Up to 50% of persons with HIV and a diagnosis of tuberculosis (TB) in Thailand die during TB treatment. In a prospective observational study, a team of physicians ascribed the cause of death after reviewing verbal autopsies (interviews of family members about events preceding death), laboratory data, and medical records. Of 849 HIVinfected TB patients enrolled, 142 (17%) died. The cause of death was TB for 38 (27%), including 6 with multidrugresistant TB and 20 with disseminated TB; an HIV-associated condition other than TB for 50 (35%); and a condition unrelated to TB or HIV for 22 (15%). Twenty-three patients (16%) were judged not to have had TB at all. Death from all causes except those unrelated to TB or HIV was less common in persons receiving antiretroviral therapy (ART). In addition to increasing the use of ART, death rates may be reduced through expanded use of modern TB diagnostic techniques.

Among HIV-infected patients with TB, extrapulmonary disease occurred in 40% of the patients, particularly in those with advanced immune suppression. Death during TB treatment was common, but the risk of death was reduced in patients who took co-trimoxazole, fluconazole, and antiretroviral therapy.

BackgroundIn Southeast Asia, HIV-infected patients frequently die during TB treatment. Many physicians are reluctant to treat HIV-infected TB patients with anti-retroviral therapy (ART) and have questions about the added value of opportunistic infection prophylaxis to ART, the optimum ART regimen, and the benefit of initiating ART early during TB treatment.MethodsWe conducted a multi-center observational study of HIV-infected patients newly diagnosed with TB in Thailand. Clinical data was collected from the beginning to the end of TB treatment. We conducted multivariable proportional hazards analysis to identify factors associated with death.ResultsOf 667 HIV-infected TB patients enrolled, 450 (68%) were smear and/or culture positive. Death during TB treatment occurred in 112 (17%). In proportional hazards analysis, factors strongly associated with reduced risk of death were ART use (Hazard Ratio [HR] 0.16; 95% confidence interval [CI] 0.07–0.36), fluconazole use (HR 0.34; CI 0.18–0.64), and co-trimoxazole use (HR 0.41; CI 0.20–0.83). Among 126 patients that initiated ART after TB diagnosis, the risk of death increased the longer that ART was delayed during TB treatment. Efavirenz- and nevirapine-containing ART regimens were associated with similar rates of adverse events and death.ConclusionAmong HIV-infected patients living in Thailand, the single most important determinant of survival during TB treatment was use of ART. Controlled clinical trials are needed to confirm our findings that early ART initiation improves survival and that the choice of non-nucleoside reverse transcriptase inhibitor does not.

IntroductionDisease-related stigma and knowledge are believed to be associated with patients' willingness to seek treatment and adherence to treatment. HIV-associated tuberculosis (TB) presents unique challenges, because TB and HIV are both medically complex and stigmatizing diseases. In Thailand, we assessed knowledge and beliefs about these diseases among HIV-infected TB patients.MethodsWe prospectively interviewed and examined HIV-infected TB patients from three provinces and one national referral hospital in Thailand from 2005–2006. At the beginning of TB treatment, we asked patients standardized questions about TB stigma, TB knowledge, and HIV knowledge. Responses were grouped into scores; scores equal to or greater than the median score of study population were considered high. Multiple logistic regression analysis was used to identify factors associated with scores.ResultsOf 769 patients enrolled, 500 (65%) reported high TB stigma, 177 (23%) low TB knowledge, and 379 (49%) low HIV knowledge. Patients reporting high TB stigma were more likely to have taken antibiotics before TB treatment, to have first visited a traditional healer or private provider, to not know that monogamy can reduce the risk of acquiring HIV infection, and to have been hospitalized at enrollment. Patients with low TB knowledge were more likely to have severe TB disease, to be hospitalized at enrollment, to be treated at the national infectious diseases referral hospital, and to have low HIV knowledge. Patients with low HIV knowledge were more likely to know a TB patient and to have low TB knowledge.DiscussionWe found that stigma and low disease-specific knowledge were common among HIV-infected TB patients and associated with similar factors. Further research is needed to determine whether reducing stigma and increasing TB and HIV knowledge among the general community and patients reduces diagnostic delay and improves patient outcomes.

Background: Of the 9.2 million new TB cases occurring each year, about 10% are in children. Because childhood TB is usually non-infectious and non-fatal, national programs do not prioritize childhood TB diagnosis and treatment. We reviewed data from a demonstration project to learn more about the epidemiology of childhood TB in Thailand.