Last Modified:                                                                                                



Homepage / Publications & Opinion / Silicon.com

Peter Cochrane's Uncommon Sense: Medical Records
Poor management of data in healthcare - and so many opportunities being lost

Every medical system in every country on this planet seems to be in trouble for the same basic set, or sub-set, of reasons. Ageing populations, extended lifetimes, rising patient expectations, more chronic and long lasting illnesses, more people in need of long-term care, a lack of carers, limited medical staff, limited facilities, rising operating costs, increasing amounts of equipment available with shorter replacement times and higher overall costs, rising drug and medicinal costs, fixed and/or limited investment, uncontrolled migrant populations, increasing amounts of bureaucracy, litigation - I won't go on.

But looking back at the history of healthcare we can see that the biggest gains have actually arisen from the simplest of revolutions. In engineering this is known as the 80/20 rule: you get 80 per cent of the result for only 20 per cent of the expenditure and it is that last 20 per cent of the result that will consume most of the time and the money.

So what might be done? Where is the next big gain going to come from?

For all nations the single biggest advances in the health of their populations occurred in the following manner: the provision of clean water supplies, sanitation and sewage treatment, followed by antibiotics. Everything else, before and since, palls into insignificance compared to these major advances.

In my view the next really big advance is already sitting in files all over the planet waiting to be tapped. The data mining of patient records going back some 100 to 150 years probably holds the key to the next giant step forward for everyone's health prospects and a revolution for the medical profession.

While this profession employs some of the most kind, compassionate and caring individuals at all levels, it is an industry reigned back by the insulation of data and isolation of the players dealing directly with the patients. Because of the nature of medical records it is impossible to correlate experiences, patient histories, treatment regimes and outcomes in anything but small and isolated experiments. As a result medicine remains an experience-, folk lore- and anecdote-dominated sector.

And without a revolution in the way patient notes are recorded, stored and accessed, nothing will change. In this 21st Century you might have expected that a doctor and/or consultant would be presented with a standard patient record system on a PC screen. At the time of a patient interview and examination a series of questions would be asked, tests performed and simple 'tick boxes' marked up at the click of a mouse. Blood pressure, glucose level, skin condition, reflexes are all recorded by pull-down selection and a click, and of course the occasional typed notes to amplify and explain points and decisions.

But it's not so. By a very large margin patient notes are still committed to loose leaf sheets of paper by fountain pen, ballpoint or felt tip.

Moreover, the style and format is generally up to the individual medic who may choose long or short sentences, acronyms, rough sketches and more that vary in style from patient to patient, day to day. At best you sometimes see a pre-printed form used with ticks and comments - and that is about as advanced as it gets for the vast majority.

Why are these medical records taken and stored at all? About 20 per cent of the reason is to aid the medical staff in their work, to chart progress, to record what has been done and what the conditions were - to generally look after the patients' long-term well-being.

And the remaining 80 per cent of the reason? To provide a legal canopy in the event that something does go wrong and litigation follows. Some of the multiple checking and correlation between the staff working on a case I can truly applaud as it serves to limit errors and omissions while facilitating a valuable cross checking mechanism. But for the most part there is a lot of skilled (and in short supply) time and talent wasted serving an uncaring bureaucracy.

I tend to think of these medical systems like a child's lexicon puzzle but without an empty square and so nothing can advance or move! Just pouring more money into the system will not help. Nothing will change. However, inflicting a single common recording system for all medical records would release vast amounts of time and talent. But more importantly it would allow the correlation of all patient types with diagnosis, treatment and outcomes. Entire histories could be rapidly run side by side for the first time. And for the first time all the medical staff would have at their fingertips the diagnostic probabilities at the time of decision.

In short, being able to slice and dice, display and decide, on the basis of a huge recorded database would bring massive rewards. And, patient by patient, there would be a progressive honing of diagnosis, treatment decisions and outcomes.

If this is such a great idea, and so obvious, why isn't it being done? In short: people. We are the limiting function. Individuals and organisations, especially very big ones, can't do anything really fast when they are locked into old and established systems.

My guess is that it will take another 20 years for a new generation of tech savvy medics and administrators to rise to power and refuse to perpetuate the old ways of doing things. For sure, all the medical records up to that time will continue to fester on shelves, contributing nothing to the general well-being of mankind and, worse, they may never do so. Their format and immensity probably means they will ultimately be destroyed and the data and wisdom lost forever.

For this not to be the case would require an army of people and an army of machines to read and reformat what has been written by hand. An unlikely prospect I think - so we still have to reach Year 0 and press the GO button!

This column was typed on BA 458 flying London to Madrid, revised on my G4 laptop and despatched to silicon.com from a free Wi-Fi site provided by some kind office or home knowingly or otherwise on a warm Spanish summer morning.