Reported Tuberculosis in the United States, 2018
Executive Commentary
Introduction
This 2018 edition of Reported Tuberculosis in the United States marks a milestone as the first one to be published exclusively on the Internet as part of the Centers for Disease Control and Prevention’s (CDC) Digital First initiative.
As with past editions, this year’s report includes information about incident cases of TB disease reported from each of the 50 U.S. states, the District of Columbia, five U.S. territories (American Samoa, Commonwealth of the Northern Mariana Islands, Guam, Puerto Rico, and the U.S. Virgin Islands), and three independent countries that are in compacts of free association with the United States (Federated States of Micronesia, Republic of the Marshall Islands, and Republic of Palau). Information in this commentary is from the 50 states and DC, except where otherwise indicated. Small variations in historical data included in this edition, compared with previous editions, are attributable to updated information submitted in the interim by reporting areas.
Overall Trends
During 2018, the United States again reported the lowest number of TB cases (9,025) and lowest incidence rate (2.8 cases per 100,000 persons) on record (Table 1). Except for 2015, the U.S. TB case count and incidence rate have declined every year since 1992 (Table 1). However, the annual percentage decrease in case count (−0.7%) and incidence rate (−1.3%) from 2017 to 2018 is the smallest year-to-year decrease, excluding the 2015 increase, observed since 1993 (Table 1).
The National Vital Statistics System reported 515 TB-related deaths (0.2 deaths per 100,000 persons) for 2017, the most recent year for which data are available (Table 1). This represents a 2.5% decrease in deaths and a 3.1% decrease in the mortality rate from 2016, although 2017 deaths remain higher than the historical low of 470 deaths (0.1 deaths per 100,000 persons) reported in 2015 (Table 1).
Among U.S. states, most TB cases continued to be reported from four states: California (23.2%), Texas (12.5%), New York (8.3%), and Florida (6.5%) (Table 28). However, when considering incidence rates by reporting area, Alaska had the highest TB rate (8.5 cases per 100,000 persons), followed by Hawaii (8.4), New York City (6.7), California (5.3), the District of Columbia (5.1), and Texas (3.9) (Table 28). Among the U.S. territories, incidence rates ranged from 0.0 (U.S. Virgin Islands) to 90.4 (Commonwealth of the Northern Mariana Islands). For the three freely associated states, reported incidence rates were 54.0 (Federated States of Micronesia), 65.5 (Republic of the Marshall Islands), and 94.4 (Republic of Palau) (Table 28).
Origin of Birth
Demographic characteristics of TB patients remain similar to previous years, with the majority of reported TB cases occurring among non-U.S.–born persons (6,335 cases; 70.2%); however, the incidence rate among non-U.S.–born persons continues to decrease, with the 2018 rate (14.3 cases per 100,000 non-U.S.–born persons) representing the lowest rate on record (Table 5). TB cases among U.S.-born persons remained virtually level from 2017 to 2018 at 2,666 cases (29.5%) and 1.0 cases per 100,000 U.S.-born persons (Table 5).
The most common countries of birth among non-U.S.–born TB patients remained like previous years, with Mexico (18.9%) the most frequently reported country of birth, followed by the Philippines (12.3%), India (9.9%), Vietnam (8.0%), and China (6.0%) (Table 6A). In a new report table this year, U.S. TB incidence rates by country of birth are presented (Table 6B). Those birth countries with the highest U.S. incidence rates were the Republic of the Marshall Islands (160.4 cases per 100,000 Marshallese), followed by the Republic of the Congo (156.8 cases per 100,000 Congolese), Somalia (100.4 cases per 100,000 Somalians), and Bhutan (90.8 cases per 100,000 Bhutanese). A substantial proportion (14.6%) of cases among non-U.S.–born persons are diagnosed within 1 year of arrival in the United States, which raises potential concern that these patients might have TB disease at the time of arrival (Table 19). Countries of birth with particularly high percentages of cases diagnosed among persons who arrived in the United States <1 year before diagnosis include the Republic of the Congo (40%), the Dominican Republic (35.7%), Afghanistan (34.1%), and Honduras (34.0%) (Table 19).
