A $2.5 Million Misdiagnosis

A woman misdiagnosed with AIDS and made ill by the medicines prescribed to her for 9 years wins $2.5 million in court. Some details from AP:

[Audrey] Serrano's attorney, David Angueira, said Dr. Kwan Lai, who treated his client at the University of Massachusetts Medical Center in Worcester's HIV clinic, repeatedly failed to order definitive tests even after monitoring of Serrano's treatment did not show the presence of HIV in her blood.

"It is one of the clearest cases of misdiagnosis that I have ever seen and it's based in part on a presumption that people who engage in certain types of conduct are more likely to have HIV and AIDS than other people without really listening to the patient," Angueira said after the verdict.

Lai testified last week that Serrano told her she had worked as a prostitute, her partner had AIDS, and that she had suffered three bouts of a type of pneumonia typically associated with those infected by the virus.

Serrano has denied she had ever been a prostitute. She confirmed that her former boyfriend tested positive for HIV/AIDS, but disputed the claim that she told the doctor that she had suffered bouts of Pneumocystis pneumonia.

Michael Moynihan blogged last month on millions of other people (mostly theoretical, still…) who were said to have AIDS but don't really.

Comments

[Michael Pack]

The problem I've had with the AIDS panic is the fact they have expanded the list of symptoms to include just about everything.Add the constant drumming that we are all at risk[we're not]and over stating the number of people who will die and it looks more like a political agenda.I also saw Magic Johnson on TNN before a NBA game and he looked quite well.Wasn't he held up as an example of how wide spread and deadly this illness is?

[Shannon Love]

I've pretty much come to the conclusion that education in Bayes' theorem should be compulsory. Everyone from the ACLU going hysteric over airport screenings to employers testing for drugs to doctors running medical test could benefit from a sound understanding of the concept.

To help promote knowledge of this important idea, here is a link to a little applet that teaches how the concept works.

[de stijl]

The problem I've had with the AIDS panic is the fact they have expanded the list of symptoms to include just about everything.

Huh?

[de stijl]

Dr. "Cookie" Kwan Lai sez:

"Stay away from the west side!"

[Shannon Love]

Michael Pack,

The problem I've had with the AIDS panic is the fact they have expanded the list of symptoms to include just about everything.

AIDS stands for Acquired Immune Deficiency Syndrome. The actual virus process itself has not overt symptom cave a drop in the white blood cell count. The overt symptoms that occur result from secondary infections that take advantage of the depressed white blood cell count. So, a person with AIDS can exhibit any disease or disease process that results from any infection or lack of immune function.

Add the constant drumming that we are all at risk[we're not]and over stating the number of people who will die and it looks more like a political agenda

That is definitely true. It was known within a couple of years of identifying the disease that it did not spread easily through heterosexual contact in the developed world but many kept harping on the idea apparently in the bigoted belief that most people would not bother to support treating a disease that did not threaten them personally.

.I also saw Magic Johnson on TNN before a NBA game and he looked quite well.

The HIV virus appears to undergo a period of evolution within each individual victim. It rapidly mutates trying to find a means of propagating throughout the white blood cell population. This evolution can take years. Not until it finds a successful strategy can it wipe out the white blood cells and trigger actual AIDS.

The normal latency period for AIDS (the time from infection to overt symptoms) averages 2-3 years. In some people, however, it can run much longer. A small population of people show concrete evidence of the virus but remain symptom free for many years.

[mike]

It's too bad she didn't spill hot coffee on herself or something. She could've gotten a lot more.

[Francine]

If you've ever watched the show "Mystery Diagnosis" or if you've ever had any sort of non-obvious illness, you'd know that the vast majority of doctors can't be bothered to make any sort of diagnosis. If their first idea doesn't pan out they shrug their shoulders and push you out the door...

You pretty much need to diagnosis yourself before going to see an MD.

[J sub D]

Count me as one of those who thinks that malpractice suits are often unwarranted.

That said,

[Audrey] Serrano's attorney, David Angueira, said Dr. Kwan Lai, who treated his client at the University of Massachusetts Medical Center in Worcester's HIV clinic, repeatedly failed to order definitive tests even after monitoring of Serrano's treatment did not show the presence of HIV in her blood.

this does appear to be a justifiable verdict. I know enough to take a plaintiff's attorney statements with a healthy dose of salt, I wasn't on the jury, but still ...

[Aresen]

When I saw this story, I wondered if the woman was a victim of a "false positive" test result.

