Value of chest radiography in a tuberculosis prevention programme for HIV-infected people, Botswana.

To exclude tuberculosis, WHO/UNAIDS recommends considering medical history, symptom screen, and chest radiograph before starting tuberculosis prevention in people infected with HIV. The value of a chest radiograph for this purpose is unknown. We prospectively assessed 935 HIV-infected outpatients seeking isoniazid preventive therapy. Of 935 patients, 692 (74%) had no signs or symptoms of tuberculosis. Of these 692, 123 (18%) were lost during the chest radiograph process, and one (0.2%) of the remaining 563 was diagnosed with tuberculosis on the basis of the chest radiograph. A screening chest radiograph should not be required routinely for asymptomatic people taking isoniazid as preventive treatment in settings able to screen for signs and symptoms of tuberculosis.

Persistent low immunization coverage among inner-city preschool children despite access to free vaccine.

OBJECTIVE: To compare vaccination coverage among children 19 to 35 months of age from public housing developments where a free vaccine outreach program was in place with children residing elsewhere in the city. DESIGN: A household survey using a multistage cluster sampling method to compare community areas which accounted for 80% of measles cases during 1989 (high-risk stratum), areas which accounted for the remaining 20% of cases (low-risk stratum), and public housing developments (public housing stratum) having free, on-site vaccination services. SETTING: Inner-city Chicago households, April to May 1994. OUTCOME VARIABLES: Antigen-specific and series-specific coverage based on written records. RESULTS: Based on evaluation of 1244 children, citywide coverage for four doses of diphtheria-tetanus-pertussis vaccine, three doses of polio vaccine, and one dose of measles-containing vaccine (4:3:1) was 47% [95% confidence interval (CI), 40% to 55%]. Coverage was significantly lower among children residing in public housing (23%; 95% CI, 18% to 28%) compared with those residing in high-risk strata (45%; 95% CI, 38% to 52%) and low-risk strata (51%; 95% CI, 43% to 60%). Compared with white children (53%), coverage for the 4:3:1 series was lower among African-American children in public housing (29%) or outside public housing (36%). Moreover, 11% (95% CI, 8% to 14%) of children residing in public housing had never received any immunizations. CONCLUSIONS: African-American children throughout Chicago, particularly in public housing, remain at increased risk for vaccine-preventable diseases and should be targeted further for vaccination services. Vaccination coverage remains low several years after a major outbreak of measles and implementation of a free vaccine outreach program. Cluster surveys may be useful for monitoring vaccination coverage in high-risk urban settings.

Large outbreak of pertussis among young children in Chicago, 1993: investigation of potential contributing factors and estimation of vaccine effectiveness.

BACKGROUND: An outbreak of pertussis from July, 1993, to April, 1994, in Chicago was investigated to identify potential contributing factors. METHODS: Surveillance was enhanced to identify cases. Information from a vaccination coverage survey was used to define a retrospective cohort to estimate vaccine effectiveness of three or more doses of pertussis vaccine. RESULTS: The median age of 218 reported cases was 8 months, 46% had Hispanic surnames and cases were clustered geographically. Vaccination status was known for 173 of 191 (91%) children younger than 6 years of age. Of these 173, 90 (52%) were younger than 7 months, and 35 (20%) children at least 7 months of age had received fewer than 3 doses of pertussis vaccine. Pertussis vaccine effectiveness was 76% (95% confidence interval, 29 to 92). CONCLUSIONS: The limited ability of the current pertussis vaccination schedule to protect young infants accounted for 52% of cases, primary vaccine failure accounted for 28% of cases and failure to vaccinate children on time accounted for 20% of cases in young children. Low vaccine effectiveness did not appear to be a contributing factor.

Reduction in ventricular fibrillation using calcium-enriched Renografin 76.

Renografin 76 has been shown to have significant calcium-binding properties in vivo and in vitro. Two contrast media solutions were compared by selective injection into the right coronary artery of dogs. One set of injections was made with commercially available Renografin 76 (referred to as stock Renografin) and another set of injections was made with Renografin 76 to which had been added 24 mMols/liter of calcium chloride dihydrate. Ventricular fibrillation occurred significantly more often with the stock Renografin 76. The nature of the calcium binding and its effects are discussed.

Nosocomial transmission of human immunodeficiency virus and subsequent transmission of multidrug-resistant tuberculosis in a healthcare worker.

