Preventing antiretroviral anarchy in sub-Saharan Africa

Combination antiretroviral therapy has dramatically improved the survival of patients living with HIV and AIDS in industrialised countries of the world. Despite this enormous benefit, there are some major problems and obstacles to be overcome.(1) Treatment of HIV-infection is likely to be lifelong.(2) Unfortunately, many HIV-infected individuals cannot tolerate the toxic effects of the drugs, or have difficulty complying with treatment which involves large numbers of pills and complicated dosing schedules. Poor adherence to treatment leads to the emergence of drug-resistant viral strains that need new combinations of drugs or new drugs altogether.

What causes smear-negative pulmonary tuberculosis in Malawi, an area of high HIV seroprevalence?

SETTING

The Central Hospital and the District Tuberculosis (TB) Registry in Lilongwe, the capital of Malawi. In this setting smear-negative pulmonary tuberculosis (PTB) is diagnosed using clinical and radiographic criteria for TB, and mycobacterial cultures are not routinely available.

OBJECTIVE

To determine the proportion of patients being registered for smear-negative PTB treatment in Lilongwe who have TB that can be confirmed microbiologically.

DESIGN

Prospective cohort study of patients about to start treatment under operational conditions for smear-negative PTB in Lilongwe between October 1997 and June 1998. Patients referred to the study team underwent a detailed clinical re-assessment, testing for human immunodeficiency virus (HIV), repeat sputum smear microscopy for acid-fast bacilli and mycobacterial cultures of sputum and blood. Bronchoscopy and bronchoalveolar lavage (BAL) were performed and BAL fluid was examined for TB, Pneumocystis carinii and other fungi.

RESULTS

Of 352 smear-negative PTB suspects assessed, the diagnosis of TB was confirmed in 137 (39%) cases. Eighty-nine per cent of patients assessed were HIV-positive, of whom 81% met the expanded case definition for the acquired immune-deficiency syndrome (AIDS).

CONCLUSION

TB was the most commonly confirmed diagnosis amongst patients about to start treatment for smear-negative PTB in an area of high background HIV seroprevalence.

Mortality rates and recurrent rates of tuberculosis in patients with smear-negative pulmonary tuberculosis and tuberculous pleural effusion who have completed treatment

SETTING

Queen Elizabeth Central Hospital, Blantyre, and Zomba Central Hospital, Zomba, Malawi.

OBJECTIVE

To follow-up human immunodeficiency virus (HIV) seropositive and HIV-seronegative patients with smear-negative pulmonary tuberculosis (PTB) and pleural TB who had completed treatment with two different regimens in Blantyre and Zomba, and to assess rates of mortality and recurrent TB.

DESIGN

Patients with smear-negative and pleural TB who had completed 8 months ambulatory treatment in Blantyre or 12 months standard treatment in Zomba and who were smear and culture negative for acid-fast bacilli at the completion of treatment were actively followed every 4 months for a total of 20 months.

RESULTS

Of 248 patients, 150 with smear-negative PTB and 98 with pleural TB, who completed treatment and were enrolled, 205 (83%) were HIV-positive. At 20 months, 145 (58%) patients were alive, 85 (34%) had died and 18 (7%) had transferred out of the district. The mortality rate was 25.7 per 100 person-years, with increased rates strongly associated with HIV infection and age >45 years. Forty-nine patients developed recurrent TB. The recurrence rate of TB was 16.1 per 100 person-years, with increased rates strongly associated with HIV infection, having smear-negative PTB and having received 'standard treatment'.

CONCLUSION

High rates of mortality and recurrent TB were found in patients with smear-negative PTB and pleural effusion during 20 months of follow-up. TB programmes in sub-Saharan Africa must consider appropriate interventions, such as co-trimoxazole and secondary isoniazid prophylaxis, to reduce these adverse outcomes.

High mortality rates in tuberculosis patients in Zomba Hospital, Malawi, during 32 months of follow-up

