Costs of measures to control tuberculosis/HIV in public primary care facilities in Cape Town, South Africa

Coûts des mesures de lutte contre la tuberculose et le VIH dans les établissements de soins de santé primaires de la ville du Cap, en Afrique du Sud

Costo de las medidas de control de la tuberculosis/VIH en centros de atención primaria de Ciudad del Cabo, Sudáfrica

Harry Peter Hausler^I,1; Edina Sinanovic^II; Lilani Kumaranayake^III; Pren Naidoo^IV; Hennie Schoeman^V; Barbara Karpakis^IV; Peter Godfrey-Faussett^III

ABSTRACT

OBJECTIVE: To measure the costs and estimate the cost-effectiveness of the ProTEST package of tuberculosis/human immunodeficiency virus (TB/HIV) interventions in primary health care facilities in Cape Town, South Africa.
METHODS: We collected annual cost data retrospectively using ingredients-based costing in three primary care facilities and estimated the cost per HIV infection averted and the cost per TB case prevented. ]]> FINDINGS: The range of costs per person for the ProTEST interventions in the three facilities were: US$ 7–11 for voluntary counselling and testing (VCT), US$ 81–166 for detecting a TB case, US$ 92–183 for completing isoniazid preventive therapy (IPT) and US$ 20–44 for completing six months of cotrimoxazole preventive therapy. The estimated cost per HIV infection averted by VCT was US$ 67–112. The cost per TB case prevented by VCT (through preventing HIV) was US$ 129–215, by intensified case finding was US$ 323–664 and by IPT was US$ 486–962. Sensitivity analysis showed that the use of chest X-rays for IPT screening decreases the cost-effectiveness of IPT in preventing TB cases by 36%. IPT screening with or without tuberculin purified protein derivative screening was almost equally cost-effective.
CONCLUSION: We conclude that the ProTEST package is cost saving. Despite moderate adherence, linking prevention and care interventions for TB and HIV resulted in the estimated costs of preventing TB being less than previous estimates of costs of treating it. VCT was less expensive than previously reported in Africa.

RÉSUMÉ

OBJECTIF: Mesurer les coûts et évaluer le rapport coût/efficacité de l'ensemble d'interventions ProTEST contre la tuberculose et le virus de l'immunodéficience humaine (VIH) dans des établissements de soins de santé primaires de la ville du Cap, en Afrique du Sud.
MÉTHODES: Des données relatives aux coûts annuels ont été recueillies rétrospectivement par évaluation des coûts à partir des éléments de cette intervention. Le coût par contamination par le VIH prévenue et celui par cas de TB évité ont ensuite été estimés.
RÉSULTATS: Dans les trois établissements étudiés, le coût par personne des interventions ProTEST se situait dans les plages suivantes : US $ 7-11 pour la délivrance de conseils et le dépistage volontaire, US $ 81-166 pour la détection d'un cas de tuberculose, US $ 92-183 pour l'administration d'un traitement préventif par l'isoniazide et US $ 20-44 pour l'administration d'un traitement préventif de six mois par le cotrimoxazole. Le coût par contamination par le VIH évitée grâce aux conseils et au dépistage volontaire a été estimé à US $ 67-112. Le coût par cas de TB prévenu par ce même type d'intervention (prévention du VIH) était de US $ 129-215, par recherche intensive des cas de US $ 323-664 et par traitement préventif par l'isoniazide de US $ 486-962. Une analyse de sensibilité a mis en évidence que l'utilisation de la radiographie pulmonaire comme méthode de dépistage préliminaire à ce traitement permettait de prévenir les cas de TB dans une proportion de 36 %. Le dépistage préliminaire au traitement par l'isoniazide avec ou sans intradermoréaction par la fraction protéique purifiée de la tuberculine présentait une efficacité économique presque équivalente.
CONCLUSION: L'article parvient à la conclusion que l'ensemble d'interventions ProTEST permet de réaliser des économies. En dépit d'une adhésion moyenne, le fait de lier les interventions à visées préventives à celles délivrant des soins contre la TB ou le VIH a entraîné une baisse du coût estimé de la prévention d'un cas de TB par rapport aux estimations antérieures du coût d'un traitement antituberculeux. L'étude indique un coût des interventions de conseil et de dépistage volontaire moins élevé que les chiffres rapportés antérieurement pour l'Afrique.