A phlebotomist with nosocomially acquired human immunodeficiency virus infection developed tuberculosis 10 months after exposure to multidrug-resistant Mycobacterium tuberculosis during a nosocomial outbreak. Healthcare workers with immunosuppression are at increased risk of tuberculosis if infected and, if exposed, should be considered for preventive therapy regardless of tuberculin skin-test status.

Validation of the surveillance system for tuberculosis in Botswana.

SETTING: Gaborone and Francistown, Botswana, where surveillance data in the 1997 Electronic Tuberculosis (TB) Register suggest that 39% of pulmonary TB patients did not have pre-treatment sputum smear microscopy performed. OBJECTIVE: To determine the proportion of patients with reportedly missing pre-treatment sputum smear results in 1997 who had smears examined, and to identify stages in the system where results were lost. METHODS: Patients with pulmonary TB in 1997 who were missing pre-treatment sputum smear results in the Electronic TB Register were cross-matched with laboratory records; medical records were reviewed. RESULTS: Of 374 patients with pre-treatment sputum smear results missing, 224 (60%) actually had had a sputum smear examined in the laboratory. The proportion of pulmonary TB patients in Gaborone and Francistown who did not have sputum examined was therefore 16% instead of 39%. Most missing results (69%) had not been transcribed from the laboratory results onto the TB Treatment Card. Patients who had a negative smear result or who sought care at a clinic that was different from where their diagnostic evaluation had been initiated were more likely to have missing results. CONCLUSIONS: The actual performance of the Botswana National TB Programme with respect to sputum microscopy examination is much better than surveillance indicators suggest. In addition to sputum collection, proper recording of results needs reinforcement among health care workers to improve routine performance indicators.

Knowledge, attitudes and beliefs regarding tuberculosis preventive therapy for HIV-infected persons, Botswana, 1999.

SETTING: Botswana, 1999. OBJECTIVE: To measure knowledge, attitudes and beliefs about tuberculosis (TB) preventive therapy (PT) for persons infected with the human immunodeficiency virus (HIV). DESIGN: A systematic sample of adult clinic attendees, using a standardised questionnaire. RESULTS: A total of 275 patients at 38 clinics in five districts were interviewed. The majority were female (65%) and unmarried (84%). Knowing someone with TB or AIDS was common (78% and 53%, respectively). Respondents perceived a relationship between TB and HIV (80%), and the majority were willing to undergo tuberculin skin testing (92%). Of those, most were willing to undergo evaluation for active TB (98%), and to take PT, although willingness to take PT declined with proposed duration (97% 6 months, 90% 1 year, 81% lifetime, P < 0.01). Previous HIV testing was reported by 13%; those who had not undergone testing reported that they would if doctors could improve the quality (95%) or duration (93%) of life of persons with AIDS. The majority favoured receiving HIV test results on the day they were tested (60%). CONCLUSIONS: Most clinic attendees in Botswana were willing to undergo HIV testing if it were beneficial to do so, such as by receiving PT. Pilot PT projects should be initiated. Voluntary HIV counselling and testing services should consider rapid HIV testing methods.

Transmission of Mycobacterium tuberculosis among employees in a US government office, Gaborone, Botswana.

SETTING: A US government office located in Botswana where two office employees, one negative and one positive for the human immunodeficiency virus (HIV), were diagnosed with pulmonary tuberculosis (TB) in January 1998. One employee had been symptomatic with untreated laryngeal TB for 8 months. OBJECTIVE: To determine the extent of and risk factors for TB transmission in the office. METHODS: Office contacts were interviewed and a tuberculin skin test (TST) was performed. A positive TST was defined as > or = 10 mm induration for employees from countries where TB is highly endemic, and as > or = 5 mm induration for those from low prevalence counties. RESULTS: Of 79 office contacts investigated, 54/57 (94.7%) born in high TB prevalence countries had a positive TST compared with 4/22 (18.2%) from low prevalence countries (RR 5.1, 95% CI 2.1-12.7, P < 0.001). Of 20 US-born contacts, three (15%) had documented TST conversion, two of whom were co-workers of the employee with laryngeal TB. Isolates of Mycobacterium tuberculosis from the TB cases had matching DNA fingerprints. CONCLUSION: Delayed diagnosis in a setting of high TB prevalence may have contributed to transmission within a US government office located in Botswana. Transmission may have been underestimated due to the high background prevalence of tuberculous infection in the population. Recent tuberculous transmission to persons living with HIV infection may be playing an important role in the escalating TB epidemic in Africa.

Low levels of drug resistance amidst rapidly increasing tuberculosis and human immunodeficiency virus co-epidemics in Botswana.