There is little information about long-term follow-up in patients with smear-negative pulmonary tuberculosis (PTB) or extrapulmonary tuberculosis (EPTB) who have been treated under routine programme conditions in sub-Saharan Africa. A prospective study was carried out to determine outcome 32 months from start of treatment in an unselected cohort of 827 adults TB inpatients registered at Zomba Hospital, Malawi, in 1 July-31 December 1995. By 32 months, 351 (42%) patients had died. Death rates were 30% (95% confidence interval [95% CI] 25-35%) in 386 patients with smear-positive PTB, 60% (95% CI 53-67%) in 211 patients with smear-negative PTB and 47% (95% CI 40-54%) in 230 patients with EPTB. Of the 793 patients with concordant HIV test results 612 (77%) were HIV seropositive: 47% HIV-positive patients were dead by 32 months compared with 27% HIV-negative patients (adjusted hazard ratio [HR] 2.3; 95% CI 1.7-3.1, P < 0.001). Smear-negative PTB patients had the highest death rates during the 32-month follow-up (HR 2.7; 95% CI 2.1-3.5, P < 0.001 compared to smear-positive patients), followed by EPTB patients (HR 1.9; 95% CI 1.5-2.5, P < 0.001 compared to smear-positive patients). When analysis was restricted to after the treatment period had finished (i.e., months 12-32), the differences in mortality were maintained for HIV-serostatus and for types of TB. Low-cost, easy to implement strategies for reducing mortality in HIV-positive TB patients in sub-Saharan Africa (such as the use of trimethoprim-sulphamethoxazole prophylaxis) need to be tested urgently in programme settings.

Efficacy of an unsupervised ambulatory treatment regimen for smear-negative pulmonary tuberculosis and tuberculous pleural effusion in Malawi

SETTING

Queen Elizabeth Central Hospital, Blantyre, and Zomba Central Hospital, Zomba, Malawi.

OBJECTIVE

To evaluate treatment outcome of unsupervised ambulatory treatment (2R3H3Z3/2TH[EH]/4H) in Blantyre and 'standard' treatment (1STH[SEH]/11TH[EH]) in Zomba in human immunodeficiency virus (HIV) seropositive and seronegative patients with smear-negative pulmonary tuberculosis (PTB) and pleural TB.

DESIGN

All patients with smear-negative and pleural TB registered between 1 April and 31 December 1995 were assessed for enrolment in the study. Study patients were followed up and 12-month treatment outcomes were recorded.

RESULTS

A total of 434 patients, 296 with smear-negative PTB and 138 with pleural TB, were enrolled: 366 (84%) of patients were HIV-positive; 220 (51%) completed treatment, and 144 (33%) died by 12 months. In patients from Blantyre and Zomba, baseline characteristics were similar, apart from older age in those from Zomba, and the proportion of patients who completed treatment and who died were similar. In both sites, significantly higher case fatality rates were found in older patients, HIV-positive patients and patients with pulmonary parenchymal lung disease.

CONCLUSION

Unsupervised ambulatory treatment evaluated in this study had an efficacy similar to that of 'standard' treatment. For operational reasons, however, it will not be recommended for widespread use in Malawi's National Tuberculosis Control Programme.

Treatment outcome of an unselected cohort of tuberculosis patients in relation to human immunodeficiency virus serostatus in Zomba Hospital, Malawi

There is little information about treatment outcome in patients with smear-negative pulmonary tuberculosis (PTB) or extrapulmonary tuberculosis (EPTB) treated under routine programme conditions in subsaharan Africa. A prospective study was carried out to determine treatment outcome in an unselected cohort of TB patients admitted to Zomba General Hospital, Malawi. Eight hundred and twenty-seven adult TB patients (451 men and 376 women) were registered between 1 July and 31 December 1995. Standardized treatment outcomes of treatment completion, death, default, and transfer to another district were assessed in relation to type of TB, human immunodeficiency virus (HIV) serostatus, age and gender. Two hundred and fifty-four patients (31%) died by the end of treatment, half of the deaths occurring in the first month. Death rates were 19% among 386 patients with smear-positive PTB, 46% among 211 patients with smear-negative PTB, and 37% among 230 patients with EPTB; 77% of the patients were HIV seropositive. Among new patients, HIV-positive patients had higher death rates than HIV-negative patients (hazard ratio [HR] 2.5; 95% confidence interval [95% CI] 1.6-3.8). Smear-negative patients had the highest death rates (HR 3.9; 95% CI 2.7-5.5 compared to smear-positive patients), followed by EPTB patients (HR 2.6, 95% CI 1.8-3.7 compared to smear-positive patients). Death rates increased with age but were similar in men and women. Adult patients in Malawi with smear-negative PTB and EPTB have low treatment completion and high death rates, related to high levels of HIV infection. National TB control programmes in areas of high HIV prevalence should no longer ignore treatment outcomes in patients with smear-negative PTB or EPTB.

Case finding for pulmonary tuberculosis, Queen Elizabeth Central Hospital, Blantyre, Malawi

SETTING

Queen Elizabeth Central Hospital, Malawi.

