HERA

HEALTH RESEARCH FOR ACTION .

THE PROGRAMME TO DEVELOP INTEGRATED HIV CARE FOR TUBERCULOSIS PATIENTS LIVING WITH HIV/AIDS OF THE INTERNATIONAL UNION AGAINST TUBERCULOSIS AND LUNG DISEASE

EVALUATION OF THE PROJECT IN THE REPUBLIC OF BENIN AND THE DEMOCRATIC REPUBLIC OF THE CONGO November 2008 Laarstraat 43

tel. +32-3-8445930

B-2840 Reet Belgium

fax. *32-3-8448221

Bank No 401-2025551-15

e-mail [email protected] www.hera.eu

IHC / The Union / Benin & DRC

Acknowledgments The preparation of this report was only made possible by the splendid support I received from The Union, Alter Santé, and the National Tuberculosis Programmes in Benin and the DRC, including the IHC Project Coordinators. The list of people to whom to express my gratitude is long, but I am especially grateful for the support during my field missions by Dr. Capo-Chichi in Benin and by Drs. Ruppol, Ngoma, and Kambale in the DRC. At the time of preparing the final version of this report, the war in the Region of Nord Kivu in the DRC has again flared up. Many of the health workers I met and interviewed just a few weeks ago are now living the reality of a brutal armed conflict. My heart goes out to these courageous women and men.

Josef Decosas

HERA / November 2008

IHC / The Union / Benin & DRC

Table of Contents Abbreviations and Glossary ________________________________________________ 3  1 

Executive Summary ___________________________________________________ 5 

2 

Main Findings and Recommendations ____________________________________ 7  2.1  Findings __________________________________________________________ 7  2.1.1  The IHC Project of The Union in Benin and the DRC __________________________ 7  2.1.2  The IHC Project in Benin _______________________________________________ 9  2.1.3  The IHC Project in the DRC ____________________________________________ 11  2.2  Recommendations _________________________________________________ 15  2.2.1  Funding HIV and Tuberculosis Programmes by the European Commission _______ 15  2.2.2  The IHC Programme of The Union _______________________________________ 15  2.2.3  The IHC Project in Benin ______________________________________________ 15  2.2.4  The IHC Project in the DRC ____________________________________________ 16 

3 

Introduction and Context ______________________________________________ 3.1  Introduction to IHC _________________________________________________ 3.2  Objectives of the Evaluation _________________________________________ 3.3  Objectives of IHC __________________________________________________ 3.3.1  Discussion of the Objectives ___________________________________________

19  19  19  19  20 

4 

Piloting Integrated HIV Care Services ____________________________________ 4.1  Description and Objectives __________________________________________ 4.1.1  Discussion of the Objectives ___________________________________________ 4.2  Status and Achievements ___________________________________________ 4.2.1  Benin _____________________________________________________________ 4.2.2  DRC ______________________________________________________________

23  23  24  25  25  29 

5 

Laboratory Research and Capacity Development __________________________ 5.1  Description and Objectives __________________________________________ 5.1.1  Discussion of the Objectives ___________________________________________ 5.2  Status and Achievements ___________________________________________ 5.2.1  Benin _____________________________________________________________ 5.2.2  DRC ______________________________________________________________

35  35  35  37  37  38 

6 

Social and Economic Research _________________________________________ 6.1  Description and Objectives __________________________________________ 6.1.1  Discussion of the Objectives ___________________________________________ 6.2  Status and Achievements ___________________________________________ 6.2.1  Benin _____________________________________________________________ 6.2.2  DRC ______________________________________________________________

41  41  42  43  43  44 

Annex 1. 

Calculation of Cohort Survival Curves in the DRC ___________________ 1 

Annex 2. 

Patient Record Card Developed by the IHC Project __________________ 5 

HERA / November 2008

1

IHC / The Union / Benin & DRC

2

HERA / November 2008

IHC / The Union / Benin & DRC

Abbreviations and Glossary AIDS

Acquired Immunodeficiency Syndrome

Alter

Alter Santé Internationale & Développement (health consulting company based in France)

ART

Anti-Retroviral Therapy

ARV

Anti-Retroviral Drug

BCZS

Bureau de coordination de zone de santé (DRC)

CBCA

Communauté des Eglises Baptistes au Centre d’Afrique (DRC)

CD4 count

Number of lymphocytes with a specific surface marker per millilitre of blood. Used as an indicator of the level of immune depletion in HIV infection.

CDT

Centre de Dépistage et Traitement (tuberculosis clinic - Benin)

CDV

Conseil et dépistage (du VIH) volontaire (see VCT)

CIF-Santé

Communication, Information, Formation (local NGO based in Goma, DRC)

CIPEC

Centre d’Information de Prospection et de Conseil sur IST et Sida (Regional HIV Counselling and Testing Centre, Benin)

CNHPP

Centre National Hospitalier de Pneumo-Phtisiology (national chest hospital - Benin)

CPLT

Coordination provinciale de lutte contre la tuberculose (DRC)

CPP

Centre de Pneumo-Phtisiology (chest hospital – Benin)

CSC

Centre de Santé Communautaire (Second level of primary care in Benin)

CSDT

Centre de Santé de Dépistage et Traitement (tuberculosis clinic – DRC)

CyFlow®

A flow-based automatic CD4 assay system manufactured by Partec (headquarters in Germany)

DALY

Disability-adjusted life year

DMISI

Département Maladies Infectieuses et Santé Internationale / Université Montpellier

DOTS

Directly Observed Treatment, Short Course (for tuberculosis)

DRC

Democratic Republic of the Congo

Dynabeads®

A Dynal assay system manufactured by Invitrogen Life Science (headquarters in California)

Dynal assay

A manual CD4 assay method using a light microscope and magnetic beads

EC

European Commission

EDF

European Development Fund

ELISA

Enzyme-Linked Immunosorbent Assay

GFATM

Global Fund to Fight AIDS, Tuberculosis and Malaria

HERA / November 2008

3

IHC / The Union / Benin & DRC

HIV

Human Immunodeficiency Virus

IHC

Programme to Develop Integrated HIV Care for Tuberculosis Patients Living with HIV/AIDS (the abbreviation is used by The Union to refer to both, a programme of work in five countries, and to country projects implemented under this programme)

INH

Isoniazid

IUATLD

International Union Against Tuberculosis and Lung Disease – (in the text of the report usually referred to as “The Union”)

MAP

Multisectoral AIDS Programme (World Bank)

MoU

Memorandum of Understanding

NACP

National AIDS Control Programme

NGO

Non-Governmental Organisation

NTP

National Tuberculosis Programme

PCR

Polymerase Chain Reaction (a method to detect and quantify nucleic acid molecules in a biological specimen)

PNLS

Programme National de Lutte Contre le Sida (see NACP)

PNMLS

Programme National Multisectoriel de Lutte Contre le Sida (DRC)

PNT

Programme National Contre la Tuberculose (see NTP)

PTB

Pulmonary Tuberculosis

RDC

République Démocratique du Congo (see DRC)

TB

Tuberculosis

UICTMR

Union Internationale Contre la Tuberculose et Maladies Respiratoires (see “The Union”)

UNAIDS

Joint United Nations’ Programme on HIV/AIDS

Union (The)

The International Union Against Tuberculosis and Lung Disease

USD

United States Dollar

VCT

Voluntary (HIV) counselling and testing

WHO

World Health Organisation

4

HERA / November 2008

IHC / The Union / Benin & DRC

1 Executive Summary The Integrated HIV Care Project (IHC) is an operational research project of the International Union against Tuberculosis and Lung Disease financed by the European Commission and implemented in Benin and in the Provinces of Bas Congo and Nord Kivu in the DRC. It is part of a larger programme of The Union to apply important lessons from the development of the DOTS strategy to the care of people living with HIV. After three years of implementation the project will end in November 2008, but the programme in Nord Kivu will continue under a second project also financed by the EC. IHC is a partnership project of The Union supported by Alter Santé Internationale & Développement (Alter), a consulting company based in France, and the Département Maladies Infectieuses et Santé Internationale of the Université Montpellier (DMISI). Implementation of the project at country level is through the National Tuberculosis Programmes (NTPs) of the Ministries of Health in Benin and the DRC. It is governed by National Steering Committees that include the national TB and HIV programmes and other senior representation of the Ministry of Health. Research on social and economic issues is sub-contracted by The Union to local institutions. The project was designed to deliver three main results: 1. To establish the practice of integrated HIV and tuberculosis care in a limited number of health facilities in the three project zones, and to continuously evaluate the practice in order to “learn by doing”; 2. To assess the cost-effectiveness of providing integrated HIV and tuberculosis care; 3. To assess the sustainability of providing integrated care for tuberculosis and HIV. Pursuing the integration of tuberculosis and HIV care is not simply an issue of establishing greater collaboration between the health sector programmes for two diseases that affect an overlapping population of patients. It is an issue of changing the paradigm of the health sector response to HIV to bring it in line with the reality of anti-retroviral therapy. HIV care in developing countries needs to adopt a model of chronic disease care, a model for which there is limited experience in most developing countries, except in the care and treatment of tuberculosis and leprosy. A key to understanding the IHC project is the understanding of the concept of systemic piloting as opposed to technical piloting. The project was designed to be implemented within national health systems, governed by national health systems stakeholders. This is essential for piloting systemic change in contrast to piloting technical solutions for the delivery of care. The approach is based on the experience of The Union in developing the DOTS tuberculosis strategy as the foremost approach to tuberculosis control globally. This orientation of the IHC project was not always understood by the financial donor and by some of the national stakeholders who sometimes focused too narrowly on controlling inputs in order to maximise short-term treatment results for a limited number of patients. As the IHC project is coming to an end of its funding period, The Union requested an external evaluation to obtain an independent analysis of the results achieved and recommendations for future initiatives. The focus of this evaluation is on strategic issues. The main findings and recommendations are summarised in the following chapter. Because of national governance of the project, and because of differences in health systems development and resource flows in the two countries, the project took on different forms and generated different results in each of the project areas. The advantages of national governance of the project are clear. System changes can only be achieved by those who are controlling and managing the system. It is also clear that decentralisation of governance constrains project implementation and creates problems of accountability. It separates implementation authority from responsibility towards the financial donor, something that development financing organisations should understand well.

HERA / November 2008

5

IHC / The Union / Benin & DRC

The project has been very successful in opening a dialogue between national tuberculosis and national AIDS programmes in Benin and in the DRC. It has demonstrated convincingly that quality HIV care can be delivered by non-specialist staff with minimal access to technology in first and second line health care facilities. What it requires, similar to the delivery of tuberculosis care, is tight monitoring and supervision, a functioning supply chain, and staff that is sufficiently supported with training and remuneration. The demonstration of this successful model is only a first step. The next step should be a process of appropriation by The Union (in order to build HIV care into its core programme of cooperation with national tuberculosis programmes), by the Ministries of Health, the National Tuberculosis Programmes and the National AIDS Programmes in Benin and the DRC (in order to continue to scale up the process of integration of care), and by international financial donors (including the EC) to understand the burden that countries currently have to bear because of the practice of disease-specific international health development financing. The process of appropriation is not very advanced but there are some encouraging signs, especially in the DRC. The reprogramming of the World Bank MAP grant integrating some of the experiences of the IHC project is a major achievement. Nevertheless, the end of the IHC project in Benin and in Bas Congo creates an urgent need to find interim solutions to support the project achievements until systemic changes can support the scale up of integrated care. It would be unacceptable by any ethical standard and also by a public health rationale to discontinue the treatment of patients who started ART under the IHC project. The continuation of the project in a second phase in Nord Kivu provides an opportunity to consolidate some of the project results and create an evidence base that is even more robust. The introduction of manual CD4 counting technology in the DRC is a success, but there are still more questions to be answered. The project rationale which was at times imposed by the financial donor needs to be overcome in the second phase in order to allow effective piloting of integrated HIV and tuberculosis care. It is indefensible that most facilities supported by the IHC project continue to provide differential HIV care to patients who entered through the TB clinic, and to those who entered through other services. The second phase of the IHC project may also provide an opportunity to review the social and economic research component. Clearly, HIV infection and tuberculosis affect not only the lungs and lymph nodes, but also family and community relationships, and the psychological, emotional, spiritual, social, and economic health of patients. Information about these dimensions remains largely anecdotic, but a programme to pilot treatment options cannot ignore them. The execution of the longitudinal studies to explore the social and economic dimension of illness in the first phase of the IHC project was flawed. But this should be a lesson on what to do better in the second phase. The IHC project in the DRC has established ARV treatment cohorts that will have two years of follow up when the project ends. They are a valuable resource to study the real benefits of universal access to anti-retroviral therapy, a goal that receives massive international financial support on the basis of little evidence about its effectiveness. The second phase of the IHC project has the potential of documenting the effectiveness of ART in resource-limited settings with an unprecedented level of precision and validity. The component of laboratory capacity building of the IHC project has achieved a significant result in establishing the feasibility of manual CD4 counting technology in peripheral health facilities in the DRC. The introduction of the technology in Benin was not successful. The anticipated results to introduce routine virological testing and anti-retroviral drug resistance monitoring were not achieved because the technology of specimen transport using dried blood spots requires further research and development. Nevertheless, the project will generate important information about the adequacy of current treatment strategies in suppressing the viral load of patients on ART in Benin and the DRC, and about the current profile of anti-retroviral resistance in the two countries.

6

HERA / November 2008

IHC / The Union / Benin & DRC

2 Main Findings and Recommendations 2.1 Findings 2.1.1 The IHC Project of The Union in Benin and the DRC HIV care Health workers interviewed during the evaluation in Benin and the DRC were unanimous in their statements that the integration of HIV care has improved the outcome of tuberculosis treatment. They usually cited better cure rates, higher survival, and less relapses. The project did not systematically track its impact on tuberculosis care, but an analysis of patient records in Nord Kivu showed a progressive increase of treatment success rates during the project period, primarily because of a decrease in patients who were lost to follow-up.1 The IHC project did attempt to systematically assess the outcome of HIV care. This assessment requires a longer follow-up of treatment cohorts than is usual for tuberculosis. It was only possible in the DRC. Here the project recorded a mortality of 155 to 250/1000 person-years for all co-infected patients, and of 138 to 225/1000 person-years for the subgroup placed on anti-retroviral therapy. These results are comparable to results obtained in other HIV treatment programmes in Africa.2 Social and economic research The Union missed a major opportunity to use the IHC project in Benin and the DRC to increase knowledge about the economic and social impact of HIV and tuberculosis, and about the real costs of the response to the national health system. The planned study was scaled back and under-resourced in terms of financial and technical support. The three longitudinal studies were intended to be cohort studies, selecting a population by exposure (tuberculosis and HIV care) and following it for outcome (social and economic situation). The two research partners in the DRC appear to have understood this, I am not certain about the partner in Benin. However, their focus on reaching a targeted number of interviews in each survey round seems to have overtaken the priority of keeping the cohort intact. The results are three sets of six cross-sectional questionnaire surveys that may yield some longitudinal data, however with an enormous waste of data collection and analytical work done by the partners.3 The experience in Nord Kivu in this context is enlightening. It is nearly impossible to find any information in the analysis of the questionnaire surveys of patients and health care staff by CIF Santé that is not reported in more detail in the single anthropological study conducted by C. Escoffier in 2008. This raises a question about the added value of the longitudinal studies in the form they were conducted. Laboratory research and capacity building The IHC project in the DRC has demonstrated the feasibility of introducing the method of manual CD4 counting in health facilities with low volumes of HIV care. The failure to establish the technology in Benin, on the other hand, demonstrated that successful introduction of the technology depends on its acceptability by laboratory staff and on high level national policy and programme support. While the project was very successful in this respect, some questions remain: What are the advantages of performing manual CD4 counts compared to providing ART on the basis of 1

Boillot F, Kabuayi J-P, Kiputsu K A, Ndogosieme A, Dlodlo R. Contribution of tuberculosis control programmes in scaling up HIV care in low-income countries: Preliminary results of the IUATLD Integrated HIV Care project in the Democratic Republic of the Congo. (submitted for publication)

2

Rosen S, Fox MP, Gill JC. Patient retention in antiretroviral programs in sub-Saharan Africa: a systematic review. PLoS Med 2007; 4(10):1691-1791 www.plosmedicine.org

3

There is still an opportunity for corrective action in Bas Congo where only three interview rounds have been done, and where there is still a possibility to focus on the established cohort rather than on trying to achieve a fixed number of interviews.

