How did the study come about?

The HIV/AIDS pandemic is a public health emergency in many low and middle-income countries. Out of the estimated 33.3 million people with HIV at the end of 2009, 22.5 million were in sub-Saharan Africa and the majority of these were women.1 The introduction in 1996 of combination anti-retroviral therapy (ART) led to a substantial reduction in morbidity and mortality in high-income countries.2,3 In more recent years, efforts by governmental programmes such as the President’s Emergency Program for AIDS Relief (PEPFAR) and The Global Fund as well as non-governmental programmes have resulted in the scale-up of ART in resource-limited settings: at the end of 2009, 5.25 million people were reported to be receiving ART therapy in low- and middle-income countries.1

Although still far from achieving universal coverage,1 the massive concerted scale-up of ART is unprecedented in global health. The long-term outcomes of ART in Africa and other regions are, however, not well defined. Poor retention in care, limited access to second-line ART regimens, co-infections and comorbidities of HIV infection, for example tuberculosis and cancer, and the emergence of drug resistance and toxicities are important challenges to long-term programme effectiveness in resource-limited settings.7 The World Health Organization (WHO) advocates a public health approach to ART in resource-limited settings, to maximize benefit in a setting of low levels of training for health-care workers, high patient burden and limited availability of drugs. Key characteristics include the standardization of first-line and second-line ART regimens, simplified clinical decision-making and standardized clinical and laboratory monitoring.4–6

In 2005, the National Institute of Allergy and Infectious Diseases (NIAID) sought applications for a global consortium structured through regional centres to pool existing clinical and epidemiological data on HIV-infected people, and particularly patients on ART: the International epidemiological Databases to Evaluate AIDS (IeDEA, see Figure 1 for logo).8 Funding of IeDEA has recently been extended to 2011–2016. The seven regions included in IeDEA are North America, Caribbean/Central and South America, Asia/Pacific and four regions in sub-Saharan Africa. A Coordinating Centre (currently at RTI International, NC, USA) provides logistical and data management and harmonization support to the regional networks. Two of the IeDEA cohorts have previously been described in the journal.9,10 The objective of the present report is to describe the four African IeDEA regions: West Africa, Central Africa, East Africa and Southern Africa.

Figure 1

Logo of the IeDEA

Figure 1

Logo of the IeDEA

What does it cover?

The African networks of IeDEA aim to inform the scale-up of ART in sub-Saharan Africa through clinical and epidemiological research. The specific research questions differ by region, but the objectives are similar and cover all populations, including pregnant women, infants, children, adolescents and adult patients. They can be summarized as follows:

  1. To provide robust evaluation of the delivery of ART in children, adolescents and adults in sub-Saharan Africa, with a focus on long-term programme effectiveness and outcomes.

  2. To describe the long-term and temporal trends in regimen durability and tolerability and to examine monitoring strategies.

  3. To describe important comorbidities and co-infections of HIV infection, including malaria, tuberculosis and cancer.

  4. To examine the pregnancy- and HIV-related outcomes of women initiating ART during pregnancy and of infants exposed to HIV or ART in utero.

  5. To develop and apply novel statistical methods to deal with missing data, loss to follow-up, competing risks and time-dependent confounding.

  6. To establish procedures to link the HIV cohort data with other databases, at local or national level, for example routine mortality data or tuberculosis and cancer registries.

Who is in the sample?

A total of 183 clinics providing ART in 17 countries in sub-Saharan Africa participate in IeDEA’s four African regions (Figure 2). Most sites are located in urban areas, operate at the primary care level, are led by nurses or clinical officers rather than physicians and are part of the public health care system of the country (Table 1). About two-thirds of sites have the capacity to measure CD4 cell counts, and a third can do HIV-1 RNA tests. Virtually all sites provide adherence support to patients, screen and provide treatment for tuberculosis. ART provision at most sites started during or after 2004 with the exception of several clinics in West Africa, which introduced ART earlier. Figure 3 shows the cumulative number of treatment sites and adult patients starting ART. There has been a rapid increase in facilities and patients contributing data to IeDEA which is for the most part due to rapid scale-up by a few countries in each region. For example, South Africa and Zambia in Southern Africa and Nigeria in West Africa had the most rapid scale-up in their region. All three contributing countries in East Africa have expanded coverage. Central Africa faces unique challenges in that Rwanda was the only PEPFAR focus country and other countries in the region have extremely weak health systems and government responses to the epidemic. Central Africa has had only modest scale up of treatment and increases in the research database are the result of active data collection and data system strengthening.