Race/Ethnicity
Non-Hispanic Asians continued to represent the largest proportion of TB patients (35.3%), followed by Hispanics (29.0%), non-Hispanic blacks (19.9%), and non-Hispanic whites (11.9%) (Table 2). However, when expressed as incidence rates, non-Hispanic Native Hawaiians/other Pacific Islanders had the highest incidence rate (20.0 cases per 100,000 Native Hawaiians/other Pacific Islanders), followed by non-Hispanic Asians (17.0 cases per 100,000 Asians) (Table 2). Non-Hispanic blacks (4.4 cases per 100,000 blacks), Hispanics (4.4 cases per 100,000 Hispanics), and non-Hispanic American Indians/Alaska Natives (4.3 cases per 100,000 American Indians/Alaska Natives) were essentially tied for third place, with multiple race (0.8 cases per 100,000 persons) and non-Hispanic whites (0.5 cases per 100,000 whites) having the lowest incidence rates (Table 2). Downward trends continued among non-Hispanic Asians and non-Hispanic blacks (Table 2). Of concern, however, incidence rates remained essentially unchanged from 2017 among all other racial/ethnic groups (Table 2).
The distribution of TB patients by race/ethnicity continued to differ markedly by origin of birth (Table 3). Among U.S.-born TB patients, the largest racial/ethnic group was non-Hispanic blacks (35.7%), followed by non-Hispanic whites (30.3%), Hispanics (22.1%), and non-Hispanic Asians (5.1%) (Table 3). Approximately half of TB cases reported among non-U.S.–born persons occurred among non-Hispanic Asians (48.1%), followed by Hispanics (31.9%), non-Hispanic blacks (13.4%), and non-Hispanic whites (4.2%) (Table 3).
Age
Distribution of TB patients by age group remains similar to past years, with a plurality of cases occurring among persons aged 25–44 years (30.2%), followed closely by persons aged 45–64 years (29.9%) and persons aged ≥65 years (26.1%) (Table 4). In contrast, only 13.8% of reported TB cases occurred among children and young adults aged <25 years (Table 4). Incidence rates by age group did not follow the proportionate distribution, however. The oldest age group (≥65 years) had the highest incidence rate overall (4.5 cases per 100,000 persons), and the incidence rate generally decreased with decreasing age. However, the youngest group (0–4 years) had an incidence rate approximately twice that of the second youngest group (5–14 years) (Table 20). This observation might be attributable to cohorts with increased risk for TB exposure and infection, compared with the present, moving through time into older age groups. Of note, among non-Hispanic Native Hawaiians/other Pacific Islanders, the incidence rate for persons aged ≥65 years is lower than the next 2 younger age groups (24.1 cases per 100,000 persons aged 45–64 years and 21.7 cases per 100,000 persons aged 25–44 years) (Table 20).
Sex
Men continued to represent the majority (61.2%) of TB patients overall, although the male:female ratio is close to 1 among children aged ≤14 years. This might indicate that the factors that cause men to be disproportionately represented among TB patients might become more influential in adulthood (Table 20). The overrepresentation of men also was somewhat greater among U.S.-born persons (64.5% male), compared with non-U.S.–born persons (59.9% male) (Tables 21 and 22).
Case Verification Criteria
The majority of U.S. TB cases continued to be verified through positive culture (78.2%), with other laboratory-confirmation methods (i.e., nucleic acid amplification or smear microscopy) only representing a limited proportion of verified cases (3.5%). In the absence of laboratory confirmation, 13.9% of cases were confirmed by meeting the clinical criteria for a verified TB case; this percentage continued to decrease as proportionately more cases were verified through laboratory techniques (Table 7).