HERA / November 2008

7

IHC / The Union / Benin & DRC

clinical decisions alone, especially among tuberculosis patients? What are the limiting conditions under which manual counting is no longer feasible and the introduction of automated CD4 counting technology should be considered? The IHC project generated important evidence about the ability of current anti-retroviral treatment programmes in Benin and the DRC to provide effective viral load suppression. The high level of virological failure observed among patients in Cotonou is an issue of concern. The capacity for virological testing in both countries was strengthened, although the desired result to make these tests generally accessible was not achieved because the problem of appropriate and convenient transport of laboratory specimen has not yet been solved. Similarly, important information was generated about the profile of anti-retroviral drug resistance in both countries. Systematic access to resistance testing in Benin and the DRC will have to await solution of the specimen transport problem. Programme management The IHC project was designed to be implemented within national health systems, governed by a National Steering Committee. At the same time, it was an operational research project, designed to test approaches, methods and tools for care delivery. The decision to proceed in this manner was based on the experience of The Union in developing the DOTS tuberculosis strategy as the foremost approach to tuberculosis control globally. It was a rational choice, given the objective to pilot systemic changes rather than technical solutions. Nevertheless, it created a complex management task for the financial donor (the EC), The Union, and the National Tuberculosis Programmes. Because of national governance, and because of differences in health systems development and resource flows in the two countries, the project took on different forms and generated different results in each of the project areas. The advantages of national governance are clear. System changes can only be achieved by those who are controlling and managing the system. But the decentralisation of governance constrained project implementation and created problems of accountability. It located the primary project decision authority at the level of the National Steering Committees, while the responsibility to account for the proper use of funds remained with The Union headquarters in Paris. Secondly, the focus on implementation at national level sometimes distracted from the fact that this was a research project. The primary aim was not to provide good quality HIV care to a limited number of patients with tuberculosis, the primary aim was to use the experience and structural development of tuberculosis control services to improve HIV care in Benin and the DRC; in the first instance for co-infected patients, but in the end in a much more generalised fashion. The complexity of this management situation was not always understood by the European Commission. During the implementation of the project, there were long delays because of protracted negotiations on issues such as the change of the laboratory research protocol (because of unanticipated difficulties in applying the dried blood spot specimen transport method), and on issues such as the use of project resources for the HIV care and treatment of persons who had completed their tuberculosis therapy or for family members of tuberculosis patients. The complexity of the situation is, nevertheless, an unalterable part of the IHC programme. From a management position, it would be simpler to focus, on the one hand, on financing the delivery of services of known value, using financing instruments such as the Global Fund, and, on the other hand, maintain a strict research development financing stream. However, simple solutions usually fail in the face of complex issues. The disconnect between evidence (i.e. the result of health research), and health policy is well known. The IHC project has demonstrated that it can be overcome if the partners are willing to manage the complexity of the approach.

8

HERA / November 2008

IHC / The Union / Benin & DRC

2.1.2 The IHC Project in Benin Programme collaboration There is solid evidence that the IHC project in Benin has brought together two health programmes that had previously been managed in splendid isolation. That such an effort was even necessary should surprise, since tuberculosis clinics are the places where people living with HIV are concentrated, and HIV infection has a major impact on the outcome of tuberculosis treatment. The achievement of IHC in Benin is highly significant. Like all good ideas1, it did not need any special effort for “scaling up” from the pilot phase. Provider initiated HIV testing, counselling and co-trimoxazole prophylaxis for all HIV positive tuberculosis patients was quickly adopted by all tuberculosis clinics nationwide. Health facilities offering tuberculosis services moved to the top of the list of institutions being considered for the expansion of anti-retroviral treatment services. Staff working in TB care understood quickly that if they wanted their patients to survive and be cured of tuberculosis they needed to pay attention to the patients’ HIV status and engage in HIV care. A staff manual for the care and monitoring of tuberculosis and HIV co-infection was produced and widely distributed. There is another side to this equation that was not part of the IHC project nor of this evaluation. To what extent are people who test positive for HIV in voluntary testing and antenatal clinics in Benin systematically screened for tuberculosis? There is anecdotal evidence that there has been an increase in referrals from these services to tuberculosis clinics. But if any pre-screening is practiced at this level, it is based on questioning about cough. From a public health perspective this is useful because it increases the detection rate of infectious tuberculosis. From the perspective of providing adequate care to people living with HIV it is insufficient, because it misses those suffering from extra-pulmonary tuberculosis. There are still many challenges to overcome. There are problems in the supply chain for laboratory reagents, co-trimoxazole, and anti-retroviral drugs in Benin managed by the NACP. Although sufficient funds are available to procure these inputs, there are weaknesses in managing the procurement process. The IHC project has, until now, succeeded in assuring the continued supply of critical inputs to the participating tuberculosis clinics by employing short-term and ad hoc procedures and by using the tuberculosis drug logistics system. If the project should end, the health facilities will have to rely on the capacity of the Ministry of Health and the NACP to supply drugs and reagents in a timely manner. HIV care The IHC project in Benin has demonstrated convincingly that it is possible to decentralise clinical care of people living with HIV to non-specialised health facilities. This will be a key factor in the effort to increase the coverage of HIV care and treatment in Benin, especially outside major urban centres. The performance of nurses and laboratory staff in the tuberculosis clinics in testing, counselling and providing care for co-infected patients is exemplary. Staff and patient satisfaction are high, and if logistic support, supervision, and training for this initiative continue, an increasing number of people living with HIV in Benin will receive adequate care and treatment. The decision by the NACP and the NTP to limit HIV care in the tuberculosis clinics to the duration of TB treatment creates a challenge. It requires a well managed process of patient transfer after completion of tuberculosis treatment to avoid ARV treatment interruptions and loss to follow up. The process appears to function relatively smoothly in the health centres that have tuberculosis and HIV services co-located in the same facility. But in many instances patients are referred to another facility in a process that is very little systematised, involves considerable effort by the patient, and only works when it is supported by the personal engagement of tuberculosis clinic staff, well beyond their regular call of duty. 1

E.g. the cultivation of corn, cellular telephones, and rap music

HERA / November 2008

9

IHC / The Union / Benin & DRC

It is therefore not surprising that the cure of tuberculosis for many patients means the end of HIV care. We do not know how many patients are affected because the health information system does not capture the referral process. But early data from a yet to be completed follow-up study by the IHC project in Cotonou indicate that a significant proportion of patients receive no further HIV care once discharged from the tuberculosis clinic. Health information The decision to transfer co-infected patients after completion of tuberculosis treatment has made it impossible for the IHC project to monitor the effectiveness of HIV care provided under the IHC project. This would have required the long-term monitoring of treatment cohorts, a process not yet established by the National AIDS Programme. For the project itself the majority of patients leave the cohorts by transfer once they have completed their tuberculosis treatment and are essentially lost from sight. The IHC project in Benin can be credited with some success in re-aligning the health information system to allow the national authorities to follow the outcome of HIV and tuberculosis care more systematically. This will be critical in enabling the Ministry of Health to project the financial, human resource, and infrastructure needs for providing HIV care. The patient registers to permit such tracking of HIV care have been produced and distributed by the NACP in close collaboration with the NTP, but staff training has not yet taken place. It remains to be seen whether the new registers will change the paradigm of monitoring HIV care from the practice of tracking the virus to a more appropriate practice of tracking patients. Laboratory capacity building The Ministry of Health of Benin decided not to support the introduction of the manual CD4 counting technology and opted for an expansion of automated flow cytometry. This does not appear to have been a deliberate decision, but rather the de facto consequence of the resistance of laboratory staff to work with the Dynabead® system. Whether this decision was cost-effective and technically sound is not known. The main laboratory partner for the IHC project in Benin is the National Tuberculosis Reference Laboratory rather than with the National HIV Reference Laboratory. This apparently was a decision by the National Steering Committee. It is somewhat problematic. The outcome of the only collaboration with the HIV laboratory, to establish local capacity for monitoring the national HIV testing algorithms, is uncertain. At the same time, the choice to establish HIV virology testing capacity in the tuberculosis laboratory may have compromised the development of a closer collaboration with the HIV reference laboratory, which, after all, is responsible for decisions on what tests will be used for HIV diagnosis, what technology will be employed for CD4 monitoring, and how viral load testing will be introduced into the clinical services for people living with HIV. It appears certain that the collaboration of the DMISI with the National TB Reference Laboratory in Benin will result in some highly relevant studies that will provide important information about HIV care in Benin.1 However, it is unlikely that the collaboration will have a systemic impact on the routine practice of immunological and virological monitoring of people being treated for HIV infection in Benin. Social and economic research I was not able to gather much information about this component in Benin because there was not much documentation available. The persons responsible for implementation in Benin could not be reached. During the early phase of the IHC project, the expectations of results from this study were progressively reduced. My impression is that the studies in Benin will not be able to meet even these reduced expectations. From what I have been able to see, and I may not have been able to see everything, there is little indication that this research will produce useful results in Benin. 1

None of these studies were concluded at the time of this evaluation.

10

HERA / November 2008

IHC / The Union / Benin & DRC

2.1.3 The IHC Project in the DRC Programme collaboration As in Benin, the IHC project in the DRC has opened a previously non-existent dialogue between the National AIDS and the National Tuberculosis Programmes. The National AIDS Programme and the Endemic Disease Directorate of the Ministry of Health have endorsed many IHC instruments and procedures. The models of patient care, laboratory support, record keeping, supervision, and training introduced by the IHC project in the DRC have been largely accepted by the Ministry and the National AIDS Programme. The National HIV Reference Laboratory actively participates in training and supervision of laboratory staff for manual CD4 counting, and provides quality control for HIV testing performed at the IHC pilot sites. Clinical training of physicians and nurses is organised jointly, and the Director of the NACP has participated in joint HIV and TB programme supervision missions. The recent reprogramming of the World Bank MAP grant is informed by the experience of the IHC project, adopting for instance the procurement planning tools developed by IHC, as well as the suggestion to use the tuberculosis supply and distribution system for anti-retroviral drug logistics. The NACP has endorsed the option to establish the capacity for manual CD4 counting in all peripheral laboratories. Until now, the commitment by the National AIDS Programme to work towards an integrated care system for HIV using the tuberculosis model has not involved a major engagement of resources. The DRC does not have a well developed public HIV care infrastructure where these resources could be applied, and most international funds are channelled to NGOs. It is important that this commitment is not forgotten if and when major external funds for HIV care by the public sector in the DRC are mobilised. A test case will be the Province of Bas Congo where IHC financing is about to end but funds will become available through the reprogrammed World Bank MAP grant. Paying for services The main issue raised at each field visit site concerns financing of HIV and tuberculosis services. This is often misrepresented as an issue of staff bonuses (“primes”). In fact it is a much more fundamental issue of health sector financing in the DRC. The salaries of Government health workers in the DRC are negligibly small. The real staff salaries are the “primes” paid by institution with income derived from patient charges. When services are mandated to be free of charge to patients, they deprive the institution and the staff of income, even though drugs, reagents, and other inputs may be provided by the State. This loss of income by the institution not only affects staff salaries, it also affects the ability to pay for fuel to run the generators needed to perform CD4 assays, and to reimburse the travel and communications expenses of staff to trace defaulting patients. Services that are provided to patients free of charge are often seen by health administrators as extra-professional activities of staff members; something that is not part of their job and for which they should not be paid from the pool of institutional revenues. The National Tuberculosis Programme provides a payment to institutions on the basis of number of smear positive TB cases detected. This allows the institutions to provide some staff payments, but the amount is not commensurate with the work load. There are initiatives in the DRC to reform health sector financing, in part supported by the EC. These will take many years before they are implemented country-wide. Until then, health care in the DRC will continue to be delivered on a fee for service basis. Health administrators will continue to run the facilities as not-for-profit enterprises within very tight financial margins. There is no room for a no-cost service in such a system. Somebody will always have to pay the cost. In the IHC project, this cost has been paid by nurses, laboratory technicians, and doctors in the form of foregone salaries and personal expenditures for cellphone air-time units and motor-taxi fares to trace patients. This is clearly not sustainable. In order to assure the delivery of quality care for patients with HIV and with tuberculosis, the National AIDS and Tuberculosis Programmes will have to find

HERA / November 2008

11

IHC / The Union / Benin & DRC

a way to pay health institutions for the care they are providing free of charge to patients. This may be a temporary solution, but it is essential if HIV care is to be scaled up, and it is urgent if the IHC project is to continue in Nord Kivu. HIV care The outcome of HIV care as monitored by the IHC project in the DRC is already mentioned above under 2.1.1. But there is an additional issue that was frequently raised in interviews during the field visits. There are clear indications that nutritional support would improve treatment outcomes for many patients living in abject poverty, be they tuberculosis patients, HIV patients, or co-infected patients. Adding the task of providing nutritional support to an already overstretched health service may not be a feasible option. However, in many locations there are religious groups or associations of “amies de l’hôpital” providing social support to indigent in-patients. This is an institutional platform of civil society that could be further developed to provide food supplements to people in need. International funding for social support of people living with HIV and for tuberculosis patients is available. It would take some visionary leadership by institutions such as the PNMLS and the Country Coordinating Mechanism of the Global Fund to access these funds and engage them for the social support to patients based on need rather than pathology. Health information management The clinical record cards introduced by the IHC project in the DRC have proven themselves very useful. They are not specific to tuberculosis patients and their utility is appreciated by the Ministry of Health at all levels. They have not been introduced into any HIV services outside the IHC project, but that appears to be primarily an issue of resources. The IHC project has introduced project registers which are, at the moment, the only available registers that allow long-term monitoring of HIV care. Meanwhile, the NACP has received the templates for cohort monitoring of HIV care from the WHO but has not yet proceeded to adapt them to the national context. Several health workers interviewed have complained about the heavy load of paper work in the IHC project. The project demands up to five different quarterly reports, some of them quite extensive. These reports have allowed the project to perform some very detailed analyses about the evolution of patient and staff parameters. It is, however, time that this data collection be rationalised. Laboratory capacity building The IHC project in the DRC has demonstrated convincingly that manual CD4 counting using the existing tuberculosis laboratory personnel and infrastructure is feasible. It is the most appropriate option in health facilities that are following small numbers of people living with HIV. The current practice of using this test only for patients who entered into care through the tuberculosis clinic is unacceptable and defies all rationale. However, when this test is offered to all HIV patients, the case load in some facilities may increase to a level where it is no longer feasible in terms of laboratory staff time. In some locations automated cytometry may become the better option. This needs to be evaluated case by case. The support provided by the IHC project to increase the capacity for CD4 counting also raises a question of priorities. Although the CD4 count is a useful and desirable test for the care of people living with HIV, blood biochemistry tests are equally desirable, especially when patients are placed on anti-retroviral therapy. These tests are often not available. Where they are, patients have to pay for them (up to USD 50 for a full blood chemistry profile in Goma). Starting all HIV positive tuberculosis patients on anti-retroviral therapy without performing a CD4 count may not change the clinical outcome, yet the availability of liver function tests for patients placed on Nevirapine may prevent some potentially fatal complications of ART.

