Mortality Among Patients with Tuberculosis and Associations with HIV Status --- United States, 1993--2008

Please note: An erratum has been published for this article. To view the erratum, please click here.

Worldwide, tuberculosis (TB) incidence increased from 125 cases per 100,000 population in 1990 to 142 cases per 100,000 population in 2004, primarily because of the human immunodeficiency virus (HIV) epidemic (1). Persons with HIV are at increased risk for TB disease, and those with TB have a high risk for death. This is documented most clearly in resource-limited settings, where limited access to antiretroviral therapy (ART) and other health-care services contribute to the elevated mortality (1). The impact of HIV on patients with TB is less clear in resource-rich nations such as the United States. To understand the impact of HIV on the risk for death during TB treatment in the United States, data were analyzed for all culture-positive patients with TB from 1993 to 2008, and the proportion that died was determined and stratified by HIV test result. Mortality data were restricted to patients reported before 2007. The proportion of all patients with TB who died during TB treatment decreased from 2,445 of 13,629 (18%) in 1993 to 682 of 7,578 (9%) in 2006. Among patients with TB and HIV, 950 of 2,337 (41%) died during treatment in 1993; this proportion declined to 131 of 663 (20%) in 2006. The proportion of patients with TB and HIV who received their TB diagnosis postmortem dropped from 191 of 2,927 (7%) in 1993 to 32 of 768 (4%) in 2006; 624 of 10,468 (6%) persons with TB and unknown HIV status received their TB diagnosis postmortem in 1993, and this proportion did not decline. Further reductions in mortality can be achieved by enhanced TB/HIV program collaboration and service integration.

Since 1993, all cases of TB diagnosed in the United States have been reported to CDC and entered into the National TB Surveillance System (NTSS), a comprehensive database that contains demographic, clinical, and outcome data. All culture-confirmed cases of TB were reviewed by CDC to determine 1) the proportion of cases diagnosed postmortem and 2) the proportion of cases in persons who were alive at diagnosis and who died during TB treatment; results then were stratified by HIV status (i.e., HIV infected, HIV uninfected, or HIV status unknown). The HIV-unknown category included patients with indeterminate or unknown results as well as patients who were not offered or refused testing. Rates of HIV test reporting during 2007--2008 were stratified by selected demographic characteristics. Mortality analyses were restricted to patients reported before 2007 (to allow 2 years for treatment outcomes to be reported) and to those whose outcomes were known (excluding patients who moved, were lost to follow-up, were uncooperative with treatment, or whose outcome was missing or listed as other). Because California reports HIV test results only for patients who receive diagnoses of acquired immunodeficiency syndrome (AIDS), and does not report the HIV status of those who test negative, all data from California were excluded.

The proportion of patients with TB who had documented HIV test results increased substantially, from 6,015 of 16,507 (36%) in 1993 to 6,234 of 7,872 (79%) in 2008 (Figure 1). The proportion of patients with TB who had a known outcome and were alive at diagnosis but died during TB treatment decreased from 2,445 of 13,629 (18%) in 1993 to 682 of 7,578 (9%) in 2006 (Figure 2). Among patients with TB and HIV, 950 of 2,337 (41%) died during treatment in 1993; this proportion declined to 299 of 1,393 (21%) in 1997 and later to 131 of 663 (20%) in 2006 (Figure 2). By contrast, the proportion of TB patients without HIV who died during treatment decreased from 213 of 2,705 (8%) in 1993 to 281 of 5,315 (5%) in 2006. For patients with unknown HIV status, 1,282 of 8,587 (15%) died in 1993, with no decrease in proportion observed over the study period (Figure 2). Among patients with HIV who received diagnoses of TB, 191 of 2,927 (7%) received their TB diagnosis postmortem in 1993, which decreased to 32 of 768 (4%) in 2006. Among culture-confirmed cases of TB that occurred in persons who were HIV uninfected, 53 of 3,080 (2%) received their TB diagnosis postmortem in 1993, a proportion that decreased to 31 of 5,762 (1%) in 2006. Of those with unknown HIV status, 624 of 10,468 (6%) received their TB diagnosis postmortem; that proportion did not decline.

Among those with known HIV status, 2,932 of 6,015 (49%) patients with TB had HIV infection in 1993 and accounted for 950 of 1,163 (82%) deaths during treatment and 191 of 244 (78%) patients who received a TB diagnosis postmortem. In 2006, 769 of 6,533 (12%) patients with reported status had HIV, but accounted for 131 of 412 (32%) and 32 of 63 (51%) of those who died during treatment and those who received a TB diagnosis postmortem, respectively.

HIV testing during 2007--2008 was lower in certain demographic groups than the overall sample, notably, 102 of 201 (51%) patients aged ≤4 years, 95 of 144 (66%) patients aged 5--14 years, 1,824 of 3,253 (56%) patients aged ≥65 years, and 2,154 of 3,056 (70%) non-Hispanic white patients had HIV test results reported (Table).

Reported by

S Shah, MD, Dept of Medicine, Albert Einstein College of Medicine, Bronx, New York. K Cain, MD, S Marks, MPH, MA, Div of TB Elimination, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention; J Cavanaugh, MD, EIS Officer, CDC.

Editorial Note

This analysis demonstrates a substantial reduction in case-fatality rate among patients with TB in the United States from 1993 to 2006, a decline that occurred almost exclusively in persons with HIV and corresponded to an increase in reported HIV test results and broader availability of highly active ART. In 2008, however, 21% of patients with TB still had unknown HIV status, and this proportion was even higher in certain demographic groups. This is unacceptable given that knowledge of HIV status is essential for appropriate treatment and that current guidelines recommend HIV testing for all patients with TB in the United States (2). A larger proportion of patients with TB were tested for HIV in some countries with a much higher burden of HIV and TB than the United States and far fewer resources, such as Kenya.*

In resource-limited settings, studies have demonstrated that without concurrent treatment of HIV, up to 50% of persons with HIV who develop TB will die during the 6- to 8-month course of TB treatment, many of them in the first 2 to 3 months (3,4). When patients with TB and HIV are treated with ART and prophylactic therapy for opportunistic infections as recommended (5), the proportion of patients who die during TB treatment can be reduced to less than 10% (4).

Recent research from New York City showed acceptable TB treatment success in patients with TB and HIV only when they received ART and directly observed therapy (6), underscoring the critical importance of these two treatment modalities. In this analysis, mortality declined steeply among patients with TB and HIV after highly active ART became widely available during 1995--1996. Data such as ART use, CD4 count, and specific cause of death are not reported to NTSS, and the impact of each of these could not be directly assessed; however, highly active ART use likely was an important factor in reducing mortality and, of course, can only be provided to those whose HIV infection is known.

A substantial proportion of culture-confirmed TB diagnoses among persons with either documented HIV infection or unknown HIV status were made postmortem. Research has demonstrated that when patients with TB and HIV die from TB, it is often because diagnosis is delayed (7), and these deaths might have been prevented if TB disease had been diagnosed and treated earlier. Screening persons with HIV for TB at regular intervals in accordance with current recommendations (8) allows for earlier diagnosis and treatment of TB and has been shown to lower mortality (9).

Treatment of latent TB infection and use of ART have been shown to substantially reduce the risk for TB disease in persons with HIV (10). Increasing HIV testing of the general population will help identify those for whom early ART initiation and treatment of latent TB infection might prevent TB before it develops (10).

The findings in this report are subject to at least two limitations. First, California accounts for approximately 20% of the patients with TB in the United States, and excluding those data might affect generalizability if those patients differed from other patients with TB in key ways. Second, outcome data were missing for 10% of all patients included in this analysis, and NTSS does not document cause of death for those who died; knowledge of mortality concerning these patients is limited.

Much progress has been made in reducing mortality among patients with TB and HIV in the United States since 1993. Further reductions in mortality can be achieved by enhanced TB and HIV program collaboration and service integration, including 1) providing HIV testing to all patients with TB; 2) screening all persons with HIV for TB disease and infection regularly; and 3) providing early and appropriate TB and HIV treatment to all patients with TB and HIV.† States and local health-care organizations should analyze their own data to determine how to best target interventions aimed at increasing HIV testing. In addition, studying the specific causes of death in patients with TB and HIV would facilitate development of additional measures to decrease the risk for death.

References

1. World Health Organization. Global tuberculosis control 2010. Geneva, Switzerland: World Health Organization; 2010. Available at http://www.who.int/tb/publications/global_report/2010. Accessed on November 19, 2010.
2. CDC. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR 2006;55(No. RR-14).
3. Mukadi YD, Maher D, Harries A. Tuberculosis case fatality rates in high HIV prevalence populations in sub-Saharan Africa. AIDS 2001;15:143--52.
4. Manosuthi W, Chottanapand S, Thongyen S, Chaovavanich A, Sungkanuparph S. Survival rate and risk factors of mortality among HIV/tuberculosis-coinfected patients with and without antiretroviral therapy. J Acquir Immune Defic Syndr 2006;43:42--6.
5. Sterling TR, Pham PA, Chaisson RE. HIV infection-related tuberculosis: clinical manifestations and treatment. Clin Infect Dis 2010;50(Suppl 3):S223--30.
6. King L, Munsiff SS, Ahuja SD. Achieving international targets for tuberculosis treatments success among HIV-positive patients in New York City. Int J Tuberc Lung Dis 2010;14:1613--20.
7. Cain KP, Anekthananon T, Burapat C. Causes of death in HIV-infected persons who have tuberculosis, Thailand. Emerg Infect Dis 2009;15:258--64.
8. CDC. Guidelines for prevention and treatment of opportunistic infections in HIV-infected adults and adolescents: recommendations from CDC, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America. MMWR 2009;58(No. RR-4).
9. Reid A, Scano F, Getahun H. Towards universal access to HIV prevention, treatment, care, and support: the role of tuberculosis/HIV collaboration. Lancet Infect Dis 2006;6:483--95.
10. Akolo C, Adetifa I, Shepperd S, Volmink J. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database Syst Rev 2010;Jan 20:CD000171.

All MMWR HTML versions of articles are electronic conversions from typeset documents. This conversion might result in character translation or format errors in the HTML version. Users are referred to the electronic PDF version (http://www.cdc.gov/mmwr) and/or the original MMWR paper copy for printable versions of official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

**Questions or messages regarding errors in formatting should be addressed to [email protected].