Cameroon suffers from a generalized HIV epidemic, with an estimated average HIV prevalence of 5.5% in the adult population (15–49 years of age) and up to 10.7% and 11.9% in adult women, in the most affected areas (the Northwest, Southwest and East provinces and the capital city of Yaoundé) . In 2008, approximately 91 000 adults and 10 000 children were estimated to be eligible for antiretroviral therapy (ART) . The national ART programme was initiated in 2001 and has already obtained a significant accomplishment as its rate of coverage (58% as of June 2008) is one of the highest in western and central Africa . A key factor in the rapid diffusion of access to ART has been the use of the pre-existing decentralized framework of the Cameroonian health system to deliver HIV care at the district level, because the national rate of coverage was only 22% in 2005 when such a decentralized approach was first implemented.
As early as 1992, the healthcare system had been reorganized on the basis of the district hospitals, whose capacity comprises between 100 and 150 beds and a minimum of one doctor on the permanent staff, and which play a reference role for primary healthcare centres at a more local level. Bigger hospitals (200 beds or more) located in the 10 capitals of the provinces and the eight national hospitals in the two main cities of Yaoundé and Douala are in charge of supervising the districts . As of June 2008, ART delivery was based in 24 accredited treatment centres (ATC) located in the main hospitals of Douala (Littoral province) and Yaoundé (Centre province) and in each of the capitals of the eight other provinces, which serve as mentors and reference centres for 108 HIV management units at the district level; overall, ART delivery facilities were available in 106 out of the 174 districts . It should also be noted that 35 (26%) ART delivery centres are located in private healthcare facilities, mostly not-for-profit hospitals managed by faith-based organizations. All sites have access to antiretroviral drugs from the National Centre for the Provision of Essential Drugs, the public procurement agency, which has a regulatory monopoly for imports of antiretroviral drugs. Management units and ATC are supplied with four first-line regimens composed of two nucleoside reverse transcriptase inhibitors and one non-nucleoside reverse transcriptase inhibitor, whereas second-line regimens are only available in ATC . All ATC, but only some management units, are equipped with FACSCOUNT machines for CD4/CD8 cell counts. Recommended 6-monthly follow-up biological and immunological tests are thus only systematically available in ATC. Management units, however, can systematically refer complicated cases and patients needing second-line regimens to their mentor ATC. Mentor ATC also help to carry out biological tests when necessary, assist in capacity building and oversee the decentralization process in their geographical area of competence. Special incentives have been allocated to healthcare providers involved in the whole process. In order to be accredited, management units must fulfill operational criteria defined by the Ministry of Public Health, which include the capacity to confirm HIV diagnosis, to have a tuberculosis prevention and treatment programme in place, to carry out minimum biological routine tests, to have a pharmacy capable of managing antiretroviral drugs and other essential drugs within the national system managed by the National Centre for the Provision of Essential Drugs, to have a registration and notification system, and finally to have personnel trained in the global management of HIV including ART  (see Appendix 1).
The independent evaluation of the national ART programme requested by the Ministry of Public Health of Cameroon, and jointly carried out by researchers from the Universities of Yaoundé and from the French National Agency for Research on AIDS and Hepatitis (ANRS), gave us the opportunity to assess ART outcomes according to the level of decentralization in HIV care delivery. Our results may also contribute to the current debate on synergies between HIV-targeted interventions, which benefit from financial support of global health initiatives, such as the Global Fund to Fight AIDS, Tuberculosis and Malaria, and health system strengthening.
Materials and methods
A cross-sectional survey (EVAL, ANRS 12-116) was carried out between September 2006 and March 2007 in a random sample of patients having been diagnosed HIV-positive for at least 3 months, being over 21 years old and attending HIV services in eight national ATC in Douala and Yaoundé (Littoral and Centre provinces), six ATC in the province capitals (Littoral, West, Southwest, Northwest and Far North) and 13 district hospitals (management units). Location of the survey's healthcare facilities participating in the survey, with regard to the distribution of the HIV prevalence rate in the country is illustrated in Fig. 1. In each facility, survey participants were randomly selected among eligible patients by healthcare workers, and refusals to participate were recorded. Patients who agreed to participate had to fill in a written informed consent form. At the end of their consultation, they were referred to a trained interviewer, independent from the medical staff, who administered a face-to-face questionnaire including: sociodemographic and economic characteristics of themselves and their households, disease history, perception of treatments and medical follow-up, adherence to ART, detailed healthcare expenditures of the previous month, perception of health status and quality of life, as well as social relationships. After the interviews, blood samples were collected by the hospital laboratory in order to assess patients' CD4 cell count at the time of the survey. Clinical data from both clinical examination of patients and retrospective medical files were collected by care providers using standardized medical questionnaires.
