]
- 13.Yuen CM, Kammerer JS, Marks K, Navin TR, France AM. Recent transmission of tuberculosis—United States, 2011–2014. PLoS One 2016; 11(4): e0153728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.France AM, Grant J, Kammerer JS, Navin TR. A field-validated approach using surveillance and genotyping data to estimate tuberculosis attributable to recent transmission in the United States. Am J Epidemiol 2015; 182(9): 799–807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cegielski JP, Kurbatova E, van der Walt M, et al. Multidrug-resistant tuberculosis treatment outcomes in relation to treatment and initial versus acquired second-line drug resistance. Clin Infect Dis 2016; 62(4): 418–430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cegielski JP, Dalton T, Yagui M, et al. Extensive drug resistance acquired during treatment of multidrug-resistant tuberculosis. Clin Infect Dis 2014; 59(8): 1049–1063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Smith SE, Ershova J, Vlasova N, et al. Risk factors for acquisition of drug resistance during multidrug-resistant tuberculosis treatment, Arkhangelsk Oblast, Russia, 2005–2010. Emerg Infect Dis 2015; 21(6): 1002–1011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mohd Shariff N, Shah SA, Kamaludin F. Previous treatment, sputum-smear nonconversion, and suburban living: the risk factors of multidrug-resistant tuberculosis among Malaysians. Int J Mycobacteriol 2016; 5(1): 51–58. [DOI] [PubMed] [Google Scholar]
- 19.Munje R, Deshmukh R, Tumane K. Multidrug-resistant TB among previously treated TB cases: a retrospective study in Nagpur, India. Indian J Tuberc 2015; 62(4): 207–210. [DOI] [PubMed] [Google Scholar]
- 20.Zhao Y, Xu S, Wang L, et al. National survey of drug-resistant tuberculosis in China. N Engl J Med 2012; 366(23): 2161–2170. [DOI] [PubMed] [Google Scholar]
- 21.Asgedom SW, Teweldemedhin M, Gebreyesus H. Prevalence of multidrug-resistant tuberculosis and associated factors in Ethiopia: a systematic review. J Pathog 2018; 2018: 7104921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Powell KM, VanderEnde DS, Holland DP, et al. Outbreak of drug-resistant Mycobacterium tuberculosis among homeless people in Atlanta, Georgia, 2008–2015. Public Health Rep 2017; 132(2): 231–240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Elmi OS, Hasan H, Abdullah S, Mat Jeab MZ, Bin Alwi Z, Naing NN. Multidrug-resistant tuberculosis and risk factors associated with its development: a retrospective study. J Infect Dev Ctries 2015; 9(10): 1076–1085. [DOI] [PubMed] [Google Scholar]
- 24.Faustini A, Hall AJ, Perucci CA. Risk factors for multidrug resistant tuberculosis in Europe: a systematic review. Thorax 2006; 61(2): 158–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Fregona G, Cosme LB, Moreira CMM, et al. Risk factors associated with multidrug-resistant tuberculosis in Espirito Santo, Brazil. Rev Saude Publica 2017; 51: 41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.AnChiu C, Xian W, Moss CF. Flying in silence: Echolocating bats cease vocalizing to avoid sonar jamming. Proc Natl Acad Sci USA 2008; 105(35): 13116–13121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Greenland S, Mansournia MA, Altman DG. Sparse data bias: a problem hiding in plain sight. BMJ 2016; 352: i1981. [DOI] [PubMed] [Google Scholar]
- 28.Dean AS, Cox H, Zignol M. Epidemiology of drug-resistant tuberculosis. Adv Exp Med Biol 2017; 1019: 209–220. [DOI] [PubMed] [Google Scholar]
- 29.Smith SE, Pratt R, Trieu L, Barry PM, Thai DT, Ahuja SD, Shah S. Epidemiology of pediatric multidrug-resistant tuberculosis in the United States, 1993–2014. Clin Infect Dis 2017; 65(9): 1437–1443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bamrah S, Yelk Woodruff RS, Powell K, Ghosh S, Kammerer JS, Haddad MB. Tuberculosis among the homeless, United States, 1994–2010. Int J Tuberc Lung Dis 2013; 17(11): 1414–1419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Demile B, Zenebu A, Shewaye H, Xia S, Guadie A. Risk factors associated with multidrug-resistant tuberculosis (MDR-TB) in a tertiary armed force referral and teaching hospital, Ethiopia. BMC Infect Dis 2018; 18(1): 249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Desissa F, Workineh T, Beyene T. Risk factors for the occurrence of multidrug-resistant tuberculosis among patients undergoing multidrug-resistant tuberculosis treatment in East Shoa, Ethiopia. BMC Public Health 2018; 18(1): 422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cai X, Zhang D, Yan Y, Tan D, Xu Y. [Meta-analysis on risk factors of multidrug resistant tuberculosis in China]. Zhonghua Liu Xing Bing Xue Za Zhi 2015; 36(12): 1424–1429. [PubMed] [Google Scholar]
- 34.Gaborit BJ, Revest M, Roblot F, et al. ; GERICCO MDR-TB study group. Characteristics and outcome of multidrug-resistant tuberculosis in a low-incidence area. Med Mal Infect 2018; 48(7):457–464. [DOI] [PubMed] [Google Scholar]
- 35.Gunther G, van Leth F, Alexandru S, et al. Multidrug-resistant tuberculosis in Europe, 2010–2011. Emerg Infect Dis 2015; 21(3): 409–416. [DOI] [PMC free article] [PubMed] [Google Scholar]