Another view: Neil Paul

In my last column, I talked about automatic ways of detecting computer performance. That got me thinking about user performance; and how to improve it.

Your software could track which bits you are using and offer training in the other bits. It might detect that you are doing something in a long-winded way and show you a short cut. Or it might show you some expert tips when you are ready for them.

This need not be intrusive. The middle of a 10 minute consultation might not be the best time for your software to offer you a video on how to do repeat dispensing; but you could imagine it building up a log and having a training section.

It might even be able to identify struggling users; of which there are probably more than those of us who are comfortable with computers are aware.

As an employer, knowing what level my employees are performing at is very important. Having more information would help me target training. Or recognise a highly achieving user, and promote them.

More enterprise features needed

In general, I think that the GP software I’m using is lacking in enterprise features. Every year, we create a flu macro for clinicians to use if they are giving the jab in their rooms. It enters the correct code and batch number.

One of our administrators has to run to each machine at lunchtime to load it. Why can’t we just push it out to them?

When we do this kind of exercise, we find other problems. Perhaps pop-up boxes aren’t working on someone’s machine – but they have either not noticed or have neglected to tell us.

Why can’t we see centrally if all the machines in the practice are working correctly? Is it that hard? A simple report of what machines are patched to what version might be useful.

And more data

At a nearby practice, a new salaried doctor didn’t know he had to check his results himself; presumably this had always been done for him or, as a locum, he had never had to do it.

No one realised for about a month, mainly because there was no way of seeing how much outstanding work people had to do.

Meanwhile, it recently transpired that a local GP who went off with stress had something like 200 notes active and unanswered (I usually have <20). Knowing this GP was performing outside the norm might have given an early clue to their problems.

Some overview of how many letters people have that are waiting for results, the time taken to answer them, and the number of abnormal results seen might be useful appraisal or performance data.

Knowing who is following our formulary would definitely be helpful. My practice has made great progress in cutting our prescribing bills, and out forecast for the year is now to be £100K under budget when we started >£100K over.

However, it has become clear that while we have tried to be as sensitive and inclusive as we can, people slip.

They missed the meeting at which a therapy was discussed. They forget what was decided when we agreed a limited formulary. They get pressurised into bad habits again.

Currently, it takes ages for us to find these few off-formulary scripts. We either have to run lots of searches or wait for historical pact data. It would be great if alarm bells went off.

More seriously, some sort of adherence report would be useful. We could then have a sensible conversation about why someone wasn’t following our collectively-agreed guidance.

Another big area of possible variance – which might become more interesting as we locally move away from a block contact on diagnostics – is who is ordering what tests on whom. I’m guessing there are some huge variations.

Don’t appraise: help

Some might be scared of ‘big brother’ if we got our software involved like this, as a lot of what I’ve mentioned is performance related. However, I think that if it was used correctly it could be very helpful.

I’m having to collect data for yet another pointless appraisal. It would be really nice if the system into which I spend most of my life typing data just fed back to me how I was doing as I went along.