In addition, data about the characteristics of the 27 health facilities participating in the survey were obtained by each supervisor of the survey through interviews of the manager of the hospital and of nursing staff and through cross validation with data included in the administrative reports sent to the National AIDS Control Committee. Data about doctors' characteristics, knowledge of ART management, and perception of their working conditions were collected using a self-administered questionnaire given to all medical staff in charge of HIV care in the facilities (response rate 92%, n = 97).
Main outcome variables and statistical analysis
Comparison of characteristics of facilities, of their medical staff and of ART-treated patients according to the three levels of healthcare delivery were performed using chi-square or Fisher's tests for categorical variables and the Kruskall–Wallis test for continuous variables.
In the present analysis, two main criteria were used to assess ART outcomes according to the level of healthcare delivery in the subsample of patients who had been ART treated for 6 months or more: average monthly gain in CD4 cells/μl since initiation of treatment and adherence to treatment.
Adherence to ART was measured using a validated list of questions concerning dose taking during the previous 4 days, as well as respect of the time schedule and occurrence of treatment interruptions during the previous 4 weeks that allows a score to be computed distinguishing between “high”, “moderate” and “low” adherence to antiretroviral drugs [8,9]. This score had already been validated in samples of HIV-infected patients in Francophone Africa . For the purpose of multivariate analysis, the adherence score was dichotomized into two categories (high versus moderate/low adherence).
The continuous response variable (the average monthly gain in CD4 cells/μl) was transformed to yield a distribution close to Gaussian, using a general transformation based on the inverse cumulative distribution function of the standard normal distribution, which has proved to perform better in general settings than ad-hoc transformations such as the logarithm .
First, the association between the two main outcome variables and the level of healthcare delivery was tested using univariate two-level models (mixed effects regression) for hierarchically structured data (patients nested within care centres), which allow to take into account the correlation between individuals within each hospital-level unit. We used a mixed-effects linear regression model for the continuous variable and a mixed-effect logistic regression model for the binary adherence variable. Multivariate analyses were then conducted to test if the results were confirmed after adjustment for the other predictors of the considered dependent variable. All factors identified in univariate analyses as being associated with the dependent variable at the P < 0.20 level were included in the multivariate analysis. Final models were obtained by stepwise backward selection procedures. The mixed models were estimated using the restricted maximum likelihood method, as implemented in the Stata command xtmixed for the linear model and xtlogit for the logistic model .
Table 1 compares the characteristics of the 27 healthcare facilities participating in the survey and of their medical staff between ATC in the two main cities, ATC in the capitals of provinces and management units at the district level. Not surprisingly, the total workload (number of new ART initiations per month and of ART-treated patients followed up in the facility) was significantly lower at the district level than in ATC, as was the workload per medical staff and the availability of laboratory equipment. It must also be noted that the workload per doctor tended to be higher in provincial ATC than in the ATC of Douala and Yaoundé, although this difference did not reach statistical significance. Except for gender (the proportion of male doctors was significantly higher in provincial ATC), no significant differences in characteristics of doctors and their knowledge of HIV infection and management of ART were found between the different types of facilities. Most doctors (93.8%) delivering ART treatment were not devoting their labour time in the facility exclusively to HIV-infected patients, but either booked some specific consultation days for them or organized mixed consultations for both HIV-positive and HIV-negative patients. In all types of facilities, an overwhelming majority of doctors expressed dissatisfaction about their current level of remuneration. Doctors at the district level, however, reported a greater degree of task shifting to nursing staff in the care of ART-treated patients and tended to express a greater degree of satisfaction with the national policy for implementation of the ART programme.
Among the 3488 patients approached, 3170 (90.8%) agreed to participate in the survey, and 99% filled out the questionnaire leading to a total sample of 3151 respondents (global response rate 90.3%). Among respondents, 2466 (78.3%) were ART treated at the time of the interview, including 1985 patients (63.0% of total sample) who had been treated for 6 months or more; among the non-ART treated, only 180 patients (5.7% of total sample) presented with clinical and immunological characteristics that made them eligible for treatment according to existing guidelines in Cameroon.
