When the “3 by 5” initiative was launched in 2003, around 400,000 people were estimated to be receiving antiretroviral therapy (ART) in low-income and middle-income countries (LMIC).1 At the end of 2008, more than 4 million people (4,030,000 people, 3,700,000-4,400,000) were receiving ART, representing a 10-fold increase in 5 years and a 35% increase in 1 year.2,3 This massive increase has been made possible by strong political and financial commitment at all levels and efforts by multiple stakeholders coupled with price reduction in antiretroviral medications and a public health approach to treatment.
ART is a life-long intervention and determining the long-term outcomes and impact of treatment programs is necessary to monitor and improve program performance and sustain national and global political commitment to program efforts. Documenting the percentage of adults and children with HIV who continue to receive treatment at yearly intervals after initiation of ART is useful to give programs a more thorough understanding of how effective they are at retaining patients over time.
This article describes trends in ART program retention up to 48 months based on national program data reported by LMIC for the year 2008 to WHO/UNICEF/UNAIDS. It discusses the major determinants associated with retention on ART and the possible programmatic factors to improve it.
To document program retention over time, the “percentage of adults and children with HIV known to be on treatment 12 months after initiation of ART” and at yearly intervals thereafter (24, 36, 48 months….) is recommended as a core indicator of the United Nations General Assembly Special Session (UNGASS) Declaration of Commitment.4 It should be generated per calendar year by selecting all patients starting ART during a given year (denominator) and reporting those who are still alive and on ART at the time-point analyzed (numerator)-that is, excluding those who were known to have died, stopped ART, or were lost to follow-up (LFU). For reporting in 2008, retention at 12 months should be analyzed among patients who started in 2007, retention at 24 months should be analyzed among those who started in 2006, and similarly for longer follow-up.
This information is generated from treatment facilities based on the cohort monitoring system in place. The chronic nature of HIV disease raises specific monitoring issues. Measuring attrition-or the number of people who discontinue ART-requires the cohort monitoring to capture information on deaths, LFU, and treatment stop with a standardized approach over time and across facilities. In particular, the period used for defining LFU may vary from a last missed appointment to a period of several months. WHO has developed a standardized paper-based patient monitoring system and recommends that a patient is classified as LFU if there was no contact for 90 days after the last missed appointment for ARV refill.5 Patients transferring between facilities to continue ART are another concern as they might be misclassified as LFU if they are not properly recorded.
Data on patient retention are aggregated at the facility level and reported to the national program where they are compiled in a national summary. These national data (numerator and denominator) are reported annually to WHO/UNICEF/UNAIDS for monitoring the health sector response to HIV/AIDS in LMIC. For the reporting period 2008, data were received from national programs by March 2009. A process of review and interaction with country-level counterparts was then undertaken when necessary to clarify data sources, methods, and reporting periods. However, valid mechanisms for data quality control are still lacking for indicators of patient retention on ART. Numerators and denominators reported by countries for each indicator at 12, 24, 36, and 48 months were aggregated to produce regional and global figures and are presented here.
In 2008, among 149 LMIC expected to report, 70 (47%) countries (representing 32% of patients on ART at global level in 2008) reported information on retention at 12 months, 54 (36%, representing 21% of patients on ART) at 24 months, 38 (26%, representing 11% of patients on ART) at 36 months, and 30 (20%, representing 7% of patients on ART) at 48 months. Reporting rates over yearly intervals depend the maturity of national programs; for example, only those established in 2004 or before were able to report at 48 months. However, an increasing number of countries reported data on retention after 12 months in 2008 as compared with 2007. In particular, among 47 countries in sub-Saharan Africa, 26 countries reported data on retention at 12 months, 17 at 24 months, 14 at 36 months, and 8 at 48 months in 2008, whereas in 2007, the number of countries reporting data on the same indicators were respectively 24, 2, 1, and 0 countries. However, data from 13 countries (representing 2.3% patients on ART at global level) were further excluded from the analysis due to clear inconsistencies in results and failure to obtain clarification through direct communication with countries. These 13 countries include 8 countries, which reported 100% retention for 1 of the indicators, 1 country which reported 0% retention (Iraq), 2 countries which reported data beyond 12 months without reporting results at 12 months, 1 country which reported only proportions without numerators and denominators, and 1 country with inconsistent denominators over time.