Site of Disease
A vast majority of U.S. TB cases had pulmonary involvement (79.6%) (Table 7). Among the 20.4% of U.S. TB cases with only extrapulmonary involvement, TB of the lymphatic system remained most common (36.9%), followed by TB of the pleura (16.9%) and TB of bones and joints (9.6%) (Table 15). TB meningitis, a particularly serious form of the disease, continued to decrease, with only 3.8% of extrapulmonary cases involving the meninges (Table 15).
Prevention
Resistance to anti-TB drugs is a serious problem both in the United States and globally. The best way to prevent drug resistance is to use the recommended combination therapy for treating TB disease and to use directly observed therapy (DOT) for ensuring adherence to the medication regimen. The most common initial drug regimen for TB disease in which the infecting organism is not known to be resistant to anti-TB drugs is a 4-drug combination of isoniazid, rifampin, pyrazinamide, and ethambutol (HRZE); however, other multidrug regimens also might be effective initial treatment, depending on the case.1 During 2018, 81.5% of all reported TB patients were started on HRZE, and an additional 11.9% of patients were started on a different 4-drug regimen (Table 10). In certain cases (e.g., when the likely drug-resistance pattern is already known), treatment with <4 drugs might be effective (Table 10). Fortunately, only 0.2% of TB patients were treated with only one drug, which can lead to drug resistance (Table 10). During 2016, the most recent year with treatment completion data available, 64.0% of cases were treated exclusively by using DOT, whereas an additional 30.3% of patients received a combination of DOT and self-administered treatment (Table 10). The majority (89.2%) of 2016 patients who were eligible to complete treatment within 1 year of diagnosis did so; an additional 6.4% of these patients completed treatment >1 year after diagnosis (Table 10).
Prevalence
Resistance to isoniazid, one of the most common anti-TB drugs, can be a precursor to multidrug-resistant (MDR; defined as resistance to at least isoniazid and rifampin) TB. During 2018, 605 isoniazid-resistant TB cases were reported in the United States, a 3% decrease from 625 cases during 2017. However, as a percentage of all TB cases, the proportion that were resistant to isoniazid has remained relatively steady at approximately 9% (Table 8). This trend was the same for both U.S.-born and non-U.S.–born populations. However, a proportionate increase occurred in isoniazid resistance among U.S.-born persons with a prior history of TB disease, and a concurrent proportionate decrease occurred in isoniazid resistance among non-U.S.–born persons with a prior history of TB disease (Table 8). These proportionate changes are based on relatively small shifts in case counts; therefore, additional years of data will be needed to determine if these changes represent a new trend. Volatility associated with small case counts is also a concern for MDR TB cases reported in the United States. During 2018, a previous notable increase in MDR TB cases (from 97 cases during 2016 to 128 cases during 2017) was reversed, with 98 MDR TB cases being reported for 2018 (Table 9). The percentage of all cases occurring among persons with no previous history of TB disease that was MDR TB (i.e., primary MDR TB) has remained steady for the past several years at approximately 1% (Table 9).
Comorbid Conditions
Coinfection with HIV is a major risk factor for progression of latent TB infection (LTBI) to TB disease. Starting with this edition of the report, we are limiting HIV coinfection trend data (Table 11) to 2011–2018, because HIV status data before 2011 had <90% completeness and data were most likely not missing at random, which can result in overestimation of HIV coinfection. Among 2018 patients who were alive at diagnosis, HIV status was known for 87.9%, and 5.1% of persons with known HIV status were coinfected with HIV (Table 11). Among TB cases diagnosed among persons aged 25–44 years, 92.4% had known HIV status, and 8.3% of these persons were HIV-positive (Table 11).
Among reported risk factors for TB, diabetes mellitus (19.8%) was most commonly reported, followed by having been a close contact of a person with infectious TB (7.6%) or an immunocompromising condition other than HIV (7.3%) (Table 35). Having been a contact of a person with infectious TB was proportionately more common among U.S.-born persons (15.9%), compared with non-U.S.–born persons (4.2%) (Table 23). Diabetes mellitus, however, was proportionately more common among non-U.S.–born persons (22.0%), compared with U.S.-born persons (14.6%) (Table 23). Although treatment with tumor necrosis factor-alpha (TNF-α) inhibitors is a well-established risk factor for progression of latent TB infection (LTBI) to TB disease, only 0.8% of all 2018 TB patients reported this risk factor (Table 35). This might be attributable to health care providers being more aware of the risk and therefore likely to order testing for TB infection and to prescribe LTBI treatment for patients identified as being infected before prescribing TNF-α medications.