RESUMEN

OBJETIVO: Medir los costos y estimar la costoeficacia del paquete ProTEST de intervenciones contra la tuberculosis/virus de la inmunodeficiencia humana (tuberculosis/VIH) en establecimientos de atención primaria de Ciudad del Cabo, Sudáfrica.
MÉTODOS: Reunimos retrospectivamente datos anuales sobre costos utilizando sistemas de determinación de los costos basados en componentes en tres establecimientos de atención primaria, para poder así estimar el costo asociado a cada infección por VIH evitada y el costo por caso de tuberculosis prevenido. ]]> RESULTADOS: Los intervalos de los costos por persona para las intervenciones de ProTEST en los tres centros fueron los siguientes: US$ 7 - 11 para el asesoramiento y pruebas voluntarias (APV), US$ 81 - 166 para la detección de un caso de tuberculosis, US$ 92 - 183 para el régimen completo de terapia preventiva con isoniazida (TPI) y US$ 20 - 44 para el tratamiento preventivo con cotrimoxazol durante seis meses. El costo estimado por infección por VIH evitada mediante APV fue de US$ 67 - 112. El costo por caso de tuberculosis prevenido mediante APV (gracias a la prevención de la infección por VIH) fue de US$ 129 - 215, mediante la búsqueda intensificada de casos, de US$ 323 - 664, y mediante TPI, de US$ 486 - 962. El análisis de sensibilidad demostró que el uso de radiografías torácicas en el cribado para TPI reduce en un 36% la costoeficacia de ésta como medio de prevención de los casos de tuberculosis. El cribado para TPI tuvo casi la misma costoeficacia con o sin cribado mediante derivados proteínicos purificados de la tuberculina.
CONCLUSIÓN: Llegamos a la conclusión de que el paquete ProTEST permite ahorrar costos. Pese a la moderada observancia de los tratamientos, vinculando las intervenciones de prevención y atención para la tuberculosis y el VIH, los costos estimados para la prevención de la tuberculosis fueron inferiores a las estimaciones anteriores de los costos asociados al tratamiento de la misma. Las medidas de APV fueron menos costosas de lo que hasta ahora se había señalado en África.

Introduction

With an antenatal human immunodeficiency virus (HIV) prevalence of 29.5% and an estimated 6.29 million people infected,¹ South Africa has the largest number of people living with HIV/acquired immunodeficiency syndrome (AIDS) in the world.² HIV increases tuberculosis (TB) incidence by reactivation of latent infection³ and rapid progression of recent infection.⁴ With increasing HIV prevalence, TB incidence has risen throughout sub-Saharan Africa.^5,6 In South Africa, the incidence of TB increased from 187/100 000 in 1989⁷ to 599/100 000 in 2004.⁸

Following the recommendations by national reviews for improved collaboration between the TB and HIV/AIDS programmes in South Africa,^9,10 four TB/HIV Pilot Districts were initiated in 1999. These districts participated in ProTEST,¹¹ a WHO supported package of TB/HIV interventions by providing voluntary counselling and testing (VCT) with rapid HIV testing, screening for TB through intensified case-finding (ICF), isoniazid preventive therapy (IPT), cotrimoxazole preventive therapy (CPT), and improved management of opportunistic infections. ProTEST aimed to decrease the transmission of HIV through VCT, decrease the transmission of TB through ICF and prevent the reactivation of TB through IPT.¹²

Cost and cost-effectiveness data for ProTEST interventions are important for programme managers to decide what is affordable for expanded implementattion. The data are relevant in the era of antiretroviral treatment (ART) programmes because VCT is necessary to identify HIV-infected persons and ICF, IPT and CPT remain part of the comprehensive package of HIV care. There are few studies in developing countries on the cost-effectiveness of VCT,¹³ rapid HIV testing,¹⁴ IPT^15–19 and CPT.²⁰

We measured the costs and estimated the cost-effectiveness of the ProTEST package of TB/HIV interventions in Cape Town, South Africa.

Methods

Setting

The Central District of Cape Town, with a population of 296 000, consists of urban/peri-urban areas with vast socioeconomic disparities. The antenatal HIV prevalence was 17% in 2001 and the TB incidence was 488/100 000 in 2002.²¹

Using purposive sampling, we chose three public primary health care facilities — a community health centre (CHC), a primary health care (PHC) clinic and a sexually transmitted infections (STI) clinic — from the 12 facilities that participated in ProTEST (Table 1, web version only, available from: http://www.who.int/bulletin). All facilities promoted VCT to self-presenting and antenatal clients as well as TB and STI patients, and provided improved management of HIV-related infections. The CHC and PHC clinic also offered ICF (TB symptom screening for HIV-positive patients, and sputum smear investigations as well as chest X-ray for TB symptomatics), IPT (isoniazid 300 mg daily for six months for HIV-positive patients with no TB symptoms, a normal chest X-ray, and a positive tuberculin skin test) and CPT (life-long cotrimoxazole 480 mg daily for patients with HIV/AIDS, WHO clinical stage III or IV) (Fig. 1). We evaluated only VCT in the STI clinic and the complete ProTEST package in the CHC and PHC clinics.

