Provision and Access

A systematic review of 69 peer-reviewed papers and 20 abstracts from 2002 to 2010 found that the roll-out of ART treatment was successful in numerous sub-Saharan African countries  (Angola, Botswana, Burkina Faso, Cameroon, Côte d'Ivoire, Kenya, Malawi, Mali, Mozambique, Nigeria, Rwanda, Senegal, South Africa, Tanzania, Uganda, Zimbabwe and Zambia) in terms of virological suppression during a two year period.  Viral load measurement of at least once after starting ART and minimum follow-up of three months after initiation of ART was required for inclusion. Virological success was considered in patients in whom HIV-1 RNA level was brought down to fewer than 400 copies. Virological failure was defined as viral RNA level of above 1000 copies. Drug resistance reports for sub-Saharan Africa were summarized and mutations were analyzed according to the International AIDS Society 2008 update. Virological data of 63,684 patients from more than 11 sub-Saharan Africa countries’ HIV/AIDS treatment programs were included. The proportion of women patients in most of the studies was above 50%. Eight percent (5,021) of the patients participated in clinical trials. The follow-up period ranged from a three months to 5 years. Most patients started ART at an advanced stage of disease. The HAART regimen for 58,608 (92%) patients was available. Ninety-nine percent of patients were treated with a regimen consisting of two NRTIs with either nevirapine or efavirenz. Only 6% of the patients were on PI based regimen. After six months of ART, virological success was achieved in 10,351 (78%) out of 13,288 patients based on on-treatment analysis. Seventy-six percent (7,413 out of 9,794 patients) had virological success after one year of ART. At 24 months 3,840 (67%) of 5,690 patients had virological success in on-treatment data.  Only 5 studies reported virological success data at three years in which it was achieved in 712 (67%) of 1,062 patients. When a lower cut-off value for virological success was used (fewer than 50 copies) the proportion of patients with virological success at 6, 12 and 24 months was 67%, 66% and 67% respectively. These finding were comparable to the data from developed countries. The overall proportion of virological failure was 15%. Eighty percent of the programs with on-treatment analysis and 63% with intention-to-treat achieved the target of virological suppression rate of above 70% at one year after initiation of ART. However, data to evaluate the long-term success rates is scarce (Barth et al., 2010). 

Data from 5 demographic surveillance sites collecting longitudinal data in Malawi, South Africa, Tanzania, and Uganda found that excess mortality in people living with HIV decreased by over 50% after the introduction of ART; however, mortality rates in adults living with HIV in these communities was still 10 times higher than the mortality rate in the HIV-negative population.  Surveillance data was utilized to estimate mortality attributable to HIV before and after ART implementation. HIV-attributable mortality was 45 to 88 deaths per 1,000 person-years before ART introduction, and 14 to 46 deaths per 1,000 person-years after ART introduction. A total of 244,269 adults, of which 127,585 knew their HIV status, contributed 1,149,484 person-years of follow-up.  The data was divided between pre-ART introduction (5 years preceding ART introduction), rollout of ART (where ART was assumed to not be fully available), and widespread ART availability.  Declines in mortality were seen in all HIV-positive mortality rates, except for men at one site. The HIV-negative mortality changed little over the study period.  The greatest reduction in mortality rates was seen in older age groups; in some 15-19 year old cohorts the mortality attributed to HIV increased slightly after ART implementation. However, in the rollout period, the mortality reduction was greatest at younger ages (less than 35 years). After ART was implemented, women below age 30 had higher rates of mortality than men, whereas after age 30, men had higher mortality rates than women. Women over 40 have a mortality risk that was 60% or less of the mortality risk of men. The significant reduction in mortality before and after ART initiation was seen in both men and women.  For example, in the cohort in Uganda, the HIV-positive mortality rate for men decreased from 88.0 deaths per 1,000 person-years pre-ART to 43.6 deaths per 1,000 person-years post-ART implementation.  For women, the mortality rate decreased from 91.5 deaths per 1,000 person-years pre-ART to 30.3 deaths per 1,000 person-years post-ART implementation.  This occurred over a period of time when the HIV-negative mortality rate stayed nearly constant; however, the HIV-positive mortality rate was still significantly higher than the 3.8 and 2.3 deaths per 1,000 person-years for HIV-negative men and women, respectively, post-ART implementation.

A longitudinal cohort study following 20,000 people in Uganda from 1999 to 2009 found that the introduction of ART resulted in a decrease of overall mortality in adults by 37% in women and 32% in men. The study evaluated the mortality rates of those living with HIV and those who were HIV-negative people before and after the introduction of ART. ART was introduced in January 2004. In addition, the analysis showed a 27-fold increased mortality risk in HIV-positive compared with HIV-negative individuals aged 15-59 before ART, which fell after ART introduction to 21-fold in the first year after introducing ART and fell nine-fold in the subsequent 4 years. The ART period data is divided into the first year after implementation (January 2004 to January 2005), and the following four years (2005-2009). National guidelines had ART initiated when CD4 count dropped below 200 cells/µl. The introduction of ART significantly brought down the rates of mortality among the entire population by lowering mortality rates of those living with HIV.  In the five years prior to 2004, when ART was provided, the probability of dying from any cause between age 15 and 60 was 51% for men and 44% for women.  In the entire post-ART period, those rates decreased to 38% for men and 32% for women. The pre-ART mortality rate for HIV-negative individuals was 4.0 deaths per 1,000 person-years, while the mortality rate for individuals living with HIV was 116.4 deaths per 1,000 person-years. The mortality rate for individuals living with HIV fell dramatically in the first year after ART was introduced to 87.4 deaths per 1,000 person-years, and the rates decreased further in the following four years to 39.9 deaths per 1,000 person-years.  During this time, the mortality rate among the HIV-negative population did not change significantly, and there was no difference in mortality trends between men and women. The impact was greatest among the individuals aged 30-44 years. Further, the drop seen in mortality among this population is associated with the drop in mortality rates among individuals living with HIV once they have access to ART.

Serological, behavioral, and demographic surveillance data from 2000 to 2009 study of 30,000 people in Tanzania found that the overall fall in female mortality rates was from 8.8 deaths per 1,000 person years before 2005 (pre-ART) to 6.5 deaths per 1,000 person-years after 2005 (post-ART), nearly a 30% decrease in mortality.  The study assessed the impact of free ART on adult mortality rates.  Free HIV care including ART has been available in the area since 2005.  Analysis of the population before and after the introduction of ART show that for women, the crude death rate fell for both women living with HIV and HIV-negative women. Although the death rate among women living with HIV was much higher than their HIV-negative counterparts, the mortality rate among both groups declined with the introduction of free ART.  From 2000 to 2009, the adjusted mortality rate was 10.4 times higher for women living with HIV than for HIV-negative women. The mortality rate among men did not change significantly over the two time periods – at 9.1 deaths per 1,000 person-years in the pre-ART period and 8.5 deaths per 1,000 person-years in the post-ART period.  ART uptake among the study population increased steadily after 2005, with the number of women on treatment quadrupling and the number of men on treatment doubling.

