A Sample of One

AIDS researchers in New York apparently cried wolf. The Los Angeles Times reported:

The announcement from New York health officials in early February was chilling: A single patient had progressed from HIV infection to AIDS in months rather than years, and his strain of the HIV virus seemed impervious to normally effective medicines.

The patient, a gay man in his 40s, had unprotected anal intercourse with scores of partners. Headlines of a potential new killer spread around the world.

"This case is a wake-up call," Dr. Thomas R. Frieden, New York City's health commissioner, said at a news conference where he issued a warning for physicians to prepare for a possible new phase in the epidemic.

Yet several AIDS experts immediately questioned the importance of the case and the strategy of publicizing it so widely.

Months later, those doubts seem to have been confirmed.

No super-strain has emerged. The patient, whose name has been withheld, has responded to drug therapy. No one — not even the man's known sexual partners — was found to be infected with the same HIV strain.

The problem was there was only one patient that sparked the panic. The LA Times continues:

Some AIDS specialists now say the New York announcement was scientifically naive and needlessly alarmist — risking the effectiveness of future prevention efforts.

"Does it do good to [mislead] people and exaggerate?" asked Dr. Robert Gallo, co-discoverer of the virus that causes AIDS. He condemned Frieden's far-reaching conclusions as "scientifically, completely invalid, without a shred of evidence."

Dr. Gallo has an excellent point. Using a sample of one is not scientifically valid. Given the danger of a super-strain of AIDS I understand why the researchers dropped the ball. Nevertheless, this story is a cautionary tale about having too small a sample set.

Earlier on the Blinne Blog I commented on the politics of the movie, Privileged Planet, but I refrained from commenting on the content of the speculation. (I call it a speculation rather than a hypothesis because I don't know how this is falsifiable.) Like the AIDS researchers, much of the idea of a rare earth depends on a sample size of one. How do we know that life needs Carbon or water? Given extremophiles (life that lives in extreme environments, e.g. ingests arsenic, lives at extreme temperatures or without oxygen) how do we know that Earth is in a Goldilocks-style sweet spot? The answer is we don't because we have a sample of one. We can give all the probabilities in the world but they mean nothing if they are plus or minus 100%.

Is life unique on Earth? We don't know and the only way we will know if we find life somewhere else and thus prove life is not unique. SETI has the same problem as the Privileged Planet. The fact that E.T. hasn't phoned home means absolutely nothing. One of the motivations for Privileged Planet was to refute Carl Sagan's notion of our planet's mediocrity. Carl Sagan was in love with SETI. It's somewhat ironic that both fell into the same pit.