Figure 2

Map of 183 ART facilities participating in the four African regions of the International epidemiological Databases to Evaluate AIDS

Figure 2

Map of 183 ART facilities participating in the four African regions of the International epidemiological Databases to Evaluate AIDS

Figure 3

Cumulative number of treatment sites (a) and cumulative number of patients starting antiretroviral therapy (b) in the four African regions of the IeDEA

Figure 3

Cumulative number of treatment sites (a) and cumulative number of patients starting antiretroviral therapy (b) in the four African regions of the IeDEA

Table 1

Characteristics of 183 facilities providing ART in the African regions of the IeDEA

 West Africa Central Africa East Africa Southern Africa All regions (%) 
No. of facilities 15 10 32 126 183 (100) 
Location      
    Urban 15 18 69 110 (60.1) 
    Semiurban 10 20 32 (17.5) 
    Rural 37 41 (22.4) 
Level of care      
    Primary 10 102 131 (71.6) 
    Secondary 15 15 33 (18.0) 
    Tertiary 19 (10.4) 
Type of facility      
    Public 12 28 108 152 (83.1) 
    Private not for profit 12 25 (13.7) 
    Private for profit 6 (3.3) 
    Patient contributions to costs of care 71 74 (40.4) 
Availability of laboratory tests      
    CD4 cell count 15 10 12 71 108 (59.0) 
    Total lymphocyte count 15 12 42 76 (41.5) 
    HIV-1 RNA 10 50 64 (35.0) 
    TST (tuberculin skin test) 15 64 81 (44.3) 
    Sputum smear 15 31 50 100 (54.6) 
    Culture for Mycobacterium tuberculosis 15 23 40 (21.9) 
    Chest X-ray 32 15 55 (30.1) 
Other interventions and services      
    Nutritional support 10 32 69 111 (60.7) 
    Voluntary testing and counselling 15 10 32 62 119 (65.0) 
    Adherence support 15 10 32 123 180 (98.4) 
    Screening for Tb 32 126 167 (91.3) 
    Treatment for Tb 29 121 154 (84.2) 
Visits after starting ART      
    Weekly 7 (3.8) 
    Two weekly 23 69 92 (50.3) 
    Monthly 15 10 58 85 (46.5) 
Visits when stable on ART      
    Monthly 15 13 49 78 (42.6) 
    Every 2 months 10 19 (10.4) 
    Every 3 months 69 87 (47.5) 
Tracing of patients not returning      
    Phone calls 10 26 125 157 (92.4) 
    Home visits 10 29 88 129 (70.5) 
    Community-based organizations 10 42 52 (28.4) 
    Other 10 (5.5) 
    No routine tracing 27 27 (14.8) 
 West Africa Central Africa East Africa Southern Africa All regions (%) 
No. of facilities 15 10 32 126 183 (100) 
Location      
    Urban 15 18 69 110 (60.1) 
    Semiurban 10 20 32 (17.5) 
    Rural 37 41 (22.4) 
Level of care      
    Primary 10 102 131 (71.6) 
    Secondary 15 15 33 (18.0) 
    Tertiary 19 (10.4) 
Type of facility      
    Public 12 28 108 152 (83.1) 
    Private not for profit 12 25 (13.7) 
    Private for profit 6 (3.3) 
    Patient contributions to costs of care 71 74 (40.4) 
Availability of laboratory tests      
    CD4 cell count 15 10 12 71 108 (59.0) 
    Total lymphocyte count 15 12 42 76 (41.5) 
    HIV-1 RNA 10 50 64 (35.0) 
    TST (tuberculin skin test) 15 64 81 (44.3) 
    Sputum smear 15 31 50 100 (54.6) 
    Culture for Mycobacterium tuberculosis 15 23 40 (21.9) 
    Chest X-ray 32 15 55 (30.1) 
Other interventions and services      
    Nutritional support 10 32 69 111 (60.7) 
    Voluntary testing and counselling 15 10 32 62 119 (65.0) 
    Adherence support 15 10 32 123 180 (98.4) 
    Screening for Tb 32 126 167 (91.3) 
    Treatment for Tb 29 121 154 (84.2) 
Visits after starting ART      
    Weekly 7 (3.8) 
    Two weekly 23 69 92 (50.3) 
    Monthly 15 10 58 85 (46.5) 
Visits when stable on ART      
    Monthly 15 13 49 78 (42.6) 
    Every 2 months 10 19 (10.4) 
    Every 3 months 69 87 (47.5) 
Tracing of patients not returning      
    Phone calls 10 26 125 157 (92.4) 
    Home visits 10 29 88 129 (70.5) 
    Community-based organizations 10 42 52 (28.4) 
    Other 10 (5.5) 
    No routine tracing 27 27 (14.8) 

Patients are consecutively included as they initiate HIV care at a participating clinic or programmes until the capacity of the site is reached. At most sites, data collection starts when a patient initiates ART; however, some programmes also collect data on patients in the pre-ART period (not yet eligible for ART or eligible but waiting to be treated). Data are stripped of identifiers at the clinic level and all analysis is performed with de-identified data. Given the use of de-identified data, most patients are not individually consented to participate. However, when additional data is collected from patients outside of routine clinical practice, informed consent is sought. All research is overseen by Institutional Review Boards (IRBs) in the countries where data are collected, and additionally by IRBs with oversight over the analytical teams. In West, East and Southern Africa data are obtained from existing clinical databases. In Central Africa, the study team created prospective cohorts as existing health records were not sufficient for epidemiologic research. Table 2 shows the characteristics of the 286 803 adult patients on ART recorded in the African IeDEA databases as of the end of 2010. In all regions, most patients were 30- to 40-years old, female, and started ART with advanced immune-deficiency and advanced clinical stage (WHO stage III or IV). A substantial minority of patients started ART without a recent CD4 cell count, many patients did not have a CD4 count measured around 6 months and only few patients had viral load measurements at baseline or 6 months. In all regions, the most commonly used first-line ART regimen was lamivudine (3TC), stavudine (d4T) and nevirapine (NVP), and the majority of patients were started on one of three regimens (Table 3). The African IeDEA regions also include smaller cohorts of paediatric patients, which will be described in a separate report.