A cohort study conducted with data from six ART programs in South Africa found that patients on ART had life expectancies that were 62-75% of HIV-negative adults, and 70-86% of HIV-negative adults when they started treatment before their CD4 count dropped below 200 cells/µl. Life expectancy was also 15-20% higher in patients who survived 24 months after initiation of ART as compared to patients just starting treatment. The study estimated the life expectancy of patients on ART. The data from 37,740 adults who started ART between 2001 and 2010 was analyzed. ART was initiated when the patient’s CD4 count dropped below 200 cells/µl or when the disease had progressed to the clinical stage IV. Individuals were grouped into CD4 count categories at initiation: less than 50 cells/µl, 50-99 cells/µl, 100-199 cells/µl, and more than 200 cells/µl. A total of 2,066 deaths were recorded in the patient record system, and 16,250 patients were lost to follow-up.  Of the 16,250 patients lost to follow-up, 13,968 were in the national population register and 2,947 deaths were recorded through this method. In total, there were 5,782 deaths in 69,514 person-years of follow-up. A recorded 13.2% of patients had initial CD4 counts above 200 cells/µl, and 61.5% of the patients were women. The mortality rate was 99.8 per 1,000 person-years for men and 72.6 per 1,000 person-years for women.  Mortality rates were highest in the first year of treatment.  The most significant factor determining life expectancy of treated patients was age at ART initiation. A man starting treatment at age 20 had an average of 27.6 additional years while a man starting treatment at age 60 had an average of 10.1 additional years. A woman starting treatment at age 20 had an average of 36.8 additional years while a woman starting treatment at age 60 had an average of 14.4 additional years. Patients who had initial CD4 counts of below 50 cells/µl had life expectancies that were 48-61% of those of HIV-negative adults, a significant reduction as compared to patients who initiated treatment at higher CD4 counts.  Patients who started treatment after 2006 also had higher life expectancies than patients who started ART in 2006 or before.

A cohort study to determine the effects of the provision of ART on adult life expectancy was conducted from 2000 to 2011 in South Africa and found that adult life expectancy increased from 49.2 years in 2003 (the year before ART was introduced) to 60.5 years in 2011. This 11.3-year gain can mostly be attributed to the introduction of ART as data collected on non-HIV-related (HIV-cause-deleted) adult life expectancy remained almost constant over the entire study period.  Based on the standard monetary valuation of life, the provision of ART was found to be very cost-effective.  The study included a total of 101,286 people, of whom about 60,000 were 15 years old or older at any given time and therefore included in the biannual household survey. During this study period, 29% of adults were living with HIV in the geographic area.  Over half of women ages 25-29 and over 40% of women and men 30-34 were living with HIV.  Before the introduction of ART, over half of all deaths were attributed to HIV.  From 2000 to 2003, the adult life expectancy decreased from 55.4 to 51.3 years for women and from 49.0 to 46.9 years for men.  In 2004, South Africa began to provide ART to adults with a CD4 count below 200 cells/µl. In 2010, eligibility for ART was extended to HIV-positive pregnant women and patients with active TB with CD4 counts below 350 cells/µl, and in 2011 these eligibility requirements were extended to all HIV-positive patients. From 2004 to 2011, the adult life expectancy increased to 64.5 for women and 55.9 for men. Verbal autopsy data was also collected to estimate HIV-cause-deleted adult life expectancy. This measure, when compared to the standard adult life expectancy, showed the effects of a particular condition (in this case, HIV/AIDS) on life expectancy as a whole. The HIV-cause-deleted adult life expectancy did not change significantly from 2000 to 2011, which shows that the changes in the overall life expectancy could be largely attributed to HIV and the implementation of ART.  The cost-effectiveness ratio was $1,593 per life year saved, less than a quarter of South Africa's 2011 per-capita gross national income, making this intervention very cost-effective.

A review of 14 studies found that providing ART in humanitarian settings, while challenging, is feasible, with a pooled mortality of 7.6% at six months and 9% at we months. Loss to follow up at six months was 6.3% at 12 months was 8.1%. Mortality, loss to follow up and adherence was comparable to stable setttings.  Six studies were carried out in armed conflict settings; five studies reported outcomes from a setting of post-election violence and three studies from natural disaster settings. Studies took place in Burkina Faso, Democratic Republic of Congo, Ghana, Kenya, Thailand, Côte d'Ivoire, Uganda, India, Sudan, and Haiti. No studies reported virological suppression as an outcome.

A cohort study was conducted in Myanmar (Burma) from 2003 to 2009 in an ART program headed by Médecins Sans Frontières (MSF) and found that 72% of patients were retained in care at 5 years following initiation; 13.8% of the patients died and 6.5% were lost to follow-up over the study period. Attrition (death and loss to follow-up combined) was nearly 4 times higher in the first 6 months than in the following period of 7-36 months.  Older age, being a man, WHO stage IV disease at initiation of ART, and body mass index (BMI) of less than 16 kg/m2 were all predictive of attrition. The study followed all adult patient who initiated ART in the program between 2003 and 2007, a total of 5,963 adults with 17,581 person-years of follow-up.  The median age of patients at baseline was 33 years, 61% were men, and 45% were in WHO stage IV.  For those with a CD4 count available, the median CD4 count was 71 cells/µl. Medical professionals as well as lay counselors and support workers provided services. Adherence support was provided at the clinic and in the community by outreach workers.  Local food support was provided for the first 6 months of treatment and exceptional cases (patients with disabilities) were provided financial support for transportation. CD4 testing was limited throughout the study period; therefore ART was initiated based on WHO clinical staging criteria alone. The strongest predictor of attrition was BMI of less than 16 kg/m2; these patients were lost from the program at more than 4 times the rate of those with a normal BMI.