Of 61 LMIC, which contributed data at 12 months in 2008, 10 (16%) reported a 12-month retention rate greater than 90% and 30 (49%) exhibited rates higher than 80%. Overall, regional variations notwithstanding, 51 (84%) countries reported 12-month retention rates greater than 70% (Fig. 1). With the exception of North Africa and Middle East (with 4 country reports at 12 months all showing rates of over 80%), the proportion of countries reporting a retention rate at 12 months below 70% ranged from 13% in sub-Saharan Africa to 23% in East, South, and South East Asia. On the other hand, the proportion of countries reporting a retention rate at 12 months over 90% ranged from 8% in East, South, and South East Asia to 17% in Latin America and Caribbean.
Twenty-five countries reported disaggregated retention rates at 12 months by sex and 21 countries by age. Aggregation of individual country data shows that there was no difference in retention on ART at 12 months between men [77.3% (6960 of 9001)] and women [76.8% (7186 of 9351)] and between children less than 15 years old [77.9% (959 of 1231)] and adults/adolescents [78.3% (16896 of 21585)].
Table 1 and Figure 2 show regional trends for the retention on ART from 12 to 48 months by aggregating country data. Although the analysis of retention at 12 months was based on around 300,000 patients in 61 LMIC, the analysis at 48 months was limited to 25,000 patients from 22 countries; this is also related to the maturity of national programs. Trends showed that most attrition from ART occurred within the first year after initiation of therapy and a slow downward trend in retention with more extended reporting periods (Fig. 2). Among countries of sub-Saharan Africa, retention was estimated at 75.2% at 12 months (22 of 47 countries reporting), at 66.8% at 24 months (13 of 47 countries reporting), and remained at a similar level thereafter. When limited to the 22 countries that reported data for each yearly interval, the analysis showed similar trends in retention: 85.6% at 12 months, 79.3% at 24 months, 76.0% at 36 months, and 73.1% at 48 months.
Measuring retention on ART is critical for determining the effectiveness of programs and inferring their impact. In routine monitoring, retention on ART is used as a proxy for the quality of care. The results presented here, based on routine national program data, focused on trends in program retention over 4 years after treatment initiation. The quality of reported national data depends on the performance of the cohort monitoring system being used in treatment facilities and on how well it can capture attrition. When assessing the numerator, it is not always clear how patients LFU are considered. Some countries may have produced a proxy of survival by excluding only patients deceased and consequently overestimating the true retention on ART. On the other hand, if recorded inaccurately, patient transfers between facilities may cause retention to be underestimated. A meta-analysis of 17 studies from sub-Saharan Africa, tracking outcomes among a total of 6420 patients LFU, found that 34% were not traceable mostly due to registration problems, 29% were deceased, and 37% were alive. Among the latter group, many, ranging from 12% to 54% depending on the study, had actually transferred out to another clinic.6 In case of the denominator, some countries may produce this indicator by selecting only a sample of ART facilities whose national representativeness may be questionable; the completeness of sites reporting was unknown. Others may produce a cumulative indicator for all patients ever started on ART, which would not reflect improvements in the program's overall performance over time. These issues highlight the difficulties that program managers and researchers face when monitoring retention on ART. Further, the number of countries reporting and the number of patients assessed at each yearly interval depend the maturity of the national programs. The overall country response rate was low, and in particular, the 5 countries with the highest burden of patients (South Africa, Kenya, Nigeria, India, Zambia)2 did not report on these indicators. Consequently, these results may not be directly generalizable; retention on ART may be overestimated considering that the programs reporting may be better organized and with better performance.