Cost analysis

Following the Costing guidelines for HIV/AIDS prevention strategies developed by UNAIDS,²² we collected the costs incurred by public and nongovernmental organization (NGO) health-care providers retrospectively, using ingredients- based costing, i.e. costing each component of an activity, including capital and recurrent costs for one financial year. Financial costs represented actual expenditure, while economic costs were financial costs plus the estimated value of goods or services with no financial transactions and some adjusted financial costs when the price paid did not reflect the cost of using it elsewhere.

We did not include the costs of research. We could not measure the costs of drugs used in the treatment of opportunistic infections because they could not be separated from drugs that were dispensed for other infections. The costs of diagnostic tests for HIV and TB and the costs of prophylactic drugs were included. Start-up costs, including initial training costs, were regarded as capital costs because the effect of the activities lasted for more than a year. We annuatized the capital costs using a discount rate of 8% (the discount rate most widely used in South Africa for that time),²³ assuming that the life-span of buildings is 30 years, furniture 10 years, equipment and vehicles five years and initial training five years. Life-spans were estimated based on consultations with district health officials.

Sources of recurrent cost data included financial records and interviews with project staff. Costs are presented in US$ (US$ 1 = R 9.28, exchange rate for South African rand for the period April 2001 to March 2002).²⁴

Total costs were apportioned to the following project activities: health education, pre-test counselling, HIV testing, post-test counselling, screening for IPT/CPT, follow-up for IPT/CPT, management of opportunistic infections (OIs) and supervision/training/mentorship. All counsellor salary costs were allocated to VCT. We calculated the weighted average personnel cost per minute from estimates made by clinical staff of the proportion of time that they spent on ProTEST. We also interviewed clinical staff to estimate the average amount of time they spent on screening and follow-up of a client for prophylaxis. This time multiplied by the cost per minute multiplied by the number of clients gave the costs for screening and follow-up. Other costs (such as buildings, furniture, equipment, vehicles and maintenance) of the health services were multiplied by the proportion of all clinic visits that were for ProTEST to determine the amount that should be allocated to ProTEST. We divided these costs equally between project activities.

Since we could not measure the efficacy of ProTEST interventions directly, results from recently published efficacy studies were used to estimate the impact of interventions, with special attention to studies done in African countries with a high burden of TB and HIV.

To estimate the cost per HIV infection averted, we used the results of a recent multicentre randomized controlled trial of VCT, which found that VCT decreased risk behaviours and estimated that for every 100 people accessing VCT,10 HIV infections are averted (in sensittivity analyses, this ranged from 1 to 24 HIV infections averted).¹³

We developed a model that used the risk of primary TB disease after infection and risk of reactivation of latent TB infection in HIV-positive and HIV-negative adults,^25–31 for estimating the number of TB cases averted each year (over a period of 10 years), by preventing HIV infection. We estimated that for every 100 people accessing VCT, 10 HIV cases and 5.2 TB cases would be prevented over 10 years.

For estimating the cost per TB case prevented through ICF it was assumed that every TB case infects 10–14 people per year and results in one more TB case,³² that ICF might decrease the infectious period from 9.6 months in HIV-infected people³³ by 30%, and that 85% of detected cases will remain noninfectious (successfully treated or died).³⁴ For every 100 TB cases detected by ICF, 25 TB cases would be prevented.

For IPT using tuberculin purified protein derivative (PPD) tests for screening, the estimated cost per TB case averted was based on the following conservative assumptions: IPT decreases TB incidence by 60% (95% confidence interval (CI): 35–76%) for two years in PPD-positives (results from a meta-analysis of intention to treat clinical trials with 60–80% adherence),³⁵ annual incidence of TB in PPD-positive HIV-posittive people is 8%; and each HIV-positive TB case causes one other case. Assuming similar efficacy to clinical trials, for every 100 people completing IPT using PPD screening, 19 (95% CI: 11–24) TB cases are averted.

Sensitivity analysis

We considered the cost-effectiveness of different screening protocols (including the WHO recommended protocol³⁶) in the sensitivity analysis and included discounting the cases of TB prevented in the future by VCT in the sensitivity analysis at a discount rate of 5% cumulatively over 1 to 10 years.^37,38 We also conducted univariate sensitivity analyses with several variables for the PHC clinic.