SETTING: Botswana, southern Africa, where the tuberculosis (TB) case rate increased by 120% from 1989 to 1996 in spite of a decade of implementation of the directly observed therapy, short-course (DOTS) strategy. OBJECTIVE: To determine prevalence of and risk factors for drug-resistant tuberculosis in an epidemic setting. DESIGN: Systematic national random survey of newly diagnosed pulmonary TB and all patients with TB requiring retreatment during 1995-1996. Interviews were conducted, human immunodeficiency virus (HIV) testing was offered, and drug susceptibility testing was performed for isoniazid, rifampicin, streptomycin and ethambutol. RESULTS: Resistance to at least one drug was identified in 16 (3.7%) new cases and 18 (14.9%) retreatment cases. One (0.2%) new and seven (5.8%) retreatment cases had resistance to at least both isoniazid and rifampicin (multidrug-resistant TB). Retreatment cases with multidrug-resistant TB were significantly more likely to have worked in the mines in South Africa than were cases with fully susceptible isolates (6/7 [85.7%] versus 32/ 103 [31.1%], odds ratio 13.3, 95% confidence interval 1.5-311.0, P = 0.007). Of 240 patients tested for HIV, 117 (48.8%) were positive; prevalence was similar among new and retreatment cases, and was not a risk factor for drug resistance in either group. CONCLUSION: During the HIV and TB co-epidemics in sub-Saharan Africa, DOTS may help to control drug-resistant TB. However, the TB case rate can be expected to continue to climb in spite of the implementation of the DOTS strategy.

Tuberculin reactivity in a pediatric population with high BCG vaccination coverage.

SETTING: The tuberculin skin test (TST) is often included in diagnostic algorithms for tuberculosis (TB) in children. TST interpretation, however, may be complicated by prior Bacillus Calmette-Guerin (BCG) vaccination. We assessed the prevalence of and risk factors for positive TST reactions in children 3 to 60 months of age in Botswana, a country with high TB rates and BCG coverage of over 90%. METHODS: A multi-stage cluster survey was conducted in one rural and three urban districts. Data collected included demographic characteristics, nutritional indices, vaccination status, and prior TB exposure. Mantoux TSTs were administered and induration measured at 48-72 hours. RESULTS: Of 821 children identified, 783 had TSTs placed and read. Of the 759 children with vaccination cards, 755 (99.5%) had received BCG vaccine. Seventy-nine per cent of children had 0 mm induration, 7% had > or =10 mm induration ('positive' TST), and 2% had > or =15 mm. A positive TST was associated with reported contact with any person with active TB (odds ratio [OR] 1.9; 95% confidence interval [CI] 1.02-3.6), or a mother (OR 5.1; 95% CI 2.1-12.4) or aunt (OR 5.3; 95% CI 2.0-14.0) with active TB. TSTs > or =5 mm (but not > or =10 mm) were associated with presence of a BCG scar. Positive reactions were not associated with age, time since BCG vaccination, clinical signs or symptoms of TB, nutritional status, crowding, or recent measles or polio immunization. CONCLUSION: The TST remains useful in identifying children with tuberculous infection in this setting of high TB prevalence and extensive BCG coverage.

Cost-effectiveness of home-based care versus hospital care for chronically ill tuberculosis patients, Francistown, Botswana.

SETTING: Francistown, Botswana, 1999. OBJECTIVE: To determine the affordability and cost-effectiveness of home-based directly observed therapy (DOT) compared to hospital-based DOT for chronically ill tuberculosis (TB) patients, and to describe the characteristics of patients and their caregivers. DESIGN: Costs for each alternative strategy were analysed from the perspective of the health system and caregivers, in 1998 US dollars. Caregiver costs were assessed using a structured questionnaire administered to a sample of 50 caregivers. Health system costs were assessed using interviews with relevant staff and documentary data such as medical records and expenditure files. These data were used to calculate the average cost of individual components of care, and, for each alternative strategy, the average cost per patient treated. Cost-effectiveness was calculated as the cost per patient compliant with treatment. The characteristics of caregivers and patients were assessed using demographic and socio-economic data collected during interviews, and medical records. RESULTS: Overall, home-based care reduced the cost per patient treated by 44% compared with hospital-based treatment (dollars 1657 vs. dollars 2970). The cost to the caregiver was reduced by 23% (dollars 551 vs. dollars 720), while the cost to the health system was reduced by 50% (dollars 1106 vs. dollars 2206). The cost per patient complying with treatment was dollars 1726 for home-based care and dollars 2970 for hospitalisation. Caregivers were predominantly female relatives (88%), unemployed (48%), with primary school education or less (82%), and with an income of less than dollars 1000 per annum (71%). Of those patients with an HIV test result, 98% were HIV-positive. CONCLUSION: Home-based care is more affordable and cost-effective than hospital-based care for chronically ill TB patients, although costs to caregivers remain high in relation to their incomes. Structured home-based DOT should be included as a component of the National Tuberculosis Control Programme in Botswana.