Despite these limitations, the results are consistent with those obtained from cohort studies. In a review of results from 33 cohorts comprising a total of 74,289 patients who started ART in 13 countries in sub-Saharan Africa, the proportion of patients continuing on ART was estimated at 79.8% at 6 months, 75.1% at 12 months, and 61.6% at 24 months.7 A number of programs have published data on outcomes based on longer follow-up from routine monitoring data that included baseline patient characteristics. In Bostwana's national treatment program, among 633 patients started on ART in 2002, the median CD4 count was 67 cells per cubic millimeter and 85% of patients were at WHO clinical stages 3 or 4. At 5 years, survival was estimated at 79.0% and the probability of LFU stood at 21.8%-in other words, around 57% were on ART after 5 years.8 The Province of Cape Town in South Africa published program results from all of its treatment sites including a total of 12,587 adults and 1709 children. Among adults, 22.7% started treatment with a CD4 count lower than 50 cells per cubic millimeter. Retention was 76% at 4 years among adults and 82.4% at 3 years among children.9 Thailand's national treatment program analyzed outcomes of 58,008 patients who started ART between 2000 and 2005. At baseline, the median CD4 count was 41 cells per cubic millimeter and 51% had AIDS. Survival at 5 years was estimated at 78%; in addition, 8.8% were LFU and 0.8% permanently stopped.10
Attrition from ART programs is high during the first year, and these results show that it declined thereafter, in particular after 24 months on ART. These trends corroborate that attrition is due to the high mortality during the first year of ART, which is mainly related to late access to ART. Various cohort studies in LMIC have shown that most deaths among patients on ART occur in the early months after treatment initiation and that mortality declines substantially thereafter.9,11,12 Evidence suggests that early high mortality rates are likely to depend not only on the quality of care delivered by programs but also more fundamentally on the stage of disease at the time of enrollment in the program and the quality of preceding health care.12 Indeed, the baseline CD4 count has been shown to be the most important predictor of survival11 and of immunological recovery.13 Even in high-income countries, the proportion of patients at advanced disease at the time of HIV diagnosis-defined by clinical AIDS conditions or by CD4 counts below 200 CD4 cells per cubic millimeter-varied from 24% to 45%.14 Access to ART at an advanced stage of disease is still common in most national programs, largely due to late presentation for HIV diagnosis. ART-LINC, a collaboration of 17 cohorts in LMIC, showed trends towards earlier initiation of ART at higher CD4 levels over time among 36,715 adults started on ART between 2001 and 2006 but with a vast majority of patients in the recent years still starting below a CD4 count of 200 cells per cubic millimeter.15
Beside a timely start of ART, further therapeutic considerations for ART may improve survival and retention on ART.16 In addition, a number of programatic factors are also likely to have an impact on retention. Fee-for-service and charges paid by patients for biological monitoring and other services are related to lower retention.7,11,17 With scaling up of ART services, incomplete strategies to cope with the growing workload may interfere with the quality of services and contribute to attrition. The geographical location of treatment centers can also increase barriers faced by patients.18 Ensuring an uninterrupted supply of drug is essential to retain patients on ART; in 2008, 34% of LMIC reported having experienced at least one stock out of an antiretroviral drug.2 Reinforcing HIV care component, with prevention, diagnosis, and treatment of opportunistic infections including referral mechanisms with tuberculosis services, remain important to improving the survival of patients on ART. Cotrimoxazole prophylaxis has been showed to reduce the mortality rate by half during the first 12 weeks of ART.19
To conclude, data from national programs show that after a high attrition during the first year, retention on ART tends to stabilize. High attrition during the first year is likely to be related to the advanced disease stage at which patients start ART due to the late presentation for HIV diagnosis. Late diagnosis remains a major challenge for treatment programs, as it continues to prevent patients from initiating treatment when its impact on survival would be highest. Countries must redouble efforts to make adequate testing and counseling available to all, with immediate and regular screening for ART eligibility.
