« April 2006 | Main | June 2006 »

May 29, 2006

DNA Computing

Diagram of DNA moleculeI often write as though one day IT will replace clinicians as repositories of medical knowledge freeing them to address the human side of care. But what if the silicon chip itself could be replaced?

In the May 2006 edition of Scientific American Bring DNA Computers to Life foresees a biomedical computer acting as a intracellular GP by sensing molecular clues and sending out appropriate signals or drugs.

In 1936 Alan Turing suggested a universal computing machine with three parts:

  • An input device;

  • A set of internal states;
  • and
  • An output device.

Such a machine resembles the way proteins are synthesized in living cells. Messenger RNA carries gene transcripts in codons each of which corresponds to an amino acid. The ribosome reads the information on the messenger RNA, like a cook a recipe, collecting ingredients, chopping and combining them to produce the protein.

A biomolecular computer would be slower than a chip, and no known enzymes can do the reading, chopping and combining needed to create it. However, while we are waiting for the enzymes to be synthesised, the article's authors suggest that a two state diagnostic device indicating "yes" and "no" may suffice.

For example, many cancers are characterized by abnormal levels of certain proteins in the cell. The gene activity needed to make them could trigger the production of molecules that would signal their presence.

May 28, 2006

In Lobe with IT: neural enhancement

Scan of a Brain.

Cat's foot iron claw
Neuro-surgeons scream for more
At paranoia's poison door
Twenty-first century schizoid man

I first heard Greg Lake yelling these lyrics to King Crimson's 21st Century Schizoid Man at Alan Fairbairn's house while skiving school games one Wednesday afternoon. They disturbed me then and seem eerily prophetic now.

In 20 May 2006's New Scientist Neurosurgeon Katrina Firlik predicts the development of a whole new field of brain surgery implanting electrical stimulators to treat conditions like depression and language disorders.

According to LiveScience.com European scientists have created a neuro-chip by merging neural proteins with a silicon chip. Though it could be decades before a living computer is realised, such technology could be used now by pharmaceutical companies to assess the effect of drugs on nerve tissue.

More about Pete Sinfield and his lyrics.

May 21, 2006

Blundering the NPfIT?

Last week, I attended a presentation by David Craig (a pseudonym) who wrote Plundering the Public Sector, which criticises the cost, justification and management of NHS Connecting for Health's National Programme for IT (NPfIT). Book and presentation contain nuggets of wisdom tarnished by uneveness and inaccuracy.

In the book, Mr. Craig claims that the total cost of NPfIT could be as high as £50bn. Even if this is correct (and I'll address that shortly) that is £5bn a year from a total NHS budget of about £70bn year. I consider a 9 percent IT spend (adding non-NPfIT) modest for a sector whose basic currency is information. At least 12 percent would be in line with similarly dependent sectors like banking and finance.

He also states NPfIT's headline cost of £6.2bn is dwarfed by the likely costs of implementation. This may be true, but I suspect his estimates include the full replacement cost of staff during training. In my experience, such costs are economic rather than real, because staff are often not replaced for short periods and their costs are already met. Further, the NHS already spends about £1bn a year on IT, at least some of which could be diverted to implementation.

Mr. Craig is scathing of Choose and Book (C&B) an IT system that will allow GPs to make direct outpatient appointments at a patient's choice of hospital.

He estimates the costs of C&B at £200m for development, £122m annual running costs and £100m for existing systems modification. In the book, he says that amounts to £125 a booking (page 216). However, in last week's presentation he revised that to a more arithmetically correct £15—I presume by using 9.4m GP bookings a year for ten years as the denominator for total costs for the same period.

However, his arguments seem tendentious when he makes no assessment of the likely cash-releasing benefits of C&B, such as reducing admin costs in the current system of letter and telephone tag, or the potential for transformation of the acute sector that using C&B may have when combined with Payment by Results and Practice Based Commissioning .

In the presentation, Mr. Craig asserted that the NHS could have saved on C&B by buying an airline booking system and adapting it. Further, most of the systems procured by NPfIT are available off-the-shelf in the US, he says. In this he seems to rely on an early report, apparently written by McKinsey, that concludes the whole lot could be done for £2bn by such adaptation—though we are not told the basis of that estimate either.

Now, I favour an off-the-shelf IT product meeting most user needs that can be adapted, rather than a development from scratch, but Mr. Craig's assessment seems superficial. IT systems are available in the US, but invariably need modification for the NHS by the suppliers who invariably underestimate the effort needed. This leads to extra cost for them (at least in my procurements) and long delays for users.

Moreover, even the US does not use healthcare IT well, for example read this assessment on Dale Hunscher's blog.

Mr. Craig has it in for management and IT consultants—blatant in his previous book Rip Off! In the presentation, he exposed their sales techniques, which he rather incongruously referred to as "all great fun".

I spent 7 years in sales and marketing (not for management consultancy) and recognise many of these assessments and techniques. I believe most are not exclusive to consultancy and are used by many service industries.

Nonetheless, he gives sound advice on employing consultants: define scope and client and consultant responsibilities, insist on skilled staff, pay by deliverables and so on. But this advice can be had free in publications from the UK's Office of Government Commerce (OGC) and Audit Commission.

