The young pharmacist was clearly unhappy. He was spending more and more time writing, but hated it. Why? He found it hard to get started; once started he found it hard to stop; once he finished the first draft he wanted to change it; and once he showed the draft to others they were quick to tell him how he should have done it better.

He was typical of the growing number of pharmacists who have found themselves on one of my "effective writing" courses. They have been reassured to find that they are not alone, and that they share these problems with other pharmacists, and with many other health professionals. To be honest, though, pharmacists do have one or two special problems of their own.

What they share with other health professionals is a lack of any formal writing training since "O"-level English. Any writing training since will have focused on two tasks: passing written examinations and securing publication in scientific journals. Few will have had training in the skills needed to write for different audiences in different formats. In general, they will plan too little, write too early, rewrite endlessly — and be unable to give a reasonable definition of effective writing anyway.

They will be working in a negative environment. Take the writing of scientific papers, which for many in the health services is the epitome of "good writing practice". The task is usually handed down to the most junior member of the team, who is given a vague brief, no additional time and certainly no additional resources. Once the author has completed the first draft, everyone is encouraged to point out, usually with a red pen, what is wrong with it. Being able to criticise content (which is easy) has become more important than recognising whether that content is likely to be communicated from one person to another.

This peculiar — and picky — culture has moved medical writing away from the tenets of good style advocated by George Orwell and many others (see Panel overleaf) and towards a style that values pomposity and obscurity. "You will have to take the tablets for a long time" becomes "long-term medication is predicated" and "swallow the medicine" becomes "it is recommended that this preparation be consumed orally". The culture favours the posh overcoat and, not surprisingly, the Plain English Campaign has had little impact here. The dense style of medical journals remains highly valued, though the articles remain largely misunderstood and unread. Articles in medical newspapers and magazines become derided as "Janet and John" or "journalese" — even though they are read and acted upon.

In this climate pharmacists become particularly vulnerable to the games that medical professionals often play. One is the ritual humiliation by red pen, described above. Another is "Spot the missing reference", when those who have not written the article try to discredit it by finding obscure experiments that have not been included (or deliberately excluded, which is different). Not only do they rarely make a difference, but this behaviour ignores the fact that the hard thing about writing is not putting things in, but leaving them out.

A more serious problem for pharmacists is that their mistakes may have dreadful consequences. Sometimes an error in the copy may be amusing, as in the patient leaflet on Deflatine that claimed that patients would be able to pass their wind "discretely" (they meant "discreetly"). On other occasions a mistake could be fatal: moving a decimal point could literally decimate populations.

Unfortunately the response has been a particularly defensive kind of writing, in which all data are included and every word pored over by committee. The writing may be safe, but it is also unreadable. Some firms take this to a fine art: letters of inquiry are no longer answered by one person writing to another, but by one person assembling an answer from a database of stock sentences, all of which have been previously approved by the marketing department and lawyers.

What is the remedy?

The remedy to all this is not too difficult to find. It involves going back to basic principles and here are some golden rules.

1. Be absolutely clear why you are writing — and how you will judge your success The real test of a good piece of writing is whether it has achieved what you set out to achieve. If your letter has succeeded in telling the GP not to prescribe the drug (provided that is what you want) then you will have written effectively. Anyone who complains about a split infinitive or a missing reference is missing the point.

2. Define your audience The more tightly you can define the audience the better. Aiming at two audiences at the same time means you are unlikely to please either. If you can write for one person (such as the GP who wrote the letter, the pharmacist who is editing the journal, the accountant who is chairing the finance committee) then your document is more likely to work. If you need to get your message across to more than one audience, then write it more than once — a scientific report for your pharmacy colleagues, for example, and an executive summary for the accountant.

3. Work out what you really want to say Define carefully the key message — not as a title ("Hexaplin and the common cold") but as a message in sentence form with verb ("Hexaplin can cure the common cold"). Make sure this message is in a prominent place, preferably at the start of your document (unless you are writing a scientific paper, in which case the convention is to put it at the end).

4. Use the language with which your readers will be comfortable If you know your target audience, you can use the words and construct the sentences in a way that is most likely to get the message across. Try the "pub test": how would you explain this to your target audience over a glass of beer? This should sort out all problems of jargon — and give you the courage to shed that posh overcoat.

5. Be robust Of course you will make some errors; of course you will leave out some references; of course you will offend some readers. But keep focused on what you are trying to do — and get your satisfaction from that. You will never write the perfect text.