12

HERA / November 2008

IHC / The Union / Benin & DRC

The collaboration of the IHC project with the National HIV Reference Laboratory is exemplary. The reference laboratory is involved in all aspects of training and quality control in support of laboratory services at IHC pilot sites, and is leading on several of these. The National HIV Reference Laboratory already had a very high level of capacity for virological testing and research, a capacity that has been strengthened further by the collaboration with DMISI. Social and Economic Research Both research partners in the DRC appear to have understood the methodology of a longitudinal (cohort) study and have attempted to follow cohorts of patients. This was somewhat more successful than in Benin because patients in the DRC continue their HIV care in the tuberculosis clinic where they started. However, even in the DRC the researchers appeared to be preoccupied with reaching a predetermined sample size in each interview round, rather than with maintaining the cohort. This has led to much unnecessary and meaningless analytical work, comparing data from different survey rounds. There is still a chance of recovery because the last survey rounds in the two research areas have not yet been conducted. The findings of the studies in the Congo are as expected: Patients are already poor at the start. As they fall ill, they lose their income and spend money on seeking care. They sell their belongings, accumulate debts, and become destitute. This seems to be more marked in Nord Kivu than in Bas Congo where the family support net is able to soften some of the economic blows. As patients enter into tuberculosis and HIV care, their personal expenditures on health care decrease, but there is little information about the development of their general economic and social status. The staff surveys confirm high job satisfaction, a manageable work load, but much dissatisfaction with the level of remuneration. The in-depth socio-anthropological study conducted in North Kivu in the beginning of 2008 confirms these general findings and documents them in an even more comprehensive and contextualised manner.

HERA / November 2008

13

IHC / The Union / Benin & DRC

14

HERA / November 2008

IHC / The Union / Benin & DRC

2.2 Recommendations 2.2.1 Funding HIV and Tuberculosis Programmes by the European Commission The European Commission (and other development financing institutions) The EC 1. should take note of the considerable progress that has been made by the IHC Project in Benin and the DRC in developing approaches and instruments to integrate HIV and tuberculosis care in the two countries. It should also take note of the programmatic requirements of achieving this progress, and of the need for further development. It should consider continued funding of this type of operational research and systemic piloting. This can neither be achieved through scaling up service delivery, for which the Global Fund provides a suitable financing instrument, nor through the current mode of health research financing and technical piloting, which is usually too far removed from national health systems governance and management. 2.2.2 The IHC Programme of The Union The Union 1. Tuberculosis programmes have a well established methodology to monitor treatment outcomes. This methodology, however, must be adapted in order to apply it to HIV care. This was only achieved by the IHC project in the DRC. The results are quite unique in the context of HIV care in Africa, but they are still based on small numbers and short observation periods. The effort to collect this information should continue with a thorough review of the methodology for data collection and analysis. The Union 2. The decision of The Union to include a component of social and economic research in the project was a good decision that unfortunately suffered from being progressively scaled back and receiving insufficient technical support. It would be useful to include the longitudinal follow-up of the social and economic situation of patients in future IHC pilot projects, as well as to include a cost comparison between different models of delivering HIV care. However, this would need to be done more methodically, drawing lessons from the experience of the IHC social and economic research in Benin and the DRC. The Union 3. The Union should use its experience in the DRC to continue to promote and support the introduction of manual CD4 counting technology in situations where testing volume is low and where the technology is acceptable to national health systems managers and local laboratory staff. However, in future IHC projects, The Union should also attempt to generate more information about the limits of the technology, and about cost-benefits compared to the two alternatives: automated CD4 testing, and no CD4 testing. 2.2.3 The IHC Project in Benin Benin 1. The IHC Project in Benin has achieved significant results in terms of changes in national health policy and has reached the limits of what can be achieved through a pilot project. It should be discontinued after the current financial envelope has been exhausted. Benin 2.

However, the good practice of providing integrated tuberculosis and HIV care in tuberculosis clinics that was introduced by the IHC project in Benin must not be abandoned and will need further financial and technical support. It should become a major focus of an application for a Global Fund grant, preferably already in Round 9. This means that the development of the grant proposal should start immediately.

Benin 3.

At the same time, more flexible and timely financial support for integrated tuberculosis and HIV care needs to be sourced to avoid interruptions of essential inputs in terms of training, supervision, drugs, and laboratory

HERA / November 2008

15

IHC / The Union / Benin & DRC

supplies. This should be supported by the core programme of cooperation between The Union and the National Tuberculosis Programme. In current jargon, this would be called “mainstreaming HIV” in the core support of The Union to Benin. Benin 4.

The very fruitful collaboration between the National Tuberculosis and the National AIDS Programmes in Benin should continue. This will depend primarily on national political will, but it could also benefit from international support. There are important decisions to be made by the Ministry of Health about the accreditation of further health facilities providing tuberculosis care for anti-retroviral therapy, about the mechanism of patient transfer after the completion of tuberculosis treatment, and most of all about the development of a patient-centred information system that will be able to adequately monitor patients once they enter into HIV care, be it through tuberculosis clinics, antenatal clinics, or any other service. This is an issue of health systems development that could also be the subject of a Global Fund application, with possible technical support provided by The Union.

2.2.4 The IHC Project in the DRC DRC 1. With the closure of the IHC project in Bas Congo, the continuation and scaling up of HIV care in this region becomes a major concern. Continued financing of HIV care in this region is apparently available through the World Bank MAP grant. The application of these funds in a manner that is informed by the IHC experience will require a close collaboration between the NTP and the NACP, and would benefit from technical support by The Union. DRC 2.

The National Tuberculosis Programme and the National AIDS Programme should find a common mechanism of financing the costs of services that are free of charge to patients. This may be an interim solution until a nationwide health systems financing reform takes effect. It should be developed and administered jointly by the two programmes. It should be adapted to current financial management practice in health institutions. This would require that it be designed as closely as possible to a system of third party payments of fees for service. The resources to finance such a scheme should be sought through an application for a Global Fund grant.

DRC 3.

The Multisectoral National AIDS Programme (PNMLS), the National Tuberculosis Programme, the Ministry of Social Affairs, and key civil society stake holders should start discussing the development of a social and nutritional support programme for patients in need. Such a programme should be accessible to all patients who are in urgent need of social and nutritional support irrespective of their diagnosis. A programme of this nature would be eligible for funding by the Global Fund, as long as there is a convincing argument that the majority of beneficiaries would be tuberculosis and HIV patients.

DRC 4.

The adoption of uniform patient record cards and of a uniform register for all patients in HIV care should be a priority for the Ministry of Health and the NACP. The format of the registers should be based on the WHO template with modifications that are informed by the experience of the IHC project. Until a new register is in place, the tuberculosis clinics providing HIV care should continue to use the IHC project registers. The volume of quarterly report forms, however, needs to be reduced. The reports should be limited to information that can be obtained from the registers and that can be verified by inspecting the registers.

DRC 5.

Support for the expansion of manual CD4 counting capacity in the DRC should continue using the existing TB control microscopy infrastructure in

16

HERA / November 2008

IHC / The Union / Benin & DRC

more than 1,500 health centres. The National AIDS Programme and the National Tuberculosis Programme should find a common solution to assure that these tests are available to all patients in HIV care in low volume clinical sites, regardless of their portal of entry. There is also a need to decide on priorities between creating access to CD4 counts and access to blood biochemistry tests without charge to patients on anti-retroviral therapy. DRC 6.

The IHC in its second phase in Nord Kivu should re-examine the decision to withdraw CD4 counting support from the Centre de Santé de Référence Maboya in Katwa Zone de Santé, Nord Kivu. The current case load of patients with HIV followed at the Centre is low, but this is precisely the situation where the availability of manual CD4 testing would be most useful. The Centre has trained staff and main-line electricity. It also has an extraordinarily busy obstetric service which provides an opportunity to expand HIV care beyond the tuberculosis clinic and include the ante-natal clinic as a major portal of entry into HIV care.

DRC 7.

There are two outstanding interview rounds of the social and economic research studies in the DRC, Round 4 in Bas Congo and Round 6 in Nord Kivu. Both are scheduled for November 2008. They provide an opportunity to take corrective action by selecting only individuals for interviews who are part of an existing cohort, i.e. who have been followed for at least one year. This would decrease the number of interviews and the cost to researchers and to patients. It may allow some recovery of the longitudinal research design. At the same time, the two research teams should only analyse cohort data rather than comparing cross-sectional results of different interview rounds. These analyses are quite meaningless because the sample of 60 patients in each round is not representative of the population in treatment.

HERA / November 2008

17

IHC / The Union / Benin & DRC

18

HERA / November 2008

IHC / The e Union / Ben nin & DRC

3 Inttroducttion and d Contex xt 3.1 In ntroduction to IH HC The Inte egrated HIV V Care Proje ect (IHC) is an operatio onal researcch project o of the Intern national Union against a Tube erculosis an nd Lung Dissease (IUAT TLD – referred to as “T The Union) financed d by the Eu uropean Com mmission (E EC) under the t project title t “Sante//2004/078-5 547 Program mme to Devvelop Integrrated HIV Care C for Tuberculosis Patients P Livin ng with HIV V/AIDS”. The pro oject is imple emented in three proje ect areas in Benin, the DRC/Bas C Congo, and the DRC/No ord Kivu, ea ach with its own field co oordinator. The cost off the projectt over three e years is estim mated at € 4..8M. The EC C approved d a grant forr € 4.3M in support s of tthe project in Decemb ber 2004 bu ut the start of o field implementation n was delaye ed and the project was s thereforre extended d to the end of Novemb ber 2008. For The e Union, the e IHC projecct is a comp ponent of a programme e of work im mplemented in Benin, Myanmar, M U Uganda, the e DRC, and Zimbabwe and funded d from seve eral sources s, includin ng a second d EC Grant (IHC ( 2) for work w in the DRC/Nord Kivu and Z Zimbabwe. The Union’s Prog gramme to Develop In ntegrated HIV H Care for Tuberculosis Patien nts Living with w HIV/AIDS

IHC is a partnershiip project. The T lead partner is The e Union, sup pported by A Alter Santé Internattionale & Dé éveloppeme ent (Alter), a consulting g company based in Frrance, and the t Départe ement Maladies Infectie euses et Sa anté Interna ationale of th he Universiité Montpelllier (DMISI)). Implemen ntation of the project att country lev vel is throug gh the Natio onal Tuberc culosis Program mmes (NTP Ps) of the Ministries of Health H of Be enin and the DRC. It iss governed by Nationa al Steering Committees C s that includ de representation of the e national A AIDS and tuberculosis progra ammes and d of other relevant depa artments of the Ministryy of Health. al and econo omic issuess is sub-con ntracted by The T Union iin three disttinct Researcch on socia contractts for each of the three e sites, to th he Ecole de Santé Pub blique de Kin nshasa for Bas B Congo, CIF-Santé,, a local NG GO, for Nord d Kivu, and to M. Tousssaint Dan, a an independ dent n. researccher in Benin

3.2 Objective O es of the Evaluatiion As the IHC project is coming to t an end off its funding g period, The Union has requested d an external evaluation n of the project to obtain an indepe endent anallysis of the results achieved and reccommendatiions for futu ure direction ns of the pro ogramme fo or integratio on of TB and d HIV care.

3.3 Objective O es of IHC The objective of the e IHC progrramme is to o reduce the e burden off HIV and tu uberculosis in low a to assisst the integra ation of care of person ns living with h HIV into general g income countries and health services. s Sp pecifically, the project in Benin and d the DRC aimed a to exxplore the fe easibility

HERA / November N 200 08

19

IHC / The Union / Benin & DRC

of delivering integrated health care for tuberculosis and HIV. The project was designed to deliver three main results:1 4. To establish the practice of integrated HIV and tuberculosis care in a limited number of health facilities in the three project zones, and to continuously evaluate the practice in order to “learn by doing”; 5. To assess the cost-effectiveness of providing integrated HIV and tuberculosis care; 6. To assess the sustainability of providing integrated care for tuberculosis and HIV. 3.3.1 Discussion of the Objectives Health care for HIV infection was developed within the paradigm of treatment and control of sexually transmitted diseases. This seemed sensible as HIV is indeed a sexually transmitted infection. With the advent of anti-retroviral therapy, however, HIV infection is increasingly turning into a long-term chronic condition.2 This is a development which should lead to a review and revision of treatment strategies and the necessary health systems infrastructure. Health care for people living with HIV requires screening of asymptomatic individuals, lifelong provision of medication, close and regular surveillance of treatment responses, and continuous support and control of treatment adherence. This is much more akin to the treatment of a non-communicable chronic condition such as hypertension than to the control of a sexually transmitted disease such as gonorrhoea. Health care systems in most developing countries are notoriously poor at treating chronic diseases. Although conditions such as hypertension, diabetes, or epilepsy are occupying an increasingly greater space in the overall burden of ill health in these countries, they are usually treated inadequately, if at all. The only area where most developing countries have successfully introduced systems that approach adequate care of chronic diseases is in the treatment and control of tuberculosis and leprosy.3 However, in the international discussions these programmes still suffer under the stigma of being “vertical”. Despite the fact that tuberculosis and leprosy care is fully integrated in primary care at the service level, it does require a specialist support structure. This will probably be the case for all chronic disease management programmes, including HIV. Development donors appear to have great problems in funding support programmes for the treatment of chronic diseases. We can speculate why this is the case. Probably the most important factor is the short horizon of most international health financing programmes which compels donors to treat even systemic initiatives such as health sector financing or health sector decentralisation in a verticalised fashion, i.e. management focused on meeting short term indicator targets. The integration of HIV and tuberculosis care is therefore timely and carries much promise. It should be explored by establishing operational pilot or learning-sites, but at the same time significant work needs to be done at the national system and policy level. Integration of treatment services can only take place if it is supported by the integration of national health information systems, reference laboratory services, procurement services, and health service management policies and practices. Integration at these levels requires some technical solutions, but most of all it requires policy reforms. These reforms will have to build on current programme management and financing practices. They will therefore invariably take different forms in different countries, and they will progress at different rates determined by a number of factors, including the political will to change.