Table 2

Baseline characteristics of patients starting antiretroviral therapy at sites participating in the African regions of the IeDEA

 West Africa Central Africa East Africa Southern Africa 
Total No. of patients 33 368 8902 60 137 184 386 
No. of female 21 057 (63.1) 6252 (70.2) 40 531 (67.4) 116 349 (63.1) 
Gender unknown 145 (0.43) 3 (0.03) 303 (0.50) 4 (0.00) 
Age (years) 40.6 (34.1–42.8) 38 (32–46) 35.8 (30.2–42.6) 29 (24–36) 
Unknown 355 (1.1) 7 (0.08) 3680 (6.1) 564 (0.31) 
Weight (kg) 57 (49–65) 60 (52–70) 55.4 (49.0–62.5) 55 (48–62) 
Not measured 3367 (10.1) 27 (0.30) 2937 (4.9) 15 107 (8.2) 
Height (cm) 165 (159–170) 164 (159–170) 164.5 (159.5–170.5) 164.0 (158–170) 
Not measured 12 340 (37.0) 43 (0.48) 21 964 (36.5) 71 141 (38.6) 
Advanced clinical stage (WHO stage III/IV) 12 713 (38.1) 5990 (67.3) 32 237 (53.6) 102 178 (55.4) 
WHO stage unknown 6271 (18.8) 29 (0.33) 4973 (8.3) 22 702 (12.3) 
Active Tuberculosis 2042 (6.1) 1892 (21.3) 14 326 (23.8) 11 771 (6.4) 
Unknown 7517 (22.5) 128 190 (69.5) 
Haemoglobin (g/dl) 10.2 (9.0–11.6) 11.5 (9.9–13.2) 11.1 (9.5–12.7) 11 (9.4–12.3) 
Not measured 10 506 (31.5) 4197 (47.2) 34 770 (57.8) 72 340 (39.2) 
CD4 count (cells/μl)a     
    At baseline 145 (62–237) 211 (110–335) 130 (54–211) 126 (62–192) 
    Not measured 8245 (24.7) 3070 (34.50) 19 590 (32.6) 39 694 (21.5) 
    At 6 month 274 (173–402) 318 (213–481) 254 (162–380) 253 (169–362) 
    Not measured 18 555 (55.6) 8428 (94.7) 38 570 (64.1) 103 791 (56.3) 
HIV–1 RNA (log copies/ml)a     
    At baseline 5.11 (4.07–5.63) 1.97 (1.70–2.40) 5.12 (3.94–5.54) 11.14 (9.74–12.39) 
    Not measured 32 175 (96.4) 8513 (95.6) 59 819 (99.5) 155 906 (84.6) 
    At 6 month 2.48 (2.00–2.48) 2.0 (2.00–2.00) 2.60 (2.60–2.60) 4.79 (3.53–5.56) 
    Not measured 32 114 (96.2) 8880 (99.7) 59 689 (99.3) 155 135 (84.1) 
 West Africa Central Africa East Africa Southern Africa 
Total No. of patients 33 368 8902 60 137 184 386 
No. of female 21 057 (63.1) 6252 (70.2) 40 531 (67.4) 116 349 (63.1) 
Gender unknown 145 (0.43) 3 (0.03) 303 (0.50) 4 (0.00) 
Age (years) 40.6 (34.1–42.8) 38 (32–46) 35.8 (30.2–42.6) 29 (24–36) 
Unknown 355 (1.1) 7 (0.08) 3680 (6.1) 564 (0.31) 
Weight (kg) 57 (49–65) 60 (52–70) 55.4 (49.0–62.5) 55 (48–62) 
Not measured 3367 (10.1) 27 (0.30) 2937 (4.9) 15 107 (8.2) 
Height (cm) 165 (159–170) 164 (159–170) 164.5 (159.5–170.5) 164.0 (158–170) 
Not measured 12 340 (37.0) 43 (0.48) 21 964 (36.5) 71 141 (38.6) 
Advanced clinical stage (WHO stage III/IV) 12 713 (38.1) 5990 (67.3) 32 237 (53.6) 102 178 (55.4) 
WHO stage unknown 6271 (18.8) 29 (0.33) 4973 (8.3) 22 702 (12.3) 
Active Tuberculosis 2042 (6.1) 1892 (21.3) 14 326 (23.8) 11 771 (6.4) 
Unknown 7517 (22.5) 128 190 (69.5) 
Haemoglobin (g/dl) 10.2 (9.0–11.6) 11.5 (9.9–13.2) 11.1 (9.5–12.7) 11 (9.4–12.3) 
Not measured 10 506 (31.5) 4197 (47.2) 34 770 (57.8) 72 340 (39.2) 
CD4 count (cells/μl)a     
    At baseline 145 (62–237) 211 (110–335) 130 (54–211) 126 (62–192) 
    Not measured 8245 (24.7) 3070 (34.50) 19 590 (32.6) 39 694 (21.5) 
    At 6 month 274 (173–402) 318 (213–481) 254 (162–380) 253 (169–362) 
    Not measured 18 555 (55.6) 8428 (94.7) 38 570 (64.1) 103 791 (56.3) 
HIV–1 RNA (log copies/ml)a     
    At baseline 5.11 (4.07–5.63) 1.97 (1.70–2.40) 5.12 (3.94–5.54) 11.14 (9.74–12.39) 
    Not measured 32 175 (96.4) 8513 (95.6) 59 819 (99.5) 155 906 (84.6) 
    At 6 month 2.48 (2.00–2.48) 2.0 (2.00–2.00) 2.60 (2.60–2.60) 4.79 (3.53–5.56) 
    Not measured 32 114 (96.2) 8880 (99.7) 59 689 (99.3) 155 135 (84.1) 