A study in Uganda conducted with HIV-positive adults who were lost to follow-up to determine the rates of mortality, loss to follow-up, and transfer of care among all those classified as lost to follow-up after being enrolled in care at a CD4 count higher than 350 cells/µl and being unable to initiate ART found that more than half of the patients who were tracked down were retained in care at another facility, whereas slightly less than half were not retained in care. There were 6,473 patients who made a visit to one of two clinics in the study between 2008 and 2011 who were enrolled in care at CD4 counts of more than 350 cells/µl. One thousand ninety-four patients (20%) were lost to follow-up.  In this sample, 71% were female, the median age was 29 years old, and the median CD4 count was 550 cells/µl.  A random sample of the patients lost to follow-up was sought in the community by peer educators.  They were to measure the patients’ vital statistics and current care status.  Patients were considered to be retained in care if they had seen a doctor or nurse for HIV care in the last 6 months.  Two hundred and seven (16% of those lost to follow-up) were randomly selected to be sought in the community. Of those patients who were interviewed directly, 51% were seeing a provider at a new clinic and 49% were no longer seeing an HIV provider.  For those patients no longer in care, the most commonly reported reasons was that work responsibility kept from them seeking care or that they had moved to an area not serviced by an HIV clinic.  For those that had transferred care, the most commonly reported reasons were that the new clinic was closer to their home or work, that they preferred the new clinic, or that they had a conflict with the staff at the clinics under study.  Of those lost to follow-up with updated information, 11% had died.  Among the patients lost to follow-up, 35.4% had initiated ART after 2.5 years.  The rate of death increased over time for patients enrolled in care with CD4 counts of more than 350 cells/µl, from 0.6% at 1 year, to 1.6% at 2 years, and 2.5% at 2.5 years.  The recorded retention in care for these clinics was 69.5% in the 2.5 years measured.  However, when the proportion of patients who were retained in care through other clinics is added, the retention in care measure rises to 88.2%.  These numbers demonstrate that reported numbers of lost to follow-up for those with CD4 counts above initiation levels may be misleading and that transfers in care may be a significant proportion of patients who are classified as lost to follow-up.

A study in South Africa of 3,162 patients initiated on ART found that patients had a one year mortality rate of 7.9% despite many having advanced immunodeficiency, with CD4 counts at initiation at 87 in the period 2002-2004 and 121 in the period from 2007 to 2008. The cumulative probability of death after six years was 15.2%. In addition, over 93% of patients had excellent virological suppression of under 400 copies per milliliter at 16 weeks but the probability of virological failure in the whole cohort of all patients at six years was 23.1%.

A retrospective longitudinal study in Ethiopia of 37,466 patients from 30 hospitals and 25 health centers which provided ART for more than 200 patients by 2008 found that the median CD4 count was 125 when patients were initiated on ART and increased to CD4 counts of 242 at six months, 269 at 12 months and 316 after 24 months on ART.  The number of patients on ART increased from 900 in 2005 to 180,000 by the end of 2008. However, after 24 months on ART, health facilities retained only 68% of their patients.

An observational study done in Chile found that expanded access to ART improved survival of patients in the long term. A total of 5,115 ART naïve patients (26% women) were followed between 2002 and 2007.  At 12 months, 72% of the patients were alive and maintained on the first line-regimen. Virologic suppression at last visit was 73.4% for the 94.8% of patients who had viral load results available. Survival rate was 88.4% at the end of the six-year follow-up period. The mortality rate was 2.55 per 100 patient-years. Inferior survival rates were noted in patients who started ART at a lower CD4 cell counts or at advanced clinical disease.

A study using routinely collected program data from 2004 to 2005 of 972 patients at three government clinics in India found that of the 927 patients for whom treatment outcomes were available, 71% were alive after two years of treatment. The majority of deaths occurred within the first six months of treatment.

Data from the Ministry of Health, Jamaica found that public access to antiretroviral therapy (ART) in 2004 coupled with improved lab capacity and support services resulted in a 40% decreased in AIDS deaths between 2004 and 2008.

A systematic comparison of antiretroviral therapy on mortality of HIV-positive patients in both low-income and high-income countries found that antiretroviral therapy is feasible and effective in low-income settings.  Mortality was higher in the first few months of treatment for patients in low-income settings. Those in low-income settings started treatment with considerably more advanced immunodeficiency than those from industrialized countries, but virological and immunological response to HAART were similar in both settings.  The study compared 4,810 treatment-naïve adult patients (51% female) from 18 HAART programs in Africa, Asia and South America (low-income settings) with 22,217 treatment-naïve adults (25% female) in 12 HIV cohort studies from Europe and North America (high-income settings) and compared baseline characteristics and outcomes during the first year of HAART.

A review by WHO of one randomized controlled trial and 13 observational studies from the USA and other developed countries found a decreased risk of death in persons who initiated ART at CD4 counts of at least 350.

A Cochrane review that analyzed data from two randomized clinical trials found that initiation of ART at CD4 levels above 200 or 250 reduced mortality. A combined total of 1,065 ART naïve, asymptomatic adults above 15 years of age were included in the analysis. Risk of death was reduced by 74% in patients who initiated ART at CD4 counts at 350 (SMART 2008) or between 200 and 350 (CIPRAHT001-Haiti). Risk of TB was reduced by half in the group of patients starting ART at levels above 200 or 250 and at 350. In the SMART study sub-group (249 patients), starting ART when patients had CD4 counts of 350 rather than waiting until the CD4 count decreased to 250 reduced the risk of disease progression by 70%.

A randomized clinical trial with 4,685 men (half MSM) and women (27% women) above age 18 found that over an average of three years, the risk of AIDS, other serious illnesses or death was reduced by 53% among those who started ART immediately (41 events of AIDS, other serious illnesses or death) compared to those who were randomized to only start ART after their CD4 count dropped to 350, (86 events). The benefits of early treatment were similar for the 2,350 participants from low- and middle-income countries and the 2,155 participants from high-income countries: (USA, Europe, Australia, Argentina, Brazil, Chile, Czech Republic, Estonia, India, Malaysia, Mali, Mexico, Morocco, Nigeria, Peru, Poland, South Africa, Thailand and Uganda). The risk of developing serious non-AIDS related events, such as hospitalizations, quality of life, etc., was also lower in the early treatment group compared to those patients who started ART at CD4 counts below 350.  Participants with a CD4 cell count of over 500 were randomized to one of two equal sized groups, where one group initiates ART immediately and the other initiates ART once the CD4 count has decreased to under 350. The original sample size was increased to 4,600 because the baseline CD4 count was much greater than originally assumed (median CD4 cell count about 650 compared with 566 cells as originally assumed).