Table 2 compares socioeconomic and clinical characteristics of ART-treated patients at the three different levels of healthcare delivery. According to the national estimates, the proportion of women is higher than men in the total HIV-infected population as well as in the population eligible for ART (approximately 62%) ; in addition, ART coverage is significantly higher (62.5%) among female than male patients (51.1%) . This logically explains why the majority of ART-treated patients in the survey sample were women (70.8%). The socioeconomic status of ART-treated patients tended to decrease with the level of decentralization of healthcare at which they were taken care of, with the higher proportion of rural workers and people getting income from the informal sector of the economy at the district level (Table 2). Table 2 also shows that monthly out-of-pocket payments for HIV care and ART were significantly lower at the district level. Because household incomes were significantly lower at the provincial and district levels, the proportion of income devoted to HIV care, as well as the proportion of households incurring “catastrophic” health expenditures, were not different between the three levels of care. Quite interestingly, adherence to ART was significantly higher at the provincial and district levels, and episodes of antiretroviral drug stock-outs were less frequent at the district level. Patients' satisfaction with quality of care (including shorter waiting times for consultation) was also significantly higher at the provincial and district levels. Delays between HIV diagnosis and first consultation in the health facility had also been significantly shorter and CD4 cell counts at initiation of ART significantly higher for patients followed at the district level. Moreover, it is noticeable that first-line regimens containing efavirenz were more often prescribed in the ATC of Yaoundé and Douala, probably due to the high co-infection rate of HIV/tuberculosis in these areas (estimated co-infection rate of 36–43%) . Conversely, Triomune (nevirapine/stavudine/lamivudine), the less expensive treatment at the time of the survey, was first prescribed in district management units where HIV-tuberculosis co-infection was considerably lower (monthly ART prices were established for all HIV treatment centres until May 2007 at 3000 francs de la communauté financière Africaine (FCFA), or approximately €4.5, for Triomune, and 7000 FCFA, or approximately €10.5 for other treatments; at €1 = 655.957 FCFA).
When comparing the 481 patients whose ART had been initiated for less than 6 months with the 1985 who has been ART treated for 6 months or more, the former had logically shorter time periods since HIV diagnosis [median of 7.0 months (interquartile range [IQR] 5.0–17.8) versus 26.3 (15.6–45.1)], shorter time periods since ART initiation [median of 3.9 months (IQR 2.4–5.0) versus 20.2 (11.6–32.1)] and were more likely to have CD4 cell counts of 200 cells/μl or less at the time of the survey (31.2% versus 19.9%, P < 0.001). No significant differences were observed, however, between these two groups of patients regarding the main sociodemographic characteristics, including gender, regarding CD4 cell counts at ART initiation, as well as for all other variables already mentioned in Table 2.
As a result of missing data in the value of CD4 cell counts at ART initiation (329 missing values on 1985 patients treated for at least 6 months; i.e. 16.6%) and in the value of CD4 cell counts at the time of the survey (129, i.e. 6.4%), analysis of monthly CD4 cell gain has been conducted on the subsample of the 1547 patients treated for 6 months or more, for whom data were available. The main sociodemographic and clinical characteristics of patients excluded from the analysis (n = 438) and of the 1547 patients included showed no significant differences. The median (IQR) of CD4 cell counts at ART initiation among the 1547 patients was 123 (52–192).
Detailed results of univariate and multivariate multilevel analyses in the subsample of patients who have been ART treated for 6 months or more at the time of the survey are presented in Table 3. After adjustment for all other explanatory factors, immunological improvements (measured through monthly gains in CD4 cells/μl) were similar in patients followed at both the central (Yaoundé and Douala) ATC and district (management units) levels, but were significantly lower at the intermediary provincial ATC level. The proportion of patients with high adherence to ART in the previous 4 weeks was significantly higher at both the provincial and district levels. Similar results were obtained when using self-reported treatment interruptions (defined as continuous intervals greater than 48 h without taking antiretroviral drugs) in the previous month as an alternative adherence measure [16,17].