No. of patients or median (interquartile range) are shown.

aBaseline was defined as the measurement closest to the start of therapy within a window of 3 months before and 1 week after starting therapy. At 6 month was defined as the measurement closed to 6 months within a window of 3–9 months.

Table 3

The three most common antiretroviral first-line regimens used in facilities enrolled in the four African regions of the IeDEA, 2000–10

 West Africa
 
Central Africa
 
East Africa
 
Southern Africa
 
Anti-retroviral regimen Rank No. on regimen (%) Rank No. on regimen (%) Rank No. on regimen (%) Rank No. on regimen (%) 
3TC-d4T-NVP 10 098 (30.3) 4374 (49.2) 36 418 (60.6) 66 369 (36.0) 
AZT-3TC-NVP 5033 (15.1) 1745 (19.6) 7486 (12.4)   
AZT-3TC-EFV 4992 (15.0) 696 (7.8) 4730 (7.9) 29 098 (15.8) 
3TC-d4T-EFV       45 421 (24.6) 
Other – 13 255 (39.6) – 2087 (23.4) – 11503 (19.1) – 43 498 (23.6) 
 West Africa
 
Central Africa
 
East Africa
 
Southern Africa
 
Anti-retroviral regimen Rank No. on regimen (%) Rank No. on regimen (%) Rank No. on regimen (%) Rank No. on regimen (%) 
3TC-d4T-NVP 10 098 (30.3) 4374 (49.2) 36 418 (60.6) 66 369 (36.0) 
AZT-3TC-NVP 5033 (15.1) 1745 (19.6) 7486 (12.4)   
AZT-3TC-EFV 4992 (15.0) 696 (7.8) 4730 (7.9) 29 098 (15.8) 
3TC-d4T-EFV       45 421 (24.6) 
Other – 13 255 (39.6) – 2087 (23.4) – 11503 (19.1) – 43 498 (23.6) 

3TC, lamivudine; d4T, stavudine; NVP, nevirapine; AZT, zidovudine; EFV, efavirenz.

How often are patients followed up and what is measured?

As IeDEA is based on routine clinical records, the patient follow-up reflects the standards of care in the participating clinics. For most patients, visits are initially bi-monthly or monthly, and then drop to every 2–3 months as therapy is stabilized (Table 1). Clinics have various methods for tracing patients who miss visits including mobile phone calls/SMS or home visits. Some sites also involve volunteers from community-based organizations to track patients. Patient tracing is clinic specific, and the methods and capacity for tracing patients is heterogeneous within the regions.

Most data are collected in the context of routine care at baseline and each follow-up visit, including socio-demographic data, contact details to facilitate the tracing of patients, the date of starting ART, type of treatment initiated, and, where available, CD4 counts and HIV-1 plasma RNA levels at baseline and during follow-up. The switching to second-line ART regimens is recorded in all sites, and the reasons for switching in some sites. Resistance testing is not routinely available in any of the programmes, but is done at some sites in a research context. Important comorbidities and co-infections, including malaria, tuberculosis and cancer, are recorded in most sites.

A survey of sites is conducted regularly to collect data on (i) level of care (primary, secondary, tertiary), points of entry to programme, typical travel time of patients to clinic, costs to patients; (ii) availability of laboratory tests and radiology, availability of other services (family planning, nutritional support), level of staffing; (iii) eligibility and preparation of patients for ART, waiting times, first-line and second-line ART regimens, and monitoring of ART; (iv) follow-up and assessment of adherence, transfers, tracing of patients lost to follow-up, ascertainment of deaths; (v) and the diagnosis and management of HIV-associated complications, including tuberculosis, cryptococcus, cytomegalovirus and malaria.

