Clinicians Often Fail to Test for HIV After Diagnosing AIDS-Defining Illnesses, U.S. Study Suggests

An Interview With Judy Chen, M.D., M.S.H.S.

There's nothing like hearing the results of studies directly from those who actually conducted the research. In this interview, you'll meet one of these impressive HIV researchers and read her explanation of a study she presented at CROI 2009.

My name is Judy Chen. I'm an internist and a health services researcher. I work at Health Benchmarks, Inc., which is located in Woodland Hills [California]. This poster is entitled "HIV Testing and Monitoring in Privately Insured Members Recently Diagnosed With Potential AIDS Defining Events."1

In 1993, the U.S. Centers for Disease Control and Prevention [CDC] published surveillance case definition for AIDS-defining events. These are 25 sets of diagnoses, including lymphoma, encephalopathy, invasive cervical cancer, yeast infection of other than the oral cavity, histoplasmosis, wasting, PCP [pneumocystis pneumonia], Kaposi's [sarcoma], etc. These are diagnoses which, the Centers for Disease Control says, if you have them, most likely you have AIDS -- not just HIV, but AIDS.

Judy Chen, M.D., M.S.H.S.
Judy Chen, M.D., M.S.H.S.

CDC recommended that clinicians should be aware of these clinical conditions, which are highly suggestive of HIV [infection], and the need for prophylaxis and therapeutic interventions. The objective of this poster is to assess the rate of HIV screening and monitoring among insured members diagnosed with potential AIDS-defining events, but without an existing HIV diagnosis.

We used 2006 administrative claim data for eight commercial health plans. Most are PPO [preferred provider organization] health plans, but we did have one Medicaid health plan. Included in our study sample are members who were insured for one year and 60 days after the potential AIDS-defining event. This is just to make sure we don't have any missing data from our population. We excluded everybody with a history of AIDS, immunodeficiency or organ transplants, [as well as everybody who was] in hospice care, on immunosuppressive medication or had prolonged steroid use.

The main thing we looked at is, in this population -- without HIV and without a history of immunosuppression -- if they did have a potential AIDS-defining event, what is the rate of having an HIV screening test, CD4 test or HIV viral load test in the 150 days before the event or 60 days after the event?

Judy Chen et al. CROI 2009; abstract 1044.
Judy Chen et al. CROI 2009; abstract 1044. Reprinted with permission.
Click here to view the full presentation.

The rates of having an HIV screening test, CD4 count or HIV viral load were very low in these patients. That's actually somewhat shocking. For example, patients that were diagnosed with [PCP], which is diagnostic of HIV and AIDS -- which nobody would actually get without HIV and AIDS -- there were about 48 of these patients in our sample. Only 10% of these individuals actually had HIV antibody screening tests, CD4 counts or viral loads sent in the 150 days before their diagnosis and 60 days after their diagnosis.

What percentage of the patients with this pneumonia were older than 50? Do you have the characteristics of that particular group?

Not right off the bat. The mean age for the whole population was around 50 years. This is a commercially insured population, so we didn't have [people on] Medicaid [or] too many people above 65; we are skewed towards a younger population anyway. But the mean age was already 50, so my guess is that these were already the older population.

It's interesting: 59.3% of the study sample were in the U.S. South.

The majority of our claims base for this particular sample was from the South. Maybe if we had more data from the West, the number could have been different, but this study reflected mostly what happened in the South and in the Midwest.

Judy Chen et al. CROI 2009; abstract 1044.
Judy Chen et al. CROI 2009; abstract 1044. Reprinted with permission.
Click here to view the full presentation.

We had a multivariate analysis show that if you are hospitalized during the year prior to your diagnosis, you're more likely to actually have HIV screening tests. [Also,] if you were in the middle age group, 18 to 50 years, you were more likely to be screened, but if you were older or younger, you were less likely to be screened.

Could you go through the potential AIDS-defining events a little bit more? I see that 25% had CMV [cytomegalovirus] pneumonia or retinitis. That's not seen in the general population, is it?

No; actually, CMV pneumonia or retinitis is almost exclusive in the AIDS. Especially progressive multifocal leukoencephalopathy, PML, that's almost exclusively seen in the AIDS population. [But] nobody was screened for HIV.

Quite shocking results. They figured out what it was; they could diagnose the AIDS-defining event, but they couldn't figure out that it was related to HIV.

Right, exactly. That's the shocking part. You'd think that if they diagnosed [the event] that they would have automatically sent a test with it to verify the origin or the associated causes with the diagnosis.

Why was the study sample so much older? Tell me a little bit more about the population characteristics and how you came to study this population.

Initially, we had intended for this not to be so much a research study, but a quality indicator: If you had certain characteristics, did you receive the quality of care? We developed it as a quality indicator that says if you have a potential AIDS-defining event, that means you have a diagnosis that could flag you as having AIDS. You didn't already have a history of HIV, but did your clinician think to send a HIV test or even a company test such as CD4 count or HIV viral load?

Judy Chen et al. CROI 2009; abstract 1044.
Judy Chen et al. CROI 2009; abstract 1044. Reprinted with permission.
Click here to view the full presentation.

As far as why the age group is older, my guess is that these patients probably have other medical problems along with [their AIDS-defining event]. Maybe the doctors haven't thought that the other medical problems are contributing to the fact that they have these diagnoses. But we did rule out the most obvious reasons that you would have other immunosuppressions as severe as AIDS. We ruled out if you had an organ transplant -- we already tossed you out. If you were on an immunosuppressive agent, such as for immunologics, or even prolonged steroid use, we already tossed you out.

So I think it is older. The comorbidity index [on my poster] is people with existing comorbidities. Among the younger population, we have zero -- that means they don't have other underlying comorbidities. This population is a little bit sicker. I think our comorbidity index, 45% have two or more comorbidities.

I see that the mean income is pretty high. $48,432 is the mean annual income.

The main income is higher in this population because most of them are commercially insured with PPO plans, so they are working.

Do you hope that there are going to be decisions or general recommendations made to change this situation?

I think because this is claims data and the health plans have access to the claims data, interventions are needed by commercial health plans to insure that members with potential AIDS-defining events are screened for HIV. Health plans can give feedback to their physicians and allow them either an opportunity to comment on why it wasn't sent or the [opportunity] to send the HIV test once [these potentially HIV-infected patients are] identified.

Were these people eventually diagnosed with HIV?

That remains to be seen. We haven't followed up the patients [over] time, but it will be interesting to assess in the future.

To say the least.


OK, thank you.

You're welcome.

This transcript has been edited lightly for clarity.


  1. Chen J, Tian H, Dahlin-Lee E, et al. HIV Testing and Monitoring in Privately Insured Members Recently Diagnosed With Suspected AIDS-Defining Events. In: Program and abstracts of the 16th Conference on Retroviruses and Opportunistic Infections; February 8-11, 2009; Montréal, Canada. Abstract 1044.
    View poster: Download PDF.

Download the audio file