The Doctor's viewpoint

Intelligent medical IT

"The real problem is not whether machines think but whether men do."
BF Skinner (US psychologist).

My personal view is that medical IT is very different to IT in different sectors. For a start, the fact that clinical staff have to sit in front of a computer in order to engage with the IT system, taking them away from the real focus of their work, seems a bit bizarre.

This is exacerbated by the increasing dependence on computer systems, which means that more time needs to be spent in front of the computer, with an obvious expense to the patients.

We are clearly in the birth stages of IT systems in the NHS and there have been notable successes in what is the biggest civilian IT project in the world. However, there are also large numbers of teething problems, which erode the clinicians' faith in IT. We are becoming dependent on IT and we realise its usefulness for patient care but we have to be careful regarding its implementation.

Basic principles

From my point of view, as someone who has designed successful and unsuccessful IT projects in trusts, I think there are five basic principles to the design of any IT system. These should be the founding cornerstones of medical IT:

1. The system supports the staff, not the other way around

Computers are faster than humans at many things, so the introduction of an IT system would speed things up. The system becomes a victim of its own success and often a new IT implementation heralds the recording of a lot more data.

Instead of IT providing more results quicker as a by-product from the data already created during everyday work, we have to input more data to record more stuff. In this case, the results of data churning becomes the focus rather than the speedier recording of useful data. The focus of an IT system needs to be helping get things done more quickly.

2. The solution fits the problem, not the other way around

I always believed that tools were supposed to help their workers. There is always a risk that the excitement and investment surrounding a new IT system means that we have to make it work at all costs. That all too often translates into more work and the IT solution forgets the problem it was supposed to solve.

As an example, I recently worked in a trust which implemented electronic discharge summaries. Originally the summary would take 20 minutes to complete in its paper format. When electronic, it took one hour per patient. It became so bad that the juniors on the team would stop going on ward rounds so they could do the discharge summary.

This is not an unusual example and highlights my point. The system had been hailed as great as we could produce legible discharge summaries, whereas the truth was that the handwritten summaries usually are legible and meant the juniors could look after the patients they were supposed to be responsible for.

3. The solution is simple

As much as I am addicted to computers, I can always appreciate a simple program. The backend can be a roaring Heath-Robinson contraption running on coal for all I care, but the user interface needs to be simple.

Most NHS staff struggle to use 5% of the functions in Microsoft Word, so then producing a program with eight hundred buttons is going to impress no-one. Google have highlighted the correct path. For all their complicated algorithms, they have still maintained one input box with one attached button.

4.The solution is beautiful

I think this is self-explanatory. By beauty I mean to look at, as well as how it works. I would say that the implementations of PACS I have worked with have been beautiful. They have used simple interfaces, the interfaces have been considered for a basic radiology user, and it is fast. Most importantly the interface should be intuitive so I don't have to spend hours with a manual learning basic functions.

5. Standardised interfaces

This is a call for the IT industry as a whole to put heads together. Imagine my frustration when I move trust every six months only to have to learn how to use yet another electronic discharge summary, and PACS system, and PAS, and workflow management.

I am aware that standardisation is a buzzword at Connecting for Health but the interface standardisation is not yet apparent so that the system is intuitive enough not to have to go to endless training days.

I think IT is abused. Most people don't understand its capabilities and limitations. Most people don't understand the problem it is trying to solve. Most people see it as a cool toy and when it breaks they blame everyone else. Medical IT systems need to be designed by clinicians and managers with an understanding of IT and an understanding of what they are trying to solve.

Once the problem is isolated, then the solution should follow.