Findings

All costs are given in US dollars (US$). The unit costs were US$ 1.07 for an HIV screening test, US$ 2.01 for an HIV confirmatory test, US$ 3.29 for an HIV enzyme-linked immunosorbent assay (ELISA), US$ 2.20 for a sputum smear examination for TB, US$ 5.12 for a sputum TB culture, US$ 14.43 for a chest X-ray, US$ 0.73 for a PPD test; US$ 0.27/month for isoniazid; and US$ 0.40/month for cotrimoxazole.

The summary of total costs is given in Table 2 (web version only, available from: http://www.who.int/bulletin). We found that the total economic cost for one year of ProTEST activities was US$ 21 623 in the CHC, US$ 47 280 in the PHC clinic and US$ 36 575 in the STI clinic. The highest costs were associated with the management of OIs in facilities offering comprehensive clinical services followed by VCT services (the sum of pre- and post-test counselling and testiing). The combined cost of screening and follow-up for IPT and CPT was similar to the cost of TB/HIV/STI education. Start-up and coordination costs were low. Financial costs were slightly higher than economic costs in the PHC clinic and STI clinic because NGO counsellor salaries (actual costs, considered "financial") were higher than the government counsellor salaries (costs required for scaling up, considered "economic").

Personnel costs accounted for a much higher proportion of the total costs than the cost of supplies: 82% versus 12% in the CHC, 85% versus 11% in the PHC clinic and 78% versus 17% in the STI clinic. Other capital costs (1% in all), and recurrent vehicle and building costs (1% in the CHC and PHC clinic and 2% in the STI clinic) were small.

The STI clinic had the largest number of VCT clients (Table 3, web version only, available from: http://www.who.int/bulletin). At all sites, most people who received pre-test counselling were tested for HIV (97–99%). HIV prevallence was lower in the CHC (20%) and STI clinic (21%) than in the PHC clinic (27%). The 34 TB cases identified by ICF represented 4% of the 781 cases registered at the PHC clinic over the same period. All TB cases diagnosed at the CHC were referred to a TB clinic; the numbers referred were not recorded. The PHC clinic screened the highest numbers for prophylaxis and achieved better adherence rates than the CHC. The proportion of screened HIV-positive clients who started IPT was 15–16% and those who started CPT was 38–57%.

Our results showed that the unit costs were similar for VCT but lower for ICF, IPT and CPT in the PHC clinic compared to the CHC (Table 4). The cost per six person-months of providing prophylaxis after screening was US$ 6–9 for IPT and US$ 6–8 for CPT.

The cost per person completing VCT ranged from US$ 7 to US$ 11 (Table 5). The cost per TB case detected and cost per person completing six months of prophylaxis were about half as expensive at the PHC clinic than at the CHC.

Estimating impact

We found that the estimated cost per HIV infection averted through VCT ranged from US$ 67 in the STI clinic to US$ 112 in the CHC. The estimated cost per TB case prevented was US$ 129–215 by VCT, US$ 323–664 by ICF and US$ 486–962 by IPT (Table 5).

Sensitivity analysis

Follow-up interviews with managers and staff showed that staff over-estimated the time spent for each intervention when they used time sheets. The sensitivity analysis showed higher costs for all interventions when using time sheets (Table 6, web version only, available from: http://www.who.int/bulletin).

Lay counsellors and rapid HIV testing were more cost-effective as our results showed that using nurse counsellors increased the cost per person post-test counselled by 182% and using ELISA increased the cost by 23%.

The cost per person completing IPT and the cost per TB case prevented were affected by changes in the screening protocol. Removing chest X-ray from the IPT screening protocol decreased the cost per TB case detected by 40% and decreased the cost per person completing IPT by 36% (Table 6).

Discussion

Cost analysis

Our findings show that total costs varied widely among the facilities and reflected the number and category of staff involved, the services offered, HIV prevalence and the number of clients. Personnel accounted for the highest propportion (78–85%) of total costs due to the labour-intensive nature of VCT and HIV clinical care. Cost of supplies was a much lower proportion of total costs (11–17%) reflecting the low cost of rapid HIV tests, isoniazid and cotrimoxazole. The high proportion of total costs atttributable to personnel and the fact that salary costs are lower in many other African countries should be considered when assessing the affordability of these interventions in other settings.

