Monday, October 24, 2011

Bad Science, Politics and the Need for Clinical Data

Ben Goldacre, Saturday 31 October 2009, The Guardian.

Every now and then it’s fun to dip into the world of politics and find out what our lords and masters are saying about science. First we find Brooks Newmark, Conservative MP for Braintree, introducing a bill to reduce the age for cervical cancer screening to 20. The Sun has been running a campaign to lower the screening age, on the back of Jade Goody’s death at 28 from cervical cancer, and gathered 108,000 signatures on a petition. The Metro newspaper have commissioned a poll showing that 82% of 16 to 24-year olds in England agree with lowering the screening age.

“Cervical cancer may be rare in women under 25,” says Mr Newmark: “but it is inexcusable to dismiss the cases that occur as negligible statistics.” Oh, statistics. “We have a vaccination programme that ends at the age of 18 and a screening programme that begins at the age of 25. That leaves young women between the ages of 18 and 25 caught in a medical limbo, eligible for neither vaccination nor screening.”

Somebody should do something: an intuition which you will find at the bottom of many calls to extend screening programmes beyond the population in which they can provide useful information, and into low risk populations where they simply waste resources, or do more harm than good.

If screening worked, you would expect to see a reduced incidence of cervical cancer diagnoses in people who have been screened, compared with people who have not been screened, in the 5 years after screening: because precancerous lesions will have been detected and dealt with before they got to a more advanced stage.

In August 2009 the British Medical Journal published a large study examining this very question. It found that screening was associated with an 80% reduction at age 64, 60% at age 40, and so on. But cervical screening in women aged 20-24 has little or no impact on rates of invasive cervical cancer in the following 5 years. Only the Liberal Democrat MP Evan Harris introduced these findings to the debate (with the rather excellent line: “The honourable Member for Braintree cited evidence from The Sun, so I want to refer to a recent edition of the British Medical Journal”).

Meanwhile on the very same day David Tredinnick, Conservative MP for Bosworth, stood up to speak on medicine. Scientists and doctors who doubt the efficacy of alternative therapies are superstitious, ignorant, and racially prejudiced, he explained. “It is no good people saying that just because we cannot prove something, it does not work… I believe that the Department needs to be very open to the idea of energy transfers and the people who work in that sphere.”

He went on. “In 2001 I raised in the House the influence of the moon, on the basis of the evidence then that at certain phases of the moon there are more accidents. Surgeons will not operate because blood clotting is not effective and the police have to put more people on the street.”

Where does this moon stuff come from? “I am talking about a long-standing discipline—an art and a science—that has been with us since ancient Egyptian, Roman, Babylonian and Assyrian times. It is part of the Chinese, Muslim and Hindu cultures… Criticism is deeply offensive to those cultures,” says Tredinnnick: “and I have a Muslim college in my constituency.”

Any attempts to challenge Tredinnick’s ideas are based, he explains, on “superstition, ignorance and prejudice” by scientists who are “deeply prejudiced, and racially prejudiced too, which is troubling.” So I hardly dare to mention that Tredinnick tried and failed to claim £125 in parliamentary expenses for attending an intimate relationships course teaching how to “honour the female and also the male essence and the importance of celebrating each”, run by a homeopath.

Meanwhile the flag-bearers for conservatism at the Spectator are now promoting climate change denialism, as George Monbiot has pointed out, and Aids denialism, under the tedious flag of “only starting a debate”, even in their print edition. And finally, the NextLeft blog recently pointed out that of all the top ten conservative blogs, every single one is sceptical about man-made climate change. It could be an interesting five years ahead.

Ben Goldacre from the Guardian does a great job of exposing the world of politics and politicians as they step into the world of science.
As always Ben does a great job of exposing the lack of science and data as some politicians jump onto a personal hobby horse.

As he points out in the attempt to introduce Cervical Cancer Screening he suggests that despite "Cervical Cancer being rare in women under 25" and suggesting we "do something".

The BMJ published a study in 2009 that did! And it demonstrated the value of screening in different age groups with "little or no impact" in the 20 - 24 age group.

Progress depends on data and introduction of new treatments, diagnosis and thinking should be based on scientific analysis of data and not on hunches.

The foundation of this is generating clinical data that can be analyzed and while our medical records are chocked full of data that remains locked in narrative blocks that are inaccessible to computer analysis without the extraction or abstraction of that information typically through manual steps


In a recent presentation at the Partners for Connected Health Symposium in Boston Last week Janet Dillione the Executive Vice President and General Manager, Healthcare Division, Nuance presented

The Voice of Healthcare, The Value of Understanding (Imperial)
- Janet Dillione, Executive Vice President and General Manager for Healthcare, Nuance Communications, Presented "The Voice of Healthcare, The Value of Understanding" that highlighted the potential for bridging this gap with technology that takes the narrative and turns it into clinically actionable data using advanced NLP technology; Clinical Language Understanding. This is the first step on what will be a critical pathway to the future of medicine based on data and science.

Posted via email from drnic's posterous