Substandard tuberculosis drugs on the global market and their simple detection.

SETTING: The prevalence of substandard anti-tuberculosis drugs is unknown. To maximize the effectiveness of tuberculosis (TB) control efforts, simple, inexpensive drug quality screening methods are needed. DESIGN: Isoniazid (INH) and rifampin (RMP) single- and fixed-dose combination (FDC) formulations were collected from selected TB programs and pharmacies in Colombia, Estonia, India, Latvia, Russia and Vietnam. Samples were screened using a recently developed thin-layer chromatography (TLC) kit. All abnormal samples and a 40% random sample of normal formulations were further analyzed using confirmatory techniques. Samples outside of 85% to 115% of stated content, and/or containing compounds other than the stated drug, were defined as being substandard. RESULTS: Overall, 10% (4/40) of all samples, including 13% (4/30) RMP samples, contained <85% of stated content. More FDCs (5/24, 21%) than single-drug samples (2/16, 13%) were substandard. A comparison of TLC with the confirmatory analysis for RMP analysis showed a sensitivity of 100% (4/4), a specificity of 92% (24/26), a positive predictive value (PPV) of 67% (4/6), and a negative predictive value (NPV) of 100% (24/24). An analysis of INH showed a specificity of 90% (9/10). However, sensitivity, PPV, and NVP could not be determined. CONCLUSION: A substantial number of anti-tuberculosis drugs from several countries, in particular FDCs, were found to be substandard. Such drugs may contribute to the creation of drug-resistant TB. TLC is an effective, convenient, and inexpensive method for the detection of substandard drugs.

Use of a computerized tuberculosis register for automated generation of case finding, sputum conversion, and treatment outcome reports.

SETTING: Tuberculosis (TB) rates in southern Africa have increased dramatically in recent years. Provision of accurate data for surveillance, program management, and supervision is increasingly essential. OBJECTIVE: To develop software that would provide more efficient collection, compilation, and analysis of TB data on an ongoing basis. DESIGN: The 'Electronic TB Register' is a user-friendly, Epi-Info based software program based on the WHO/IUATLD format of recording and reporting. Individual records from the TB registry are entered in a program that provides interactive support. The software provides several patient management and supervision functions, such as lists of defaulters. Finally, it generates standard quarterly and annual reports on case-finding, sputum conversion, and cohort analysis, and provides graphs of trends and maps of TB indicators. RESULTS: The 'Electronic TB Register' software has been successfully implemented in five pilot projects in southern Africa. User acceptance has been high and quality of data has improved, although timeliness remains unchanged. Factors critical for success include a functioning, paper-based system, involvement of staff from the TB program, health information systems, and health facilities, ongoing training, and backup support. CONCLUSIONS: The 'Electronic TB Register' is a potentially powerful tool for surveillance, management, and supervision for countries with well-functioning paper-based recording and reporting systems.

Control of anti-tuberculosis drug resistance in Botswana.

SETTING: Botswana, where in 2000 the prevalence of human immunodeficiency virus (HIV) infection among adults was 38%, and the tuberculosis (TB) rate was 591/100,000. A 1995-1996 survey demonstrated low levels of anti-tuberculosis drug resistance. OBJECTIVE: Because TB drug resistance may increase rapidly in HIV-infected populations, a second survey was undertaken in 1999 to determine any increase in anti-tuberculosis drug resistance. DESIGN: Sputum specimens positive for acid-fast bacilli from patients without prior TB treatment (new patients), and all sputum specimens from patients reporting prior TB treatment (retreatment patients) were collected nationwide. Specimens were cultured for Mycobacterium tuberculosis and tested for resistance to isoniazid, rifampicin, ethambutol, and streptomycin. RESULTS: From January to May 1999, 783 patients were consecutively enrolled from all districts. Of these, 483 (61.7%) were male, the median age was 33 years, and 82% were new patients. Drug resistance occurred in 6.3% of new patients (95 % confidence interval [CI] 4.6-8.6) and 22.8% of retreatment patients (95% CI 16.5-30.1). Resistance to at least isoniazid and rifampicin was found in 0.5% of new (95% CI 0.1-1.3) and 9.0% of retreatment patients (95% CI 5.1-14.5). CONCLUSION: Anti-tuberculosis drug resistance remains relatively low in Botswana, probably as a result of a well-functioning TB program. Periodic surveys will be essential to adequately determine any significant trend.

