IT appears to be one of the simplest problems to treat.

A woman rings in with 'cystitis' and we prescribe a three-day course of antibiotics.

But shouldn't the doctor do more? Shouldn't we be seeing the patient and sending urine samples to the lab?

In 'cystitis' the bladder becomes inflamed. The woman feels as though she wants to pass urine all the time and, when she does, it is like passing razor blades.

There can even be some blood in the urine. Although most patients talk about 'cystitis' doctors usually call it a urinary tract infection or UTI.

Cystitis literally means inflammation of the bladder but, in UTIs, the inflammation affects far more of the urinary tract than just the bladder.

The reason women suffer UTIs is because the urine is infected with bacteria.

These are often bacteria which live harmlessly in the bowel but spread to the bladder and cause problems.

We now know that the situation is not so simple. More than one third of women with symptoms of a UTI have a clear urine with no infection. Also many women who have a proven infection get better without antibiotics. If we give antibiotics to every woman with symptoms of a UTI then one in ten of the adult female population would receive antibiotics every year.

This would lead to resistant bacteria so the antibiotics would not work when they are really needed.

But it is difficult for a doctor on the front line to refuse antibiotics to a woman in pain. And we do know that nearly two thirds of women in this situation do have an infection which needs treating.

There has been some recent research from Southampton which might help solve this problem. Women with symptoms of a UTI were treated in one of five ways:

• just given antibiotics

• given antibiotics and only told to take them after 48 hours if they still have symptoms

• given a questionnaire asking about cloudy urine, smell and symptoms and only given antibiotics if there was a high score

• given antibiotics if a urine dipstick showed an infection. This is the management recommended in South Devon

• given antibiotics after sending a urine sample to the lab and waiting for a positive result.

The result? There was no difference in the duration of symptoms or harm to the patients in any of the approaches. Of the women who were told to wait 48 hours and only take the antibiotics if they still had symptoms – 80 per cent ended up taking the antibiotics.

The only difference between the groups is that fewer antibiotics were taken in every group except the first, the group just given antibiotics with no further questions asked.

But there was another, unexpected, result. If the doctor was assertive and clear that the outlook was good the burning pain did not last as long. They got better quicker. In purely scientific terms this is illogical. We have a physical disease caused by bacteria and being treated with antibiotics. Why should the approach of the doctor affect the speed of action of antibiotics on bacteria? And this flies in the face of modern teaching. Doctors are now taught to be 'patient centred': to involve the patient in management and point out risks and side effects.

They are not taught to be dogmatic and tell people they will get better. But it works.

I now know that when I see a woman suffering from cystitis I may or may not need to see her. I may or may not need to ask detailed questions about her symptoms. I may or may not need to dip stick her urine or send it to the lab. But, whatever I decide I need to be confident and tell her she will get better. It reassuring to find scientific evidence that an old-fashioned approach works.