Linkages to routine data sources have been conducted in the Republic of South Africa. For example, cohort data were linked with the database of the South African National Health Laboratory Services (NHLS) to obtain additional CD4 cell counts, which were not recorded in the HIV database.11 Similarly, cohorts were linked to the routine mortality data to improve ascertainment of deaths.11–13 Linkages with cancer registries are under way.

What is attrition like?

Retention in care is an important issue for the African sites participating in IeDEA, and for treatment programmes in resource-limited settings in general.14,15,16 In a recent IeDEA study of 11 treatment programmes in 10 countries (Botswana, Côte d’Ivoire, Kenya, Malawi, Rwanda, Senegal, South Africa, Uganda, Zambia, Zimbabwe) loss to follow-up at 1 year ranged from 2.8% to 28.7%. In this study, a patient was considered lost to follow-up if the last visit was >9 months before the closure date for that site, with the closure date defined as the most recent visit date recorded in the database.17 A study from the Central African region found that rates of lost to follow-up, defined as not attending the clinic for six months or longer, were 35% in the Democratic Republic of Congo, 38% in Burundi and 27% in Cameroon.18 Using the same definition, an analysis of IeDEA West Africa found that among patients with at least one follow-up visit 20% of patients were lost to follow-up at 1 year.19 The most appropriate definition of loss to follow-up was examined in the Ministry of Health-Centre for Infectious Disease Research in Zambia (MoH-CIDRZ) programme, the largest cohort in the Southern African region.20 The definition that minimized misclassification was ‘at least 2 months late for the last scheduled clinic appointment’.20 Efforts to standardize definitions within and across IeDEA regions are now under way.21

The successful treatment of individual patients and the monitoring and evaluation of ART programmes both depend on regular and complete patient follow-up. Programmes with high rates of loss to follow-up and poor ascertainment of deaths in patients lost will underestimate mortality of all patients starting ART. A meta-analysis of studies tracing patients lost to follow-up found that these patients experience high mortality22 compared with patients remaining in care.23 Standard survival analyses, which censors lost patients, will underestimate overall clinic mortality as censored mortality is estimated from the mortality of patients remaining in care. Analyses of the determinants of survival may also be biased, as empirically demonstrated in an analysis from East Africa.24

IeDEA investigators developed approaches for more accurate and less biased measurements of mortality and determinants of survival. East Africa used methods based on the concept of ‘double sampling’25,26 to adjust mortality estimates based not on those in care, but instead on a subset of patients lost to follow-up whose status was ascertained through extensive tracing efforts.27 In another analysis, the same region used patient tracing data to construct weighted Kaplan–Meier curves, which assign the proper weight to deaths discovered through patient outreach.28 A study based on data from three regions filled the missing survival times of patients lost to follow-up by multiple imputation, using estimates of mortality from studies that traced patients lost to follow-up.29 The Southern Africa region extended these methods to create a simple nomogram and web calculator (see www.iedea-sa.org) which can be used by programme managers to correct mortality estimates for loss to follow-up.17

Key findings and publications

Here, we provide an indicative summary of some of the major research themes. A complete list of publications and presentations from the different IeDEA regions can be found at www.iedea-hiv.org. Mortality and retention in care in children and adults are central to evaluating ART programmes in resource-limited settings, and have been the focus of several analyses. Analyses have considered the first year of ART,15,17,30–33 or the first few years,34–36 and documented high early mortality and loss to follow-up (LTFU), and very high mortality in patients waiting to be treated.11 Significant for programme evaluation, IeDEA Southern Africa found that estimates of adult mortality in South Africa substantially increased after data from the Free State Province,11 the Khayelitsha programme,12 or the Themba Lethu Clinic cohort13 were linked with the South African death registry and deaths among patients LTFU included. Analyses of the South African IeDEA data contributed to evaluating the National Antiretroviral Treatment Programme, both for adults32,33 and children.35

Using data for adult patients who started ART in four scale-up programmes in Côte d’Ivoire, South Africa, and Malawi from 2004 to 2007, IeDEA investigators developed two prognostic models to estimate the probability of death in patients starting ART in sub-Saharan Africa.23 One model with CD4 cell count, clinical stage, bodyweight, age and sex (CD4 count model); and one that replaced CD4 cell count with total lymphocyte count and severity of anaemia (total lymphocyte and haemoglobin model), because CD4 cell count is not routinely available. Probability of death at 1 year ranged from 0.9% to 52.5% with the CD4 model, and from 0.9% to 59.6% with the total lymphocyte and haemoglobin model. Both models accurately predicted early mortality in patients starting ART in sub-Saharan Africa compared with observed data. A web calculator is available at www.iedea-sa.org.