For national programmes, measuring the rates of patient retention on ART will be crucial to sustain political commitment and financial investment. As programs extend over time, it will be increasingly important to document long-term patient outcomes; yet many countries are still unable to report short-term outcomes and notably those with the highest number of patients. Monitoring retention will become increasingly difficult with growing cohorts of patients receiving ART and place a higher workload on health workers. Innovative approaches, such as the use of a representative sample of patients to measure program outcomes, need to be envisaged to support national programs in their continuous efforts to improve cohort monitoring and data quality.
The authors would like to thank all those, specially in national HIV programs, who contributed to the collection, validation, and analysis of data for annual monitoring and reporting on the health sector response toward universal access in 2009.
1. World Health Organisation and UNAIDS. Treating 3 Million by 2005: Making It Happen
. Geneva, Switzerland: WHO; 2003.
2. WHO, UNAIDS, UNICEF. Toward Universal Access. Scaling Up Priority HIV/AIDS Interventions in the Health Sector. Progress Report
. Geneva, Switzerland: WHO; 2009.
3. Mahy H, Tassie J-M, Ghys P, et al. Estimation of ART coverage: methodology and trends. Curr Opin HIV AIDS
4. UNAIDS. Monitoring the Declaration of Commitment on HIV/AIDS-Guidelines on Construction of Core Indicators 2010 Reporting. Geneva, Switzerland: UNAIDS; 2009.
5. WHO, UNAIDS, PEPFAR, USAID. Patient Monitoring Guidelines for HIV Care and Antiretroviral Therapy
. Geneva, Switzerland: WHO; 2006.
6. 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
7. Rosen S, Fox MP, Gill CJ. Patient retention in antiretroviral therapy programs in sub-Saharan Africa: a systematic review. PLoS Med
8. Bussmann H, Wester CW, Ndwapi N, et al. Five-year outcomes of initial patients treated in Botswana's National Antiretroviral Treatment Program. AIDS
9. Boulle A, Bock P, Osler M, et al. Antiretroviral therapy and early mortality in South Africa. Bull World Health Organ
10. Chasombat S, McConnell MS, Siangphoe U, et al. National expansion of antiretroviral treatment in Thailand, 2000-2007: program scale-up and patient outcomes. J Acquir Immune Defic Syndr
11. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet
12. Lawn SD, Harries AD, Anglaret X, et al. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. AIDS
13. Nash D, Katyal M, Brinkhof MW, et al. Long-term immunologic response to antiretroviral therapy in low-income countries: a collaborative analysis of prospective studies. AIDS
14. Fisher M. Late diagnosis of HIV infection: major consequences and missed opportunities. Curr Opin Infect Dis
15. Keiser O, Anastos K, Schechter M, et al. Antiretroviral therapy in resource-limited settings 1996 to 2006: patient characteristics, treatment regimens and monitoring in sub-Saharan Africa, Asia and Latin America. Trop Med Int Health
16. Bartlett JA, Shao JF. Successes, challenges, and limitations of current antiretroviral therapy in low-income and middle-income countries. Lancet Infect Dis
17. Souteyrand YP, Collard V, Moatti JP, et al. Free care at the point of service delivery: a key component for reaching universal access to HIV/AIDS treatment in developing countries. AIDS
. 2008;22(Suppl 1):S161-S168.
18. Posse M, Meheus F, van AH, et al. Barriers to access to antiretroviral treatment in developing countries: a review. Trop Med Int Health
19. Walker AS, Ford D, Ssali F, et al. Does cotrimoxazole prophylaxis improve outcomes after ART initiation in HIV-infected African adults? A causal analysis using marginal structural models. Presented at: 15th Conference on Retroviruses and Opportunistic Infections; February 3-6, 2008; Boston, MA.