A follow-up study on the HPTN 052 randomized control trial (Cohen et al., 2011a) and found that early initiation of ART delayed the time to, and significantly reduced the incidence of, AIDS-defining events, tuberculosis, and WHO stage 2 and 3 events. A group of 1,763 HIV-positive people with a serodiscordant partner from sites in Botswana, Brazil, India, Kenya, Malawi, South Africa, Thailand, and Zimbabwe were randomly assigned to early ART treatment or to delayed treatment.  Eight hundred eighty-six were assigned to the early treatment group (initiation of ART at enrollment) and 877 were assigned to the delayed treatment group (initiation when CD4 dropped below 250 cells/µl or with the development of an AIDS-related illness).  The median age of the participants was 33 years, half the study population were women, and the median baseline CD4 count was 436 cells/µl.  The median CD4 count at initiation of ART for the delayed treatment group was 230 cells/µl.  The cumulative probability for developing a primary outcome event over 2 years was 4.8% in the early treatment group and 7.9% in the delayed treatment group.  Primary outcomes included death, onset of WHO stage 4 disease, development of tuberculosis, bacterial infections, serious cardiovascular events, liver and renal disease, development of diabetes, and non-AIDS defining malignant diseases.  There was no difference in development of a primary outcomes between the early and delayed treatment groups when the data was stratified by age, geography, sex, and baseline CD4 count.  The cumulative probably of having an AIDS event over 2 years was 3.3% in the early treatment group and 6.0% in the delayed treatment group.  The cumulative probably of developing tuberculosis over 2 years was 1.2% in the early treatment group and 3.7% in the delayed treatment group.  The study, however, “did not record an effect on all cause mortality,” (Grinsztejn et al., 2014: 9).  Secondary outcomes were also measured, which included stage 2 and 3 events and other conditions including malaria, chronic renal insufficiency, and hypertension (among others).  In total, 34% of patients in the early treatment group and 36% of patients in the delayed treatment group experienced secondary outcomes.  Primary and secondary outcomes were not concentrated among participants with low CD4 counts; most events were recorded when the most recent CD4 count was higher than 350 cells/µl.  Overall, 24% of patients in the delayed treatment group ultimately began ART.  The median change over 2 years in CD4 count for the early treatment group was an increase of 225 cells/µl, while the median for the late treatment group was a decrease of 37 cells/µl.  After treatment was initiated, the median change over 2 years in the late treatment group was an increase of 246 cells/µl.  Although the increase was similar in both groups, the initially lower CD4 cell counts for patients in the delayed treatment group did not reach the same levels as the early treatment group reported, which may suggest that delayed treatment can impede or curtail restoration of immunity. [Note: Others argue that HPTN 052 cited here was not powered to examine mortality (De Cock and El-Sadr, 2013a).]

A randomized controlled trial conducted in Haiti showed that earlier initiation of antiretroviral therapy improved the rate of survival and quality of life of patients. A total of 816 patients (58% women) with CD4 counts between 200 and 350 were recruited between 2005 and 2008 and followed for a total of nearly two years. Patients were randomized to 1) a treatment group in which ART was initiated within two weeks of enrollment to the study, referred to as the early treatment group; or 2) deferred in accessing ART until their CD4 count fell at or below 200 or until an AIDS defining illness developed according to the national guideline at the time (standard treatment group), referred to as the standard treatment group. Only 1% in the early treatment group died compared to 6% in the standard group in the median 21-month follow-up. At 36 months 98% of the participants in the early treatment group were alive versus 93% in the standard treatment group. The risk of dying for a patient in the standard treatment group was four times greater than those in the early treatment group. There was only one death from infectious cause in the early treatment group compared to 17 in the standard treatment group. A fifty percent reduction in the incidence of tuberculosis was also noted in the early treatment group.

Data from the CAPRISA 002 cohort study following women living with HIV in South Africa found an improvement in overall quality of life, and in the sub categories of physical well-being, emotional well-being, and social well-being with the initiation of ART. No negative impact was found on the other two quality of life measurements recorded in the study (functional and global well-being and cognitive functioning). This study included a group of 160 women living with HIV recruited in 2004 and followed for up to nine years.  Participants completed a baseline assessment when enrolled in the study, which included the Functional Assessment of HIV Infection. The assessment measured self-reported quality of life under five subcategories: physical well-being, emotional well-being, function and global well-being, social well-being, and cognitive functioning.  For all participants, this assessment was done at baseline, at three months, and then biannually. Through the period of follow-up, 51 of the 160 participants initiated ART following South African guidelines. Of the ART cohort, the mean age was 25.9 years old, 75% were married or had a stable partner, and 13.7% reported being involved in sex work.  The mean CD4 count was 488 cells/µl at enrollment.  Comparisons between the ART and the non-ART group showed that the self-reported quality of life was higher in the group of women on ART.  About one-third of women in the ART group had long-term gains in quality of life after they initiated ART. Stable partnership was also strongly associated with high self-reported quality of life.

A retrospective cohort study from 2001 to 2010 in four HIV programs in Malawi, Uganda, and Kenya supported by Médecins Sans Frontières (MSF) found that patients with a CD4 count falling below 500 cells/µl experienced higher mortality than patients with a CD4 count consistently above 500 cells/µl. A total of 24,037 patients were followed for 69,516.2 person-years of follow-up.  Of these patients, 68% were women, 2.4% died, and 10.3% were lost to follow-up.  CD4 counts were measured upon initiation of ART and were routinely monitored every 6-12 months.  CD4 counts were classified in five categories: less than 50, 50-199, 200-349, 350-499, and 500 cells/μl or more. The study found higher survival rates of patients at each higher increment of immune response.  The mortality rates were 0.36 deaths per 100 person-years for a CD4 count of at least 500 cells/µl, 0.58 deaths per 100 person-years for 350-499 cells/μl, 0.88 deaths per 100 person-years for 200-349 cells/μl, 1.91 deaths per 100 person-years for 50-199 cells/μl, and 7.43 deaths per 100 person-years for less than 50 cells/µl. The study also found that women had better survival rates than men.  Without the extra support of MSF, "the survival benefit achieved by reaching 500 CD4 cell counts would be even higher," among other sub-Saharan public health programs.

An observational study in Sub-Saharan Africa, with 22,315 patients (69.4% women) who started antiretroviral therapy between January 2000 and 2010 in Uganda found that likelihood of survival increased in patients with higher CD4 counts on initiation of treatment. Baseline CD4 counts were categorized into seven groups, with the lowest CD4 count of less than 50 and the highest CD4 count of at least 300. The Ugandan cut-off for initiation of antiretroviral therapy was a CD4 count under 250). The study showed the highest mortality among patients whose CD4 count was less than 50 at initiation of therapy (risk of dying was 75% likely compared to 41% for those with at least 250 cells). The highest death occurred in the first year of initiation of antiretroviral therapy in all categories. Survival generally decreased with CD4 count and the difference was significant between categories over time.