The durability of first-line ART regimens and switching to second-line ART has been another focus. A recent analysis of the United States Agency for International Development–Academic Model Providing Access to Healthcare (USAID-AMPATH) partnership, a large treatment programme in western Kenya, found that ART discontinuation was more common among patients with advanced disease and those receiving a zidovudine-containing regimen.37 A further analysis of data from all four African IeDEA regions found that many patients did not switch to a second-line regimen, despite developing treatment failure.38 Unsurprisingly, these patients experienced high mortality.38

IeDEA has also supported public health programmes through analyses for UNAIDS or WHO. African IeDEA data were used to parameterize the Spectrum projection package, used to estimate the impact of HIV in low and middle-income countries.39 Similarly, IeDEA data were used to evaluate different sampling strategies to assess programmatic indicators of the quality of care.40 A study comparing mortality of HIV-infected patients starting ART in sub-Saharan Africa with background mortality in Côte d’Ivoire, Malawi, South Africa and Zimbabwe used estimates of HIV-unrelated mortality rates from WHO’s Global Burden of Disease project.36 Finally, IeDEA West Africa documented important differences in treatment response in patients infected with HIV-1 and HIV-2, with implications for future treatment guidelines.1

What are the main strengths and weaknesses?

The IeDEA networks in Africa provide a unique platform for operational and clinical research that is highly relevant to the scale-up of ART in sub-Saharan Africa. The large number of participating sites and large number of patients followed in high burden countries are important strengths allowing for determination of outcomes at the individual and programme level. The data reflect routine care across the range of care settings: urban and rural clinics, large and very large programmes run by national health systems, smaller clinics run by non-governmental organizations, and private clinics. The data undergo considerable data cleaning, and the regional teams work closely with clinic staff to understand and correct data quality problems. The AMPATH programme in East Africa provides next day access to study data and individual patient temporal trend graphs, for instance weight and CD4 count, to improve clinical management. These graphs are put in patient charts after data entry so that they are readily available and interpretable. By making data relevant to clinic personnel, the consortium strengthens the relevance of data and increases clinical commitment to collection. The service delivery models, clinical protocols, monitoring schedules and efforts in place to trace patients lost to follow-up vary widely between sites. This diversity of data gives IeDEA substantial ability to generalize findings across different care delivery settings.

The data collected within IeDEA are observational and causal inferences are challenging. Furthermore, participation in IeDEA indicates that the facility has a certain level of capacity in data management and patient follow-up. The participating programmes are more likely than non-participating facilities to be equipped with electronic medical record systems and to have access to CD4 cell counts and second-line therapy, which may reflect a higher clinical capacity generally. Understanding the contextual variables which differentiate IeDEA sites from other care facilities is therefore important. Other weaknesses relate to the quality of the data, with missing data in key variables, varying definitions and data collection protocols, and a high rate of loss to follow-up. In the years to come, the African regions of IeDEA will address these challenges by an iterative process of quality improvement where those variables found to be the most important are harmonized first and their quality emphasized in data improvement efforts. IeDEA will also perform dedicated multicentre studies to address specific questions, and continue to develop advanced statistical methods that can account for missing data, loss to follow-up and competing risks, and time-dependent confounding.

Where can I find out more?

The participating sites sign an agreement to allow their data to be used in IeDEA, however, data ownership remains at the clinic level and all analyses have to be approved by the regional Steering Groups and, if analyses involve several regions, by the Executive Committee of IeDEA. The IeDEA Executive Committee is charged with facilitating data access for all worthy research projects. We welcome collaborations with other cohort studies or cohort collaborations, and other interested parties, for example mathematical modelers or colleagues working in international organizations. Readers who wish to find out more should visit the IeDEA website (www.iedea-hiv.org) where they will find contact details and links to the websites of the different regions.

Funding

The study is funded by the National Institutes of Allergy and Infectious Diseases (NIAID).

References

1
Global report: UNAIDS report on the global AIDS epidemic 2010
 
2010. Geneva, Joint United Nations Programme on HIV/AIDS (UNAIDS), 2010. http://www.unaids.org/globalreport/Global_report.htm (17 March 2011, date last accessed).
2
Egger
M
Hirschel
B
Francioli
P
, et al.  . 
Impact of new antiretroviral combination therapies in HIV infected patients in Switzerland: prospective multicentre study
BMJ
 , 
1997
, vol. 
315
 (pg. 
1194
-
99
)
3
Palella
FJ
Delaney
KM
Moorman
AC
, et al.  . 
Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection
N Engl J Med
 , 
1998
, vol. 
338
 (pg. 
853
-
60
)
4
Gilks
CF
Crowley
S
Ekpini
R
, et al.  . 
The WHO public-health approach to antiretroviral treatment against HIV in resource-limited settings
Lancet
 , 
2006
, vol. 
368
 (pg. 
505
-
10
)
5
World Health Organization
Antiretroviral therapy of HIV infection in infants and children: towards universal access. Recommendations for a public health approach: 2010 revision
2010
Geneva
World Health Organization
 