Over the past few years, controversy has been growing about the effects on the health systems in general of scaling up interventions to deal with specific diseases, notably HIV/AIDS that has attracted the major bulk of increase in foreign aid devoted to health since the beginning of the new century and that has been the focus of multilateral health initiatives such as the Global Fund to fight AIDS, Tuberculosis and Malaria [18–23]. Early findings from studies focusing at the district level in Ethiopia, Malawi and Zambia suggest that advances in scaling up HIV/AIDS services are accompanied by a fairly complex range of positive and negative impacts on other health services and the system in general (see Global HIV/AIDS Initiatives Network www.ghinet.org). One major concern is that HIV-targeted programmes may be less effective in delivering health services than integrated programmes. Another concern deals with equity in access to care in the HIV-infected population itself , the risk being that access may be restricted to more developed urban areas and the more privileged [25,26].
The independent evaluation of the national ART programme in Cameroon suggests that polarization of expert opinions on these two issues should be, at least partly, overcome, to the extent that scaling up access to ART is effectively decentralized. In spite of their more limited human and technical resources, of the use of simplified algorithms with restricted availability of complementary biological examinations for the management of care, and of a higher degree of task shifting from doctors to nurses , ART delivery centres based in decentralized district hospitals present similar and sometimes better performance than referral centres. Although such findings had already been observed at the small-scale level of individual facilities , they are confirmed at the scale of the global ART programme in Cameroon. A non-inferiority randomized controlled trial (STRATALL – ANRS 12-120 project) is currently ongoing in district hospitals of the Centre Province in order to assess the clinical effectiveness of the simplified approach used for the management of ART at district level . In addition, our results suggest that the closer proximity of district hospitals to local communities seems to facilitate patients' adherence to treatment and earlier access to treatment after HIV diagnosis. This latter result was also suggested by another study conducted using the EVAL data, which showed that among eligible patients, those who were diagnosed for HIV in a healthcare facility not agreed to deliver ART were less likely to be ART treated, probably as a result of difficulties in the referral system .
Moreover, although the whole programme has been able to guarantee gender equity, decentralized delivery of ART at the district level clearly allows increased access to treatment for the poorest sectors of the HIV-infected population as well as those living in rural areas. In Cameroon, as in many other African countries, the chances of poor individuals being exposed to HIV in the first place are not necessarily greater than for wealthier individuals , but it is undisputable that the former are likely to be hit harder by the downstream impacts of AIDS. In the epidemiological situation of Cameroon, where better-educated individuals tend more likely to be HIV infected , decentralization of access to treatment may be one of the only opportunities to promote equity of access and to limit the capture of a disproportionate portion of public health resources by some groups with higher social capital .
Some limitations of the EVAL survey should be acknowledged. First, although it was carried out in one of the largest samples ever of the HIV-infected population attending health facilities for HIV care in Africa, and it covered more than a quarter of all ART delivery centres at time of data collection, the survey was only representative of six out of the 10 provinces in Cameroon. Moreover, three of the four provinces that were not included (Adamawa, South and East) had the lowest (<40%) rate of ART coverage although they had higher HIV prevalences than some of the provinces included in the survey (see Fig. 1). In any case, variability of coverage across provinces and sites remains a major challenge for the national ART programme. Another challenge, which national statistics reveal  and which the current evaluation did not fully investigate, is the total proportion of patients (approximately 12%) that are lost to follow-up by the ART delivery centre in which they had initiated their treatment. This rate seems quite comparable to the retention rate of 86% observed in another large scaling up programme of sub-Saharan Africa in Kenya , suggesting, as did the meta-analysis published by Rosen et al. , that the incidence of loss to follow-up did not increase with ART scaling up. Moreover, to date, antiretroviral drug resistance among ART-treated patients has remained relatively low in Cameroon [35,36].
Second, although some retrospective data had been collected, the cross-sectional design of the evaluation did not allow for an in-depth assessment of the long-term sustainability of the ART programme. Domestic funding for AIDS expenditures has increased from less than US$1 million before 2004 to US$9.6 million in 2007, but this has only represented approximately 15% of the total cumulative AIDS budget during the 2004–2007 period (US$139.2 million); and it should be noted that most of the increase in “domestic” funding has come from the allocation to the National AIDS Control Committee of 5% of the global resources made available in 2006 for Cameroon through debt relief in the context of the highly indebted poor countries initiative . In addition, 70% of total purchases of antiretroviral drugs had been supplied by generic manufacturers (mainly from India). This very high dependency of the ART programme on both external funding and the procurement of generic drugs raises strong uncertainties for its forthcoming future. To date, only 1% of treated patients had needed a switch to more expensive second-line regimens, but this proportion will ineluctably increase in the near future.