1

See Logical Framework Analysis, EC Project Santé/2004/078-547 Programme to Develop Integrated HIV Care for Tuberculosis Patients Living with HIV/AIDS, 2004. 2

The Antiretroviral Therapy Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy in highincome countries: a collaborative analysis of 14 cohort studies. Lancet 2008; 372:293-99

3

Harries AD, Jahn A, Zachariah R, Enarson D. Adapting the DOTS framework for tuberculosis control to the management of non-communicable diseases in sub-Saharan Africa. PLoS Med 2008; 5(6):e124.

20

HERA / November 2008

IHC / The Union / Benin & DRC

Much of the health system work of integrating treatment services for HIV and tuberculosis is included in the activity package listed under the first result of the IHC logical framework. The pilot sites for joint HIV and tuberculosis care simply cannot work unless there is a commitment and action by the Ministry of Health to orient tuberculosis and HIV programmes towards the establishment of common information systems, laboratory support, drug and supply procurement and management support, health service financing, quality control, monitoring and supervision, human resources management, and other systemic health service functions. The remaining two results, to assess the cost effectiveness and the sustainability of the integrated care approach, are primarily designed to explore and support the necessary systems and policy changes. Cost-effectiveness is a parameter of comparison. It requires a comparison of costs among different approaches that have the same effect. In the context of medical care for HIV, the desired effect is the long-term disability-free survival of people living with the virus. The real cost of different approaches1 in the developing country context is unknown. A three year project like the IHC is not likely to generate useful data on the cost of keeping people alive who may require services for 30 or more years. Sustainability is a concept with many different meanings. It has to do with the cost of an activity and available financing options. It is related to needs and resources, and it also has much to do with governance, policies and programme structures. It seems obvious that operational research on the integration of HIV and tuberculosis care aims to find approaches that can be sustained by existing health systems with or without international assistance. The question of sustainability therefore needs to be addressed at the level of each proposed action. Finally, we should not forget that this is an operational research project and not a service delivery project. The possibility that a proposed action is not feasible, is not desirable, or cannot be implemented for one reason or another must be accepted from the start. If the possibility of failure were excluded, the scope of “research” would be very limited indeed.

1

Delivery of care through dedicated ART clinics operated by different entities, delivery through general hospital services, etc

HERA / November 2008

21

IHC / The Union / Benin & DRC

22

HERA / November 2008

IHC / The Union / Benin & DRC

4 Piloting Integrated HIV Care Services 4.1 Description and Objectives The main objective of piloting integrated HIV and TB services by the IHC project is “learning by doing” as captured under the first result of the logical framework: “IHC is implemented and continuously evaluated ('learning by doing') in the partner countries.”1 The list of activities included under the first result of the logical framework matrix reveals that the purpose of setting up the pilot sites is not restricted to testing the technical feasibility of integrating HIV and tuberculosis services. It includes a review and revision of national antiretroviral treatment guidelines, HIV testing policy, and health information systems related to anti-retroviral therapy. The activities are guided by a national steering committee which can only function if it has support by the highest level of the Ministry of Health and includes the most senior cadres of the National AIDS and National Tuberculosis Programmes. Of course, the other two results (assessing the cost effectiveness and the sustainability) also require that the services are piloted in some form. Under these two result areas are additional activities that are directly related to the piloting of integrated HIV services. The following is a summary of the activities most pertinent to the piloting of services as listed in the logical framework matrix: 1. Identify pilot areas and plan how full pilot area coverage will be achieved. 2. Develop an HIV related drugs and supplies logistic support system based on the antiTB drug management system. 3. Ensure that the national ART guidelines define standardised treatment and retreatment ARV regimens and that no undesirable drug interactions exist between antiTB drugs and selected ARVs. 4. Integrate national VCT policy with TB patient management. 5. Integrate ART recording, reporting and outcome cohort analysis system with the existing TB and health information system of the partner countries. 6. Carry out an IHC training needs analysis among health staff in the pilot areas, develop and pilot IHC training materials, train health facility staff in the pilot areas, and organise annual refresher courses on programme aspects that require strengthening. 7. Develop a system for intensive staff follow up after training sessions and regular supervision thereafter based on NTP supervision. 8. Develop an IHC procedure manual for patient management. 9. Based on the IHC procedure manual, prepare a standard Memorandum of Understanding (MoU) between pilot area health teams and the national planning and steering committee. 10. Start patient recruitment in the pilot areas, carry out staff supervision according to the agreed schedule, collect and analyse IHC quarterly reports, and provide regular feedback to the centres. 11. Develop data collection software. 12. Analyse treatment outcome cohort results at 6, 12, 18, 24 and 36 months of patient follow-up. 13. Analyse ARV stock management at health facility level. 1

Logical Framework Analysis, EC Project Santé/2004/078-547 Programme to Develop Integrated HIV Care for Tuberculosis Patients Living with HIV/AIDS, 2004.

HERA / November 2008

23

IHC / The Union / Benin & DRC

14. Compile a summary of ARV adverse effects using patient treatment cards and registers. 15. Analyse benefits of early HIV treatment under programme conditions. 16. Assess the impact on TB patient management. 4.1.1 Discussion of the Objectives There are different ways of piloting health care programmes. Most anti-retroviral treatment services piloted by NGOs in Africa, for example, focus primarily on documenting technical feasibility. The services are established with high levels of input and protected from the constraints of the national health system. Once they have shown to produce the desired results, integration into the health system is attempted. This last step is not always smooth and not always feasible. Many “successful” pilot projects never develop beyond the pilot phase. This has generated some controversy and suspicion about pilot initiatives. They are not always welcome by national authorities. The IHC project has learned from this experience and, from the beginning, strived to make the pilot project activities part of the national health system. This required that the Ministry of Health and its programmes on tuberculosis and AIDS were fully in charge of technical and policy decisions through the National Steering Committees, and that the pilot activities were implemented by regular health care staff. The activities of the IHC project are focused on building capacity and on generating systemic improvement rather than on improving patient treatment results directly. This approach was not always fully understood by the financial donor. In practice, the distinctions between service delivery and service capacity building, or between treatment success and systemic improvement are not as stark as they are sometimes presented. Health system improvements have to make sense to those who are providing the services. Their focus is on the well-being of patients, and they therefore need to see an improvement in care in order to retain their motivation. A functioning supply chain for drugs and laboratory reagents, for instance, is essential to assuring adequate treatment for TB and HIV. It is also the Achilles Heel of many national health services. When there are breaks in the supply chain, the services are threatened, as is the success of pilot projects that are integrated in these services. There is often the need to fill the gap through short-term direct procurement and distribution. It is, however, important that this is conceived and executed as a temporary stop-gap activity and not as a parallel activity to the national procurement and supply systems. Similar arguments also apply to other health systems building blocks such as human resources or information systems. The IHC project clearly opted for “systemic piloting” rather than “outcome-directed piloting”. This has several important implications: 1. The governance of the project has to be at the national level, i.e. the National Steering Committee. Although international experts mobilised by The Union or recruited to the International Advisory Committee can make suggestions and develop models, the final decisions of what will be implemented are made by national authorities at the national level. 2. Services provided under the project have to rely on national staff without additional salaries or bonuses. In each country the IHC maintains a small staff complement to manage the project inputs, but all service delivery activities are assured by regular staff of the Ministry of Health. The rationale for this approach and its advantages are obvious. They are related to sustainability and to the recognition of local ownership of national development. But there are trade-offs.

24

HERA / November 2008

IHC / The Union / Benin & DRC

1. National governance means that national systems and policies are pre-eminent. The profile of the project will therefore differ from country to country. What may seem like ingenuity at international level, for instance the introduction of manual CD4 counting technology, may not be acceptable at national level. The devolution of governance to the national level also creates a problem of accountability. The Union is accountable to the financial donor for the delivery of results. But if key decisions on the production of these results are devolved, it no longer controls everything it is accountable for. This requires some understanding and flexibility by the donor. 2. The human resources constraints that affect almost all health systems in developing countries also affect the project. They include staff shortages, low levels of staff satisfaction, labour disputes, high levels of staff mobility, and low productivity related to low salaries or non-payment of salaries. There is often an expectation by staff that some of these shortcomings are made up by direct payments of bonuses from project funds. Many projects are fulfilling these expectations although the negative effects of these bonus payments on national health systems are well known. The IHC project chose not to pay bonuses to service staff. This has become the main sustainability issue in the DRC where the salary of health workers is almost exclusively generated by fee for service income by institutions, and paid in the form of monthly bonuses. Since the project activities generate no user payments, staff participation in the project has been largely volunteer work, something that cannot be sustained 3. The project relies primarily on the national procurement and supply chains for TB and HIV drugs and supplies. Although IHC has filled gaps in the supply chain through project procurement, there are limits. For instance it cannot assure the continuous long-term supply of appropriate anti-retroviral medication. 4. In its first year, the IHC developed a very comprehensive approach to health information to be applied to a number of tasks ranging from calculating drug inputs to following treatment outcomes of TB and HIV therapy. These efforts, however, had to be subordinated to national health information systems. Some approaches of the IHC project were adopted in some countries, and some were not. A process of negotiation was unavoidable if IHC aimed at having an impact on health systems. But it did not always result in the adoption of the most efficient information system. 5. Finally, we need to keep in mind that all countries involved in the IHC programme of The Union are receiving other international grants in support of their responses to HIV and TB. These grants come with their own procedures and policies in critical areas such as procurement, human resources and information systems. They overlap with IHC projects in terms of geography and content. This is bound to influence the functioning of IHC project pilot sites.

4.2 Status and Achievements Both country visits were severely constrained by health worker strikes. 4.2.1 Benin Context - HIV Benin has a mixed HIV epidemic with one component concentrated in the urban commercial sex environment and a second component of a low level generalised epidemic that varies considerably among regions in the country. The epidemic is generally more severe in the South than in the North of the country. National adult HIV prevalence is estimated at 1.7 percent. In 2007 an estimated 78,000 adults and children were living with HIV in Benin.1

1

République du Bénin, Programme National de Lutte contre le SIDA. Rapport de surveillance de l’infection par le VIH et de la Syphilis au Bénin – Année 2007. Cotonou, Avril 2008

HERA / November 2008

25

IHC / The Union / Benin & DRC

According to the second quarterly report of the NACP for 2008, approximately 11,000 persons living with HIV were receiving anti-retroviral therapy at 57 accredited treatment sites throughout the country.1 This represents about 14% of all people believed to be living with HIV in the country and constitutes a massive achievement. Assuming very optimistically that half of all HIV positive Beninois know their serological status, it would indicate that one third of them are currently receiving anti-retroviral treatment. This is approaching universal coverage. There are, however, signs that the roll-out of anti-retroviral therapy is not proceeding without problems. According to the quarterly report cited above, approximately 900 new patients were placed on ART during the second quarter of 2008, while 400 had abandoned treatment, 300 were lost to follow-up, and 100 had died. This leaves only an increase of 100 patients to the active ARV treatment register in the three month period. The information is difficult to interpret as these patients belong to different treatment cohorts. Another worrying information comes from a yet to be published virological study of 125 patients on ART in Cotonou conducted within the context of the IHC project in the first quarter of 2008. The patients had a median anti-retroviral treatment duration of 3 years. On average, they had achieved good immunological recovery. However, one third had signs of virological treatment failure.2 Research is continuing to determine what proportion of this failure is attributable to anti-retroviral drug resistance. What emerges is the impression that with international assistance3 the health system in Benin is able to achieve high anti-retroviral treatment rates in the short term. However, it needs to develop a much stronger focus on long-term patient retention if it is to make a real difference in the life of people living with HIV in Benin. With the exception of the IHC project, international support for the response to HIV seems to have overlooked this dimension. Context - Tuberculosis The National Tuberculosis Programme supports a TB treatment case load of 3,500 to 4,000 patients per year. In 2008, the WHO estimated that the case detection rate for smear positive TB in Benin was 86 percent in 2006, and that the treatment success rate for the 2005 treatment cohort was 87 percent.4 These rates are considerably above the African average and are an indicator of good programme performance. Tuberculosis diagnosis and treatment is provided through 54 diagnostic and treatment centres (CDT) that are supported and supervised by the NTP. Most are fully integrated in general hospitals or health centres. The notable exceptions are two specialist chest hospitals in Cotonou and Porto Novo that treat about one third of patients with tuberculosis in the country. The prevalence of HIV infection among incident TB cases is estimated at 15 percent based on the testing of approximately 3,300 TB patients in 2006. IHC project achievements There are indications that the IHC project has contributed significantly to improving the care of people living with HIV in Benin. Before 2005, there was no substantive dialogue between the NTP and the NACP. Tuberculosis clinics in Benin did not examine the HIV status of new patients. According to staff interviewed during the field visits, treatment success rates were low because many patients died or continued to be incapacitated although their tuberculosis was cured.