6
World Health Organization
Antiretroviral therapy for HIV infection in adults and adolescents
 
Recommendations for a public health approach: 2010 revision. http://www.who.int/hiv/pub/arv/adult2010/en/index.html (17 March 2011, date last accessed).
7
Egger
M
Boulle
A
Schechter
M
Miotti
P
Antiretroviral therapy in resource-poor settings: scaling up inequalities?
Int J Epidemiol
 , 
2005
, vol. 
34
 (pg. 
509
-
12
)
8
National Institute of Allergy and Infectious Diseases (NIAID)
Request for Applications. International Epidemiologic Databases to Evaluate AIDS (IEDEA). Bethesda, NIDAID
2005
 
9
Gange
SJ
Kitahata
MM
Saag
MS
, et al.  . 
Cohort profile: the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD)
Int J Epidemiol
 , 
2007
, vol. 
36
 (pg. 
294
-
301
)
10
McGowan
CC
Cahn
P
Gotuzzo
E
, et al.  . 
Cohort Profile: Caribbean, Central and South America Network for HIV research (CCASAnet) collaboration within the International Epidemiologic Databases to Evaluate AIDS (IeDEA) programme
Int J Epidemiol
 , 
2007
, vol. 
36
 (pg. 
969
-
76
)
11
Ingle
SM
May
M
Uebel
K
, et al.  . 
Outcomes in patients waiting for antiretroviral treatment in the Free State Province, South Africa: prospective linkage study
AIDS
 , 
2010
, vol. 
24
 (pg. 
2717
-
25
)
12
Boulle
A
Van Cutsem
G
Hilderbrand
K
, et al.  . 
Seven-year experience of a primary care antiretroviral treatment programme in Khayelitsha, South Africa
AIDS
 , 
2010
, vol. 
24
 (pg. 
563
-
72
)
13
Fox
MP
Brennan
A
Maskew
M
MacPhail
P
Sanne
I
Using vital registration data to update mortality among patients lost to follow-up from ART programmes: evidence from the Themba Lethu Clinic, South Africa
Trop Med Int Health
 , 
2010
, vol. 
15
 (pg. 
405
-
13
)
14
Rosen
S
Fox
MP
Gill
CJ
Patient retention in antiretroviral therapy programs in sub-Saharan Africa: a systematic review
PLoS Med
 , 
2007
, vol. 
4
 pg. 
e298
 
15
Brinkhof
MW
Dabis
F
Myer
L
, et al.  . 
Early loss of HIV-infected patients on potent antiretroviral therapy programmes in lower-income countries
Bull World Health Organ
 , 
2008
, vol. 
86
 (pg. 
559
-
67
)
16
Geng
EH
Nash
D
Kambugu
A
, et al.  . 
Retention in care among HIV-infected patients in resource-limited settings: emerging insights and new directions
Curr HIV/AIDS Rep
 , 
2010
, vol. 
7
 (pg. 
234
-
44
)
17
Egger
M
Spycher
BD
Sidle
J
, et al.  . 
Correcting mortality for loss to follow-up: a nomogram applied to antiretroviral treatment programmes in sub-Saharan Africa
PLoS Med
 , 
2011
, vol. 
8
 pg. 
e1000390
 
18
Losoma
JA
Kiumbu
M
Azinyue
I
Mukumbi
H
Ryder
R
Losses to follow-up in 3 Central African country non-PEPFAR ART delivery programs
5th IAS Conference on HIV Pathogenesis, Treatment and Prevention Cape Town
 (pg. 
19
-
22
July 2009; Abstract WEPED176
19
Ekouevi
DK
Balestre
E
Ba-Gomis
FO
, et al.  . 
Low retention of HIV-infected patients on antiretroviral therapy in 11 clinical centres in West Africa
Trop Med Int Health
 , 
2010
, vol. 
15
 
Suppl. 1
(pg. 
34
-
42
)
20
Chi
BH
Cantrell
RA
Mwango
A
, et al.  . 
An empirical approach to defining loss to follow-up among patients enrolled in antiretroviral treatment programs
Am J Epidemiol
 , 
2010
, vol. 
171
 (pg. 
924
-
31
)
21
Chi
B
Westfall
A
Fox
M
, et al.  . 
Empirically defining lost to follow-up for antiretroviral therapy programs in Southern Africa
XVIII International AIDS Conference, Vienna
 (pg. 
18
-
23
July 2010
22
Brinkhof
MW
Pujades-Rodriguez
M
Egger
M
Mortality of patients lost to follow-up in antiretroviral treatment programmes in resource-limited settings: systematic review and meta-analysis
PLoS One
 , 
2009
, vol. 
4
 pg. 
e5790
 
23
May
M
Boulle
A
Phiri
S
, et al.  . 
Prognosis of patients with HIV-1 infection starting antiretroviral therapy in sub-Saharan Africa: a collaborative analysis of scale-up programmes
Lancet
 , 
2010
, vol. 
376
 (pg. 
449
-
57
)
24
Geng
EH
Glidden
DV
Emenyonu
N
, et al.  . 
Tracking a sample of patients lost to follow-up has a major impact on understanding determinants of survival in HIV-infected patients on antiretroviral therapy in Africa
Trop Med Int Health
 , 
2010
, vol. 
15
 
