We didn’t introduce drug tests as a stand-alone measure. We looked closely at our whole drugs policy and tried to improve our education programme; the testing was just one part of that. We were also clear that there would be no punitive element to the scheme. Children who test positive are offered help and advice. Dealing drugs, of course, is treated differently.

We were one of the first schools to use random testing and we worried that we might be infringing the Human Rights Act. But we decided that we could get round that by making the tests voluntary. Both the child and their parents have to agree to the tests. The first year, there were 16 parents who objected, the second year just two. It’s even more rare for children to object; in fact, only one pupil has ever refused the test. They see it as part of school routine now. We use a urine-based test, which detects a range of drugs, including opiates and cannabis. The test is 98.5 per cent accurate and gives instant results, so children aren’t kept waiting. It can detect cannabis for up to three weeks after use. That means that there are only two weeks a year, at the start of the summer holidays, when children can smoke cannabis without risking being caught.

We only test in Years 9, 10 and 11, because statistically younger children are unlikely to be using drugs. Pupils are picked at random from the roll and asked to give a urine sample. A member of staff - male for boys, or female for girls - waits outside the cubicle. They then check that the sample is fresh using a temperature gauge. They also check it’s urine and not warmed-up lucozade. (We took advice from Nottingham Prison drugs unit, who told us all the possible tricks that can be used.) The sample is then put in contact with a detection strip, and if four pink lines appear, then the test is negative.

At every stage of the test, the child signs a document to confirm that everything has been carried out correctly. The police have vetted our methods, and approved our way of doing things. If the first test is positive, we run a second test, though it usually isn’t necessary as most children confess at this point. We then call in parents, and ask the child to sign up to a two-month course of counselling and support, which involves an outside agency.

Is it really part of a teacher’s job to do this kind of thing? I don’t know, to be honest. I admit I have some doubts about that. But if we can stop one child from going down the road of becoming a drug addict, then I’m not going to worry too much about job demarcations. This is more important.

We’ve carried out nearly 400 random tests and there have been eight positives. In the first year, six pupils tested positive for cannabis.

After the course, four of them stayed completely clean for the rest of the year, the other two fitfully so. This year only two children have tested positive.

It seems odd that children who have smoked cannabis still agree to take the test, when they could just refuse. Perhaps they don’t want to attract suspicion, perhaps they think they won’t be picked up, or perhaps they want to be caught so they can get help. It’s not easy being a young person.

There are so many temptations, but at least the random tests give children an excuse to say no. If their friends ask them to try cannabis, they can say “Forget it - knowing my luck I’m bound to be tested.”

Is what children do at weekends the school’s business? Of course it is. It affects the way they behave in school, and that puts a burden on staff. The programme costs us pound;1,500 a year, and we think that is money well spent, to keep our school free of drugs and free of the aberrant behaviour that comes with drug taking.

Peter Foster is deputy head at The National school, Nottingham. He was talking to Steven Hastings