1

République du Bénin, Programme National de Lutte contre le SIDA. Rapport du monitoring des activités du deuxième trimestre 2008. Abomey, Juillet 2008 2

Personal communication – Prof Eric Delaporte, DMISI

3

Benin has benefited from two five-year Global Fund grants for HIV. A grant for USD 17 M in Round 2 (implementation started in Jul 2003) and a grant for USD 59 M in Round 5 (implementation started in Jan 2007) 4

WHO Report: Global Tuberculosis Control. Annex 3: http://www.who.int/tb/publications/global_report/2008/pdf/annex3_afr.pdf (accessed 18/08/08)

26

HERA / November 2008

IHC / The e Union / Ben nin & DRC

Once th he IHC proje ect started, and once staff s in the tu uberculosiss clinics werre trained an nd applied the new su urveillance and a treatme ent guideline es develope ed with assiistance of th he IHC project1, patient reccovery and survival sta arted to imp prove. These e anecdotes are suppo orted by the evid dence of an increase in n tuberculossis treatmen nt success in Benin tha at started be efore the project, but app peared to gain momentum with the 2005 trea atment coho ort, the first who benefite ed from IHC C. 2 DOTS Treatment Success s Rate in Benin B

At the tiime of the evaluation, e t there was evidence e of a national effort e to acccredit health h facilitiess that were designated tuberculosis treatmen nt centres (C CDTs) to alsso offer antiiretrovira al therapy. Of O the 20 IH HC pilot site es, 13 had received succh accredita ation from th he NACP. Significantlyy, all 54 CD DTs in the country were e said to have adopted d the IHC prrotocol of provvider-initiated HIV coun nselling and testing, pro ovision of co o-trimoxazo ole prophyla axis to co-infeccted patientss, and, if they were nott accredited d for ART prrovision, refferral of HIV V positive e tuberculossis patients to t the neare est clinical facility f providing ART. Staff in all centres c I visited during d the evaluation e c credited the training and supervisio on provided d by the IHC C project with the e rapid natio onal uptake of integrate ed care of co-infected c p patients. At the same s time, HIV H testing and treatment centres s were also said to have adopted the t practice e of systema atically referring people e living with h HIV for tub berculosis sscreening. I was not able to verify v this, but b it was re eported at several s sites s. Neverth heless, there e are some problems. The pro oblem of refferral The pro oblem of refe erral arises in three diffferent conte exts 1. Tub berculosis treatment ce entres in faccilities accre edited for ART A provide e this service e only until the treatm ment of tube erculosis is completed. The patien nts are then referred to the nea arest clinica al service prroviding HIV V care. The referral pra actices differ among the e clinics and d are not alw ways succe essful. There e is no systtem in place e to assure the seamless con ntinuation off ART, nor any a system m of confirma ation of succcessful patient referrall. In Porto P Novo,, for example, the physsician in cha arge of the chest c hospiital commun nicates perrsonally by cell c phone with w the trea ating physic cian at the Regional R Ho ospital to as ssure that the patients who have e completed d tuberculosis treatment are acce epted in the ART clin nic. At the CNHPP C in Cotonou, C on the other hand, h many patients wh ho were started on ART either refused the tra ansfer to an nother clinic c, or were no ot able to fin nd another service. At the t time of the t mission, between 20 2 and 30 patients p who o had comp pleted their tuberculosis tre eatment continued to receive r their ART at the e CNHPP.

1

Ministère e de la Santé. Guide de surveillance épidémiologique et de la prise en charge e de la co-infecttion tuberculose e/VIH au Bénin. Octtobre 2006. 2

WHO Re eport: Global Tuberculosis Conttrol. Annex 3: http://www.who.i h int/tb/publication ns/global_reporrt/2008/pdf/anne ex3_afr.pdf (accessed 18/08/08)

HERA / November N 200 08

27

IHC / The Union / Benin & DRC

The referral of those co-infected patients who are not eligible for ART1 appears to be even more dysfunctional. At the time of my mission, the CNHPP had started a follow-up study of co-infected patients registered between 2005 and 2007 who had been discharged after completion of tuberculosis treatment. At this time, the study had traced 27 patients. Of the 22 among them who were still alive, most had dropped out of HIV care. They either never presented to the HIV service or they had abandoned treatment because they were dissatisfied. Only a very small number continued co-trimoxazole prophylaxis, and only two were on ART, one of them at the CNHPP because of a relapse of tuberculosis. 2. Tuberculosis treatment centres in facilities not accredited to provide ART refer HIV positive tuberculosis patients for simultaneous clinical management of HIV infection to the nearest facility. The success of this referral appears to depend primarily on the dedication of the CDT staff. In Comé, the laboratory technician of the tuberculosis clinic takes HIV positive tuberculosis patients on his motorcycle to the nearest hospital for HIV treatment. In other clinics, the success of the referral is entirely left to the patients’ initiative, and the tuberculosis clinic receives no information to confirm whether the patient is being followed for HIV infection or not. In all these cases the referral demands considerable effort by the patient. One example is the health centre in Abomey. The laboratory of the health centre performs a rapid HIV test together with the sputum exam. Those tuberculosis patients who are found to be HIV positive are then sent to the CIPEC laboratory (in the same compound) for CD4 testing. They are then referred to the Regional Hospital, several kilometres away, for clinical follow-up of their HIV infection, while they are being treated for tuberculosis at the health centre. If they require ART, they have to return to the CIPEC for blood chemistry and haematology testing. These tests are available at the hospital, but only the CIPEC has the authorisation to perform these tests free of charge to people living with HIV. All in all the care involves multiple waiting lines and considerable travel. The risk of loss to follow-up is high. 3. The referral of close contacts of co-infected patients is even less systematic. The standard of care for tuberculosis in Benin is to ask patients about family contacts, and to place close contacts under 5 years of age on INH prophylaxis. The CDTs I visited do not test family contacts of co-infected patients for HIV, but counsel index patients to discuss their HIV status with their family, and encourage contacts to seek testing. They have no information about the uptake. The health information system The Benin Ministry of Health and the National AIDS Programme in Benin did not accept the format of the patient records, registers, information system, and procurement planning system developed by the IHC with the exception of the laboratory register and request form. However, the National AIDS Programme produced and distributed new registers with columns for treatment information about both diseases based on a template distributed by the WHO. There are two types of registers, one for patients living with HIV who are not on ART, and one for those who are on ART. The NTP collaborated closely with the NACP in the adaptation of the WHO templates in Benin, and the printing of the registers was paid for by the IHC project. The new registers were available in all facilities I visited that were accredited by the NACP to provide ART. But staff training on the use of the registers had not yet taken place. Consequently, they were used differently and sometimes inappropriately. Staff training is planned to commence in the second half of 2008. Once the training is complete, the registers will be a step towards a nationwide HIV treatment cohort information system. However there 1

According to the national ARV treatment guidelines, patients with pulmonary tuberculosis and no other AIDS-defining illness 3 (i.e. patients in WHO stage 3 of HIV infection) who have a CD4 count above 350/mm are not eligible to receive ART.

28

HERA / November 2008

IHC / The e Union / Ben nin & DRC

is still much m work to o be done in terms of developing d systems an nd procedurres for reporting, supervission, and da ata analysiss. These are e areas whe ere the NTP P through th he IHC proje ect in Benin has h considerable experrience that could c be mo obilised. It remains to be seen whetherr the introdu uction of the e new registters will allo ow the Natio onal AIDS Programme to t start mon nitoring the roll-out of anti-retrovira a al therapy b by collecting g survival data on tre eatment coh horts. This information will be of in ncreasing im mportance for f health care c plannin ng as the nu umber of pe eople on antti-retroviral therapy is e expected to o keep growing g. As of the second trim mester 2008 8, this inform mation was not availab ble to the NA ACP.1 The trim mester report states tha at a study of the one-ye ear survivall rate of peo ople on ART T is being planned. Butt a one time e study is sttill far from a system off monitoring g treatment cohorts. ent cohort monitoring m iss a well esttablished routine of the e National T Tuberculosis s Treatme Program mme. While e it has some applicatio on to the mo onitoring of ART, this a application is limited as the cohorts c are only followe ed for an avverage of six months. Anti-retrovir A ral treatmen nt cohorts, however, need to be monitored for f years an nd decades. C project ha as, to some extent, introduced an ART cohortt analysis among patie ents The IHC followed d in tubercu ulosis clinicss. Among th hese patientts, the NTP monitors survival, loss s to follow-u up, and treatment disco ontinuation until the end of the tub berculosis trreatment. However the coho ort is then truncated t be ecause the surviving patients are referred ou ut and presu umably enter into the inform mation systtem of the NACP. N g an integra ated system m for plannin ng and Overall,, the resultss of the IHC in terms off introducing monitorring tubercu ulosis and HIV H servicess in Benin are disappointing. The n new registers may create some s greate er congruen nce, but the e optimistic hope h that le essons from m monitoring g tuberculosis treatm ment could be b adapted and applied d to anti-retroviral thera apy services remainss somewhatt unfulfilled in Benin. 4.2.2 DRC Contextt - HIV The pro ofile of the HIV H epidemiic in the DR RC is not cle ear. There are a many co onflicting sta atistics publishe ed by nation nal and inte ernational so ources. In 2007, 2 the Na ational AIDS S Programm me reported d a nationall adult HIV prevalence p rate of 4.1 percent. Th his estimate e is based on o the sentinell surveillancce of ante-n natal clinics in 22 sites throughout the countryy.2

1

Républiq que du Bénin, Programme P National de Lutte co ontre le SIDA. Rapport R du mon nitoring des activvités du deuxièm me trimestre 2008. 2 Abomey, Juillet J 2008 2

Républiq que Démocratique du Congo, Programme P National de Lutte contre c le SIDA. Passage P de la ssurveillance sen ntinelle du VIH chez les femmes encceintes fréquenta ant les servicess de CPN, Kinsh hasa, Décembre e 2007

HERA / November N 200 08

29

IHC / The Union / Benin & DRC

The range of reported sero-prevalence at these sites is from 2% to 6.9%. Some doubt was thrown on these data by the 2007 Demographic and Health Survey which reported a national HIV prevalence of 1.3 percent. This value is also reported by UNAIDS and WHO.1 However, there may have been serious sampling errors in this study, and the real prevalence may be somewhere between 1.3 and 4.1 percent. The HIV prevalence observed by the IHC project among tuberculosis patients indicates that the general HIV prevalence in the country is closer to two than to four percent. The HIV epidemic in the DRC appears to have been relatively stable over the past 20 years, with a doubling of the estimated number of people living with HIV between 1987 and 2007. This increase approximately mirrors the rate of population growth. Nevertheless, the NACP reports an expansion of the epidemic in rural areas, and an increasing feminisation of the epidemic.2 Because of the uncertainty about the national HIV prevalence, the estimates of the current number of people who require anti-retroviral treatment vary widely between 100,000 and 340,000. Between 27,000 and 30,000 people are reported to be on treatment3, but there is also a high degree of uncertainty surrounding this estimate. The institutional structure of the national response to HIV is weak. Many international donor agencies bypass the Government and finance programmes of international NGOs and faithbased organisations directly. The National AIDS Programme has only limited information on how many clinics in the country are providing ART, which drugs they use, and how they are monitoring the treatment. The most reliable information about HIV care in the DRC is presently provided by the IHC project, which however only covers two provinces. The Government response itself is divided between the National AIDS Programme (PNLS) of the Ministry of Health, and the Multisectoral National AIDS Programme (PNMLS) attached to the Presidency. The PNLS and the PNLMS have overlapping responsibilities, including for clinical HIV services. In addition, there is the 4ème Direction of the Ministry of Health, in charge of the control of endemic diseases, including AIDS and tuberculosis. The PNLS, however, has the status of an autonomous agency within the Ministry of Health and does not report to this Directorate. Finally, there is the Country Coordinating Mechanism of the Global Fund which has major influence over the resource flow for the response to HIV and tuberculosis in the DRC, but poorly defined formal institutional links to the Ministry of Health, the PNMLS, and the PNLS. Attempts are underway to reform the health sector and to organise it according to functional service and administrative units.4 These initiatives are supported by the EC and by several international cooperation agencies. However, as long as the major funding channels remain disease-specific, it will be difficult to establish health services that are patient oriented rather than problem oriented. A more achievable objective that is also being pursued actively by an institutional audit is to rationalise and clearly define the functions of the PNLS, the PNMLS, and the 4ème Direction. Context – Tuberculosis The prevalence and incidence of tuberculosis in 2006 in the DRC are reported by WHO as 390,000 and 240,000 respectively.5 The calculated incidence rate of 392/100,000 places the DRC among the countries with serious tuberculosis epidemics in Africa. However, these 1

UNAIDS/WHO. Report on the global AIDS epidemic, July 2008.

2

République Démocratique du Congo, Ministère de la Santé Publique. Plan stratégique de lutte contre le VIH et le SIDA, 2008 – 2012. 3

WHO/UNAIDS/UNICEF. Epidemiological Fact Sheet on HIV and AIDS, Democratic Republic of the Congo, 2008 http://www.who.int/globalatlas/predefinedReports/EFS2008/full/EFS2008_CD.pdf (accessed 26/09/08) 4

Centre de Santé – Hôpital de Référence - Zone de Santé - Inspection Provinciale de la Santé - Secrétariat Général de la Santé 5

WHO Report: Global Tuberculosis Control. Annex 3: http://www.who.int/tb/publications/global_report/2008/pdf/annex3_afr.pdf (accessed 18/08/08)

30

HERA / November 2008

IHC / The e Union / Ben nin & DRC

numberrs are extrap polated from m a tubercu ulosis prevalence surve ey that is mo ore than 20 0 years old, and d the rate is calculated with an unccertain deno ominator be ecause therre has not been b a population census in the DRC C for many years. y There e is thereforre a high levvel of uncertainty about th he national case detecttion rate wh hich is reporrted as 39% % (61% for ssmear posittive tuberculosis) by WHO. mber of new w cases of tuberculosis t s diagnosed d each yearr in the DRC C increased rapidly The num from lesss than 20,0 000 in 1990 0 to more than 100,000 0 in 2003. Since then th he numbers s have remaine ed static wh hich may ind dicate that the t health system has reached the e limit of its ability to detecct new case es, or that th he case detection rate is considera ably higher than reportted by WHO. dent, that th here has been a major improveme ent in the pe erformance of tuberculo osis It is evid control in the DRC. Of the 199 95 treatmen nt cohort, on nly about 20 0% of newlyy diagnosed d smear positive e tuberculossis patients were w evalua ated and de eclared cure ed, while for the 2006 cohort c this proportion wass 81%. Therre is, howevver, conside erable varia ation in the T TB cure rate e among the 11 provvinces of the e DRC as illustrated in this graph:1

Among patients witth tuberculo osis in the tw wo regions of the IHC project, the e HIV prevallence 2 n the 13 pilo ot sites in Nord N Kivu to o 17% in the e 10 pilot sittes in Bas Congo C . ranges from 11% in Nationw wide, the WH HO reports an HIV pre evalence ratte among in ncident tube erculosis patients of 9 percent which is likely an un nder-estimate.3 ulosis care and control in the DRC C is delivere ed through 1545 1 accred dited health h service Tubercu facilitiess that are pa art of the pu ublic health infrastructu ure and rece eive additio onal supportt and supervission from th he NTP for their t activities in tuberc culosis conttrol. Direct ssupervision is provided d by nurse--supervisorss placed in 515 zonal health h office es (Bureau d de coordina ation de zone de e santé - BC CZS) which are in-turn supported and supervvised by 24 provincial coordina ation officess, soon to become b 264 (Coordinattion provincciale de lutte e contre la tuberculose – CPLT). A centra al unit in Kin nshasa man nages tuberrculosis con ntrol in the country, c me assuress the link to the Endem mic Diseasess Directoratte of the Min nistry of Health (the 4èm Directio on) and to th he many NG GO partnerss who partic cipate in tub berculosis ccontrol in the e DRC.

1

Mbulula M-L. M Surveillancce epidémiologique PNT-RDC (slide show, no date)

2

Bas Cong go: Nord Kivu u:

10 sittes, Jul 2006 to Jun 2008: 2252 2 tuberculosis patients p tested; 387 3 found HIV p positive 13 sittes, Jul 2006 to Jun 2008: 3321 1 tuberculosis patients p tested, 350 3 found HIV p positive.