Finally, the main goal of the evaluation was to assess the efficiency and equity of the national ART programme and its contribution to the fight against the HIV epidemic, but the evaluation was not specifically designed to measure its direct and indirect consequences for other segments of the health system. The intrinsic limitations of observational data do not allow measuring the impact of a counterfactual scenario (what would have happened if the additional resources from multilateral and bilateral donors targeted for ART have been used differently?).
In spite of these limitations, findings of the evaluation of the ART programme in Cameroon suggest that it has benefited, as well as induced, positive synergies with the process of decentralization of health services delivery, but that these positive effects may soon reach a critical reversal point if specific initiatives and policy changes are not introduced to limit the “verticalization” of the programme .
It should be remembered that decentralization of the health system was initiated in Cameroon during the 1990s in a context that was characterized by repeated economic crises, which were associated with a decline in mean incomes as well as adverse distributional shifts contributing to a significant increase in poverty , and by structural adjustment programmes recommended by international institutions that strongly limited the availability of public budgets. Until now (see www.who.int/whosis/), per capita government expenditures on health at an average exchange rate remain as low as US$15, and still represent less than one third (28.0%) of total national health expenditures (5.2% of gross domestic product). As in many other African countries, existing evidence suggests that this economic context had a strong detrimental impact on the health status of the population and on the access to health care of the lowest income groups , as well as on the quality of care provided by the health system. One study found that only a small proportion of healthcare workers were aware of current evidence on maternal and child health care, and that even fewer of them applied it to their clinical practice . Government health personnel had experienced large cuts in salaries, allowances and incentives, and frequently reported additional “survival strategies” such as legal after-hours income-raising activities, parallel selling of drugs, requesting extra charges for services and running private practices during work hours; they were also often less motivated than their counterparts in the private sector . In such a context, decentralization could be de facto assimilated with deregulation and limitation of government's involvement , notably in the field of health care .
The introduction of the national ART programme since 2001 and its rapid scale-up since 2005 has been an opportunity to remobilize the public health sector, to build new ways of collaboration with the private not-for-profit hospitals and to provide additional resources and technical equipment at the district level.
Qualitative studies carried out in parallel to the EVAL – ANRS survey  have, however, pointed out that the success of the increase in ART coverage fuels a trend towards recentralization of the governance of the health system, with potential negative spillovers for the continuation of effective decentralization of healthcare delivery. This is clearly illustrated by the spontaneous tendency to de facto establish a separate chain for antiretroviral drug procurement with tighter direct control of national authorities, which may delay the establishment of an effective procurement and supply system in order to ensure the continuous availability of all essential medicines and technologies . The EVAL survey also points out the persisting dissatisfaction of the medical personnel involved in HIV care, notably at central and provincial levels, which may fuel an increasing pressure to benefit from “exceptional” financial incentives and consequently a growing gap with other categories of doctors and personnel in the same facilities, if answers are not proposed for solving the human resources crisis in the health sector as a whole. Last but not least, the policy of free antiretroviral drugs introduced in May 2007 has not solved all financial difficulties in access to care . In the EVAL sample, ART-treated patients still devoted 4000 FCFA (US$8) in median per month for health expenditures other than antiretroviral drugs, meaning that even with free antiretroviral drugs, approximately a quarter of them still face catastrophic healthcare expenditures. The sustainability of the free antiretroviral drugs policy at the point of delivery and its potential extension to other components of HIV/AIDS services will necessarily imply the improvement of tax-based funding, or health insurance risk pooling mechanisms, or both , as well as the allocation of a greater share of public resources to the health system. This issue may be exacerbated in the near future by the growing needs for second-line regimens for patients in first-line regimens experiencing failure, as well as by the effective scaling up of national guidelines, which recommend undertaking viral load tests once a year. To guarantee its political feasibility, as well as equity with sectors of the population affected by other diseases, a reform of healthcare financing would have to reduce significantly the share of households' out-of-pocket payments in total health expenditures (currently 80%) for the whole population.
For positive synergies between the national ART programme and health system strengthening to be pursued, and risks of jeopardizing the fruitful decentralization process of health delivery to be minimized, rigorous and innovative health policy analysis and reform  is needed more than ever in Cameroon.