A study conducted in Brazil found that timely entry to HIV care and treatment reduced mortality. Between 2003 and 2006, 115,369 adults living with HIV were included in the study. A total of 32,602 (28.3%) had CD4 counts under 200 among whom 9,870 (30.3%) had an AIDS defining illness and 12.5% died soon after entry in to HIV care. Late entry to care (i.e. CD4 counts under 350 or an AIDS-defining illness even if CD4 count was above 350) was observed in 43.6% of the patients. A total of 18,002 (15.6%) patients died in the first 12 months of HIV care, which is a 16.5% probability of death from AIDS in the first 12 months of HIV care. Ninety-six percent of the deaths occurred in the first 6 months of treatment. With increased duration of treatment death rates were observed to decrease. Nearly all the deaths (97.5%) occurred among patients who entered into HIV care late. The probability of death in the first 12 months was 36.3% for patients who entered into care late compared to 1% for patients who gained timely access to HIV care. Of the total 18,002 deaths that occurred in the first year, 17,189 could have been avoided had the patients entered into HIV care while still in the early stages of HIV infection (an attributable risk of 95.5%). The estimated number of avoidable deaths corresponded to 39.5% of the total number of deaths (43,523) recorded in the 4 year period. Annual decrease in late entry to care was observed subsequently decreasing the risk of death in the first 12 months from 20.3% in 2003 to 12.5% in 2006.  When asymptomatic individuals with CD4 counts 200 to 350 were included, the number of deaths increased by 249 but the probability of death from AIDS in the first 12 months of care decreased to 27.6% for individuals who entered care late and to 0.6% for those who had timely access to HIV care. The risk of death   attributable to late entry into HIV care increased to 97.2%. The results showed that death from AIDS in the first 12 months was strongly associated to late entry into HIV care. Late entry to HIV care increased AIDS mortality rates by more than one-third.

A study conducted between 2006 and 2010 in Lesotho with 1,177 patients (67% women) found that initiating antiretroviral therapy at CD4 counts above 200 decreased the number of deaths, hospitalization and loss to follow-up. Outcomes were assessed based on baseline CD4 count at initiation of antiretroviral therapy. Five hundred and thirty eight patients were initiated when their CD4 count was 200 or less and 639 were initiated on antiretroviral therapy when their CD4 count was greater than 200. Patients who were started on antiretroviral therapy at when their CD4 counts were above 200 were 68% less likely to die and 39% less likely to be lost to follow-up compared to those started who initiated treatment when their CD4 counts were below 200. The study also found that early initiation of antiretroviral therapy was associated with a 27% risk reduction of morbidity from diseases such as TB, cryptococcal meningitis and diarrhea. Patients who initiated antiretroviral therapy at CD4 counts above 200 had a 63% decreased risk of hospitalization by 63%.

A study of adults on ART in South Africa, initiated between 2004 and 2012, found a 46% decrease in early mortality (defined as within 91 days of ART initiation) or those who initiated ART between 2011 to 2012, when CD4 counts rose to 199 cells, as compared to earlier years, when median CD4 counts were 110 until 2010, then 145 cells in 2011 and 317 by 2012, once ART eligibility criteria changed to CD4 counts of 350 in 2011. Almost half of the adults who enrolled into care between 2011 and 2012 had a CD4 count above 350. Patients enrolled in more recent years also had lower rates of CD4 cell counts under 50. Patients with CD4 cell counts under 50 had a four fold increased risk of early mortality compared with CD4 cell counts of 201-350. Men had higher early mortality. The study was based on a review of 19,080 patients, 67.6% female. Of those who die within the first year of ART initiation, 60% die within the first three months.

The Temprano randomized trial was done in Côte d'Ivoire on 2,076 people from 2008 to 2015.  The study was evaluating the effects of immediate ART initiation as compared to initiation according to WHO criteria, as well as the use of Isoniazid preventive therapy (IPT).  The study found that early initiation of ART reduced the risk of severe morbidity by 44% and that IPT reduced the risk of severe morbidity by 35%. The WHO recommends that people living with HIV who do not have active TB receive IPT to reduce the incidence of TB.  The participants were randomized into four groups: ART initiation with WHO criteria, 6-month course of IPT and ART initiation with WHO criteria, immediate initiation of ART, and 6-month course of IPT with immediate initiation of ART.  The participants were 78% women, had a median age of 35 years, and a median CD4 nadir of 465 cells/µl.  Upon entering the study, no participants were eligible for ART according to WHO guidelines.  Only 2.2% were lost to follow-up.  In this study, immediate initiation of ART and IPT significantly decreased severe mortality for those living with HIV.

A randomized trial of 1,763 couples in nine countries – Botswana, Kenya, Malawi, South Africa, Zimbabwe, Brazil, India, Thailand and USA - in which one partner was HIV-1-positive and the other was HIV-negative, with 50% of infected partners men, found that early initiation of antiretroviral therapy at CD4 counts between 350 and 550 in 886 couples was associated with a relative reduction of 41% in the number of HIV related clinical events, which “suggests a clinical benefit for the initiation of antiretroviral therapy when a person has a CD4 count of 350 to 550 cells per cubic millimeter, as compared with therapy that is delayed until the CD4 count falls into the rage of 200 to 250 cells per cubic millimeter.

A review of 860 patients on ART from 1996 to 2006 in Côte d'Ivoire for a total of 2,789 person years found that rates of death or AIDS in patients with CD4 counts of 350-499 resulted in a decreased rate of mortality, with 10.4 deaths per 100 person-years at CD4 counts between 200 to 349 and 1.6 deaths per 100 person years at CD4 counts of 350-499. However, the rates of death or AIDS in patients with CD4 counts remained substantial and were higher than those reported from Western Europe.

An analysis of 17,517 asymptomatic patients with HIV infection in the United States and Canada who received medical care from 1996 to 2005 that compared HAART initiation when CD4 count was between 351 to 500 cells/mm3 as compared to greater than 500 cells/mm3 found that among patients who waiting to initiate treatment until their CD4 counts were between 351 to 500 cells/mm3, there was an increase in the risk of death. In the first analysis with 8,362 patients, among patients in the deferred therapy group there was an increase in the risk of death of 69% as compared with that in the early therapy group, after adjustment for calendar year, cohort of patients and demographic and clinical characteristics. In the second analysis involving 9,155 patients, there was an increase in the risk of death of 94%. Female sex was associated with an increased risk of death, but the risk was not significant after adjustment for RNA level, a history of injecting drug use or presence of HCV. The benefits of initiating antiretroviral therapy earlier after HIV infection will need to be weighed against adverse effects of treatment. [*Note that study authors receive consulting fees from many of the companies who manufacture and market antiretroviral therapy drugs. The funding for the study, however, came from NIH and other US government agencies.] 

An analysis of multiple cohort studies from Europe and North America showed that patients who initiated ART at CD4 counts of 350 or above had improved survival rates. A total of 45,691 patients of whom 26% were women were followed for those prior to access to ART and those who had access to ART. CD4 counts of patients ranged between zero and 550. Among the 24,444 patients who started ART, 37% started ART when their CD4 counts were between 201 and 350 cells and 21% started ART when their CD4 counts were above 350. Patients who were started on ART when their CD4 counts were between 351 and 450 had improved survival rates compared to patients who initiated ART when their CD4 counts were between 251 and 350. Patients who started antiretroviral treatment when their CD4 counts were between 51 and 150 were 5.67 times more likely to die or progress to AIDS than patients who initiated on ART when their CD4 counts were between 351 and 450.