DNA fingerprinting of a national sample of Mycobacterium tuberculosis isolates, Botswana, 1995-1996.

DNA fingerprinting may be useful to elucidate tuberculosis (TB) transmission in community settings, but its utility is limited if only few fingerprint patterns are observed or band numbers are low. We performed DNA fingerprinting on a national, population-based sample of Mycobacterium tuberculosis isolates from Botswana. During 1995-1996, a random sample of 213 isolates, representing 5% of all smear-positive TB cases, underwent DNA fingerprinting using restriction fragment length polymorphism (RFLP) IS6110 analysis. Eighty-two (38%) of the 213 isolates belonged to one of 18 clusters, with 2-9 isolates/cluster. The median number of bands was 10 (range 1-19); 183 (86%) had six or more bands. Sixty-three (49%) of 128 patients tested were infected with the human immunodeficiency virus (HIV). The degree of RFLP pattern heterogeneity and high band number support the feasibility of a prospective DNA fingerprinting study in Botswana.

Completeness and timeliness of treatment initiation after laboratory diagnosis of tuberculosis in Gaborone, Botswana.

SETTING: Gaborone, the capital of Botswana. OBJECTIVE: To determine the time from positive sputum smear microscopy for acid-fast bacilli (AFB) to initiation of therapy, and to identify risk factors for delays. DESIGN: Retrospective cohort study of medical records and surveillance data for patients with positive smear microscopy and newly diagnosed tuberculosis (TB) from January to May 1997. Treatment delay was defined as more than 2 weeks from the first positive sputum smear to the initiation of TB treatment. RESULTS: Of 127 patients identified, 15 (11.8%) had treatment delay, 13 (10.2%) had an incomplete workup (only one smear performed) and were not registered for TB treatment, and six (4.5%) had two or more positive smears but were not registered for TB treatment. Risk factors for treatment delay or non-registration included TB patients who had been diagnosed in a hospital outpatient setting vs. a clinic (RR 2.9, 95% CI 1.2-3.6, P = 0.02), or in a high volume vs. low volume clinic (RR 2.2, 95% CI 1.2-5.3, P = 0.01). CONCLUSION: More than a quarter of the smear-positive TB patients identified had treatment delay or no evidence of treatment initiation. Proper monitoring of laboratory sputum results and suspect TB patient registers could potentially reduce treatment delays and patient loss.

Detection of substandard fixed-dose combination tuberculosis drugs using thin-layer chromatography.

SETTING: A convenience sample of 13 fixed-dose combination (FDC) tuberculosis (TB) drugs from 'The Fixed Dose Combination Project' was analysed in laboratories at the University of Botswana and the US Food and Drug Administration (FDA). OBJECTIVE: To determine actual versus stated content of drugs in these FDCs. DESIGN: Chemical analysis was performed using thin-layer chromatography (TLC) as a screening method, and ultraviolet (UV) spectrophotometry or liquid chromatography (LC) as confirmation. FDCs with content outside of 85-115% of stated concentration were defined as substandard. RESULTS: All 13 FDCs contained the stated drugs. However, four (31%) were substandard, including two (15%) with low rifampicin content, one (8%) with excessive rifampicin, and one (8%) with excessive pyrazinamide. Both FDCs with low rifampicin contained four drugs and failed TLC screening. The FDC with excessive rifampicin was not detected by TLC screening. Using UV as the gold standard, the sensitivity of TLC for low rifampicin was 2/2 (100%), and the specificity was 9/10 (90%). CONCLUSION: This study found that 31% of the FDCs in 'The Fixed Dose Combination Project' had substandard content, irrespective of bioavailability. Low rifampicin content, which can be reliably detected by TLC screening, was identified in both four-drug FDC products and is particularly worrisome. TB drugs should be screened for quality using TLC to optimise treatment outcomes and to prevent increases in acquired drug resistance.

Risk factors for transmission of Mycobacterium tuberculosis from HIV-infected tuberculosis patients, Botswana.