Social and Behavioral Characteristics
During 2018, 3.6% of TB cases were diagnosed among residents of correctional facilities who were aged ≥15 years. Correctional facilities include federal prisons (11.9% of TB cases diagnosed among residents of correctional facilities), state prisons (17.4%), local jails (25.8%), juvenile facilities (2.3%), and other facilities (40.6%) (Table 37). Other correctional facilities include detention facilities operated by U.S. Immigration and Customs Enforcement (ICE). Of the TB cases diagnosed in correctional facilities, 44.9% occurred among persons who were in ICE custody, either at an ICE-operated facility or another correctional facility (Table 37). Another setting posing potential increased risk for TB is long-term–care facilities. Overall, only 1.7% of 2018 TB cases were diagnosed among long-term–care facility residents aged ≥15 years; however, higher proportions were reported among certain reporting areas, including South Dakota (8.3%), Oklahoma (4.3%), and Wisconsin (4.3%) (Table 39).
Persons experiencing homelessness are also at increased risk of TB. Nationally, 4.3% of all TB cases were diagnosed among persons aged ≥15 years experiencing homelessness (Table 38). This proportion varied by reporting area, with the highest percentages of homelessness among TB patients being reported by Maine (33.3%), New Mexico (16.2%), and Arkansas (14.1%) (Table 38).
Substance misuse is also a TB risk factor. Overall, 1.3% of all 2018 TB patients aged ≥15 years reported injection-drug use (IDU) during the year preceding diagnosis (Table 40). Reported use of noninjection drugs among patients aged ≥15 years was higher (6.8%) than IDU, as was excessive use of alcohol (9.3%) (Tables 41 and 42).
Notably, 21.8% of all 2018 TB patients aged ≥15 years reported being unemployed; an additional 33.7% of TB patients were not seeking employment or were retired (Table 43).
Successful therapy completion for TB patients is a major performance indicator for TB programs. Among patients during 2016 who were alive at diagnosis, 87.2% had completed TB treatment successfully (Table 12). However, 6.7% of patients died before they could complete treatment; 1.3% of patients were lost to follow-up before completing treatment; and 4.5% did not complete treatment for other or unknown/unreported reasons (Table 12). Of note, only 29 patients (0.3%) had to permanently stop TB treatment because of an adverse treatment event (Table 12). Among TB cases diagnosed during 2016, a total of 833 (9.0%) patients died, with 315 (37.8%) of those deaths attributed to TB disease or TB treatment (Table 13). Of the 833 deaths, 206 (24.7%) were dead at the time of TB diagnosis; 33.5% of those deaths were attributed to TB (Table 13). The remaining 627 (75.3%) deaths occurred after diagnosis; 39.2% of these deaths were attributed to TB (Table 13).
Time required for a patient’s positive sputum culture to convert to negative is a key indicator of treatment effectiveness. Among 5,115 cases during 2016 with positive sputum cultures, 4,215 (82.4%) had documented sputum culture conversion to negative (Table 14). Among the 809 (15.8%) cases for which sputum culture conversion was undocumented, the most common reason was that the patient had died (30.3%) before sputum culture conversion; however, a substantial proportion of these cases (36.8%) did not have a known reason reported for not having documented sputum culture conversion (Table 14).
