For the past 7 years or so, I've used professional actors working as simulated patients to test the communication skills of medical students and doctors. This training takes place in a classroom setting, and the participants know that they are being observed.
Not surprisingly, most try hard to be empathic, but there is still a wide variation in the diagnosis and treatments they offer to patients presenting with exactly the same symptoms, despite the plethora of guidelines we all have to shunt us towards best practice.
This variability is understandable when a patient has vague and confusing symptoms that don't fit into any particular pattern. But more often, it is the result of the students making basic clinical errors.
When I rather rashly explained all this to the editor of the Daily Express, she decided to see if this same variability existed in real life by sending simulated patients undercover to GPs' surgeries.
The situations chosen were very basic and pitched at medical school finals level a woman suffering from classical migraines requesting the combined pill and an actor with asthma requesting beta-blockers for his nerves.
The contraindications to both of these prescriptions are clearly stated in guidelines in the British National Formulary. And yet, in a sample of 20 GPs in London and Manchester, half handed out the requested prescriptions to the simulated patients.
To be fair, neither 'patient' walked in and said "I've got asthma, give me some beta-blockers" or "I'm having migraines with an aura, can I have the pill?" But neither did they make any attempt to hide their symptoms or past medical history.
Half the GPs very efficiently asked the pertinent questions to uncover the danger and refused to prescribe. Half didn't. It was as simple as that.
Should we be surprised?
Based on US estimates, 320 000 clinical errors are made in the UK each year, resulting in 40 000 deaths. In an NHS where effective quality control is virtually non-existent, no-one even picks up on the majority of our mistakes, let alone learns from them.
In the Express survey the simulated patients were by necessity temporary residents, and without the benefit of access to their records it's far easier for a GP to gloss over their past history and assume that their own GP has done all the groundwork.
It also adds weight to the argument that walk-in centres and telephone health advice services are more likely to make these errors (indeed, a survey of the latter by Which? magazine found similar variability).
But instead of gloating at the failings of New Labour's big ideas, we should be concentrating on putting our own house in order.
I'm not sure how easy that will be. Most of the GPs in the survey were running at least half an hour late and were clearly stressed. The vast majority had excellent communication skills and were interested in their patients.
The paper had originally wanted to name and shame the GPs, but I argued that this would miss the point entirely. All the research from America and industry shows that blaming individuals very rarely results in changes that prevent the same mistakes happening again.
Yes, we have to be accountable for our individual actions, but we need to work in a system that protects our patients and ourselves from very easily made errors and oversights.
The answer is not to have a three-mile high pile of guidelines in the treatment room and a copy of the BNF on the smear trolley, but a system which ensures that best practice is actually practised.
We have to move from knowing best to doing best, and at present we're a long way off.
Perhaps the answer lies in giving patients legal ownership of their drugs and illnesses. If you're on the combined pill and you get a severe one-sided headache preceded by loss of part of your vision, it's your duty to stop taking it immediately and see a doctor. Similarly, if you have asthma, you should know never to take beta-blocking drugs.
The survey shows that if patients don't learn the basic do's and don'ts, they can't assume that their doctor will remind them.
Maybe the only thing that'll save the NHS is patients taking on more responsibility for their health.