orgtheory.net

On scientific skepticism…why it’s best to double (and triple) check, even when the data fit the profile

The New York Times recently carried a story that highlights the importance of scientific skepticism, even in the face of strong evidence. The author, Dr. Danielle Ofri, was seeing a patient, Mr. S, who had been diagnosed with HIV years prior while incarcerated. Apparently, not only did the medical test offer a positive diagnosis, but Mr. S’s social experiences were thought to be consistent with many of the risk factors associated with HIV. He was an inmate on Rikers Island, one of the most notorious jails in the country. For a variety of reasons, inmate populations in our nation’s jails and prisons have, on average, higher rates of HIV, hepatitis, TB and other communicable diseases than the rest of the population.  The man was also a former IV drug user, and using unsterilized needles is a leading mode of HIV transmission.  In fact, whenever Mr. S went to visit his doctor, nurses struggled to find healthy veins to draw blood to test his T-cell counts and viral loads, markers of HIV progression in the body. After years of taking HIV medications and doing annual blood work (as the routine blood tests that most HIV-positive individuals have every 3-6 months were deemed impossible given the condition of his veins), a nurse practitioner approached the situation with a healthy dose of scientific skepticism and asked for a new HIV test.

Turns out, Mr. S was never HIV positive.

Rather, the first test was a false positive.  Dr. Ofri wondered, “how the misdiagnosis could have persisted for so long — why we chose to explain his vigorous T-cells by classifying him as a nonprogressor, rather than considering that his initial H.I.V. test might have been incorrect.”

Mr. S likely wasn’t retested after the positive test result, Dr. Ofri hypothesizes, partly because his social experiences confirmed a positive HIV diagnosis. In short, he fit the profile due to his drug and incarceration history.  Years and thousands of dollars in HIV drugs later, it is discovered that he’s not HIV positive after all. The “profile” didn’t fit the man.

False positives are rare in HIV testing, but they do happen. As a result, many of the respondents in my HIV research describe being tested and retested to make sure that the results are accurate.

I guess the moral of the story is…. respect the science, but test and evaluate again. This applies of course to medicine, in which biological markers and social factors are interpreted together by health professionals to offer a diagnosis. But this also applies to social scientists, so that we can catch those unexpected instances in our data when 2+2 does not in fact equal 4.

Advertisements

Written by Celeste Watkins-Hayes

June 30, 2011 at 9:30 pm

2 Responses

Subscribe to comments with RSS.

  1. “False positives are rare in HIV testing, but they do happen. As a result, many of the respondents in my HIV research describe being tested and retested to make sure that the results are accurate.”

    At least two tests per subject are mandatory for HIV research that is meant to be published.

    Like

    Guillermo

    July 2, 2011 at 1:11 pm

  2. I missed this story in the NYTimes. Thanks for the post!

    Like

    cperchesk

    July 5, 2011 at 6:10 pm


Comments are closed.

%d bloggers like this: