A survey of 258 people screened for potential primary HIV infection (newly exposed individuals) indicates that the presence of fever and rash at the same time is the best clinical guide to recent infection with HIV.
The study was carried out by Positive Health Program at San Francisco General Hospital, which advertised in the community for people with suspected high risk exposures such as unprotected sex or needle sharing with an HIV-positive partner, and/or suspected symptoms of HIV seroconversion. Individuals were tested for HIV using a variety of methods, and the researchers assessed which symptoms occurred more frequently in people who were subsequently found to have HIV infection.
The strongest predictors were:
- Fever and rash together (8.4 times more likely to occur in people with primary HIV infection)
- Fever (5.2 times more likely)
- Rash (4.8 times more likely)
- Oral ulcers (3.1 times more likely)
- Joint pain (2.6 times more likely)
- Sore throat (2.6 times more likely)
- Loss of appetite (2.5 times more likely)
- Weight loss of greater than 5lbs (2.5kg) (2.8 times more likely)
- Muscle pain (2.1 times more likely)
- Fatigue (2.2 times more likely)
- Nausea (1.9 times more likely)
Other previously reported symptoms, such as headaches, night sweats, diarrhoea, ulcers on the genitals and vomiting were just as likely to occur in people who did not have infection.
The study was not able to identify whether the presence of a greater number of these symptoms had stronger predictive power, possibly due to the relatively small number of people who were diagnosed with primary HIV infection (40 out of 258).
The nature of the rash seen in patients with suspected HIV infection was not elaborated on, but previous reports have described what is called a maculopapular rash – raised bumps that are a different colour from the surrounding skin.
There was no difference in the length of time that key symptoms lasted for between people with HIV infection and people without. On average, symptoms lasted for ten days or less. The only exception was genital ulcers, which lasted significantly longer in patients with HIV infection (27 days vs 9.5 days).
The study also looked at which diagnostic tests most accurately identified HIV infection at such an early stage.
P24 antigen testing (which looks for core proteins form the virus) was no more accurate than a third generation ELISA antibody test designed to detect antibodies earlier than standard tests. However, both methods were less sensitive and specific for the presence of the virus than viral load tests, which detected HIV in every infected person, compared to a 79% detection rate for p24 antigen and ELISA methods.
However, viral load testing also wrongly diagnosed the presence of HIV infection in a small number of individuals. The authors noted that in each case, viral load was reported to be below 3,000 copies/ml, leading them to suggest that any viral load result below 5,000 copies/ml in a patient with suspected primary HIV infection should be treated as a false positive result and the patient should be re-tested.
The authors recommend that patients with suspected primary infection should be tested with a third generation ELISA test and a viral load test, with repeat viral load testing if the ELISA test is negative. If the viral load result remains higher than 5,000 copies/ml while the ELISA result is negative, HIV infection should be presumed but ELISA testing should be continued on a regular basis until the presence of antibodies is confirmed.
Hecht FM et al. Use of laboratory tests and clinical symptoms for identification of primary HIV infection. AIDS 16: 1119-1129, 2002.