3

WHO Re eport: Global Tuberculosis Conttrol. Annex 3: http://www.who.i h int/tb/publication ns/global_reporrt/2008/pdf/anne ex3_afr.pdf (accessed 18/08/08) 4

The Consstitution of the DRC D provides fo or an administra ative reorganisation of the coun ntry by 2009 with h an increase frrom previously 11 to 26 provin nces.

HERA / November N 200 08

31

IHC / The e Union / Ben nin & DRC

IHC proj oject achievvements The IHC C project ha as a very go ood reputation on the national n and d provincial level. In inte erviews held during this eva aluation, se enior Ministrry of Health officials sta ated that the e training, supervission, material support, and inform mation suppo ort of the prroject is the way they would w like to see s HIV care e to be orga anised in the DRC. At least l one major m financiial donor, th he World Bank, B has accepted the e IHC appro oach as a model m on wh hich to orien nt its reprogrammed MAP grrant supportt to the DRC C. Because of a healtth worker sttrike that at times took on a very militant m charracter I was not able to collect as much m field in nformation as a I wanted to, especia ally in Bas C Congo. It wa as, howeve er, quite rem markable tha at in severa al hospitals and a clinics the t only non-medical staff s who pro ovided esse ential services where th he nurses in n charge of tuberculosis care, and d this despite threats aga ainst them by b other uniion memberrs. anised. Stafff is conscientious in Generally, the pilott tuberculossis clinics arre well orga ning the reg gisters. Patiients are routinely coun nselled and d tested for HIV. For qu uality maintain control of o HIV testing, blood samples are sent as dried blood sp pots on filter paper to the Nationa al HIV Referrence Laborratory in Kin nshasa. All HIV positive tuberculosis patients s and their contacts have e CD4 counts done and d are taken into HIV ca are. Since th here is usua ally nowhere e to refer th hese patientts too, they remain in the care of the t tubercullosis team. The patient charts1 devveloped by the t project are a in use everywhere e and provide e a rapid an nd clear picture of o the treatm ment record d. The regissters allow the t tracking of HIV trea atment for several years. e manual CD D4 counting g system inttroduced byy the IHC prroject is the e only In most places, the o determine e CD4 coun nts. There have been in nterruptionss in the supp ply of available method to ory reagentss but these are now so olved. Howe ever, there is i a major isssue with th he fact laborato that the CD4 tests are only avvailable to patients ente ering into ca are through the tubercu ulosis clinic. Several S heallth facilities have otherr portals of entry. e While e they often manage to o harmonise care forr all HIV pattients, they are restricte ed by the IH HC project in the use of o the CD4 rea agents. Thiss gives the IHC projectt a bad repu utation of “vverticality”. IIt cannot be e justified d by any ratiionale except the very limited ratio onale of a project p menttality. Patient retention an nd survival Because most patients who entered HIV care throug gh the tuberrculosis clin nic have rem mained within th his service, at least for the duratio on of the IHC C project, th he project w was able to calculatte survival curves c for th hese patients. Two coh horts have been b followe ed for long enough to be ab ble to assesss the resultt of joint HIV V and tuberrculosis care e provided in the DRC, the 1

See Anne ex 1

32

H HERA / Novem mber 2008

IHC / The e Union / Ben nin & DRC

cohort that t entered d care when n patient treatment starrted in mid 2006, 2 and the cohort th hat entered into care in n the beginn ning of 2007 7. In total th hese two groups includ de 614 patie ents of whom 342 3 received d anti-retrovviral therapyy. The total number of person-yea ars of obserrvation for these two cohorts is 867, among a those receiving ART it is 400. ethod of calcculating the survival cu urves is explained in An nnex 1. The e first graph shows The me the high h and low esstimates of survival over a period of 24 months or 18 mo onths for the e two 1 groups of patients. The secon nd graph sh hows the sa ame informa ation for tho ose patients s who were pla aced on antti-retroviral therapy.

2006 Coho ort (high and low es stimates)

It is diffiicult to expla ain why the e patients on n ART in the 2007 cohort appear to have a lo ower rate of survival s than the entire e 2007 coho ort. The datta allow a crrude estima ation of morrtality of pattients per 10 000 person--years of observa ation. Depen nding on wh hether we assume a thatt all patientss lost to follow up are alive, a or that all patients p losst to follow up u have died, the observed mortallity rate of th he entire 20 006 and 2007 co ohort is betw ween 155 and 250 per 1000 perso on-years. Fo or the sub-g group of patients on anti-retroviral tre eatment the e estimated mortality is s between 138 1 and 225 5 per 1000 personyears. a very hig gh mortality rates, betw ween 10 and d 20 times higher h than the mortalitty rates These are observe ed among people p on an nti-retroviral treatment in high inco ome countriies. The life e expecta ancy study published p re ecently in th he Lancet re eports an ovverall morta ality of peop ple on anti-retrroviral treatm ment of 13 per 1000 pe erson-years s. Among th he sickest p patients, tho ose with CD4 counts of lesss than 100 at a entry, the e observed crude c morta ality rate in industrialise ed countrie es is 21.4/10 000 person years, and d among the e group mosst difficult to o treat, injec ction drug users, it was found f to be 20.4/1000 person yea ars.2 es published d a meta-an nalysis of pa atient retenttion rates on antiIn 2007, Rosen and colleague al treatmentt in Africa among 33 clinical cohorrts (74,000 patients) in 13 countrie es.3 The retrovira average e follow-up time t of thesse patients was w only ab bout 10 mon nths. Rosen n estimated d the 24 month retention r forr patients on n ART in Affrica at 60 percent, p whiich is aboutt the same result r achieve ed by the 20 006 ART cohort of the IHC program mme in the DRC (58% % retention). But a more va alid comparrison would be to only look l at the nine n cohortss in the ana alysis that provide retention results att 24 monthss (either reported or exttrapolated).4 1

The high estimate of surrvival is calculatted with the assumption that all patients lost to follow up are a alive; the low esttimate with ese patients havve died. the assumption that all the 2 The Antirretroviral Therap py Cohort Collaboration. Life exxpectancy of ind dividuals on com mbination antire etroviral therapy y in highincome countries: a collab borative analysiss of 14 cohort studies. s Lancet 2008; 2 372:293-9 99 3 Rosen S,, Fox MP, Gill JC. Patient reten ntion in antiretro oviral programs in sub-Saharan Africa: a system matic review. PL LoS Med 2007; 4(10 0):1691-1791 ww ww.plosmedicin ne.org 4 These co ohorts are (acco ording to the nam mes used for ide entification in th he paper): Botsw wana 1, Camero oon 1, Cameroo on 4, Kenya 1, Malawi 1, South Africa 3, Senegal 1, Uganda U 1, and Multi-Country M 1.

HERA / November N 200 08

33

IHC / The Union / Benin & DRC

Nine treatment cohorts cited in the meta-analysis1 2006 IHC anti-retroviral treatment cohort

abandoned treatment

lost to follow-up

Reasons for stopping treatment

death

Patients retained on treatment after 24 months

Number of patients started on ART

Treatment results after 24 months of anti-retroviral therapy

4233

68.1%

50.7%

46.2%

3.1%

143

58.4%

60.4%

37.7%

1.9%

The treatment results of the 2006 IHC cohort in the DRC are somewhat less favourable than those of other published studies in Africa, but they are certainly within the same margins of error of the estimate. Human resources and financing The main issue that was raised in each of the 11 field visit sites concerns staff payment. The discussion always focused on “primes”, a term that is usually translated as “staff bonuses”. However, this is not a discussion about bonuses that are paid in addition to salaries. In the DRC, the “primes” are the staff salaries. The basic salary paid by the State is negligibly small. The issue thus becomes a fundamental issue of health sector financing. Health care institutions in the DRC function primarily like not-for-profit enterprises. Although they receive a budget from the State, their real economy, including staff payments, is based on income from user charges. Some NGOs operating HIV services, primarily with money from the Global Fund, bypass the health sector financing discussion by paying staff salaries directly. From a health systems perspective, this is probably the worst of all options. Tuberculosis care and HIV care are mandated to be free of charge. The drugs and laboratory reagents are provided by the State or by projects such as IHC. Yet they demand considerable staff time and additional resources to operate generators, to pay transport and telephone bills to track defaulting patients, and to waive the in-patient hospital charges of indigent patients. But there is no financing available for these inputs. Health workers themselves are paying for the services in the form of foregone salaries. The NTP pays a small “bonus” based on the number of smear positive tuberculosis patient detected, but the amount of these payments is insufficient to cover the cost to the institution and the staff. It is surprising that the IHC project has been able to work for so long without addressing this issue. It attests to the high level of commitment of the health staff, and the inspiring supervisory and training work done by the project. But with the end of IHC1 in the DRC, the issue will have to be addressed in order not to lose the project achievements. In Bas Congo, a reprogramming of the World Bank MAP has been negotiated. It will include staff payments. In Nord Kivu, the hope was that the issue of health care financing and staff salaries will be addressed by the EC health programme under the 9th EDF. However, the MAP will take several months before it will become operational, and the EDF support to North Kivu targets a limited number of health zones, none of which include IHC pilot sites. I do not see much hope for the continuation of integrated HIV and tuberculosis care in Bas Congo and Nord Kivu unless the issue of staff salaries is resolved as soon as possible.

1

See text

34

HERA / November 2008

IHC / The Union / Benin & DRC

5 Laboratory Research and Capacity Development 5.1 Description and Objectives The laboratory research component of IHC implemented in partnership with the Département des Maladies Infectieuses et Santé Internationale of the Université Montpellier has two primary objectives: 1. To validate the national HIV testing algorithms (including the types of test used) and to establish national quality control systems for HIV testing in Benin and the DRC. 2. To develop the national capacity in Benin and in the DRC to routinely measure the viral load of persons receiving anti-retroviral treatment, and to diagnose resistance to anti-retroviral drugs in case of treatment failure. The contract with DMISI was not signed until the beginning of 2007. The main reasons for this delay was a renegotiation of the initial objectives between The Union and the EC when it became clear that the proposed method of transporting blood samples for virological tests in the form of dried spots on filter paper required further research and was not ready for implementation. Research, capacity building, and technology transfer activities started in February 2007 and are planned to be completed by the end of November 2008. This is a very tight timetable. Not included in the contract with DMISI are the evaluation, technology transfer, and quality control of the CD4 counting method introduced by the IHC in Benin and in the Congo. It is a manual procedure based on the use of magnetic beads examined under a light microscope (dynal assay). This is a low technology but labour intensive alternative to available automated flow cytometry methods that are the current international laboratory standard, but that have problems when used under field conditions in places with poor infrastructure and low volumes of testing. The contribution of the IHC project to the development of laboratory capacity for CD4 counting is discussed under this heading although this component is not under the responsibility of the DMISI. 5.1.1 Discussion of the Objectives Validation and quality control of national HIV testing procedures Given the large number of HIV variants circulating in Africa, and given the catastrophic personal consequences of a false positive or a false negative diagnosis of HIV infection, it is rather surprising that most National AIDS Programmes have not already established a rigorous validation and quality control system for the HIV testing algorithms used in the country. Through the project, the DMISI is assisting the national HIV reference laboratories in Benin and the DRC establish panels of locally obtained sera, and develop procedures to regularly evaluate the performance of the national HIV testing algorithms. This task is not specifically related to the issue of HIV and TB co-infection. But any HIV treatment programme has to be based on a reliable method of diagnosing HIV infection. Without this foundation all further work is of questionable value. Viral load measurements and determination of anti-retroviral resistance The WHO recommends that patients receiving anti-retroviral therapy should have at least one examination of viral load after 12 months of treatment where this test is available. A computer simulation published in 2008 suggests that in resource limited settings, treatment decisions based on clinical examination alone do not significantly compromise patient survival when compared with decisions supported by virological or immunological laboratory

HERA / November 2008

35

IHC / The Union / Benin & DRC

tests.1 The findings support the WHO recommendation of a “public health approach” to scaling up ART in resource-limited session, monitoring viral load and anti-retroviral resistance on a population level through sentinel surveys rather than on the level of individual patients. The risk of this approach is that people may stay for long periods on failing regimens with accumulation of drug resistance and the transmission of resistant strains. Virological failure, i.e. the increase in viral load, manifests itself earlier than immunological failure, i.e. the decrease in CD4 cell count, and again earlier than clinical failure, i.e. the emergence of new opportunistic infections. Viral load testing is therefore the most sensitive method to diagnose anti-retroviral treatment failure. Genetic sequencing is then the most precise way to determine whether this treatment failure is due to viral resistance. Virological tests require an advanced laboratory infrastructure. In order to offer them on a routine basis for patient management, the transport of serum to the laboratory has to be assured. Transport of frozen sera within countries is difficult, and international transport is prohibitively expensive. Spots of blood on filter paper can be used to determine the presence of viral nucleic acid and samples are routinely transported in this manner to diagnose HIV infection, for instance in infants. An early report of the use of dried plasma spots for viral load determination carried much promise.2 It sparked the idea to investigate the application of this technology for the introduction of routine viral load testing in Benin and the DRC. The project conducted intensive investigations on the use of filter paper as transport medium, resulting in at least one published scientific paper. These studies continue, but until now it is not possible to apply this method to the transport of specimen for viral load testing. In the mean time, the project focused on (1) establishing a functional facility for viral load tests in one national reference laboratory in each of the two countries, and (2) training laboratory staff from both countries in genetic sequencing techniques for viral resistance testing in Montpellier to familiarise them with the procedure. Part of this training and technology transfer were two one-time cross-sectional studies of viral load and ARV resistance in samples of sera obtained from patients in each country. These cross-sectional studies will provide the Ministries of Health in Benin and the DRC with information on the current performance of anti-retroviral therapy in their country. Immunologic laboratory tests Technology transfer and evaluation of CD4 testing methods are included in the project’s logical framework and identified as a responsibility of DMISI. They were however not included in the contract for the laboratory research component. According to the Director of DMISI they are not an area of competence of the Department. The project intended to introduce the method of manual dynal assay for CD4 counting (Dynabeads®) in most project sites. This method requires minimal equipment, most of which is readily available in a tuberculosis laboratory. It has been evaluated in several studies and was found to generate valid results (e.g.3). At the same time, large numbers of automated flow cytometry machines are being procured throughout Africa, primarily with funds that have become available thorough Global Fund grants. There are many problems with these machines. They frequently become inoperable or unreliable within a short time because of fluctuations in electric currents, lack of service, stock-out of reagents, or insufficient volumes of tests.