Any test will produce a small proportion of "false positives". Even if the rate is as low as 1%, that means thousands of people will have false positive test results for AIDS. Repeated tests would clear up most of the false positives. However, even if three tests were done, a 1% error rate would still mean that 1 person in a million would receive a false positive diagnosis that was confirmed.

[Aresen]

J sub D

Thanks, that answers my question at 11:16.

[Joe S.]

What, I'm the first person on this thread to mention the name of Peter Duesberg? The AIDS Lobby must be doing its job!

[Episiarch]

Greg House would have only had the wrong diagnosis for about 45-50 minutes.

[Tacos mmm...]

this does appear to be a justifiable verdict. I know enough to take a plaintiff's attorney statements with a healthy dose of salt, I wasn't on the jury, but still ...

Drug therapy can produce an undetectable viral load. This kind of result would also be seen in an infected person whose medication regime was efficaceous.

[R C Dean]

I might also note that the rapacious plaintiff's bar for medical malpractice has produced a pathologically risk-averse medical profession.

Dr. Lai probably figured that stopping AIDS treatment when there is a positive test result in the file would be economic and professional suicide.

[Marcvs]

Everyone needs to take a much more active role in their healthcare than they do right now. I'm not talking about insurance (that's a different issue), but the actual treatment. Speaking as one with a chronic disease, it is mind-boggling how many people have conditions and don't even know what drugs they are taking, what the side effects are, etc. Notice the line in this story: "Serrano filed the lawsuit in 2003 after she became suspicious of her diagnosis and had herself tested at another hospital." This was FIVE YEARS after she began treatment. When you are talking about that serious of a disease, GET A SECOND OPINION. Don't let doctors dictate to you, make them treat you like a customer. They don't have to be nice, but they need to answer your questions and not brush you off. YOU also have to follow up and understand your condition and treatment options. With the Internet, ignorance is not an excuse.

</soapbox>

[thoreau]

Dr. Lai probably figured that stopping AIDS treatment when there is a positive test result in the file would be economic and professional suicide.

True, it might be professional suicide to simply stop treatment if one subsequent test was at odds with the first. But there are options intermediate between "Stop all treatment" and "Do exactly what you've been doing." Ordering a variety of additional tests and trying to get to the bottom of it would make sense.

[Neu Mejican]

Shannon Love,

You seem to know the topic:

The actual virus process itself has not overt symptom save a drop in the white blood cell count.

I was under the impression that there were specific cognitive deficits directly associated with HIV+ status independent of secondary infections.

Is that not correct?

True about Bayes.

[R C Dean]

The "no-load" test result is not all that uncommon in AIDS patients who have a good cocktail working.

Not sure what additional tests would be considered the standard of care in cases like this. It might well be no tests at all, given that a no-load test generally means the treatment is working.

[thoreau]

Fair enough, R C. I admit that I don't know what the additional tests would be either. I just figure that in general there's always a third. Sort of like Kirk in that simulator.

But this could be an exception.

[the innominate one]

an initial positive ELISA should always be followed by a DNA test for the HIV virus.

The "AIDS test" one is usually given is the ELISA test, which tests only for the presence of antibodies produced by the patient's immune system in response to the presence of HIV.

Of course, false negatives could also occur for no known reason, or as a result of an immune-depressed patient not producing antibodies (e.g, if the patient was on immune-suppressing drugs following a transplant).

a DNA test for the viral DNA itself is an important tool for confirmation of an initial positive diagnosis, or even confirmation of a negative diagnosis if there is good reason to suspect the patient may have been exposed to HIV.

[Dave W.]

Do AIDS drugs really help?

[David Crowe]

Repeated testing is no solution. It's basic probability theory that if one event has a probability of 1/6 (e.g. rolling a 6) that the probability of two independent events is obtained by multiplying the odds together (i.e. the odds of rolling two consecutive sixes is 1/36 = 1/6 x 1/6).

However, antibody tests are NOT independent. For example, if someone's blood cross-reacts with a component of the test kit, the fact that the possibility of a false positive in the general population is, say, 1%, is meaningless. For that one person the possibility might be very high and repeated testing will not significantly reduce the possibility.

Another problem with HIV testing is the Positive Predictive Value predicament. Let's say that 1/1000 people in the general population are HIV infected. And let's say that the HIV test is 99% accurate. That means that if you test 1000 people you'll get about 1 true positive and 10 false positives. This is the reason why testing a low risk population is a BAD idea!

Getting away from mathematics, it's interesting to note that some of Serrano's symptoms were AIDS defining. That is, this is proof that AIDS drugs can cause AIDS.