An observational cohort study done in Europe, Australia and Canada found that initiation of HAART for individuals at CD4 counts of 350 to 500 was associated with slower disease progression. A total of 9,455 (22.1% women), ART naïve, AIDS-free individuals with CD4 under 800 cells were followed between 1996 and 2009. A cohort study was imagined in which subsequent disease progression of individuals who initiated HAART during a specific month was compared with that of patients who did not initiate HAART during this month. The survival times of patients who deferred HAART in the index month were used to represent the average population prognosis of individuals who were AIDS free and HAART naïve with a CD4 cell count in a specified stratum (0-49 cells, 50-199 cells, 200-349 cells, 350-499 cells or 500-799 cells). During the follow-up period 812(8.6%) patients developed AIDS and 544(5.8%) patients died. Initiating HAART at a given month when CD4 drops under 350 was protective compared to deferring HAART. The study found that at CD4 counts 350 to 499 there was a 25% reduction in the risk of death or progression to AIDS and a 49% reduction in mortality from all causes. There was no observed benefit of initiating HAART at CD4 counts 500 to 799 cells. The estimated number of patients who needed to be treated to prevent 1 death or AIDS decreased from 79 to 16 at 5 years. Risk reduction was one third as large for patients with CD4 counts of 350 to 499.

A retrospective observational cohort study done in the United States found that the optimal time for ART initiation to maximize patient health measured in terms of quality of life was to initiate ART at CD4 counts under 554. A total of 1,034 patients of whom 27% were women were followed between 1998 and 2007. At entry of the study patient’s CD4 counts ranged between 200 and 700. Sixty percent of patients started ART during follow-up. Outcome was measured either based on CD4 count or quality of life measured in terms of AIDS events, death, non-AIDS events and CD4 counts if asymptomatic. The study estimated that one year after a patient entered the study, patient health was maximized by starting ART within 3 months of first CD4 measurement below 554 for those patients for whom outcome was measured based only on CD4 counts. For whom quality of life was the patient outcome, patient health was maximized by initiating ART at CD4 counts of 354. The study also showed that those with pre-ART CD4 counts above 500 generally preferred initiating ART at slightly lower CD4 counts.

A review was done on 15 studies published from 2011 to 2013 in South Africa, Mozambique, Cameroon, Uganda, Tanzania, and Zimbabwe on point-of-care CD4 testing.  Compared with lab-based testing, point-of-care CD4 testing increased the probably of a patient having their CD4 count measured and of receiving the results of that test.  Time to being tested, as well as time between testing and receipt of result, was significantly reduced.  When considering all the data, ART initiation rates were higher when point-of-care CD4 testing was utilized; however, when limited to studies reporting ART initiation by eligible individuals, a statistically significant relationship was not demonstrated.  The review included studies on adults, adolescents, children, and pregnant women.  There was one randomized controlled trial, one non-randomized trial, 11 observational cohort studies, one retrospective cohort, and one cross-sectional study.  The review focused on the proportion of patients in care at each step of the care pathway, and the time it took to reach different levels of the pathway.  The steps evaluated included: (1) HIV diagnosis to CD4 testing; (2) CD4 testing to delivery of CD4 results; and (3) eligibility assessment to ART initiation for eligible individuals.  The CD4 count cutoff for ART initiation ranged from 200 to 350 cells/µl (except notably in a study including children under 3, where the initiation cutoff was 750 cells/µl).  Among the studies reporting on time between HIV testing and eligibility assessment, the probability of CD4 testing occurring was four times higher when point-of-care CD4 testing occurred, as compared to lab-based testing.  The probability of patients receiving their CD4 result also increased by nearly a factor of 3. Six studies reported on the proportion of people who initiated treatment after CD4 testing. In the randomized trial, patients receiving point-of-care testing were 3 times more likely to initiate ART than those who received lab-based CD4 testing.  A cohort study found that patients were 6 times more likely to initiate ART after point-of-care CD4 testing was introduced, as compared to before introduction when lab-based testing was standard of care.  In all studies reporting on initiation of care after CD4 testing, there was either improved rates of initiation, or no statistically significant difference between point-of-care and lab-based CD4 testing.  When considering all of the data, ART initiation was more likely among the patients tested with point-of-care CD4 testing; however, when limiting the data to only those studies that reported eligibility of all individuals, there was no statistically significant relationship between initiation of ART and point-of-care CD4 testing.  The reduction in time to ART initiation with point-of-care CD4 testing was statistically significant in one of the two studies reporting data on it.  This review also notes that simplicity of point-of-care CD4 testing, which was successfully used in five studies in rural locations and in at least one study by non-physician clinicians.

A study in South Africa which randomized 344 patients, 64.5% female, to three arms found that receipt of CD4 result at the time of HIV diagnosis were 2.6 times more likely to report for ART initiation and 2.1 times more likely to enroll for ART. Other patients received standard of care, i.e. receipt of CD4 results one week after HIV counseling and testing; or CD4 results one week after HIV testing and counseling plus an informational leaflet. ART initiation was measured by arrival at pre-ART care site within one month of HIV testing or for ART initiation within three months if CD4 counts were over 215. Of the patients that received their CD4 counts immediately following HIV testing and counseling, 47.6% reported for further care (pre-ART and ART), where as to those who received their CD4 counts one week later, 33.6% reported for further care. However, for patients not yet eligible for ART, knowledge of a CD4 count immediately after diagnosis did not increase the numbers who enrolled in pre-ART care.

Data analysis of 19,525 Thai clients who had HIV testing at a VCT center in Thailand from 2006 to 2009 found that of the 13.2% clients who tested HIV-positive for the first time, CD4 measurements were performed in 73.3% of clients living with HIV at the same center where they had their HIV test and 91.4% occurred within the first month of HIV diagnosis. CD4 count measurement was available to clients at the same VCT center where they had their HIV test. Of the Thai clients, 31.8% were women, and 13.2% of these women tested HIV-positive. “The much shorter duration between CD4 count measurement and HIV diagnosis shown in our study might be the results of our routine posttest counseling which always includes the recommendation of immediate CD4 count measurement in the same clinic” (Phanuphak et al., 2011: 250). Treatment prior to CD4 counts of below 300 increases the risk for mortality.

In Mozambique, following the introduction of point-of-care CD4 testing, the proportion of patients lost to follow up before completion of CD4 staging dropped from 57% (278 out of 492 patients) to 21% (92 out of 437 patients). The median time from enrollment to antiretroviral therapy initiation was reduced from 48 days to 20 days. The study was conducted at four primary health care clinics. Total losses between enrollment and antiretroviral therapy initiation dropped from 64% to 33%. After implementation of point of care CD4 testing, the proportion of enrolled patients who initiated ART increased from 12% to 22%. The median time between enrollment and completion of CD4 staging dropped from 32 days to 3 days.