Suppl. 1
(pg. 
63
-
69
)
25
Frangakis
CE
Rubin
DB
Addressing an idiosyncrasy in estimating survival curves using double sampling in the presence of self-selected right censoring
Biometrics
 , 
2001
, vol. 
57
 (pg. 
333
-
42
)
26
Baker
SG
Wax
Y
Patterson
BH
Regression analysis of grouped survival data: informative censoring and double sampling
Biometrics
 , 
1993
, vol. 
49
 (pg. 
379
-
89
)
27
Yiannoutsos
CT
An
MW
Frangakis
CE
, et al.  . 
Sampling-based approaches to improve estimation of mortality among patient dropouts: experience from a large PEPFAR-funded program in Western Kenya
PLoS One
 , 
2008
, vol. 
3
 pg. 
e3843
 
28
Geng
EH
Emenyonu
N
Bwana
MB
Glidden
DV
Martin
JN
Sampling-based approach to determining outcomes of patients lost to follow-up in antiretroviral therapy scale-up programs in Africa
JAMA
 , 
2008
, vol. 
300
 (pg. 
506
-
07
)
29
Brinkhof
MW
Spycher
BD
Yiannoutsos
C
, et al.  . 
Adjusting mortality for loss to follow-up: analysis of five ART programmes in sub-Saharan Africa
PLoS ONE
 , 
2010
, vol. 
5
 pg. 
e14149
 
30
Fenner
L
Brinkhof
MW
Keiser
O
, et al.  . 
Early mortality and loss to follow-up in HIV-infected children starting antiretroviral therapy in Southern Africa
J Acquir Immune Defic Syndr
 , 
2010
, vol. 
54
 (pg. 
524
-
32
)
31
Brinkhof
MW
Spycher
BD
Yiannoutsos
CT
, et al.  . 
Adjusting mortality for loss to follow-up: analysis of five ART programmes in sub-Saharan Africa
PLoS One
 , 
2010
, vol. 
5
 pg. 
e14149
 
32
Cornell
M
Technau
K
Fairall
L
, et al.  . 
Monitoring the South African National Antiretroviral Treatment Programme, 2003-2007: the IeDEA Southern Africa collaboration
S Afr Med J
 , 
2009
, vol. 
99
 (pg. 
653
-
60
)
33
Cornell
M
Grimsrud
A
Fairall
L
, et al.  . 
Temporal changes in programme outcomes among adult patients initiating antiretroviral therapy across South Africa, 2002-2007
AIDS
 , 
2010
, vol. 
24
 (pg. 
2263
-
70
)
34
Keiser
O
Orrell
C
Egger
M
, et al.  . 
Public-health and individual approaches to antiretroviral therapy: township South Africa and Switzerland compared
PLoS Med
 , 
2008
, vol. 
5
 pg. 
e148
 
35
Davies
MA
Keiser
O
Technau
K
, et al.  . 
Outcomes of the South African National Antiretroviral Treatment Programme for children: the IeDEA Southern Africa collaboration
S Afr Med J
 , 
2009
, vol. 
99
 (pg. 
730
-
37
)
36
Brinkhof
MW
Boulle
A
Weigel
R
, et al.  . 
Mortality of HIV-infected patients starting antiretroviral therapy in sub-Saharan Africa: comparison with HIV-unrelated mortality
PLoS Med
 , 
2009
, vol. 
6
 pg. 
e1000066
 
37
Braitstein
P
Ayuo
P
Mwangi
A
, et al.  . 
Sustainability of first-line antiretroviral regimens: findings from a large HIV treatment program in western Kenya
J Acquir Immune Defic Syndr
 , 
2010
, vol. 
53
 (pg. 
254
-
59
)
38
Keiser
O
Tweya
H
Braitstein
P
, et al.  . 
Mortality after failure of antiretroviral therapy in sub-Saharan Africa
Trop Med Int Health
 , 
2010
, vol. 
15
 (pg. 
251
-
258
)
39
Mahy
M
Lewden
C
Brinkhof
MW
, et al.  . 
Derivation of parameters used in Spectrum for eligibility for antiretroviral therapy and survival on antiretroviral therapy
Sex Transm Infect
 , 
2010
, vol. 
86
 
Suppl. 2
(pg. 
ii28
-
34
)
40
Tassie
JM
Malateste
K
Pujades-Rodriguez
M
, et al.  . 
Evaluation of three sampling methods to monitor outcomes of antiretroviral treatment programmes in low- and middle-income countries
PLoS ONE
 , 
2010
, vol. 
5
 pg. 
e13899
 
41
Drylewicz
J
Eholie
S
Maiga
M
, et al.  . 
First-year lymphocyte T CD4+ response to antiretroviral therapy according to the HIV type in the IeDEA West Africa collaboration
AIDS
 , 
2010
, vol. 
24
 (pg. 
1043
-
50
)