The cost per person post-test counselled was lower in the STI clinic than in the other facilities because of the higher number of persons coming for testing and the exclusive use of lay counsellors. The cost per person post-test counselled in our study (US$ 7–11) across all three sites was lower than that reported from Kenya (US$ 30) and the United Republic of Tanzania (US$ 32).¹³ This may be due to the use of lay counsellors instead of professional counsellors and rapid HIV tests instead of laboratory-based ELISAs. A study from South Africa showed that the cost per person post-test counselled almost halved from US$ 20.95 to US$ 11.30 by using rapid tests compared to ELISA.¹⁴ In our study using rapid HIV tests, almost every person tested received their HIV test results (99–100%) whereas with the use of ELISAs in the study in Kenya and the United Republic of Tanzania a smaller proportion of people (70–95%) received their results.³⁹

The cost per clinical intervention (ICF, IPT, CPT) was lower at the PHC clinic than at the CHC. This was mostly due to the larger proportion of HIV-positive clients starting and completing prophylaxis as the weighted average personnel cost per minute was found to be similar for both facilities.

The cost per person completing six months of IPT incremental to ICF (US$ 51–110) was considerably lower than the cost per person completing IPT including ICF (US$ 92–183). Thus, once a programme decides to do ICF, the additional cost of providing IPT would be relatively small.

The cost per person completing six months of IPT (US$ 92) with 57% adherence was higher than the cost reported in Uganda with 62% adherence (US$ 24)¹⁹ and from a modelling study in Zambia with an assumed 63% adherence (US$ 42),¹⁸ due to the higher cost of personnel and lower adherence to IPT in South Africa.

Estimating impact

Our estimates of cost per HIV infection averted by VCT (US$ 67–112) compares favourably to the cost per HIV infection averted by other HIV prevention interventions, such as improved management of STIs (US$ 280)⁴⁰ and nevirapine to prevent mother-to-child transmission of HIV (US$ 318).⁴¹

Our results on cost per TB case prevented through VCT (US$ 129–215), ICF (US$ 323–664) and IPT (US$ 486–962) were less than the cost of treatiing a new case of TB (US$ 823–1362) reported in a previous study from Cape Town.⁴² The cost of providing IPT is also likely to be less than the cost of treating a TB case in other African countries where a smaller proportion of IPT costs would be for personnel and a higher proportion of costs to treat TB would be for TB drugs.

Randomized clinical trials from Côte d'Ivoire reported that CPT decreases mortality in HIV-infected TB patients by 46% (95% CI: 23–62%)⁴³
and hospitalizations in symptomatic HIV-positive people by 43% (95% CI: 25–57%).⁴⁴ An observational cohort in Cape Town showed similar results with CPT decreasing mortality by 44% (95% CI: 15–67%) and the incidence of severe HIV-related illnesses by 48% (95% CI: 32–62%).⁴⁵ Another study from Cape Town has reported the number of hospitalization days for people not on ART as 1.84 per year at a cost of US$ 206.⁴⁶ Therefore, if CPT decreased the cost of hospitalization by even 25% it would be cost-saving in this setting.

Sensitivity analysis

The sensitivity analysis showed higher costs for all interventions when using time sheets, highlighting the importance of methodologies for measuring personnel time and costs. Lower salaries for lay counsellors make their involvement more cost-effective than nurse counsellors. We assumed that the quality and effectiveness of their counselling is similar to nurse counsellors as has been found previously in South Africa.⁴⁷

The adherence to IPT at the PHC clinic was 57%. Setting adherence at 62%, as for other studies, the cost per person completing IPT remained higher (US$ 85) than in Uganda (US$ 24)¹⁹ and Zambia (US$ 42).¹⁸ High adherence levels (>95%) have been obtained with antiretroviral programmes in South Africa through adherence counselling, support groups, pill boxes, drug identification charts, daily schedules, diaries and treatment literacy educational materials.⁴⁸ We suggest that the cost-effectiveness of similar interventions to improve adherence to IPT be evaluated.

Limitations

One limitation of this study was that the three facilities were purposively sampled rather than randomly selected. While assessing the generalizability of the results, the following factors should be considered: urban/periurban/rural setting, HIV prevalence, TB incidence, number of staff and salary levels. The following factors would decrease cost-effectiveness: settings with fewer patients per staff member, lower HIV prevalence, lower TB incidence and higher salaries.

Our estimates of the impact of VCT on HIV prevention were based on a study³⁹ that measured changes in risk behaviours and estimated changes in HIV incidence in Kenya, United Republic of Tanzania, and Trinidad and Tobago. We do not know if VCT in Cape Town was as effective as in that study, but a similar model of risk-reduction counselling was used and the communities had a similar HIV prevalence.

Another limitation of our study was that confidence intervals were not calculated for ascertaining the degree of certainty of cost-effectiveness estimates.