A meta-analysis including studies published from 2007 to 2010 with data from Cambodia, Ethiopia, Malawi, South Africa, Uganda, and Zimbabwe examined the effect of daily prophylactic co-trimoxazole on mortality in HIV-positive populations.  The analysis found that co-trimoxazole significantly increases the survival rate of HIV-positive individuals on ART.  Individual studies also demonstrated a reduction in the cases of malaria and diarrhea, a reduction in the cases of new or recurring WHO stage III or IV disease events, and a reduction in all-cause morbidity.  The hazard ratio associated with prophylactic co-trimoxazole was 0.42, as compared to a group on ART who did not take co-trimoxazole.  Co-trimoxazole contains two antibiotics and provides protection against bacterial and fungal infection. The WHO recommends (citing this article, among others) that all people living with HIV who have a CD4 count of below 350 cells/µl or have been classified as clinical stage III or IV be provided with prophylactic co-trimoxazole.  In settings with high HIV prevalence, high infant mortality, and limited health infrastructure WHO recommends that all HIV-positive individuals receive co-trimoxazole prophylactic treatment (WHO et al., 2013). The analysis included 8 cohort studies and 1 randomized trial. The effect co-trimoxazole had on mortality was seen if initiated before ART, with ART, or after the patient was stable on ART.  There was no evidence of change in the benefit of co-trimoxazole after CD4 count gains following initiation of ART. Providing cotrimoxazole can also, "help increase the retention rate, provide an opportunity to assess an individual’s adherence to treatment before the start of ART, and improve survival in those not on ART." 

A study in Uganda and Zimbabwe using patients from the DART Trial found that co-trimoxazole prophylaxis improved survival in patients in the first 72 weeks on ART. A total of 3,179 adult patients (65% women) were observed between 2003 and 2008. All patients had CD4 counts under 200 cells. Ten percent did not take co-trimoxazole during the follow-up, 62% were taking co-trimoxazole at ART initiation and the remaining 28% started while on ART. Co-trimoxazole use in patients who started during or before ART initiation halved mortality in the first 12 weeks on ART. The overall mortality rate was reduced by 35%. Benefit did not vary with increasing time on co-trimoxazole. On the other hand, when time on ART initiation was considered, mortality reduction was the greatest in the first 12 weeks of treatment. The reduction in mortality was sustained from 12 to 72 weeks, but not evident after 72 weeks of ART initiation. Average 5 years survival increased by 5% and 2% in patients who started ART at CD4 count 15 or 150 cells, respectively. Use of co-trimoxazole started before or at initiation of ART reduced risk of deaths potentially preventable by co-trimoxazole (malaria, toxoplasmosis, cryptococcal meningitis, diarrhea, septicemia etc.) by 21%. After 72 weeks no difference in co-trimoxazole effect was observed. The study also found that use of co-trimoxazole was associated with 26% reduction in risk of new malaria episode.

Services in rural Uganda which provided participants living with regular three month follow up, including screening for ART eligibility plus provision of co-trimoxazloe prophylaxis found that between 2007 and 2008, of 322 patients who learned their HIV-positive serostatus any time between 2004 and 2008, 79.2% had been screened for ART. Of those eligible for ART, 99% had initiated ART. Prior to ART initiation, patients received three counseling sessions, a medical exam and identified a treatment supporter. On starting ART, patients were seen at 2 weeks, four weeks and then at 3-month intervals or more often if needed.

A study done in Kenya showed that provision of free co-trimoxazole in antiretroviral therapy-ineligible clients improved survival. A total of 5,175 clients had baseline CD4 counts among the 5,854 clients enrolled in to a clinic in Nairobi between 2005 and 2007. Fifty three percent (1,024) of the clients who had WHO stage 1 or 2 disease and CD4 counts above 250 were ineligible for ART. Sixty percent of clients accessed care before co-trimoxazole was routinely offered for free and the remaining 40% before free co-trimoxazole made available. Women represented more than 70% of the clients. The one year retention in care was 84% for ART-ineligible clients enrolled when free co-trimoxazole was available compared to 63% retention for clients who were enrolled in the period when co-trimoxazole was not routinely offered for free. Clients who were enrolled to care in the period when co-trimoxazole was not free were more than 2.5 times likely to be lost to follow-up and die. Results were similar for clients who had CD4 counts between 251 and 350 cells and above 350 cells.

An observational cohort study conducted in South Africa showed that use of co-trimoxazole preventive therapy at initiation of antiretroviral therapy increased survival among HIV/AIDS patients. The use of co-trimoxazole prevented Pneumocystis jirovecii infection and toxoplasmosis. In resource limited countries, use of co-trimoxazole alleviated the burden of life threatening infections such as malaria, bacterial pneumonia and diarrhea among HIV positive patients. In South Africa even though malaria and Pneumocystis jirovecii pneumonia are not as prevalent and co-trimoxazole resistance is high, co-trimoxazole therapy reduced mortality at any CD4 count on initiation of antiretroviral therapy. A total of 14,097 patients (38% women) were recruited from 231clinics who started antiretroviral therapy between January 2003 and 2008. Co-trimoxazole preventive therapy was initiated for 53% of the patients and continued for a mean period of 9 months according to international and WHO recommendations. Reduced mortality was associated with higher CD4 count at initiation of antiretroviral therapy. Co-trimoxazole use was strongly associated with lower mortality, with a 36% decrease in mortality. On the other hand, co-trimoxazole prophylaxis was not associated with reduction in mortality among patients who had both higher CD4 count and stage 1 and 2 WHO clinical diseases at antiretroviral therapy initiation.