OBJECTIVE: To identify risk factors for transmission of Mycobacterium tuberculosis from patients with tuberculosis and human immunodeficiency virus (HIV) infection in Botswana. DESIGN: Transmission was studied in 210 children aged <10 years (contacts) of unknown HIV status exposed to 51 adults with tuberculosis (index cases), including 41/49 (83.7%) with HIV infection. METHODS: Data collected on index cases included demographics, clinical and social characteristics, sputum, HIV, and CD4 lymphocyte results. Tuberculin skin testing was performed on contacts, and their parent or guardian was interviewed. A positive test was defined as > or = 10 mm induration. Skin test results were compared with results obtained from a population survey of children of similar age from the same community. RESULTS: A positive skin test was found in 12.1% of exposed children compared with 6.2% in the community (P = 0.005). Of the infected children, 22 (78.6%) were contacts of a close female relative. The risk of transmission increased with the degree of sputum smear positivity for acid-fast bacilli among female index cases (10.8% if smear 0+, 9.3% if smear 1+,29.4% if smear 2+, 44% if smear 3+, P < 0.001). In multivariate analysis, severe immunodeficiency (CD4 lymphocyte count <200 cells/mm3) among HIV-infected index cases was protective against transmission (OR 0.08, 95%CI 0.01-0.5, P = 0.006). CONCLUSION: The intensity of exposure to tuberculosis patients and the degree of sputum smear positivity for acid-fast bacilli remain important risk factors for transmission of M. tuberculosis during the era of HIV. However, tuberculosis patients with advanced AIDS may be less infectious than patients in earlier stages of AIDS.

Pathology and causes of death in a group of 128 predominantly HIV-positive patients in Botswana, 1997-1998.

BACKGROUND: Little is known about causes of death in countries of southern Africa seriously affected by the HIV/AIDS epidemic. METHODS: After obtaining informed consent, autopsies were performed on 128 mainly hospitalised adults in Francistown, Botswana, between July 1997 and June 1998. Criteria for case selection included those who died before a diagnosis could be established, those whose condition deteriorated unexpectedly during hospitalization, and those who had respiratory disease. This represented 14% of adult medical patients who died in hospital during the study period. RESULTS: Of the 128 patients, 104 (81%) were HIV-positive. Among HIV-positive patients, the most common pathologic findings were tuberculosis (TB) (40%), bacterial pneumonia (23%), Pneumocystis carinii pneumonia (11%), and Kaposi's sarcoma (11%); these conditions were the cause of death in 38%, 14%, 11%, and 6%, respectively. Of the 40 pulmonary TB cases, 90% also had disseminated extra-pulmonary TB. Chest radiology could not reliably distinguish the pathologies pre-mortem. CONCLUSIONS: TB was the leading cause of death in our series of HIV-positive adults in Botswana, selected towards those with chest disease; in most, it was widely disseminated. Bacterial pneumonia also played an important role in mortality. Pneumocystis carinii pneumonia was present, but relatively uncommon.

The validity of HIV testing using sputum from suspected tuberculosis patients, Botswana, 2001.

SETTING: The World Health Organization recommends that sentinel HIV surveillance be conducted on tuberculosis patients. However, serum HIV testing is complicated in the TB clinic context, and may not be acceptable to patients. DESIGN: To determine the utility of the OraQuick HIV-1/2 Assay for the detection of HIV antibodies in sputum, we consecutively enrolled adult in-patients in Botswana who had sputum sent for acid-fast bacilli testing and serum sent for HIV ELISA testing. OraQuick HIV-1/2 Assay was applied to gingival secretions according to manufacturer's guidelines, and was also dipped into sputum specimens. A subset of 60 sputum specimens was also serially tested up to 72 hours after collection. RESULTS: Of 377 patients, 84% were HIV-positive by serum ELISA. Compared with serum ELISA, the OraQuick HIV-1/2 Assay detected HIV in gingival secretions with 98.4% sensitivity and 98.3% specificity (95%CI 97-99 and 92-100, respectively), and 97.1% sensitivity and 98.3% specificity on initial sputum specimens (95%CI 95-99 and 92-100, respectively). OraQuick HIV-1/2 Assay performance on sputum declined slightly when tested up to 72 hours after collection. CONCLUSIONS: When applied to sputum specimens, the OraQuick HIV-1/2 Assay demonstrates sensitivity and specificity comparable to its intended application on gingival secretions. This novel testing method will be valuable in anonymous sentinel HIV surveillance surveys among tuberculosis patients.