Eager whistle blowers apparently inundated Mr. Craig with material allowing him to draft his recent book in a few weeks. It shows. He should have taken the time to carry out a more balanced assessment and to have at least got the Programme's name right.

Also see this article by the Guardian Online's Mike Cross on E-Health-Insider.

May 15, 2006

Medicine: humanly impossible?

My wife is studying Physiotherapy and spends her life scouring publication databases on-line. If she proposes a regime of treatment to a lecturer they ask: “How do you know? Where is your evidence?” Excellent, I say as an inveterate sceptic.

Both Shahid Shah and Dale Hunscher have posted on Evidence Based Medicine (EBM). Shahid has posted slides from a presentation, and Steve Beller makes some particulary interesting comments at the end of Dale’s entry.

I’m not sure that I would entirely agree with Dale’s comment: “It's about getting all relevant facts in front of the clinician at the point of care and then trusting her to use her own judgment as to how the facts apply to the situation at hand.”

I more inclined to agree with a quote I saw in a presentation on Artificial Intelligence in healthcare: “Medicine is a humanly impossible task.” The estimated rate of medical adverse events in the US and in the UK would seem to support that view.

Integrating EBM and healthcare IT is an exciting prospect. This article: Assessment of the potential impact of a reminder system on the reduction of diagnostic errors: a quasi-experimental study concludes in support of the use of ISABEL to support clinical decision making.

However, as usual, technology will be a small part of the challenge. Experience so far indicates that disciplinary boundaries, differences in the interpretation of evidence and general reluctance (or inability) to change practice owing to the pressure of business-as-usual are likely to be more potent factors.

To make progress healthcare must adopt change and continuous improvement as givens and move away from a static, authority-driven approach.

May 10, 2006

Future Health IT in the News

FHIT is reviewed by Healthcare Today. You can download free a PDF of the April 2006 edition here. The review is at the bottom of page 21.

May 09, 2006

Scientific Metamorphoses: robots resemble humans

woman.jpgIn a recent interview for an article, Honda Motor Europe’s William de Braekeleer told me: “The long term objective of our engineers is to create a robot able to help people in their daily lives. So that is why ASIMO has been designed to walk and move just like us.”

Others have in mind resemblances beyond movement. Android Science in May 2006’s Scientific American describes an honest-to-goddess android called Repliee Q1Expo that looks like a thirty-something woman. To create it, Hiroshi Ishiguro used silicon rubber and polyurethane to cover the metal armature and Japanese newscaster Ayako Fujii as his model.

Japanese are less anxious about robots than Westerners and regard them as benign and friendly—Sony’s robot dog AIBO (see FHIT article) is treated like a family pet in many Japanese homes. Nonetheless, humans may be uneasy about robots that are similar to them, but not similar enough—what the Japanese apparently call “spooky valley”.

No matter how similar to the Pygmalion myth Repliee's creation may seem, Ishiguro does not think it necessary or possible to sculpt Blade Runner-like replicants. Rather, he sees developments like Repliee as a means for humans to understand themselves better .

May 05, 2006

Sounds Healthy: iPOD in healthcare

On the London Underground everyone seems plugged into one. Even above ground in healthcare MP3 players are becoming ubiquitous.

Take the NHS Greater Glasgow Hospitals use of Voicemap™ for staff training. Like the guided audio tours used in galleries and museums, this system helps staff to find their way around, advising then, for instance, on location-related health and safety risks.

This FHIT entry describes iPOD being applied to train students to use a stethoscope.

Learning modalities are sometimes categorised as auditory, visual and kinaesthetic. My wife, a former ballerina, is a kinaesthetic, learning best by movement. She remembers a 90 minute ballet class, whereas I—a visual learner—would be lost in the warm-up pliés. Fortunately, iPOD also offers something for us visuals.

Dr. Osman Ratib is storing and transmitting and medical images with his iPOD. He even offers free software so we can try it.

It may be cool to carry iPOD, but who would go this far? Researchers have dispensed with Bluetooth and implanted a chip in a forearm using the body as a conduit. One day implanted devices could communicate without an enabling technology.

Might the elegantly realised iPOD evolve into one of the first popular information appliances that combines IT, medicine, technology and ease of use?

May 01, 2006

SmartPill® Tracks Your Tract

In 1995 I visited the exhibition in London's Tate Gallery of the (infamous) Turner Prize shortlist. Stepping into the darkness of Mona Hatoum's installation Corps Etranger I watched a video journey through her intestine and other bodily passages. Today she could avoid the discomfort of the endoscope in the interests of Art by using the multi-vitamin-pill sized Smartpill®.

When swallowed, SmartPill records its trip through the digestive tract using its battery of on board equipment and sensors, like a thermometer, pressure gauge and acidity meter.

Unlike the miniaturised submarine carrying Raquel Welch et al in the 1966 movie Fantastic Voyage, SmartPill is propelled by peristalsis. When excreted 24-48 hours later, a doctor downloads its data in less than 10 minutes and analyses them, a process which could transform the diagnosis of conditions like dyspepsia, constipation and irritable bowel syndrome.

For more see Roger Dobson's article in the UK's Daily Mail of 25 April 2006 or visit SmartPill Corp's website.