Genotype Clusters
TB genotyping is a laboratory-based analysis used to characterize a small portion of the genetic material of bacteria belonging to the Mycobacterium tuberculosis complex. TB genotype clusters are defined in this report as ≥2 cases with matching genotypes in the same county or county-equivalent area during a 3-year period. CDC identified clusters among 19.3% of genotyped cases during 2016–2018 (Table 24). Clusters are classified into alert levels on the basis of a log-likelihood ratio (LLR) calculation; clusters with an LLR of 5–<10 are classified as a medium alert level, and clusters with an LLR ≥10 are classified as a high alert level. Clustered cases were often part of medium- (22.6%) or high-level alerts (18.6%). At the cluster level, 392 (29.1%) of 1,349 clusters identified nationally were either medium- or high-level alerts. The proportion of clusters that alerted is similar to previous years. Although not all clustered cases result from recent transmission, clusters with high or medium alert levels might represent TB outbreaks.2
Estimates of Recent Transmission
CDC attributes a TB case to recent transmission if a plausible source case with a matching genotype is identified within a specified geographic radius and has a diagnosis ≤2 years before the case (see Estimates of Recent Transmission, 2017–2018 section for more details).3 Nationally, CDC attributed 1,712 (12.6%) of 13,601 genotyped cases reported during 2017–2018 to recent transmission (Table 57; Figure 1). CDC has provided national estimates of recent transmission and extensive recent transmission throughout a 2-year period since the publication of Reported Tuberculosis in the United States, 2016. The number and proportion of cases attributed to recent transmission have declined. By comparison, estimates were 1,894 (13.7%) for 2015–2016.
The 1,712 cases were further attributed to extensive recent transmission if a plausible chain of transmission of ≥6 cases, including the plausible source case and ≥4 other cases ≤3 years before a given case in the chain, was identified (see Estimates of Recent Transmission, 2017–2018 section).4 Nationally, 589 (4.3%) genotyped cases were attributed to extensive recent transmission, although estimates vary considerably at the state (Table 57) and county (Table 58) levels. Forty-seven counties or county-equivalent areas had >5% of their genotyped cases attributed to extensive recent transmission during 2017–2018 (Table 58; Figure 2). Temporally, the number and proportion of genotyped cases attributed to extensive recent transmission have also decreased. By comparison, estimates were 690 (5.0%) for 2015–2016.
The majority (56.6%) of genotyped cases attributed to recent transmission were among U.S.-born persons (Table 59). Greater proportions of cases attributed to recent transmission and extensive recent transmission were identified among certain racial groups and among persons with social and behavioral risk factors, compared with national averages.
Although TB in the United States continues to decrease each year, the rate of decrease remains low. This low rate of decrease is most evident among the U.S.-born population, although most of reported TB cases continue to occur among non-U.S.–born persons. Half of all cases continue to be diagnosed in just 4 states. Notable new information included in this year’s report includes country-of-birth–specific incidence rates for TB reported among persons born outside the United States: these new data reveal that certain populations have particularly high rates of TB, even though the size of those populations in the United States is relatively small. Incidence rates among non-Hispanic Asians and non-Hispanic blacks continue to decrease, but the lack of concurrent decrease among other racial/ethnic groups is concerning and warrants further investigation. Continued reductions in TB/HIV coinfection are encouraging, although the continued high prevalence of diabetes among TB patients (more than double the national diabetes prevalence of 9.4%) remains concerning because diabetes can complicate TB treatment.1,5
- Nahid P, Dorman SE, Alipanah N, et al. Official American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America clinical practice guidelines: treatment of drug-susceptible tuberculosis. Clin Infect Dis. 2016;63(7):e147-e195.
- Ghosh S, Moonan PK, Cowan L, Grant J, Kammerer S, Navin TR. Tuberculosis genotyping information management system: enhancing tuberculosis surveillance in the United States. Infect Genet Evol. 2012;12(4):782-788.
- 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.
- Yuen CM, Kammerer JS, Marks K, Navin TR, France AM. Recent transmission of tuberculosis—United States, 2011–2014. PLoS One. 2016;11(4): e0153728.
- Centers for Disease Control and Prevention (CDC). National Diabetes Statistics Report, 2017. Atlanta, GA: US Department of Health and Human Services, CDC; 2017. https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf.