Missed Opportunities for Prevention of Tuberculosis Among Persons With HIV Infection
Selected Locations, United States, 1996-1997
August 4, 2000
Public health contact investigations are conducted to find persons who have been exposed to patients with active tuberculosis (TB) and to evaluate and treat those contacts for TB infection and active TB. Persons in close (i.e., prolonged, frequent, or intense) contact with patients with active TB are at high risk for TB infection. The risk for TB infection is increased greatly if the close contact is infected with human immunodeficiency virus (HIV) (1,2). Isoniazid (INH) treatment for latent TB infection (LTBI) reduces the risk for developing active TB by 41%-92% (1). This study examined the clinic records of TB programs to determine whether these programs used recommended practices to manage HIV-positive persons exposed to TB (3-8). The study suggests TB programs need to review their contact investigation policies, procedures, and outcomes to reduce missed opportunities for preventing active TB among HIV-positive close contacts.
Study investigators collected data during June 1998-January 1999 site visits. Eleven U.S. urban areas were selected by the highest number of contacts completing LTBI treatment. After case reports were linked to personal identifiers, study staff reviewed the clinic records for 6225 close contacts to 1080 sputum-smear -- positive TB patients reported to CDC during July 1996-June 1997.
Of the 6225 close contacts, HIV status was unknown for 5415 (87%). Of the 810 close contacts with known HIV status, 109 (13%) were HIV-infected, of whom 79 (72%) received a chest radiograph; 14 (13%) had TB symptoms (e.g., cough, night sweats, and weight loss); 90 (83%) received an initial tuberculin skin test (TST); and nine (8%) did not receive a chest radiograph or an initial TST. Forty (53%) of 75 TST-negative contacts did not receive follow-up TSTs; 21 (28%) received neither a follow-up TST nor a chest radiograph. Fourteen (13%) of 109 HIV-positive contacts were identified as having active TB compared with 120 (2%) of 6116 HIV-negative contacts or contacts with unknown HIV status. HIV-infected close contacts were less likely to be TST-positive than HIV-negative contacts or contacts with unknown HIV status (14% and 36%, respectively).
Among 95 HIV-infected contacts without active TB, 11 (92%) of 12 TST-positive contacts were placed on LTBI treatment compared with 19 (23%) of 83 TST-negative or TST-unknown contacts. A median of 50 days passed before starting an HIV-positive contact on LTBI treatment compared with 33 days for HIV-negative contacts or contacts with unknown HIV status. TB programs employing public health nurses to conduct investigations placed 11 (92%) of 12 TST-negative or TST-unknown contacts on LTBI treatment compared with eight (11%) of 71 at programs that employ TB outreach workers.
Of the 30 HIV-positive contacts started on LTBI treatment, approximately half (14) completed treatment. Directly observed treatment (DOT) for LTBI was given to three HIV-positive contacts; two completed treatment. During the course of LTBI treatment, 10 HIV-infected contacts had interruptions of >1 month (when treatment was self-administered) or >2 weeks (when placed on DOT); three of the 10 completed treatment. Of 16 HIV-positive close contacts who did not complete treatment, six (38%) refused or were unwilling to continue treatment, two (12%) were lost to follow-up, one (6%) had alcoholism, one (6%) could not tolerate medication, and six (38%) had undocumented reasons.
Reported by: TB programs in Los Angeles County, San Diego County, San Francisco, and Santa Clara County, California; Fulton County, Georgia; Chicago, Illinois; Newark, New Jersey; New York, New York; Shelby County, Tennessee; Houston, Texas; and King County, Washington. Prevention Effectiveness Section, Research and Evaluation Br, Div of TB Elimination, National Center for HIV, STD, and TB Prevention, CDC.
The study showed that few close contacts were assessed for HIV and that one quarter of those known to be HIV-infected were not screened completely for TB. Of eligible HIV-positive contacts, a third started and a sixth completed LTBI treatment. Because HIV positivity alters the approach to TB screening and the use of LTBI treatment, early knowledge by the health-care provider of a close contact's HIV status is essential. Active TB is curable and can be prevented in HIV-positive contacts when health-care providers know a close contact's HIV status and follow CDC guidelines for TB screening and treatment and facilitate adherence to TB treatment.
Health-care providers should assess all close contacts for HIV infection by asking about their serostatus and offering voluntary HIV counseling and testing when the status is unknown (8). TB staff should be trained to offer HIV counseling and testing to close contacts or should collaborate with HIV programs to offer these services. The use of rapid diagnostic tests may facilitate timely assessment of HIV status. All HIV-positive close contacts should be evaluated for active TB by medical history, symptom screening, and chest radiograph, and those with an abnormal chest radiograph or symptoms should receive a sputum examination (5). HIV-positive close contacts should receive an initial TST regardless of previous TST results (5); those with initial TST-negative reactions should receive a follow-up TST 10-12 weeks after last exposure to the patient with active TB (4). As soon as active TB is excluded, LTBI treatment should begin for all HIV-infected close contacts regardless of age, TST results, or history of previous LTBI treatment (5). Most HIV-positive close contacts should complete a full course of LTBI treatment (9). Because the HIV-positive population is less likely to react to TST and more likely to have atypical chest radiographs, health-care providers need to be diligent in diagnosing TB infection and active TB. Two treatment regimens, 9 months of INH (to be taken with pyridoxine to prevent peripheral neuropathy) or 2 months of daily rifampin (or rifabutin for those taking protease inhibitors or certain nonnucleoside reverse transcriptase inhibitors) and pyrazinamide, are preferred for the treatment of HIV-positive persons with LTBI (10). The use of 2-month LTBI regimens for HIV-infected adults may facilitate treatment implementation and increase completion rates (10). However, INH is the only recommended regimen for children and pregnant women (5).
The findings in this study are subject to at least three limitations. First, because the study relied on existing clinic records, documentation of HIV status often was incomplete or nonexistent. Laws restricting the recording of HIV status in databases may have affected such documentation. Second, the timing of health-care provider knowledge of HIV status and chest radiograph results was unknown because these dates were not collected and often were not recorded. Third, this study was designed to represent urban TB programs not rural programs or programs not using LTBI treatment.
These findings indicate a need for better incorporation of HIV assessment into contact investigation procedures and improved coordination between local TB and HIV programs to facilitate voluntary HIV counseling, testing, and follow-up for HIV-infected close contacts. Health-care providers and HIV-infected persons should be aware of optimal management of close contacts and of the benefits of prompt and well-supervised LTBI treatment to prevent active TB.
This article was provided by U.S. Centers for Disease Control and Prevention. Visit the CDC's website to find out more about their activities, publications and services.