OBJECTIVE

To evaluate the investigation of patients with cough who attend out-patient services and the adherence to recommended diagnostic protocols.

DESIGN

Two operational studies in 1995: 1) an audit of management of patients presenting to non-fee-paying out-patient services (OPD) with cough, and 2) an audit of the laboratory sputum register and the OPD cough register. The annual number of out-patient attendances was also recorded.

RESULTS

Of 2381 patients seen by OPD medical assistants, 438 (18.4%) complained of cough: 303 for < 3 weeks and 135 for > or = 3 weeks. Sputum smear examinations were requested in 97 patients, 79 (58.5%) with long duration and 18 (5.9%) with short duration of cough. Between May and December 1995, of the 1668 OPD patients who had sputum results in the laboratory register, 1392 (83%) had sputum results in the cough register. Of patients listed in the cough register, 98% collected their sputum smear results. In 1995, there were 395,439 OPD attendances; data extrapolation suggests that about 15,000 patients should have had sputum examined instead of the 2337 listed in the laboratory register.

CONCLUSION

A large burden is imposed on out-patient services by patients with cough. Despite recommended protocols, this investigation suggests deficiencies in case detection which require further study.

Tuberculosis in a prison population in Malawi

BACKGROUND

Much concern has been expressed about the high prevalence of tuberculosis in prisons in industrialised countries. Since there is almost no information from developing countries, we investigated the rate of pulmonary tuberculosis in a large prison in Malawi.

METHODS

Between May and July, 1996; we carried out an active case-finding survey in Zomba Central Prison, Malawi, through the National Tuberculosis Control Programme. We interviewed prisoners, and those with a cough of at least 1 week's duration were screened by sputum-smear microscopy. If microscopy was negative, prisoners underwent chest radiography. We offered HIV testing, with voluntary consent and counselling before and after tests, to all prisoners, whether positive or negative for pulmonary tuberculosis.

FINDINGS

914 (70%) of 1315 prisoners were screened (905 men, nine women; mean age 30 years [SD 11]). 47 (5%) screened prisoners (all men) had pulmonary tuberculosis: 14 were taking antituberculosis treatment and 33 were undiagnosed at the start of the study; 18 were sputum-smear positive and 15 were sputum-smear negative. 16 (73%) of 22 prisoners with previously undiagnosed pulmonary tuberculosis and 30 (75%) of 40 prisoners with cough but no evidence of pulmonary tuberculosis were HIV seropositive. In all prisoners, except one, symptoms of pulmonary tuberculosis had developed after they had entered prison.

INTERPRETATION

We found a high rate of pulmonary tuberculosis in Zomba Central Prison, which suggests active transmission of tuberculosis. As a result of this study, the National Tuberculosis Control Programme has implemented interventions in eight prisons in Malawi to improve tuberculosis control, including collection of health data, education of prisoners and clinical staff about tuberculosis, active screening of prisoners for pulmonary tuberculosis by sputum-smear microscopy, and active case-finding in the prisons.

The scourge of HIV-related tuberculosis: a cohort study in a district general hospital in Malawi

Malawi is similar to a number of other African countries in having an escalating, HIV-related, tuberculosis (TB) epidemic. A prospective study was carried out to determine the pattern of disease and HIV serostatus in unselected, adult, TB patients consecutively admitted to a large, district general hospital in Zomba (in the Southern region of Malawi). Clinical details were obtained, from the district TB register, for the 714, adult TB patients, aged > or = 15 years, who were registered with the district TB officer between 1 July and 31 December in 1995. Patients were counselled, and offered HIV testing using an ELISA and particle agglutination test. Concordant HIV-test results were available for 686 (96%) of the subjects: 547 (80%) of these were HIV-seropositive and 139 seronegative. The HIV-positive patients were significantly younger than the HIV-negative patients and significantly more HIV-positive patients were males (P < 0.05 for each). The proportions of HIV-positive subjects who were new patients, had been previously treated for TB, had pulmonary TB (PTB), had smear-positive PTB or had different types of extrapulmonary TB were similar to those of the HIV-negative. A high percentage of an unselected cohort of adult TB patients admitted to a district, general hospital in Malawi, particularly of the younger age groups was therefore HIV-positive. The pattern of disease was uninfluenced by the HIV serostatus. The large number of cases registered emphasises the severity of the current epidemic of TB in Malawi and its impact upon young adults.