Policy implications

A study from Botswana showed that of the 560 clients screened only one case of TB was detected by chest X-ray and a large proportion of clients (18%) were lost to follow-up.⁴⁹ Our sensitiviity analysis found that not using chest X-rays (with or without PPD) was the most cost-effective IPT screening protocol. It decreased the cost per TB case prevented by 36%. We recommend that the requirement for chest X-rays as part of the screening process for IPT in the WHO guidelines be removed.

Although excluding PPD decreases the cost per person completing IPT by 60%, the cost per TB case prevented decreases only by 4% because the efficacy of IPT is lower. PPD increases costs for screening (tuberculin, syringes, needles, personnel time) and for patients (time and transport to return for skin reactions to be read). There are also technical problems in administering and reading the test correctly. However, not doing PPD exposes many people to isoniazid (with its potential side effects) who might not benefit from it and increases the burden on health services (more people starting isoniazid and being followed up). Given that the cost-effectiveness of both approaches is similar, WHO's recommendations remain appropriate: PPD testing should be done, and where it is not feasible, it can be omitted when the prevalence of TB infection is greater than 30% or in high-risk groups.

To assess the affordability of interventions, we suggest that average cost data be combined with the number of people eligible for each intervention and compared with the available resources. Interventions are likely to become more cost-effective as the number of people accessing services increases.

Conclusions

We conclude that VCT using lay counsellors and rapid HIV testing is a cost-effective intervention to prevent HIV and TB in South Africa. VCT services should be expanded for prevention and to link HIV-positive clients to care and support. Cost-saving interventions such as, ICF, IPT and CPT should be offered at all primary health care facilities in South Africa for HIV-positive clients. The use of chest X-rays for IPT screening decreases the cost-effectiveness of IPT. PPD screening does not influence the cost-effectiveness of IPT to prevent TB.

Our results prompted the South African Department of Health in December 2003 to include ProTEST interventions as part of the comprehensive package of care for people living with HIV linked to provision of antiretrovirals in South Africa.

Acknowledgements

We would like to acknowledge the contribution of the late Dr Barbara Karpakis whose enthusiasm and professional excellence helped ensure the success of the pilot district activities. Special thanks go to Susan Cleary (Health Economics Unit, University of Cape Town) for assistance in data collection, Catherine Goodman (London School of Hygiene and Tropical Medicine (LSHTM)) for assistance in discounting future health benefits and to Liz Corbett (LSHTM) for assistance in modelling the number of TB cases averted by preventing HIV infections through VCT. The research could not have been done without the cooperation and assistance of managers and health workers in the City of Cape Town who delivered ProTEST services and provided information on service utilization and time allocation.

Funding: The TB/HIV Pilot Districts were funded by the South African Department of Health. Many departmental officials provided valuable inputs and guidance at national, provincial and municipal (City of Cape Town) levels. Support was also received from the Canadian Institutes for Health Research, the Department for International Development, the LSHTM and the Stop TB Department of the World Health Organization.

Competing interests: none declared.

References

1. Department of Health. National HIV and syphilis antenatal sero-prevalence survey in South Africa, 2004. Pretoria: Department of Health; 2005.        [ Links ]

2. Joint United Nations Programme on HIV/AIDS (UNAIDS). 2004 report on the global AIDS epidemic: 4th global report. Geneva: UNAIDS/World Health Organization; 2004.        [ Links ]

3. Selwyn PA, Hartel D, Lewis VA, Schoenbaum EE, Vermund SH, Klein RS, et al. A prospective study of the risk of tuberculosis among intravenous drug users with human immunodeficiency virus infection. New Engl J Med 1989;320:545-50.        [ Links ]

4. Daley CL, Small PM, Schecter GF, Schoolnik GK, McAdam RA, Jacobs WR Jr, et al. An outbreak of tuberculosis with accelerated progression among persons infected with the human immunodeficiency virus. An analysis using restriction-fragment-length polymorphisms. New Engl J Med 1992;326:231-5.        [ Links ]

5. Bleed D, Dye C, Raviglione M. Dynamics and control of the global tuberculosis epidemic. Curr Opin Pulm Med 2000;6:174-9.        [ Links ]

6. World Health Organization. Global tuberculosis control: surveillance, planning, financing. WHO Report 2005. Geneva: WHO; 2005. WHO document WHO/HTM/TB/2005.349.        [ Links ]

7. Kustner HGV. The trend of tuberculosis through the eyes of cohorts. Epidemiol Comments 1992;19:184-97.        [ Links ]

8. World Health Organization. Report of the 2005 annual TB review of the National Tuberculosis Control Programme of South Africa 2-12 October 2005. Geneva: World Health Organization and South African Department of Health; 2005.        [ Links ]

9. World Health Organization. Joint review of the South African Tuberculosis Control Programme. Geneva: South African Department of Health and WHO; 1996.        [ Links ]