1

Phillips AN, Pillay D, Miners AH, Bennett DE, Gilks CF, Lundgren JD. Outcomes from monitoring of patients on antiretroviral therapy in resource-limited settings with viral load, CD4 cell count, or clinical observation alone: a computer simulation model. The Lancet 2008, 371:1443–51

2

Cassol S, Gill MJ, Pilon R, et al. Quantification of human immunodeficiency virus type 1 RNA from dried plasma spots collected on filter paper. J Clin Microbiol 1997; 35: 2795–801. 3

Lyamuya E F, Kagoma C, Mbena EC, Urassa WK, Pallangyo K, Mhalu FS, Biberfeld G. Evaluation of the FACScount, TRAx CD4 and dynabeads methods for CD4 lymphocyte determination. Journal of immunological methods 1996; 195: 103-112

36

HERA / November 2008

IHC / The Union / Benin & DRC

Nevertheless, manual CD4 assays have not made a major impact on laboratory practice in Africa. They are rarely used outside the context of the IHC programme. There are a number of reasons, including the fact that they are very labour intensive. The IHC project in Benin and the DRC generated mixed results about the prospects of establishing the technology, a positive result in the DRC and a largely negative result in Benin. This is an issue that needs further investigation to provide a clear answer to The Union on whether and where to pursue the transfer of manual CD4 counting technology, weighing all possible alternatives.

5.2 Status and Achievements 5.2.1 Benin Validation and quality control of national HIV antibody testing algorithms According to the Coordinator of the National HIV Reference Laboratory, this work is almost completed. The serum panels have been established, a laboratory technician has been trained, and a procedure for regular quality control of HIV tests used at central and peripheral level has been drafted. The final step was to send samples that remained indeterminate after repeated retesting to Montpellier for genetic characterisation. Preliminary results suggest that the most commonly used HIV test in Benin perform adequately. The studies on samples with indeterminate results are ongoing, and final results will be available by the end of October. However, the level of engagement of the project with the National HIV Reference Laboratory appears to be low, as DMISI works primarily with the National Tuberculosis Reference Laboratory. While the results of the studies will provide important information to the Ministry of Health and the NACP in Benin, it is not certain that they will contribute significantly to long-term systematic improvement in quality control for HIV testing in the country. Viral load measurement The DMISI assisted the NTP in procuring and installing equipment to perform PCR viral load tests in the National Tuberculosis Reference Laboratory in Cotonou. DMISI also organised the training of laboratory technicians on site. The equipment is functioning and reagents and supplies are available. An initial study of viral load among 125 patients on ART in Cotonou found that 39 were experiencing virological failure. This is a troubling result, but corresponds to reports from other countries where anti-retroviral therapy is being scaled up using the WHO public health approach. The only other ability to measure viral load in the public sector in Benin is with equipment installed in the National HIV Reference Laboratory that is apparently based on outdated technology.1 I received conflicting reports about whether it was functioning or not. I was also told that the PNLS of Benin has started the procurement of two new PCR machines to be installed in the National HIV Reference Laboratory in Cotonou and in the regional laboratory in Lokossa. It still remains to be seen whether the National Tuberculosis and the National HIV Reference Laboratories will be able to collaborate effectively to introduce a coordinated approach to viral load testing in Benin. Whatever the plans, I did not see a single result of a viral load test in any of the patient charts I reviewed. As of now, the test is simply not available for routine care in Benin. It is likely that it will soon become available at the level of the National Tuberculosis Reference Laboratory beyond its use for research studies. But it will only be accessible to patients in Cotonou, and there is an unresolved question of continued supply of reagents after the completion of the IHC project.

1

Capacity for virological testing is apparently available in at least one privately operated hospital in the country.

HERA / November 2008

37

IHC / The Union / Benin & DRC

Testing for resistance to anti-retroviral drugs The analysis of mutations of the viral genotype that result in anti-retroviral drug resistance requires specialised skills and laboratory equipment that was not meant to be transferred to Benin. The initiative therefore relied heavily on the successful use of dried blood spots for the convenient shipment of samples to a regional reference laboratory or to Montpellier. When it was found that this was not yet possible, the DMISI trained a senior laboratory technician from Benin in the Montpellier laboratory (M. Faihoun). The training was completed in July using serum samples from patients experiencing virological treatment failure in Benin. Among 33 samples examined, 27 were found to contain HIV strains that were resistant to at least one drug in the current ART regime of the patient. It will not be possible to ship frozen serum samples from Benin to Montpellier routinely to investigate virological treatment failure. It is therefore clear that genotype sequencing will not be done in Benin for several years, except for occasional research studies. The stated rationale of the training was to have one person at national level in Benin become familiar with the technique to be able to advise Government and international donors. Furthermore, the training was meant to explore the feasibility of performing preliminary steps in drug resistance testing in Benin, depending on the future access to genetic sequencing technology in the country or region, and depending on future availability of better technology for the transport of specimen. The rationale is weak, especially when one considers the high international mobility of staff trained in retrovirus laboratory techniques. However, the identification of the profile of ART resistance in Benin is an important result even if it will not have the long-term systemic impact hoped for by the project. Immunologic laboratory tests CD4 counts are recorded for about two thirds of HIV positive patients registered in the tuberculosis clinics in Benin. All the test results I saw in the South of the country were obtained by automated flow cytometry available in hospitals or HIV clinics near the tuberculosis clinic sites. The National Tuberculosis Reference laboratory has a flow cytometer for CD4 assays procured by the IHC project. The most common reason for the non-availability of CD4 counts was the stock-out of reagents for the CyFlow® equipment supplied by the NACP. The IHC project introduced the Dynabead® CD4 counting system in 12 project sites. A laboratory technician of the National Tuberculosis Reference Laboratory was trained in the method in the DRC. Following his return, he trained laboratory technicians of the IHC project sites in Benin. Of the seven project sites I visited, only one was equipped to perform Dynabead® CD4 assays but used the equipment rarely. I was told that there are 4 clinics in the North of Benin that continue using this method. There is no enthusiasm for manual CD4 counting at any level of the NTP or the NACP. I was told that the NACP had already tried to introduce manual CD4 assays some years earlier but had failed. There appears to be a consensus in Benin that this method does not merit further consideration. It may provide a temporary solution for isolated clinics in the North of the country, and to fill gaps in other places when flow cytometry becomes unavailable for one reason or another. But even when this happens, the tendency is to wait for automated cytometry services to be re-established. 5.2.2 DRC Validation and quality control of national HIV antibody testing algorithms Even before the start of the IHC project, the National Reference Laboratory of the NACP had a system of validation and quality control of HIV tests, established with technical support from the US Centre for Disease Control and with the Institute of Tropical Medicine in Antwerp. A reference panel of local HIV positive sera was already in existence. According to

38

HERA / November 2008

IHC / The Union / Benin & DRC

the Director, HIV diagnostic tests used in the DRC are standardised. Generally Determine HIV1/2® is used for screening and Uni-Gold Recombigen® for confirmation at all testing facilities with few, if any, exceptions. The National HIV Reference Laboratory trained the laboratory technicians in the IHC pilot sites in HIV testing procedures and continues to supervise their work jointly with the NTP. The IHC project introduced a system of quality control of HIV testing to public facilities in the country for the first time. Until then, the exchange of filter paper blood spots for quality control and for the resolution of indeterminate test results was only established with clinics operated by some NGOs, notably Médecins sans Frontières, Médecins du Monde and the testing facilities supported by Family Health International. Viral load measurement According to the Director of the National HIV Reference Laboratory, six Provincial laboratories have the capacity to perform viral load testing. The IHC project supplied an additional PCR machine to the National Reference Laboratory and organised training in the use of this equipment. Routine viral load tests for patients on anti-retroviral treatment are accessible in Kinshasa and in the six provincial laboratories. The procurement of reagents is included in a grant from the Global Fund that has been approved but is not yet implemented. Further expansion of this service will depend on finding a successful solution to the issue of specimen transport. Testing for resistance to anti-retroviral drugs At the time of the mission, a laboratory technician from the National Reference Laboratory was in Montpellier to be trained in the technique of genetic sequencing and to conduct a study on anti-retroviral resistance in the DRC. The objectives of the study are to collect information about the prevalence of resistant HIV strains among patients on ART and among patients who had never been exposed to ART. The National Reference Laboratory has included the procurement and installation of a genetic sequencer in the approved grant application to the Global Fund. This will establish the capacity to perform anti-retroviral resistance testing in Kinshasa. Given this information, the training in genetic sequencing procedures in Montpellier is timely. Immunologic laboratory tests Manual CD4 counting using the Dynabead® system is available and practiced in all IHC pilot sites. Two laboratory technicians in each of the IHC pilot clinics were trained by the National HIV Reference Laboratory. Quality control procedures are in place and are adhered to. On the first level, this involves a double counting of all laboratory specimens by two technicians. During supervision missions by the Reference Laboratory, a sample of specimens is then retested on-site for a second level of quality control. As the graphic from Nord Kivu shows, there is a high degree of correlation between the two CD4 counts performed locally.1

1

Kambale A, Kabuayi JP, Ndongosieme A, Okenge A, Dlodlo R, Fujiwara PI, Boillot F. Assurance de qualité des comptages de th CD4 réalisés au laboratoire de la tuberculose dans le cadre du suivi de la coinfection. Poster Presentation, 39 Union World Conference on Lung Health, Paris, 2008

HERA / November 2008

39

IHC / The e Union / Ben nin & DRC

Manual CD4 countting is the only technolo ogy availab ble for CD4 testing t outsside provinc cial capitalss. The accesss, however, is limited to co-infectted tuberculosis patien nts, primarily y because e there is a scarcity of reagents. Supply S of Dy ynabead® reagents r wa as often me entioned as a pro oblem with long l delayss before arrival of the first shipmen nt and a bre eak in the su upply chain off several mo onths in 200 08. I was to old that the origin o of the ese problem ms were at the t level of The Union’s U cen ntral procure ement unit in India. Even the health faccilities that have h a veryy small ART T patient loa ad outside th he tuberculosis clinic arre only asse essing the CD4 C counts of these pa atients in exxceptional ccircumstanc ces. I was nott able to find d out wheth her this was entirely due to the sho ortage of reagents or whether w this wass on instrucction from th he IHC proje ect coordina ators. Work load was ra arely mentio oned as a limitin ng factor. According to its Dirrector, the NACP N planss to establis sh automate ed CD4 testting in all prrovincial hospitalls, but will continue c to support s manual CD4 assays a below w this level. This plan may have to be modified d according g to patient load. For in nstance the two very closely locate ed hospitalls of the Zones de San nté of Muanda and Kito ona have a combined c a active ART patient p load of 240, 2 a level at which itt may be fea asible to ins stall a share ed automate ed cytomete er. In most health facillities I visite ed, patients who entere ed into care through the e HIV diagn nostic centre (CDV) ( and not n the tube erculosis clinic were iniitiated into ART A on the basis of clinical counts. Th parame eters supporrted by the measureme ent of total lymphocyte l he question arises whetherr the qualityy of their tre eatment is any a less, and whether their t treatme ent results are a any worse because b the ey do not ha ave access to CD4 cou unts. Althoug gh CD4 dete ermination is part of a standard s of adequate care c for peo ople living with w HIV, equally important are a haemato ology and blood b bioche emistry testts for patien nts on anti-re etroviral treatme ent. Yet, these tests are e often not available. a Where W they are, they arre provided on a cost reccovery basiss at prices that t most pa atients cann not afford. I was told th hat a single liver enzyme e test in Gom ma costs USD 10, a fu ull blood che emistry proffile may cosst as much as a USD 50. Som me hospitalss, for instan nce the Gen neral Hospital in Beni, have h spectrrophotomete ers in their lab boratory, bu ut are not prrocuring rea agents beca ause the tessts are unafffordable to patientss. This is an n issue of prriority setting when planning laborratory capaccity building g for the scale-up p of anti-rettroviral thera apy.

40

H HERA / Novem mber 2008

IHC / The Union / Benin & DRC

6 Social and Economic Research 6.1 Description and Objectives The objectives of the social and economic research component of IHC as outlined in the logical framework of the project were very ambitious. Proposed activities distributed over two result areas included: 1. Analyse direct costs of IHC implementation; 2. Evaluate cost of different follow-up approaches; 3. Evaluate affordability and indirect costs (costs to patients receiving IHC services); 4. Conduct DALY analysis; 5. Describe patient adherence to treatment; 6. Analyse health staff adherence to the IHC procedure manual; 7. Analyse cost benefit under different programme and general economic circumstances in the partner countries; 8. Assess IHC acceptability among patients; 9. Develop cost projections for IHC scaling up; 10. Identify possible resources for IHC scaling up; 11. Carry out time-motion and ergometric analysis. In April 2005, M. Philippe Vinard, the consultant mobilised by Alter to provide technical support to the social and economic research component of IHC conducted an inception mission to the DRC and Benin.1 At this time it was already clear that the scope of objectives outlined in the logical framework matrix was over-ambitious and not feasible. The consultant proposed to reprogram the research component. Only parts of his proposal were accepted. The DALY analysis and the time-motion studies were omitted with eventual agreement by the financial donor. However, his suggestion to conduct a cost comparison study between different HIV clinical services in the IHC project areas was not accepted. The final portfolio for social and economic research in IHC is outlined in the application for ethical clearance dated December 2006.2 The application proposed to test five research hypotheses: 1. The integration of HIV care within tuberculosis treatment protocols allows programmes to reach the most vulnerable population subgroups. 2. The integration of HIV care within tuberculosis services generates indirect costs that are affordable for a majority of patients and makes it easier for them to participate in treatment and follow up. 3. The additional workload generated by the integration of HIV care and management into tuberculosis services remains manageable for health centre staff without compromising the quality of tuberculosis care. 4. The reorganisation of work processes allows the health centre to continue performing its routine functions, and can even have a positive impact on other activities. 5. The support provided to the initiative (training, logistics, supervision) is considered by health staff to be sufficient to allow them to perform their role satisfactorily. 1

Philippe Vinard. Programme “Integrated HIV Care for Tuberculosis Patients Living with HIV/AIDS (IHC)” Aspects socio Economiques. Rapport de mission N° 1. 4 au 14 avril 2005 au DRC; 29 avril 2005 au Bénin

2

st

EAG Application: The cost of Integrated TB and HIV care: the perspective of patients and health staff. 1 December 2006

HERA / November 2008

41

IHC / The Union / Benin & DRC

This hypothesis testing was to be achieved through the following activities: 1. A longitudinal study of 60 co-infected patients in each of the three locations (Benin, Bas Congo, Nord Kivu) followed over a period of 18 months with 3-monthly questionnaire surveys. 2. A longitudinal study of care providers in six tuberculosis clinics participating in IHC in each of the three study locations followed over a period of 18 months with 6-monthly questionnaire surveys. 3. A one-time socio-anthropological study of patients and providers (conducted only in Nord Kivu) 6.1.1 Discussion of the Objectives Researching the cost effectiveness of integrating HIV and tuberculosis care in terms of the cost per disability adjusted life year (DALY) saved and in terms of comparing the cost of integrated care with other modes of delivering HIV care should be a priority. Clinical services for HIV are expanding rapidly due to national and international commitments to reach the goals of universal access to HIV care and prevention. Large sums of international money are available to support these initiatives, and efficiency and cost control do not seem to be a major preoccupation at this time. But HIV care, if it is effective, requires a long-term engagement. The demands on national health care systems in terms of finances, human resources, infrastructure, and information management will continue to increase for many years. It appears prudent to start a systematic enquiry about the most cost effective way to deliver these services, and to obtain locally relevant information about the benefits of investing in these services in comparison to other health service priorities. The fact that the IHC initiative will not be able to generate this type of information is therefore regrettable. But it is also clear that the resources of IHC in terms of budget and technical personnel were not sufficient for this task. The limited portfolio of social and economic research included in the IHC should nevertheless be able to answer some important questions in the form of hypothesis testing. There are, however, some doubts whether the research design will actually be able to test these hypotheses. Hypothesis 1: The integration of HIV care within tuberculosis treatment protocols allows programmes to reach the most vulnerable population subgroups. The only interest here is whether integrated clinical services are better at reaching the poorest and most vulnerable populations than stand-alone HIV clinics or other forms of delivering care. That the majority of patients receiving care for tuberculosis are poor is well known and requires no further research. It would be useful to know whether there is a difference in the vulnerability profile between patients receiving HIV care through integrated tuberculosis services and patients receiving care through other modalities. But the proposed research will not be able to generate this information because there is no element of comparison in the research design. Hypothesis 2: The integration of HIV care within tuberculosis services generates indirect costs that are affordable for a majority of patients and makes it easier for them to participate in treatment and follow up. Testing this hypothesis also includes an element of comparison. “Easier” than what alternative? Since the design includes no examination of alternatives, the results of the research will hardly lead to any useful conclusion. Hypothesis 3: The additional workload generated by the integration of HIV care and management into tuberculosis services remains manageable for health centre staff without compromising the quality of tuberculosis care.