The pattern of tuberculosis in Queen Elizabeth Central Hospital, Blantyre, Malawi: 1986-1995

SETTING

Queen Elizabeth Central Hospital (QECH), Blantyre, Malawi.

OBJECTIVE

To determine the pattern of tuberculosis (TB) cases over a period of ten years.

DESIGN

TB registers for QECH for the years 1986-1995 (January 1st to December 31st) were obtained and clinical information on the type of TB in each patient was recorded. Human immunodeficiency virus (HIV) test results of adults with smear-positive pulmonary TB (PTB) and children with TB between 1993 and 1996 were recorded from case notes.

RESULTS

There were 19,377 TB cases, 10,982 men and 8,395 women, registered over the ten-year period. Of these, 4,691 (24%) cases were in children aged 0-14 years and 11,890 (61%) cases were in adults aged 15-44 years. The number of cases increased from 657 in 1986 to 2,734 in 1995, and the proportion of cases with extra-pulmonary TB (EPTB) rose from 11% in 1986 to 33% in 1995. The largest increase in cases was in children and in young adults aged 15-44 years. In all age groups, PTB was more common than EPTB. There were significant increases in the proportion of adult TB cases with pleural effusion. Of those who were tested, 72% of adults with smear-positive PTB and 64% of children with TB were found to be HIV-seropositive.

CONCLUSION

There has been a dramatic increase in cases of TB and changes in disease pattern in QECH during the last 10 years, which is related to the HIV epidemic.

Screening pulmonary tuberculosis suspects in Malawi: testing different strategies

Alternative strategies for screening tuberculosis (TB) suspects are needed in sub-saharan Africa. Ambulatory adult TB suspects who were seen in the chronic cough room of Queen Elizabeth Central Hospital, Blantyre, Malawi, were assessed with respect to appropriateness of referral. Appropriate referrals (patients with cough 3 weeks or longer, weight loss and no antibiotic response) were screened by 3 sputum specimens for microscopy and culture of Mycobacterium tuberculosis and chest radiography (CXR). Hypothetical strategy A (screening by sputum smear examination followed by CXR in patients with negative sputum smears) was compared with strategy B (screening by CXR followed by sputum smear examination in patients with a CXR consistent with TB) in terms of diagnostic efficacy and cost. Of 1127 patients referred to the cough room, 402 (38%) were appropriate TB suspect referrals. Of these, 111 (28%) were sputum smear-positive, 213 (53%) were culture-positive, and 221 (55%) had smear and/or culture-positive evidence of TB. Routine CXR was consistent with pulmonary (P) TB in 230 patients (57%). With strategy A, 243 (60%) patients were diagnosed as PTB, but 40 (25%) of those not diagnosed as PTB had positive mycobacterial cultures. With strategy B, 230 patients (57%) were diagnosed as PTB, but 53 (31%) of those not diagnosed as PTB had positive mycobacterial cultures, including 13 with smear-positive sputum. The cost per diagnosed case of PTB was US$ 4.63 with strategy A and US$ 5.44 with strategy B. Screening patients with good criteria of TB has high diagnostic sensitivity, but screening by CXR is less effective and more costly than screening by sputum smear microscopy.

Delays in diagnosis and treatment of smear-positive tuberculosis and the incidence of tuberculosis in hospital nurses in Blantyre, Malawi

There is little information about nosocomial transmission of tuberculosis (TB) in sub-Saharan Africa. A study was carried out to examine the process of diagnosis and treatment of smear-positive pulmonary TB patients in Queen Elizabeth Central Hospital, Blantyre, Malawi, and the incidence of TB in nurses working in specific departments of the hospital. Case notes of 1365 patients with smear-positive pulmonary TB, diagnosed and treated at the hospital in 1993 and 1994, were analysed. The number of qualified nurses who worked in specific departments of the hospital between 1993 and 1994 and the number who were diagnosed and treated for TB during this period were obtained from nursing records. 787 patients (58%) were diagnosed as out-patients and 578 (42%) were diagnosed in hospital wards, 544 from medical wards. In medical wards, there were long delays from the time of admission to diagnosis and start of anti-TB treatment in new and previously treated TB patients. Of 310 qualified nurses, 12 (4%) were treated for TB in 1993-1994; 4 (14%) of 29 nurses working in the medical wards developed TB. The results indicate the importance of finding simple measures in resource-poor countries to improve the diagnosis and treatment of TB in hospital patients in order to decrease the risk of nosocomial TB transmission.

