One-third of HIV patients were not aware of their risk

Should physicians screen all patients for HIV?

HIV testing remains controversial, despite clear evidence that routine testing and early diagnosis can help HIV-infected patients receive the medical care they need earlier in their disease progression, which could have a positive impact on their long-term prognoses.

New research shows that more than one-third of HIV-infected patients at two urban hospitals were not aware of their HIV risk before they were tested for the virus, especially if their source of infection was through heterosexual sex rather than homosexual sex or injection drug use.

"Acknowledging that awareness of your risk is the first step to getting tested," says Jeffrey H. Samet, MD, MA, MPH, associate professor of medicine and public health at Boston University School of Medicine and Public Health. "We’ve done a reasonable job in the last 20 years of making it clear that people who inject drugs and have unprotected sex with men are at higher risk of infection," Samet says. "But what’s happening is that more than those risk groups are at risk. Anyone having sex without condoms with a partner who is not known to be HIV-negative is also at risk."

Samet argues that because the down side of HIV testing is so negligible and the up side of knowing one’s HIV status is so significant, the medical community should have a very low threshold for HIV testing. Already, nearly one-third of adults in the United States have been tested for HIV. Some of those adults have voluntarily sought their HIV status, but for many others the testing is done routinely as part of blood donor screening, life insurance screening, and military service.1 "Those generally are people who are not in risk groups," Samet adds.

In the Boston study, investigators found that 80% of HIV-infected patients initially presented to medical care with CD4 cell counts of less than 500/µml, and 37% had counts of 200/µml or less.2 The study population included 203 outpatients at the Boston Medical Center (previously called the Boston City Hospital) and Rhode Island Hospital in Providence between February 1994 and April 1996. Although this time period encompassed the pre-protease inhibitor era, the findings would likely be the same today, Samet notes. 

"This was all post-Magic Johnson’s HIV disclosure," Samet says. When pro basketball star Magic Johnson announced that he was infected with the virus, public health officials hoped that the public would finally realize that many more people were at risk than they had believed, he adds.

The study shows that people continue to remain ignorant about their risk-taking behaviors. And although the investigators have not formally studied more recent data, it would appear that the problem still exists, Samet says. "We’ve made a little progress, but without a doubt there are many, many people coming in with opportunistic infections and still coming in quite late," he says.

The study found that HIV-infected patients who first presented with lower CD4 cell counts were more likely to have these characteristics:

  • had no or only one close friend;
  • had not been in jail in the past 10 years;
  • had been voluntarily tested;
  • had lower hope and a poor quality of life;
  • had more symptoms of HIV infection;
  • were older.2

Another interesting finding was that among the patients who knew they were at risk for HIV, they still would wait months to years before being tested. The median time lapse between when a person first felt at risk and when he or she was tested was one year. The mean time lapse was 2.5 years. Based on an analysis of the subjects’ CD4 cell counts, investigators theorized that many HIV-infected patients have had the disease for 6.0 to 11.6 years before being tested. This is their approximation of a mid-range period of delay.

This brings the issue back to medical treatment and policy, Samet suggests. "It was the heterosexual group that was least aware of HIV risk at time of testing, and that finding to me is totally compatible with what we’re seeing clinically," Samet says. The solution is for clinicians to lower their threshold for recommending HIV testing, he adds.

Rather than providing testing as a diagnostic tool in the cases of patients who have symptoms that could signify HIV disease, primary care physicians could use HIV testing as a screening tool. As such, it would be similar to Papanicolaou smears for cervical cancer or mammography of older women for breast cancer. A Pap smear is given routinely in the case of a disease that has a prevalence rate of 0.1%. By contrast, the national prevalence rate of HIV infection is 0.3%.1

Samet admits that routine universal testing, while ideal from an epidemiological perspective, will not be feasible. However, physicians could lower their threshold for when to suggest testing. Here are some possible scenarios in which routine HIV testing could be applied:

• A hospital has one or more patients with newly diagnosed AIDS per 1,000 patient population. Hospitals with an incidence rate this high could routinely test all inpatients for the disease.

• Patients presenting with varicella zoster virus, community-acquired pneumonia, tuberculosis, or hepatitis C, or who have a history of any sexually transmitted disease or recurrent vaginal candidiasis should be tested.

• Patients who have experienced sudden weight loss, unexplained lymphadenopathy, or dermatological diseases should be tested.

• Physicians should offer testing to patients who have reported on their physical report a history of alcohol dependence, cocaine abuse, homelessness, or psychiatric hospitalization.

• Clinicians could routinely approach the subject by asking new patients to consider having an HIV test if they have had any unprotected sexual contact with a person who is either HIV-positive or who has an unknown HIV status.

Clinics and clinicians who promote testing under these circumstances will undoubtedly find additional cases of HIV that otherwise would have fallen through the cracks. At least that was the experience Samet had when his hospital made a major push for HIV testing of all untested patients. Unfortunately, the time and energy needed to maintain such an effort proved too difficult to maintain, Samet says. "It’s hard to implement because people are in and out of the hospital or clinic," he explains. "But it was clearly a useful policy in our setting."


1. Freedberg KA, Samet JH. Think HIV: Why physicians should lower their threshold for HIV testing. Arch Intern Med 1999; 159:1994-2000.

2. Samet JH, Freedberg KA, Savetsky JB, et al. Understanding delay to medical care for HIV infection: The long-term non-presenter. AIDS 2001; 15:77-85.