Dr Phil Hammond, broadcaster and sessional GP in Bristol

When did you last have any training in prescribing? It's a core skill for any doctor and takes time to develop, yet rarely does it receive the attention it deserves, either in the consultation or in our postgraduate education. With only 10 minutes for a consultation—or 6 by the time the patient gets in and out of the door with a double buggy/arthritic hips/orthopaedic shoe—we have just 5 minutes to make a diagnosis and maybe 1 minute to prescribe. Small wonder accidents can happen.

The General Medical Council's major study of GP prescribing found that while the vast majority of prescriptions written by GPs are appropriate and monitored effectively, around 1 in 20 contain an error.1 This was helpfully reported on the front page of the Daily Express as: '1 in 6 patients at risk from GP blunders.'2 Oddly, you never see the headline: '1 in 6 patients at risk from the Daily Express.'

The study took place in just 15 practices, but the media loves to extrapolate. Apparently 45 million GP prescriptions contain errors each year, of which 1.8 million (1 in 550) are serious.3 Does this surprise you? Given the tiny amount of time we have to prescribe safely, it's generally left to the computer to spot dangerous drug interactions or question our dosage regimen. Indeed, I'd say our software is the best safety defence we have. Never go into a GP surgery when the computers are down. It's the most dangerous place on earth. Just don't tell the Daily Express.

I once attended a Medicine Safety in Practice conference, which discussed prescribing errors so serious that they couldn't be hidden. The first concerned a man receiving long-term warfarin therapy to reduce his risk of a stroke; he also had arthritis. Everyone in the room knew what was going to happen. He was going to take a painkiller that would interact with the warfarin, and he would have a catastrophic bleed.

So how did it happen? The patient saw a locum GP on a morning when the computer crashed. The locum didn't know him or ask about his other medications before giving him a prescription for a non-steroidal anti-inflammatory drug. His usual pharmacy had a big Christmas queue so he went to an unfamiliar chemist who also didn't check his medications. And his blood test at the warfarin clinic was delayed because of the holidays. He went home, ate some turkey, and had a massive stomach bleed.

Most examples showed how heavily GPs rely on their computer to prevent cock ups, and how easily errors can occur. For example, a consultant mumbled the name of a drug—triamcinolone—into a tape recorder and it was wrongly transcribed as methotrexate. Not even close. The letter was sent to the GP unchecked. The GP issued a prescription. The pharmacist queried it, but was satisfied when the receptionist read the letter back to her. The drug was given as an injection by the practice nurse who didn't realise it was cytotoxic and got her decimal point in the wrong place, giving 10 times the dose. The nurse realised her latter mistake, got the patient into A&E, and reported the error. The Primary Care Trust has now banned cytotoxic injections in primary care and offered medicine safety training, which GPs are too busy to attend.

I could go on but you get the picture. Prescribing errors harm and kill more patients than anything else we do, and we need to regularly refresh our prescribing skills and make time in the consultation to check what we're doing. As well as take a good history, make an accurate diagnosis, and be a model of empathy. Oh, and fix the double buggy. All in under 10 minutes. No one said this job was going to be easy.


  1. General Medical Council website. GP prescribing: a good standard but improvement possible. www.gmc-uk.org/news/13017.asp (accessed 8 May 2012).
  2. Daily Express website. 1 in 6 patients at risk from GP blunders. tinyurl.com/6r67sq2 (accessed 8 May 2012).
  3. The Telegraph website. Millions of GP prescriptions contain dangerous errors: research. tinyurl.com/c5lk3w9 (accessed 8 May 2012). G