10. Department of Health. National STD/HIV/AIDS Review. Pretoria: Department of Health; 1997.        [ Links ]

11. Hausler H. Lessons learned in South African TB/HIV pilot districts. S Afr Resp J 2002;8:17-22.        [ Links ]

12. World Health Organization. Report of a "Lessons Learnt" Workshop on the Six ProTEST Pilot Projects in Malawi, South Africa and Zambia. Geneva: WHO; 2004.        [ Links ]

13. Sweat M, Gregorich S, Sangiwa G, Furlonge C, Balmer D, Kamenga C, et al. Cost-effectiveness of voluntary HIV-1 counselling and testing in reducing sexual transmission of HIV-1 in Kenya and Tanzania. Lancet 2000;356:113-21.        [ Links ]

14. Wilkinson D, Wilkinson M, Lombard C, Martin D, Smith A, Floyd K, et al. On-site testing in resource-poor settings: is one rapid test enough? AIDS 1997;11:377-81.        [ Links ]

15. Bell JC, Rose DN, Sacks HS. Tuberculosis preventive therapy for HIV-infected people in sub-Saharan Africa is cost-effective. AIDS 1999;13:1549-56.        [ Links ]

16. Masobe P, Lee T, Price M. Isoniazid prophylactic therapy for tuberculosis in HIV-seropositive patients — a least-cost analysis. S Afr Med J 1995;85:75-81.        [ Links ]

17. Heymann SJ. Modelling the efficacy of prophylactic and curative therapies for preventing the spread of tuberculosis in Africa. Trans R Soc Trop Med Hyg 1993;87:406-11.        [ Links ]

18. Foster S, Godfrey-Faussett P, Porter J. Modelling the economic benefits of tuberculosis preventive therapy for people with HIV: the example of Zambia. AIDS 1997;11:919-25.        [ Links ]

19. Aisu T, Raviglione M, van Praag E, Eriki P, Narain JP, Barugahare L, et al. Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre. AIDS 1995;9:267-73.        [ Links ]

20. Bachmann MO. Effectiveness and cost effectiveness of early and late prevention of HIV/AIDS progression with antiretrovirals or antibiotics in Southern African adults. AIDS Care 2006;18:109-20.        [ Links ]

21. Department of Health, Provincial Administration of the Western Cape. The provincial and district HIV antenatal survey Western Cape 2002. Cape Town: Department of Health; 2002.        [ Links ]

22. Kumaranayake L, Pepperall J, Goodman H, Mills A. Costing guidelines for HIV/AIDS prevention strategies. UNAIDS Best Practice Collection — Key Materials. Geneva: UNAIDS; 2000.        [ Links ]

23. Du Preez M. The discount rate for public sector conservation projects in South Africa. Afr Dev Rev 2004;16:456-71.        [ Links ]

24. Kumaranayake L. The real and the nominal? Making inflationary adjustments to cost and other economic data. Health Policy Plan 2000;15:230-4.        [ Links ]

25. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, et al. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med 2003;163:1009-21.        [ Links ]

26. Wood R, Maartens G, Lombard CJ. Risk factors for developing tuberculosis in HIV-1 infected adults from communities with a low or very high incidence of tuberculosis. J Acquir Immune Defic Syndr 2000;23:75-80.        [ Links ]

27. Dye CS, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 1999;282:677-86.        [ Links ]

28. Vynnycky E, Fine PE. The natural history of tuberculosis: the implications of age-dependant risks of disease and the role of reinfection. Epidemiol Infect 1997;119:183-201.        [ Links ]

29. Borgdorff MW, Fine PE, Leung CC, Tam CM, Vynnycky E. Slow decline of tuberculosis notifications in Hong Kong, China: evidence for a high risk of developing reactivational disease? TSRU Prog Rep 2003;123-40.        [ Links ]

30. Sonnenberg P, Glynn JR, Fielding K, Murray J, Godfrey-Faussett P, Shearer S. How soon after infection with HIV does the risk of tuberculosis start to increase? A retrospective cohort study in South African gold miners. J Infect Dis 2005;191:150-8.        [ Links ]

31. Wilkinson D, Squire SB, Garner P. Effect of preventive treatment for tuberculosis in adults infected with HIV: systematic review of randomized placebo controlled trials. BMJ 1998;317:625-9.        [ Links ]

32. Styblo K. Epidemiology of tuberculosis. The Hague: VEB Gustav Fischer Verlag Jena; 1984.        [ Links ]

33. Corbett EL, Charalambous S, Moloi VM, Fielding K, Grant AD, Dye C, et al. Human immunodeficiency virus and the prevalence of undiagnosed tuberculosis in African gold miners. Am J Respir Crit Care Med 2004;170:673-9.        [ Links ]

