Producing patient information is fraught with pitfalls. Updating a number of factsheets fell to me over a recent wet bank holiday weekend. By Monday evening my mind was reeling from trying to achieve a balance between plain English, political correctness, evidence-based factual accuracy and being neither simplistic and condescending on the one hand, nor over-complicated and academic on the other.

The first problem was which reference books to carry home. The 'bible' most frequent ly used by our health information service is the British Medical Association's Family Health Encyclopaedia , whose editor in-chief, Tony Smith, was the College of Health's chief medical adviser from the early 1980s until last year. Its A-Z coverage of illnesses, symptoms and procedures makes it hard to beat.

But the 2000 edition had been delayed, so I was left with the 1998 edition. It does not include references, which information purists insist on.

So I turned to the BMJ Publishing Group's new clinical evidence series, billed as 'an up-to-date compendium of the best available evidence on the effects of common clinical interventions'.

But it is expensive, and the index contained only 10 of the first 50 subjects I had to cover.A problem with reference books which concentrate on the results of randomised controlled trials is that many common illnesses have never been the subject of RCTs.And many people who need the sort of patient information we produce - the young, women of childbearing age, elderly people, those with other things wrong with them - often have to be excluded from RCTs. Nor do they necessarily have evidence-based illnesses.

But a handbook I did find useful was Guidelines , published by Medendium. This summarises primary care clinical guidelines from reputable centres, and is free to GPs and organisations like the College of Health.

Next came the British National Formulary , along with the Consumers' Association's Drugs and Therapeutics Bulletin and some recent consumer versions of it, along with a precious out-of-print dictionary of rare syndromes.For good measure, I added a couple of reliable guides to home remedies, since people always want to know what they can do to help themselves , and many cannot afford to buy over-the-counter preparations.

But which Internet sites to use for up-to-date evidence-based information? I have a list of favourites. Then come the websites of condition-specific and other self-help groups and voluntary organisations, often with much more practical advice than the purely medical. We have over 2,000 of these on our national database.

By midday Saturday, 'paper overload' had set in and I could no longer see my desk.

By evening I had begun to compile a guide to political correctness and plain English. The words 'victim' and 'sufferer' come somewhere near the top of my list. 'Patient' is sometimes OK, though 'people with. . .' is usually better.Many organisations now favour 'user', though in some contexts it has negative connotations, as in 'drug user'. Also banned, as far as possible, are the 'ives' and the 'ics' - as in manic depressives and agoraphobics.

Enter political correctness. Many modern reference books still use the term 'mental handicap', although 'learning difficulties' is more favoured by patient - or do I mean user - groups? And yet, 'learning difficulty' seems somehow inadequate to describe the profound effects some syndromes can have on mental and intellectual capacity.

With nearly 20 factsheets on HIV and AIDS to check, I hit a problem.

The informative and respected Terrence Higgins Trust says early in its website literature: 'Although you may still hear the term AIDS used, it is now a term not used so frequently by doctors. They prefer to talk of later stage or advanced HIV infection.' I started to cross out 'AIDS' until I realised we would still have to refer people to the National AIDS Helpline, and that most probably do understand what is meant by AIDS rather than HIV. This is certainly true as far as the developing world is concerned since it does not have the luxury of combination therapy.

I compromised with HIV/AIDS.

Referring people to other sources of help is another area fraught with difficulty. Not so many years ago, you had to look under VD or venereal disease in the phone book to find out where you could go for help with sexually transmitted diseases. Or do I mean sexually transmitted infections? My widespread weekend reading suggests that STIs are preferable to STDs. Nevertheless, it is STDs you are most likely to find in the phone book, usually cross-referenced to genito-urinary medicine or GUM clinics. That was the case in my BT directory, but I could find nothing under either of these headings in my 2000 edition of Yellow Pages. I even looked under VD, but to no avail.

Finally, there is the vexed question of drug names. There is no problem with using ibuprofen or paracetamol rather than the many, more expensive, brand names although the temptation can be quite strong to use a name such as Calpol, which a mother will probably instantly recognise since that is what her doctor, nurse or NHS Direct might well advise her to buy.

But I was at a loss with the suggestion that women with menorrhagia who also need to use contraception might consider the intra-uterine device marketed as Mirena. As there is only one on the market, it really did seem absurd to write the BNF's correct pharmacopoeial description: 'A T shaped plastic frame (impregnated with barium sulphate and with threads attached to the base) with polydimethylsiloxane reservoir releasing levonogestrel 20 micrograms/24 hours'.

At that point in my bank holiday, all the Latin I ever knew deserted me - and I never did study Greek. So I called it a day.