The authors would like to thank ANRS and the French NGO SIDACTION for their financial support, including PhD grants. They also thank the Cameroonian Ministry of Public Health for its technical support, as well as all the 27 participating hospitals and their medical teams for their warm welcome and strong involvement in the survey. Finally, the authors would like to thank all the patients who agreed to take part in the study.
Authors' contributions: Study concept and design: SB, FE, MC, CA, SK-S and J-PM. Data collection: SB, FE, MC, CA and ON. Statistical analysis and interpretation of data: SB, FE, MC and J-PM. Writing of the manuscript: SB, FE, MC and J-PM. Critical review of the manuscript: ON, SK-S and J-PM. Overall study supervision: SK-S and J-PM.
Sponsorship: This study was supported by the French National Agency for Research on AIDS and Hepatitis (ANRS) and by the French non-governmental organization SIDACTION.
Competing interests: None.
1. INS (Institut National de la Statistique) and ORC Macro. Demographic health survey in Cameroon
[in French]. Yaoundé, Cameroun: INS, Ministère de la Planification, de la Programmation, du Développement et de l'Aménagement du Territoire; 2005.
2. Ministry of Public Health and National AIDS Control Committee. Towards universal access to care and treatment for adults and children living with HIV/AIDS in Cameroon
[in French]. Progress Report no. 11. Yaounde, Cameroon: Ministry of Public Health and NACC; September 2008.
3. UNAIDS. 2008 Report on the global AIDS epidemic
. Geneva: UNAIDS; 2008.
4. Gruenais M, ed. Changing a health system: the case of Cameroon
. APAD report no. 21. Paris: Euro-African Association for the Anthropology of Social Change and Development; 2002.
5. National AIDS Control Committee. Third progress report on the UNGASS declaration
[in French]. Yaounde, Cameroon: NACC; 2007.
6. Ministry of Public Health. National guidelines for antiretroviral treatment for people living with HIV
[in French]. Yaounde, Cameroon: Ministry of Public Health, République du Cameroun; June 2007.
7. Koulla-Shiro S, Delaporte E. The public health approach to antivetroviral treatment: the case of Cameroon. In: Coriat B, editor. The political economy of HIV/AIDS in developing countries: TRIPS, public health systems and free access. Northampton: Edward Eldar Publishing and ANRS; 2008. pp. 259–271.
8. Le Moing V, Chene G, Carrieri MP, Besnier JM, Masquelier B, Salamon R, et al
. Clinical, biologic, and behavioral predictors of early immunologic and virologic response in HIV-infected patients initiating protease inhibitors. J Acquir Immune Defic Syndr 2001; 27:372–376.
9. Spire B, Duran S, Souville M, Leport C, Raffi F, Moatti JP. Adherence to highly active antiretroviral therapies (HAART) in HIV-infected patients: from a predictive to a dynamic approach. Soc Sci Med 2002; 54:1481–1496.
10. Msellati P, Juillet-Amari A, Prudhomme J, Akribi HA, Coulibaly-Traore D, Souville M, et al
. Socio-economic and health characteristics of HIV-infected patients seeking care in relation to access to the Drug Access Initiative and to antiretroviral treatment in Cote d'Ivoire. AIDS 2003; 17(Suppl. 3):S63–S68.
11. van Albada SJ, Robinson PA. Transformation of arbitrary distributions to the normal distribution with application to EEG test-retest reliability. J Neurosci Meth 2007; 161:205–211.
12. Rabe-Hesketh S, Skrondal A, eds. Multilevel and longitudinal modeling using Stata
. Texas, USA: College Station; 2005.
13. Souville M, Msellati P, Carrieri MP, Brou H, Tape G, Dakoury G, Vidal L. Physicians' knowledge and attitudes toward HIV care in the context of the UNAIDS/Ministry of Health Drug Access Initiative in Cote d'Ivoire. AIDS 2003; 17(Suppl 3):S79–S86.
14. UNAIDS. Epidemiological fact sheets on HIV/AIDS and sexually transmitted infections
. Cameroon 2006 update. Geneva: UNAIDS; 2006.
15. Union Internationale Contre la Tuberculose et les Maladies Respiratoires. Le programme national de lutte contre la tuberculose de la republique du Cameroun
. Rapport no. 12. Yaoundé: Programme Nationale de Lutte contre la Tuberculose; December 2007.