The use of sputum induction for establishing a diagnosis in patients with suspected pulmonary tuberculosis in Malawi

SETTING

There has been a marked increase in notified cases of smear-negative pulmonary tuberculosis in Malawi since 1986. One reason for this may be related to the difficulties of getting adequate samples of expectorated sputum from patients. Sputum induction with nebulized hypertonic saline may be a simple way of obtaining a better specimen.

OBJECTIVE

To examine the value of sputum induction for detecting cases of smear-positive tuberculosis.

DESIGN

Sputum induction was performed on 82 adults presenting to the Queen Elizabeth Central Hospital, Blantyre, Malawi with clinically suspected pulmonary tuberculosis who were expectorated sputum smear-negative or unproductive of sputum. The induced sputum smear was examined for acid-fast bacilli and cultured for mycobacteria.

RESULTS

Sputum was successfully induced from 73 of the 82 patients (26 previously smear-negative and 47 previously unproductive). The induced sputum was smear-positive in 18 patients (5 previously smear-negative and 13 unproductive). Cultures were positive for Mycobacterium tuberculosis in the 18 smear-positive patients and a further 12 that had been smear-negative. 94 cases of smear-positive pulmonary tuberculosis were notified during the study period. 18 (19%) were as a result of sputum induction.

CONCLUSION

Sputum induction is a useful technique for improving the case detection rate of smear-positive tuberculosis in Malawi.

Cost-effectiveness of chemotherapy for sputum smear-positive pulmonary tuberculosis in Malawi, Mozambique and Tanzania

The cost-effectiveness of chemotherapy for pulmonary sputum smear-positive tuberculosis was examined in the national tuberculosis control programmes of Malawi, Mozambique and Tanzania. In these three programmes, routine cure rates have exceeded 80 per cent. Average, average incremental and marginal unit costs for standard, short-course and retreatment regimens with and without hospitalization have been measured. The average incremental cost per year of life saved through chemotherapy ranged from US $0.90-3.10. In all conditions, short-course chemotherapy is preferable to standard 12-month chemotherapy. When hospitalization during the intensive phase of chemotherapy increases the cure rate by 10-15 percentage points, it can be relatively cost-effective. Analysing the cost-effectiveness of short-course and standard chemotherapy, where the depth of the margin of benefit is different, illustrates some of the dangers of simplistic use of cost-effectiveness ratios.

Cost effectiveness of chemotherapy for pulmonary tuberculosis in three sub-Saharan African countries

The value of programmes to control pulmonary tuberculosis in developing countries remains the subject of debate. We have examined the cost-effectiveness of chemotherapy programmes for the control of pulmonary sputum-smear-positive tuberculosis in Malawi, Mozambique, and Tanzania. Effective cure rates of 86-90% were achieved with short-course chemotherapy and of 60-66% with standard chemotherapy. The average incremental costs per year of life saved were US $1.7-2.1 for short-course chemotherapy with hospital admission, $2.4-3.4 for standard chemotherapy with hospital admission, $0.9-1.1 for ambulatory short-course chemotherapy, and $0.9-1.3 for ambulatory standard chemotherapy. Chemotherapy for smear-positive tuberculosis is thus cheaper than other cost-effective health interventions such as immunisation against measles and oral rehydration therapy. Because the greatest benefit of chemotherapy is reduced transmission of the bacillus, treating HIV-seropositive, tuberculosis smear-positive patients would be only slightly less cost-effective than treating HIV-seronegative, tuberculosis-smear-positive patients.

Nutritional status in Malawian patients with pulmonary tuberculosis and response to chemotherapy

Anthropometric measurements, hand-grip dynamometry and serum albumin determinations were performed on 122 (73 m, 49 f) Malawian patients, aged 16-68 years, who were consecutively admitted to the chest wards, Kamuzu Central Hospital, Lilongwe, with sputum-positive pulmonary tuberculosis. Nutritional findings were compared with those obtained in healthy age- and sex-matched controls, and were also related to clinical and radiographic features. All patients received chemotherapy under supervision in hospital for 8 weeks, and nutritional status was monitored at 4-week intervals. Weight, muscle mass and function, subcutaneous fat and serum albumin were all significantly reduced compared with controls (P less than 0.001), and the nutritional deficit increased with the extent of radiographic disease. With chemotherapy there was a significant improvement at 4 weeks, and again at 8 weeks, in all nutritional indices (P less than 0.05), and this occurred irrespective of clinical, radiographic and nutritional features which were observed on admission. The results indicate that pulmonary tuberculosis is associated with a high degree of undernutrition, but this improves with effective chemotherapy.