`Test and treat` unlikely to halt Washington epidemic

This article is more than 15 years old. Click here for more recent articles on this topic

A mathematical model from the Harvard School of Public Health, based on the population of Washington, DC has found that recalling every adult for annual HIV screening and treating every HIV-positive person with antiretrovirals as soon as they were diagnosed would result in a 30% decline in the proportion of the HIV positive population who had non-suppressed viral loads (defined as over 500 copies/ml in this study), in the event of realistic takeup of tests and treatment.

In the unlikely event of 100% of adults being offered and accepting tests and 100% of those diagnosed starting treatment, there would be a 60% reduction in the proportion of people with viral loads over 500.

Universal testing and treatment with realistic uptake would result in a decline in the proportion of time spent with a detectable viral load from 65% to 44%, produce no increase or a slight decline in the absolute numbers of infectious people for 18 months after the adoption of the strategy, and would increase post-diagnosis life expectancy from 25 to 27 years.


detectable viral load

When viral load is detectable, this indicates that HIV is replicating in the body. If the person is taking HIV treatment but their viral load is detectable, the treatment is not working properly. There may still be a risk of HIV transmission to sexual partners.

linkage to care

Refers to an individual’s entry into specialist HIV care after being diagnosed with HIV. 

mathematical models

A range of complex mathematical techniques which aim to simulate a sequence of likely future events, in order to estimate the impact of a health intervention or the spread of an infection.


A healthcare professional’s recommendation that a person sees another medical specialist or service.

test and treat

A public health strategy in which widespread HIV testing is facilitated and immediate treatment for those diagnosed with HIV is encouraged.

However even annual screening with 100% recall, and treatment of every diagnosed person, would not result in a long-term decline in the infectious population. After ten years the number of people with viral loads over 500 in the population would have increased from 450 to 1100, which is the number with viral loads over 500 currently in the population under the present strategy of no universal screening and treatment for people with a CD4 count under 350.

Washington, DC has the highest HIV prevalence of any city in the USA, with 3% of the population infected. The researchers calculated what would happen to CD4 counts at diagnosis, post-diagnosis life expectancy, and the proportion of time spent with viral load over 500 in six different scenarios:

  • Same diagnosis rates as at present but no treatment
  • Current practice
  • Universal screening for every adult performed once,
  • Universal screening performed every three years,
  • Universal screening performed annually
  • Universal screening for every adult performed once and treatment for all diagnosed
  • Universal screening performed every three years and treatment for all
  • Universal screening performed annually and treatment for all.

They presumed an undiagnosed HIV prevalence in the population of 0.6%, and an annual incidence of 0.139% or one new case in every 719 adults a year.

In their initial analysis they assumed that 80% of adults would be offered a test, that 60% would accept the test offer, and that 80% who diagnosed positive would be linked to care, meaning that 38% of previously undiagnosed people in the population would be linked to care. All of these assumptions are derived from previous surveys of testing in US urban populations. They also calculated what would happen in the unlikely scenario of 100% being offered a test, 100% accepting one, and 100% linkage to care.

They found that screening every three years would increase the CD4 count at diagnosis from a current 351/mm3 to 399/mm3 and annual screening would increase it to 441/mm3. Current life expectancy after diagnosis is 25.1 years (an increase from 16.6 years if treatment was not available); this would increase to 25.8 if every diagnosed person was offered ARVs, 26.8 if screening was performed annually, and 27.2 if both were offered. The proportion of people in the population with viral loads over 500 would decrease from 65% currently to 54-55% if either annual screening or immediate treatment were offered to 44.4% if both were offered.

The number of people with viral load in the population would be held at about 900 for the next two years with annual screening and treatment below 350 CD4s, compared with about 1100 at present. After two years this would start increasing and reach current levels within six years of introduction.

Universal testing and immediate treatment would reduce the number with viral load over 500 to about 850 and this figure would stay static or even decline slightly for the next four years. However it would then start to increase and reach 1100 within ten years (by which time the number under current practice would have increased to 1650).

In the unlikely scenario of 100% detection and referral to care, the number with detectable viral load would be cut to 450 at programme initiation, but would then increase to 1100 within ten years too.

The researchers comment: “The base case is intended to project a very high level of program participation. Even with such optimism, the impact of the test and treat program is limited.

“Even frequent screening with complete participation and ART initiation immediately upon diagnosis is unlikely to eliminate the Washington DC epidemic.”