16. Oyugi JH, Byakika-Tusiime J, Ragland K, Laeyendecker O, Mugerwa R, Kityo C, et al
. Treatment interruptions predict resistance in HIV-positive individuals purchasing fixed-dose combination antiretroviral therapy in Kampala, Uganda. AIDS 2007; 21:965–971.
17. Marcellin F, Boyer S, Protopopescu C, Dia A, Ongolo-Zogo P, Koulla-Shiro S, et al
. Determinants of unplanned antiretroviral treatment interruptions among people living with HIV in Yaounde, Cameroon (EVAL survey, ANRS 12-116). Trop Med Int Health 2008; 13:1470–1478.
18. Schieber GJ, Gottret P, Fleisher LK, Leive AA. Financing global health: mission unaccomplished. Health Affairs 2007; 26:921–934.
19. Banati P, Moatti JP. The positive contributions of global health initiatives. Bull WHO 2008; 86:820.
20. Yu D, Souteyrand Y, Banda MA, Kaufman J, Perriens JH. Investment in HIV/AIDS programs: does it help strengthen health systems in developing countries? Globalization and Health 2008; 4:8.
21. Oomman N, Bernstein M, Rosenzweig S. Seizing the opportunity on AIDS and health systems. Washington DC: Center for Global Development; 2008.
22. Garrett L. The challenge of global health.Foreign Affairs
23. Gostin LO. President's emergency plan for AIDS relief: health development at the crossroads. JAMA 2008; 300:2046–2048.
24. Daniels N. Fair process in patient selection for antiretroviral treatment in WHO's goal of 3 by 5. Lancet 2005; 366:169–171.
25. Daniels N, Flores W, Pannarunothai S, Ndumbe PN, Bryant JH, Ngulube TJ, et al
. An evidence-based approach to benchmarking the fairness of health-sector reform in developing countries. Bull WHO 2005; 83:534–540.
26. Hanefeld J. How have global health initiatives impacted on health equity? Promot Educ 2008; 15:19–23.
27. Samb B, Celletti F, Holloway J, Van Damme W, De Cock KM, Dybul M. Rapid expansion of the health workforce in response to the HIV epidemic. N Engl J Med 2007; 357:2510–2514.
28. Sieleunou I, Souleymanou M, Schonenberger AM, Menten J, Boelaert M. Determinants of survival in AIDS patients on antiretroviral therapy in a rural centre in the Far-North Province, Cameroon. Trop Med Int Health 2009; 14:36–43.
29. Loubiere S, Boyer S, Protopopescu C, Bonono CR, Abega SC, Spire B, et al
. Decentralization of HIV care in Cameroon: increased access to antiretroviral treatment and associated persistent barriers. Health Policy 2009; 92:165–173.
30. Gillespie S, Kadiyala S, Greener R. Is poverty or wealth driving HIV transmission? AIDS 2007; 21(Suppl. 7):S5–S16.
31. Fortson JG. The gradient in sub-Saharan Africa
: socioeconomic status and HIV/AIDS. Demography 2008; 45:303–322.
32. Preker AS, Langebrunner J, Langebrunner J, eds. Spending wisely: buying health services for the poor
. Washington, DC: The World Bank; 2005.
33. Ochieng-Ooko V, Ochieng D, Sidle JE, Holdsworth M, Wools-Kaloustian K, Siika A, et al
. Gender and loss to follow-up from a large HIV treatment program in Western Kenya
. Bull WHO
2009; in press.
34. Rosen S, Fox MP, Gill CJ. Patient retention in antiretroviral therapy programs in sub-Saharan Africa
: a systematic review. PLoS Med 2007; 4:e298.
35. Laurent C, Kouanfack C, Vergne L, Tardy M, Zekeng L, Noumsi N, et al
. Antiretroviral drug resistance and routine therapy, Cameroon. Emerg Infect Dis 2006; 12:1001–1004.
36. Kouanfack C, Montavon C, Laurent C, Aghokeng A, Kenfack A, Bourgeois A, et al
. Low levels of antiretroviral-resistant HIV infection in a routine clinic in Cameroon that uses the World Health Organization (WHO) public health approach to monitor antiretroviral treatment and adequacy with the WHO recommendation for second-line treatment. Clin Infect Dis 2009; 48:1318–1322.
