Cross Talk on saving lives and causing deaths
There is no doubt that healthcare IT is helping to save lives every day; but sooner or later it is going to get the blame for killing someone who might have lived.
A debate at this year’s Healthcare Computing conference suggests that the issue is being taken seriously. Mike Cross says the next step is to focus on making IT systems as safe as possible.
If you offered a product or service that killed tens of thousands of innocent people a year, you would normally expect to end up in the dock.
Unless, of course, you were a popular outfit that was just doing things the way they had always been done – which is how the health service (mostly) gets away with those terrifying annual statistics about the incidence of avoidable deaths in hospital. “You can bet that a report about an innovation killing one person will generate more publicity than a report about it saving 999 lives.”
Reporting the bad news
Cutting that annual toll is a very good reason for getting passionate about healthcare informatics.
Around the world, electronic medical records have saved many lives - by alerting physicians to existing conditions for example, or ensuring that paediatric patients don't get adult doses of drugs or simply ensuring that the right procedure is carried out on the right patient.
Sooner or later, however, IT is going to get the blame for killing someone who might otherwise have lived. And you can bet that a report about an innovation killing one person will generate more publicity than a report about it saving 999 lives. “You can bet that a report about an innovation killing one person will generate more publicity than a report about it saving 999 lives.”
There's nothing particularly perverse in the fact that the media and politicians are obsessed with catastrophes – they merely reflect the way normal people view the world, which is not as a mathematical model. However, enthusiasts for healthcare informatics need to be ready for this.
Discussing safety at HC2008
There are signs that we're starting to think about it. A session on the safety aspects of clinical software was one of the highlights of this year's Healthcare Computing conference in Harrogate.
The fact that it was held at all, and that it featured thoughtful contributions from senior NHS figures, was an important step forward in a debate that has sometimes treated any discussion of IT's negative consequences as some kind of sabotage.
Recognising that risks exist is one step. The next is to start acting to minimise them. Here, as so often in IT, we can take lessons from older industries. One of these is aerospace.
At Harrogate, Kit Lewis, one of NHS Connecting for Health's user interface design architects, drew a fascinating parallel with the development of military aircraft cockpits. Like clinical IT computer interfaces, aircraft cockpits are complex systems and a mistake can be disastrous.
The modern era in cockpit design began in the US in the Second World War, when officers noticed that an alarming number of pilots were landing their planes safely - and then promptly retracting the undercarriage. As pilots say, this can ruin your whole day - not to mention a very expensive aircraft.
Investigators noticed that the accident was common to a few types of aircraft. And it turned out that what they had in common was that the lever to control the flaps, which should be retracted after landing, was next to the lever to control the undercarriage, which should not. The levers looked, and felt, identical.
The rough and ready solution was "mnemonic shapes" - replacing the undercarriage lever with one shaped like an aircraft wheel. The incidence of "wheels up on landing" accidents plummeted. “Recognising that risks exist is one step. The next is to start acting to minimise them.”
The long term solution was to lay out the cockpit so that it was nearly impossible for a pilot to confuse the two levers. That principle now guides every step of cockpit design. (And just to make sure, the mnemonic shapes survive to this day.)
Making systems safer – and sharing information
Similar thinking is now being applied to the NHS common clinical interface. For example, an instantly recognisable Stop! triangle has replaced the lines of text that previously warned users on the verge of taking a major consent decision.
Much more can, and should be learned from other safety critical industries. One speaker at the conference session, Dr Maureen Baker, the national clinical lead for clinical safety commended the oil industry's practice of freely sharing safety-critical information.
Healthcare IT firms should follow suit. It won't eliminate all accidents - healthcare involves just too many variables - but when a bad one does happen, and IT gets the blame, we'll be able to put our hands on our hearts and say we did our absolute best. That's important.
Microsoft’s Common User Interface programme
As part of its long-term commitment to the NHS, Microsoft is running the Common User Interface programme. This has a number of projects to help the NHS get the most out of its IT – and to make it easier and safer for staff to use.
Find more information about the desktop of the future and how it will make healthcare safer
About the author: Michael Cross is a freelance journalist specialising in healthcare informatics and e-government. He is a member of the British Computer Society.