42

HERA / November 2008

IHC / The Union / Benin & DRC

Hypothesis 4: The reorganisation of work processes allows the health centre to continue performing its routine functions, and can even have a positive impact on other activities. Hypothesis 5: The support provided to the initiative (training, logistics, supervision) is considered by health staff to be sufficient to allow them to perform their role satisfactorily. The design should be able to test these three hypotheses, and this would indeed provide useful information about how health systems in Benin and the DRC will be able to care for people living with HIV. However, in the Nord Kivu project area for example, information about the organisation of work processes is already being collected in the quarterly reports and analysed by Alter. Information about staff satisfaction and staff perceptions about integrated HIV and tuberculosis care is also included in the in-depth socio-anthropological study conducted in 2008. The three successive questionnaire surveys among staff conducted by CIF Santé therefore appear to be quite redundant. The intentions of the six rounds of questionnaire surveys was to follow a group of patients through their experience of illness and treatment to explore and document outcomes that were not only related to their physical health, but also to their family and community relationships, and to their psychological, emotional, social, economic, and spiritual health. This continues to be a valid line of questioning which is most appropriately answered by a longitudinal study as proposed. The rhythm of three monthly interviews was too tight and probably unnecessary, but that is a minor point. However, what appeared to have happened is that the three research partners focused predominately on trying to achieve a sample size of 60 in each interview round, and only secondarily on maintaining the cohort. In successive interviews the initial cohort was quickly lost and the analysis performed by the research partners is a trend analysis between successive cross sectional questionnaire surveys, each with a small and non-representative sample size. This is obviously meaningless and should have been picked up by The Union as early as the second interview round. However, as one of the researchers told me by email, he never received a response to any of the reports he submitted.

6.2 Status and Achievements 6.2.1

Benin

My assessment of the status and achievements of the social and economic research in Benin is based on limited information. I was not able to meet the local research coordinator in Benin (M. Toussaint Cokou-Dan) because he was out of the country. The only person I was able to interview was a member of the team conducting patient interviews. The team conducting the staff interviews was also not available. I had interviews and e-mail exchanges related to the Benin research with the principal investigator (Mme Odile Ferroussier-Davis), and I obtained copies of four of five reports of completed patient surveys prepared by M. Dan. I was not able to obtain any documentation of the two staff surveys that had been completed at the time of my mission. Because of the constraints noted above, I am not able to assess the research results related to health care personnel. I presume that the three rounds of surveys in the six health facilities will produce some information about work load, staff satisfaction, and staff attitudes towards integrated care but I am not able to assess the usefulness of this arm of the study. The four available reports of successive patient surveys (Round 1, 2, 3, and 5) confirm that a large proportion of tuberculosis patients are poor, that they spend considerable sums of money on seeking treatment before presenting to the tuberculosis clinics, and that the illness is a strain on family resources. The reports also reveal a high level of satisfaction with the integrated HIV and tuberculosis services, and recount some of the difficulties experienced by

HERA / November 2008

43

IHC / The e Union / Ben nin & DRC

patientss in situation ns where th he integratio on is not com mplete, i.e. when they are referred d for HIV carre to a differrent facility. This info ormation is presented in a more or o less anec cdotal mann ner. It may b be possible to analyse e and present this information morre systemattically once all six surve ey rounds are a completted, but there is one major m constra aint: Because the tuberculosis clin nics in Benin transferr patients on nce treatme ent is complleted, these e patients arre no longer included in the study. Only O a small number off patients were w intervie ewed after completion c o of treatmentt. There was no cohort to be e followed. Each intervview round selected a new n sample e, although a proportiion of this sample is ca arried forward to the ne ext round.

A small number of the patientss may actua ally have be een followed d for three o or even four survey rounds, i.e. for up to t one year. This inform mation is no ot included in i the surve ey reports. But B it is improba able that the e study will be able to draw d conclu usions abou ut how the illlness and the t treatme ent affects th he househo old economyy and the qu uality of life of people w who enter in nto HIV care thrrough the tu uberculosis clinic. 6.2.2 DRC onomic rese earch compo onent of the e IHC project in the DR RC suffers from f a The soccial and eco similar weakness w a in Benin, but there are as a some im mportant diffferences. Fo or a numbe er of reasonss related to the two local implemen nting agenc cies, the rhyythm of six ssurvey roun nds for patientss over 18 months was not n maintain ned. In Bas s Congo, the e Ecole de S Santé Publique de Kinshassa will only be able to complete c 4 rounds of patient p surve eys and 3 ro ounds of ca are provider surveys be etween Feb bruary 2007 7 and Novem mber 2008. In Nord Kivvu, CIF-San nte will unds, albeitt with some delays. complette all six rou The Eco ole de Santté Publique had only co ompleted th hree rounds of surveys at the time of this evaluatiion, the lastt in Decemb ber 2007. But it had be een able to maintain m a ccohort that was w somewh hat intact. Of O the 40 pa atients intervviewed in February F 2007, 31 were e re-intervie ewed in Decemb ber of the same year. It remains to o be seen how h many will w be includ ded in the la ast round of o interviewss to be cond ducted in Occtober 2008 8. According to discusssions with the principal researcher (Dr. Eric Mafuta), th he series of three rounds of patient in nterviews in n Bas Congo o show som me trends. However H (a)) it is dange erous to draw co onclusions too t early in a cohort stu udy, and (b)) the trendss as they are e presented d are not trends within w the co ohort, but ra ather trendss between each e of the three intervview rounds s which include new patien nts in each round: r •

44

Patients exxperience a decrease in n earning ca apacity and d income, att least at the e beginning of o treatmentt. The studyy of a closed d cohort forr at least 18 months wo ould be required to see if there e is a recove ery as the trreatment ta akes effect. Despite the e d decrease in n earnings, there appea ars to be no o major dete erioration off the social and e economic s status of pattients in Bass Congo. The family ne etwork offerrs some pro otection. T This is quite e different from the find dings in Norrd Kivu.

H HERA / Novem mber 2008

IHC / The Union / Benin & DRC

•

Out-of-pocket expenditures on health care as well as number of episodes of illness are decreasing significantly as the treatment takes effect.

•

Health care seeking outside the HIV and tuberculosis care appears to be shifting, primarily from traditional medicine to pharmacies and self-medication.

These trends may be confirmed or rejected by the fourth round of interviews. In any case, their validity is in doubt if they are not based on an analysis of a closed cohort. The study sample is small and there is no random selection. For the same reason, it will be quite meaningless to use the data from this study to draw conclusions about the profile of patients who access HIV and tuberculosis care in the DRC. Some of the more qualitative findings of the study in Bas Congo confirm what is also observed in Benin. Patients are very satisfied with the integration of HIV and tuberculosis care. They prefer to be treated by the same personnel for a number of reasons. They are not spending much money on TB and HIV care, although they do make health care expenditures for other illnesses. The indirect costs in terms of transport, waiting time, etc. are generally bearable; at least they are not interfering with treatment adherence. The Ecole de Santé Publique also conducted two survey rounds among health staff in Bas Congo in February and December 2007. This enquiry among approximately 20 care providers in 5 different health facilities documents that their work load has increased because of the integration of HIV and tuberculosis care, sometimes significantly. While they are dissatisfied with remuneration, as are all health workers in the DRC, it is also evident that the IHC project has significantly contributed to increasing their job satisfaction. In Nord Kivu, CIF-Santé completed the 5th round of surveys in September 2008. The report is not yet available. In addition to the completed survey reports, I obtained a preliminary (uncompleted) version of the report for the 4th round of surveys which was conducted in May 2008. As in the other two sites the survey reports of CIF Santé are extensive documents with many graphs and trend analyses of limited usefulness. The principal researcher, C. Kimanuka whom I was not able to meet, complained in an e-mail that he has never received any feedback from The Union about his reports, nor any guidance beyond a site visit conducted at the beginning of the study in early 2007. CIF Santé is aware of the objective of the longitudinal study and has tried to follow patients for at least three interview rounds. This has not been easy because apparently many patients discontinue treatment or are lost to follow-up. Among the 66 patients interviewed during the fourth round, 14 were interviewed for the 4th time, and 11 for the 3rd time. This did result in one interesting sub-sample analysis of patients who were interviewed at least three times, showing a marked decrease in personal health service expenditure among these patients.

HERA / November 2008

45

IHC / The Union / Benin & DRC

This is a significant finding. The other analyses of trends prepared by CIF Santé are much less useful because they confound patients who where interviewed once, twice, thrice, or four times. The combination of tuberculosis and HIV infection appears to have catastrophic economic effects on patients and their families. This is much more marked in Nord Kivu than in the other two study areas. The exceedingly high level of poverty in the region, combined with the prolonged situation of war, may be the main reasons why family networks are not able to soften the double blow of loss of income and increasing expenses on health care. Before reaching the public tuberculosis and HIV services, most patients spend considerable sums of money on treatment by traditional healers and on long hospitalisations. Most end up selling their fixed assets to finance this care. In general, patients are very satisfied once they receive effective treatment for tuberculosis and HIV free of charge. Direct expenditures on health care drop rapidly. The rates of adhesion to treatment are high, even among those patients who have to travel far to the nearest clinic. As in the other two IHC project areas, health care staff is dedicated and very satisfied with the positive results of treating HIV and tuberculosis jointly. They accept the fact that their work load has increased. However, even more so than in the other two areas, the economic conditions of health care staff are precarious. Despite the dedication of nurses and laboratory technicians to the job, they will not be able to continue providing these services without adequate remuneration. Nord Kivu is the only IHC project area that benefited from an in-depth socio-anthropological study conducted in April/May 2008. The study report is very rich in information about personal and community perceptions of illness, about care seeking behaviour and motivation among people with tuberculosis and HIV infection, about the social effects of the illness on the family, and about the interaction between patients and care providers from both perspectives. The study makes rare reference to results of the longitudinal study and can easily stand on its own. In fact, it raises the question whether the longitudinal series of surveys contribute any knowledge that could not have been obtained independently by this one-time study.

46

HERA / November 2008

ANNEX 1

IHC / The Union / Benin & DRC

Annex 1. Calculation of Cohort Survival Curves in the DRC All data used for the calculation of the survival curves are taken from the following table of an Excel workbook provided by the project in the DRC. cohorte 2006 suivi sans traitement DCD sans traitement suivi sous CTM arrêt CTM PDV CTM DCD CTM transférés CTM progressés CTM suivi sous ARV1 suivi sous ARV2 arrêt ARV PDV ARV DCD ARV transférés ARV Total enregistrés

entrée 1

2007 0 1 25 2 26 44 13 58 93 0 0 16 28 3 251

2008 0 1 17 3 27 46 14 61 84 0 1 20 32 6 251

cohorte 2007 suivi sans traitement DCD sans traitement suivi sous CTM arrêt CTM PDV CTM DCD CTM transférés CTM progressés CTM suivi sous ARV1 suivi sous ARV2 arrêt ARV PDV ARV DCD ARV transférés ARV Total enregistrés (CTM = co-trimoxazole)

entrée

2007

2008

248

120 2 16 27 19 64 145 0 0 10 20 4 363

168

82

251

115

363

2006 0 1 59 1 11 33 11 53 125 0 0 2 8 0 251

82 3 21 32 26 84 152 0 0 15 23 10 364 0 (1 transfert enregistré)

The observation period for the 2006 cohort is 24 months from July 2006 to June 2008 with data available at 0, 6, 18, and 24 months. The observation period for the 2007 cohort is 18 months from January 2007 to July 2008 with data available at 0, 12 and at 18 months. The low mortality estimate is calculated under the assumption that all patients lost to followup (PDV) are still alive at the end of the observation period. The high mortality estimate is calculated under the assumption that all patients lost to follow up have died (DCD) within the observation period.

HERA / November 2008 (Annexes)

A1

IHC / The Union / Benin & DRC

ANNEX 1

Full cohort - 2006, low mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-6 months

251

42

83.3%

83.3%

6-18 months

185

31

83.2%

69.3%

18-24 months

120

6

95%

65.9%

Full cohort - 2006, high mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-6 months

251

55

78.1%

78.1%

6-18 months

185

60

67.6%

52.8%

18-24 months

120

11

90.8%

47.9%

ART cohort - 2006, low mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-6 months

82

8

90.2%

90.2%

6-18 months

125

20

84.0%

75.8%

18-24 months

93

4

95.7%

72.5%

ART cohort - 2006, high mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-6 months

82

10

87.8%

87.8%

6-18 months

125

34

72.8%

63.9%

18-24 months

93

8

91.4%

58.4%

Full cohort - 2007, low mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-12 months

363

47

87.1%

87.1%

12-18 months

267

8

97.8%

85.1%

Full cohort - 2007, high mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-12 months

363

73

79.9%

79.9%

12-18 months

267

18

93.3%

74.5%

A2

HERA / November 2008 (Annexes)

ANNEX 1

IHC / The Union / Benin & DRC

ART cohort - 2007, low mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-12 months

115

20

82.6%

82.6%

12-18 months

145

3

97.4%

80.5%

ART cohort - 2007, high mortality assumption Cohort size at the beginning of the period

Number of deaths during the period

Period survival

Cumulative survival

0-12 months

115

30

73.9%

73.9%

12-18 months

145

8

94.5%

69.8%

HERA / November 2008 (Annexes)

A3

IHC / The Union / Benin & DRC

A4

ANNEX 1

HERA / November 2008 (Annexes)

ANNEX 2

IHC / The Union / Benin & DRC

Annex 2. Patient Record Card Developed by the IHC Project Front view

Back view

HERA / November 2008 (Annexes)

A5