37. Nkoa FC, Eboko F. Official development aid and financing of global public goods: the case of the fight against AIDS in Cameroon [in French]
. In: Koulla-Shiro S, Delaporte E, Eboko F, Abé C, eds. Decentralisation of access to ART: Evaluation of the Cameroonian Experience [in French]
. Paris, France: ANRS; 2009 [in press].
38. Sepulveda J, Bustreo F, Tapia R, Rivera J, Lozano R, Olaiz G, et al
. Improvement of child survival in Mexico: the diagonal approach. Lancet 2006; 368:2017–2027.
39. Baye F. Growth, redistribution and poverty changes in Cameroon: a shapley decomposition analysis. J Afr Econ 2006; 15:543–570.
40. Pongou R, Salomon JA, Ezzati M. Health impacts of macroeconomic crises and policies: determinants of variation in childhood malnutrition trends in Cameroon. Int J Epidemiol 2006; 35:648–656.
41. Tita AT, Selwyn BJ, Waller DK, Kapadia AS, Dongmo S. Evidence-based reproductive health care in Cameroon: population-based study of awareness, use and barriers. Bull WHO 2005; 83:895–903.
42. Israr SM. Coping strategies of health personnel during economic crisis: a case study from Cameroon. Trop Med Int Health 2000; 5:288–292.
43. Werlin H. Linking decentralisation and centralization: a critique of the new development administration. Pub Admin Dev 2006; 12:223–235.
44. Camara M, D'Almeida C, Orsi F, Coriat B. Procurement policies, governance models and ARV availability in French-speaking African countries: an overview. In: Coriat B, editor. The political economy of HIV/AIDS in developing countries: TRIPS, public health systems and free access. Northampton: Edward Eldar Publishing and ANRS; 2008. pp. 225–257.
45. Boyer S, Marcellin F, Ongolo-Zogo P, Abega SC, Nantchouang R, Spire B, et al
. Financial barriers to HIV treatment in Yaounde, Cameroon: first results of a national cross-sectional survey. Bull WHO 2009; 87:279–287.
46. Dussault G, Fournier P, Letourmy A. Health care insurance in Francophone Africa. Access to health care and strategies for poverty alleviation [in French]. Washington DC, USA: World Bank; 2006.
47. Walt G, Shiffman J, Schneider H, Murray SF, Brugha R, Gilson L. ‘Doing’ health policy analysis: methodological and conceptual reflections and challenges. Health Policy Plan 2008; 23:308–317.
Appendix 1. National guidelines for antiretroviral therapy delivery in Cameroon 
- Initial evaluation of HIV diagnosed patients by physical examination and CD4 cell numeration (≤200 cells/μl) or complete cell blood count in HIV service where no CD4 cell counter is available.
- Evaluation of patients' eligibility for ART using the Centers for Disease Control and Prevention classification or the World Health Organization (WHO) classification, when CD4 cell testing is not available (WHO stage III or IV and WHO stage II when total lymphocytes are less than 1200).
- For patients eligible for ART: pre-therapeutic subsidized check-up including cell blood count, transaminases (serum glutamic-oxaloacetic transaminase/serum glutamate-pyruvate transaminase), glycaemic status and CD4 cell testing if available.
- Presentation and discussion of medical files of eligible patients by a therapeutic committee, which aims to ensure that the proposed ART regimen is the best regimen to prescribe according to the patient's situation.
- Recommendation of four standard first-line regimens composed of two nucleoside reverse transcriptase inhibitors (NRTI) and one non-nucleoside reverse transcriptase inhibitor (NNRTI): zidovudine or stavudine plus lamivudine plus nevirapine or efavirenz (the less expensive and most often prescribed is a generic fixed-dose combination of nevirapine, stavudine and lamivudine – Triomune, Cipla, India, which was shown to have excellent safety and efficacy). Five alternative regimens are recommended for special cases (tuberculosis co-infection, hepatitis B/C co-infection and pregnant women) composed of two NRTI (including tenofovir or abacavir) and one NNRTI or of three NRTI (abacavir plus lamivudine plus zidovudine).
- In case of treatment failure: reference of patients to the HIV treatment centre of reference in the province, which is authorized to decide initiation of a second-line regimen (including one protease inhibitor and two NRTI).
- After ART initiation, biannual routine monitoring is recommended including the same tests as those of the pretherapeutic check-up and once a year viral load testing, when available.