A randomized controlled trial was done in Uganda following patients on ART between 2005 and 2009 in groups that were treated with either a home-based care regimen or a routine clinic-based care system.  Mortality rates were similar in both treatment groups, even among patients starting ART at CD4 counts of less than 50 cells/µl.  In both clinic- and community-based care, those patients who initiated care at a CD4 count of less than 50 cells/µl had significantly higher mortality rates, and the rates for all participants were significantly higher in the first 6 months than they were in the subsequent follow-up period.  Overall, 1,453 adult patients initiating ART between 2005 and 2006 were followed until 2009. Clusters were put together by geographic area and then randomized into either the home-based care group or the clinic-based care group.  Patients were started on ART following Ugandan guidelines of initiation at CD4 counts at or below 200 cells/µl or late stage III or stage IV disease.  About 71% of the patients were women, the median age was 37 years old, and 31% of patients had CD4 counts of less than 50 cells/µl at baseline.  Routine clinic-based care involved patients picking up their medication and receiving adherence support monthly at the clinic.  They also had clinical visits at 2 and 3 months, and then every 3 months thereafter.  Home-based care involved patients receiving lay healthcare workers at their homes monthly.  These healthcare workers would travel by motorbike to each home, would assess the patients clinically, deliver the ARVs, and provide adherence counselling. They had access to a phone so that they could contact the clinic if they had any questions.  The healthcare workers could also refer patients to the clinic if necessary, and patients would attend the clinic for clinical visits at 2 and 6 months, and then every 6 months thereafter.  Even when the patients attended clinic, they had a home visit that month to receive their ARVs.  There was an overall mortality rate of 6.36 deaths per 100 person-years of follow-up. There was no statistically significant difference between the cohorts at either CD4 count level, but there was a significant difference between the mortality of those with a baseline CD4 count of less than 50 cells/µl and a baseline CD4 count of more than 50 cells/µl.  In the first 6 months of treatment, the mortality rate for all participants was 16.5 deaths per 100 person-years, and in the remaining follow-up period the mortality rate was 3.52 deaths per 100 person-years. There was a significantly higher mortality rate in the first 6 months of ART regardless of treatment group.  Among the patients who had a CD4 count of less than 50 cells/µl at baseline, there was a higher proportion of deaths attributable to tuberculosis in the home-based care cohort and a higher proportion of deaths attributable to poor nutritional intake in the clinic-based care cohort.

A randomized trial done in Uganda found similar survival rates, plasma viral suppression and reduced cost when patients on ART were followed at home versus at a health facility. A total of 859 patients (73% women) were randomized to home-based care and 594 patients (68% women) to facility-based care and followed between 2006 and 2009. Counseling and information were provided for each patient at the initiation of ART and drugs for one month were given with a pill box. Trained officers travelled on motorcycles to deliver drugs, monitor with a checklist of signs and symptoms of disease progression and drug toxicity, and provide adherence support to patients randomized to home care. They carried mobile phones to facilitate referral when needed. All patients were invited to the clinic for routine reviews at 2 and 6 months after initiation of ART and every 6 months thereafter. In the health facility care group, patients were scheduled at 2 and 3 months after initiation of ART and every 3 months thereafter.  Rate of mortality or virological failure (viral RNA above 500 copies) was 26% both in home (11.29 per 100 person-years) and facility (11.45 per 100 person-years) cared patients.  Adherence to therapy in the past 28 days was 94% for home and 91% for facility cared patients. By the end of the study, 566 (66%) in the home and 377 (63%) patients in the facility group were alive and on follow-up with full viral suppression. Sixteen percent of patients in home care and 21% in facility care had a drug substitution at 8 months after initiation of ART. The total cost per patient per year was 793 US$ for home care and 838 US$ for facility care.

A prospective observational cohort study was conducted in Rwanda from 2007 to 2009 to compare the standard of care to a community-based program in addition to the standard of care. The study found that 85% of the community cohort and 79% of the standard of care cohort was retained in care with viral load suppression at 1 year. The community cohort had lower rates of death and loss to follow-up, and a higher probability of suppressed viral load at 1 year after multivariate analysis. A group of 610 patients were initiated according to national program guidelines at CD4 counts of less than 350 cells/µl and followed for at least a year. Standard of care included ART and cotrimoxazole provided monthly free of charge at the clinic. CD4 counts were taken every 6 months after initiation of care. Counseling was also provided. The community-based program included community health workers visiting patients in their homes every day. These health workers provided social support, monitored the health of their patients, and directly observed therapy once a day. The health worker also accompanied the patients to health visits for the first 4 monthly visits. A food ration was provided for the first 10 months of ART, and a transportation stipend was available for clinic visits. There was additional support provided to those in need, which included payment of school or insurance fees, microfinance loans, employment aid, and repairs in the household. There were 306 patients in the standard of care cohort and 304 patients in the community care cohort. Sixty-six percent of the patients in the standard of care were women, while 58% in the community cohort were women. The mortality rate in the standard of care cohort was 7.2% and 4.3% in the community cohort. A total of 3.3% of the standard of care cohort and 1% of the community cohort were lost to follow-up. Patients in community-based care had a 15% greater probably of being retained in care with suppression of viral load at 1 year. Both programs had high rates of retention to care, at 92% in the community cohort and 87% in the standard of care cohort.

A global review of 90 studies in lower and middle income countries found that community-based services were most cost-effective compared to facility based programs. The review found that an integrated continuum of care, with a formally established affiliation between facility based programs and community-based programs resulted in better HIV treatment patient outcomes with lower rates of loss to follow up. Facilities can provide ART, with community sites providing condoms, adherence counseling, psychosocial support and other care.

A study was conducted from 2011 to 2012 in South Africa to evaluate a home-based counseling and testing program that included point-of-care CD4 testing and facilitated referrals to HIV care.  The home-based counseling and testing program achieved high rates of testing, linkage to care, and uptake of ART.  Eighty-six percent of participants who were eligible initiated ART by 3 months.  For participants with a CD4 count of less than 200 cells/µl, the proportion with viral suppression increased from 20% at baseline to 80% at 6 months. A questionnaire about demographics, sexual behavior, and history of HIV testing was conducted at the initial visit on 671 adults, along with a HIV testing with pre- and post-test counseling. Thirty percent of the participants were HIV-positive; 36% of the HIV-positive individuals were newly identified. An additional questionnaire was administered to the HIV-positive participants, which included questions on sexual behavior, HIV testing and knowledge, HIV clinic visits, and ART initiation. These patients also received point-of-care CD4 testing to see if they were eligible for ART under South Africa’s initiation guidelines (CD4 below 200 cells/µl until August 2011, when it was changed to CD4 below 350 cells/µl).  The counselors encouraged the participants who were eligible for treatment to visit their local clinic.  Follow-up visits for the HIV-positive participants were conducted at 1, 3, and 6 months to assess ART initiation and to provide counseling on HIV care and adherence.  The median age of HIV-positive participants was 34 years, 82% were female, and 30% were employed.  Of the HIV-positive women, 52% did not know their partner’s status, and 22% reported having an HIV-positive or the probability of HIV-positive partner.  Ninety-two percent of those who knew they were HIV-positive reported that they had a CD4 count previously done, but only 52% reported receiving their results.  Only 57% of HIV-positive participants had visited an HIV clinic at baseline, while 96% had visited a clinic at 6 months.  Reported condom use at last sexual encounter for HIV-infected participants increased from 44% at baseline to 68% at 6 months.  Home-based counseling and testing with point-of-care CD4 testing and facilitated referrals improved rates of HIV testing, ART initiation, and viral suppression; however, there was one reported case of social harm where a participant lost their income and residence.  In this case, they returned to live at home with support within 3 months.