34. City of Cape Town. City of Cape Town/Metropole region TB control programme progress report 1997–2002. Cape Town: City of Cape Town; 2003.        [ Links ]

35. Bucher HC, Griffith LE, Guyatt GH, Sudre P, Naef M, Sendi P, et al. Isoniazid prophylaxis for tuberculosis in HIV infection: a meta-analysis of randomized controlled trials. AIDS 1999;13:501-7.        [ Links ]

36. World Health Organization. Preventive therapy against tuberculosis in people living with HIV. Wkly Epidemiol Rec 1999;74:385-400.        [ Links ]

37. Gravelle H, Smith D. Discounting for health effects in cost-benefit and cost-effectiveness analysis. Health Econ 2001;10:587-99.        [ Links ]

38. van Hout BA. Discounting costs and effects: a reconsideration. Health Econ 1998;7:581-94.        [ Links ]

39. Voluntary HIV-1 Counseling and Testing Efficacy Study Group. Efficacy of voluntary HIV-1 counselling and testing in individuals and couples in Kenya, Tanzania, and Trinidad: a randomised trial. Lancet 2000;356:103-12.        [ Links ]

40. Gilson L, Mkanje R, Grosskurth H, Mosha F, Picard J, Gavyole A, et al. Cost-effectiveness of improved treatment services for sexually transmitted diseases in preventing HIV-1 infection in Mwanza Region, Tanzania. Lancet 1997;350:1805-9.        [ Links ]

41. Marseille E, Kahn JG, Mmiro F, Guay L, Musoke P, Fowler MG, et al. Cost effectiveness of single dose nevirapine regimen for mothers and babies to decrease vertical transmission in sub-Saharan Africa. Lancet 1999;354:803-9.        [ Links ]

42. Sinanovic E, Floyd K, Dudley L, Azevedo V, Grant R, Maher D. Cost and cost-effectiveness of community-based care for tuberculosis in Cape Town, South Africa. Int J Tuberc Lung Dis 2003:9 Suppl 1:S56-62.        [ Links ]

43. Wiktor SZ, Sassan-Morokro M, Grant AD, Abouya L, Karon JM, Maurice C, et al. Efficacy of trimethoprim-sulphamethoxazole prophylaxis to decrease morbidity and mortality in HIV-1-infected patients with tuberculosis in Abidjan, Côte d'Ivoire: a randomised controlled trial. Lancet 1999;353:1469-75.        [ Links ]

44. Anglaret X, Chene G, Attia A, Toure S, Lafont S, Combe P, et al. Early chemoprophylaxis with trimethoprim-sulphamethoxazole for HIV-1-infected adults in Abidjan, Côte d'Ivoire: a randomised trial. Lancet 1999;353:1463-8.        [ Links ]

45. Badri M, Ehrlich R, Wood R, Maartens G. Initiating cotrimoxazole prophylaxis in HIV-infected patients in Africa: an evaluation of the provisional WHO/UNAIDS recommendations. AIDS 2001;15:1143-8.        [ Links ]

46. Cleary S, Boulle A, McIntyre D, Coetzee D. Cost-effectiveness of antiretroviral treatment for HIV-positive adults in a South African township. Cape Town: Health Systems Trust; 2003. Available from: http://www.hst.org.za/publications/579.        [ Links ]

47. Richter L, Durrheim K, Giresel D, Solomon V, Van Rooyen H. Evaluation of counselor service provision within the South African context (with specific reference to HIV/AIDS counseling). Pretoria: Department of Health; November 1999.        [ Links ]

48. Coetzee D, Boulle A, Hildebrand K, Asselman V, Van Cutsem G, Goemaere E. Promoting adherence to antiretroviral therapy: the experience from a primary care setting in Khayelitsha, South Africa. AIDS 2004;18 Suppl 3:S27-S31.        [ Links ]

49. Mosimaneotsile B, Talbot EA, Moeti TL, Hone NM, Moalosi G, Moffat HJ, et al. Value of chest radiography in a tuberculosis prevention programme for HIV-infected people, Botswana. Lancet 2003;362:1551-2.        [ Links ]

50. Department of Health. Operational Plan for comprehensive HIV and AIDS care, management and treatment for South Africa. Pretoria: Department of Health; 2003.        [ Links ]

(Submitted: 5 October 2004 – Final revised version received: 17 November 2005 – Accepted: 21 November 2005)

1 Correspondence to this author (email: hhausler@uwc.ac.za).