ARV roll-out in Ethiopia has reduced adult AIDS deaths by 50% in capital

John Owuor
Published: 05 March 2009

The roll-out of antiretroviral therapy has led to a decline of about 50% in adult AIDS deaths in Ethiopia's capital, Addis Ababa, over a period of five years, according the findings of a study published in the February 20th edition of the journal AIDS.

The effectiveness of antiretroviral roll-out in sub-Saharan Africa has been widely reported as encouraging despite persistent concerns about universal access and adherence. However, there are still only limited data on its effects at a population level on deaths.

In Ethiopia, antiretroviral treatment was made freely available in public hospitals from October 2005. The investigators carried out the current study to find out what effect the availability of antiretroviral treatment had had on AIDS-related mortality.

The researchers used data from burial surveillance records and 'verbal autopsy' interviews. Burial surveillance was implemented in all Addis Ababa cemeteries in 2001 and records about 20,000 deaths per year. The surveillance is undertaken by cemetery clerks who receive regular training. They record a lay report of the cause of death as narrated by close relatives or friends of the deceased and other demographic details.

Verbal autopsies are post-mortem interviews conducted by researchers with close relatives or caretakers, about the signs and symptoms they witnessed during the terminal illness of the deceased. Causes of death described in the interviews were then confirmed through physicians’ review.

Two different physicians reviewed the verbal autopsies to assign cause of death to the described symptoms. Whenever the assigned cause of death by two physicians did not match, a third physician was used to review the verbal autopsy questionnaire. The data used in the current study were derived from 413 cases involving individuals aged 20 to 64 years. Physicians assigned causes of death, classified as either AIDS or non-AIDS deaths.

Epidemiological modelling was used by the investigators to determine mortality trends in the study population.

The investigators then compared projected deaths with observed numbers from burial surveillance.

To determine possible averted AIDS-related deaths, the investigators compared the estimated with the implied numbers of AIDS deaths in population projections. They estimated HIV prevalence using UNAIDS estimation and projections package (EPP 2007).

Results showed that the ratio of observed over projected deaths in adults peaked in 2001. However, between 2001 and 2005 the ratios dropped by about 11% (from 1.92 to 1.71) for women and 20% for men (from 1.80 to 1.44).This was a period when patients part-paid for treatment. The researchers attributed this average AIDS mortality drop of about 15% to treatment effect and noted that the drop was higher in men than women partly because of sex imbalances in access to healthcare financing.

The results further showed that between 2005 and 2007, there was a decline of about 25% for women (from 1.71 to 1.28) and 21% for men (from 1.44 to 1.13). The investigators attributed this drop of over 40% to free treatment, suggesting that treatment cost is an important factor in the decline of AIDS deaths.

To confirm whether the trends noted above resulted from reduced AIDS mortality, the investigators turned to the findings from the lay reports. They found a decline from 8467 deaths in 2001 to 4230 in 2007 (about 50%). The study further found that the decline was greater between 2005 and 2007 when treatment was free.

The researchers noted that the decline observed took place during a period when mortality was supposed to be very high. HIV infections in Ethiopia peaked in the late 1990s, demonstrating the impact of treatment. Assuming that the burial surveillance coverage was 85%, the scientists estimated a reduction of about 56% in AIDS deaths by 2007.

The 56% decline compared favourably with findings from São Paulo in Brazil, which reported a decline in AIDS deaths of about 65% between 1995 and 2002, noted the investigators. They further compared their findings to a New York study which showed a higher decline (63% in two years).

The researchers said that their findings demonstrate the effectiveness of treatment coverage on averting deaths in early phases of HIV care (long-term impact is not yet known). For Ethiopia and similar settings, the researchers said that the immediate worry is about short-term mortality because treatment coverage is still at its infancy (only 2% of adult patients are on second-line regimens in Addis Ababa).

The researchers also noted that their findings raise questions about whether and how the decline in mortality can be sustained, and whether improvements in access to antiretrovirals alone can achieve this goal. More proactive attempts to diagnose people earlier and initiate treatment earlier may be necessary in order to reduce death rates further, given the continued high risk of death during the first year of antiretroviral treatment in those who start treatment at lower CD4 cell counts in sub-Saharan Africa.

The investigators acknowledged that, even though their study was based on epidemiological models, model-based and observational estimates can be very different. The findings might have been limited further by shortcomings of burial surveillance such as under-reporting, said the scientists.

However, for adaptability to similar settings, the investigators noted that burial surveillance is logistically simple to implement because it uses existing structures and is usable in settings with no population-based data.


Reniers G et al. Steep decline in population-level AIDS mortality following the introduction of antiretroviral therapy in Addis Ababa, Ethiopia. AIDS, 23:511-518, 2009.

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap

This content was checked for accuracy at the time it was written. It may have been superseded by more recent developments. NAM recommends checking whether this is the most current information when making decisions that may affect your health.

NAM’s information is intended to support, rather than replace, consultation with a healthcare professional. Talk to your doctor or another member of your healthcare team for advice tailored to your situation.