Do you have a niggling worry that you may have osteoporosis? Perhaps you are concerned about vague symptoms of bowel cancer or an enlarged prostate. Maybe you think that you don't have time to visit the GP or, if you're honest, you are too embarrassed.

These are among the reasons why sales of health self-testing kits have risen by 55 per cent in the past five years in the UK. The DIY testing market, in which you can check yourself for anything from diabetes to chlamydia, is worth £100 million a year.

Some tests, such as a cholesterol kit, can be bought off the shelf or over the internet for as little as £7. Others can cost more than £100. But increasing numbers of people see being able to “take charge” of their health diagnosis as a price worth paying.

Yet how reliable are these tests? This year Sense about Science, a science education charity, published a report warning people that DIY health-testing kits are a waste of money, often giving inaccurate results that lead either to false reassurance or to unnecessary worry.

The report Making Sense of Testing, produced with the help of the Association of Clinical Biochemistry and the Royal College of Pathologists, points out that many of the tests being offered are available on the NHS, alongside discussion of family history and lifestyle, and advice from professionals. Samples of blood or other bodily fluids sent by post may not survive the journey to the laboratory as well as they would in a hospital. Although experts agree that there are benefits to home testing, there are calls for a national system to evaluate diagnostic tests, and a publicly accessible database of evidence of their performance and usefulness.

Alastair Kent, director of the Genetic Interest Group, says that the results of such tests should be explained “in the context of a proper clinical consultation”. But time spent talking to patients is expensive. “If a company wishes to offer a test, it should offer a full service, not cherry-pick the most profitable bits of the process,” he says.

So, are many such companies essentially “snake-oil salesmen” preying on the worried well, often by selling expensive health supplements on the back of the results? Or are they a healthy addition to the market? We tried out three of the tests...

CAROL MIDGLEY

The liver test

Smug doesn't come close to how I felt when I opened my liver-test results from BUPA.Like most journalists, who count social drinking as a necessary - nay, fundamental - part of the job, I suspected that over the years I might occasionally have “pushed the envelope” on the alcohol front. But, like many people, I put my hands over my ears and said “la la la” when I thought about it, forgoing the full private health MOT that The Times offers its employees each year and adopting the full-on ostrich position.

Then my editor told me that I had to test one of the many health home-testing kits available on the market to determine its reliability. I'd been put down under “liver” and that was that. The plan was to have a test and compare the results with those of a hospital test. In a state of fascinated fear, I submitted myself to a BUPA once-over.

Reader, my liver result was fine. Better than fine. In the four categories tested, everything was comfortably within the ideal parameters. As I bragged later over a half of Carling Extra Cold, “my liver's so healthy it could do press-ups”.

Then came the bad news.

I had also sent off for a £99 Livercheck test via the internet. It arrived with an impressive panoply of equipment in a powder-blue plastic pouch: a pen, an antiseptic wipe, a device to help to puncture the skin, and a phial in which to send a sample of blood to YorkTest's laboratories. It was quite an effort getting the blood from my thumb into the tiny test tube but I posted it off immediately, as instructed.

The results came back quite quickly.

As I stood in my hall with the opened envelope, I felt as if I'd been punched in the face. Staring at me was a large rectangle consisting of coloured blocks, with green (good) at one end, followed by yellow, amber, then red (very bad) at the other. My liver, said the impersonal letter, was in the red zone. “Your overall LiverCheck score is RED,” it informed me. “This means that your liver health is less than optimal. Even mild liver- test abnormalities may be an early clue to the presence of liver disease.” I felt sick.

There was other stuff about lifestyle changes, abstaining from alcohol and seeing my GP. Finally it said: “We would advise you to retest in one months [sic] time.”

I was bewildered (though not dismissive. Anyone who enjoys a drink secretly expects to be told that the game's up). But how could a BUPA test say one thing and the LiverCheck quite another? I examined the results more closely. LiverCheck had tested my blood for only two things: AST (aspartate aminotransferase) and ALT (alanine aminotransferase), enzymes that can leak into the blood if the liver is damaged. BUPA's test was more comprehensive, also testing Gamma GTP and bilirubin (both, by the way, low). BUPA had found my AST level to be 20 (anything up to 36 is normal). But LiverCheck found it to be 70.

I phoned the “customer care team” but was held in a queue. I gave up and phoned my GP, securing an appointment for 9am on Monday. Then began a miserable, teetotal weekend in which I obsessively Googled “liver disease” and convinced myself that I had every symptom.

On Monday, faced with my two results sheets, my GP was phlegmatic. “Well, one of these is wrong and I know which I suspect it is, but let's find out,” he said. So for the third time my liver was tested, and was found to be normal, with readings very similar to BUPA's. I was grateful, but my irritation with LiverCheck was mounting.

Sense about Science has said specifically that some tests for liver disease measure just two biomarkers, whereas an NHS laboratory would carry out at least four or five (my NHS one conducted seven).

Dr Ian Watson, president of the Association for Clinical Biochemistry, who studied my results, described the rigorous standards that NHS testing labs must meet. But with a DIY test you don't always know where, or by whom, your sample is being tested.

Dr Watson explained that AST is a more sensitive enzyme than ALT but less specific to the liver. Its levels can also be raised when body tissue or another organ, such as the heart, is diseased or damaged: ie, a high reading may be nothing to do with the liver. He asked me how I did the test and I recounted squeezing the blood from my thumb. There was a possibility, he said, that I had have damaged some tissue, and that had contributed to the high reading. Or, of course, that the results weren't even mine.

For people who want to be more engaged with their health Dr Watson recommends Lab Tests Online, which offers non-commercial patient education on blood, urine and other laboratory tests to help them to understand their tests better.

I phoned the LiverCheck customer care line again, expressing anxiety about my results, and spoke to a lovely man who advised me to see my GP and gave me the number of the Liver Trust. When I asked about retesting in a few weeks' time, as the bumph advised, he offered me another test at the “special retest price” of £74. Mmm.

I know I would never buy a home-test kit again. To be told that you are unwell in a mass-produced letter is a lonely experience. Yet now, if the next DIY test told me that I was fitter than a flea, I wouldn't believe it.

If you are feeling worried or ill, see your GP. If not, forget it. My philosophy now is to repeat my grandmother's old saying: “Don't trouble trouble till trouble troubles you”.

The Times posed ten questions to YorkTest Laboratories Ltd, in York, which sells LiverCheck. The company offered no explanation of why my results were wrong. This, in summary, was its response:

“Blood samples are tested to the same stringent standards as hospital ALT and AST tests. The sample collection method, associated stability and transport of the blood samples ... have been fully validated to meet the essential requirements of the European Medical Device Directive 93/42/EEC. The accuracy of the ALT and AST tests used by YorkTest Laboratories are 98.1 per cent and 98.8 per cent respectively. YorkTest Laboratories are members of the UK NEQAS (National External Quality Assurance Scheme; www.ukneqas.org.uk) for ALT and AST.The LiverCheck test provides a ‘first stage' for those concerned about their liver health.”

CAROL MIDGLEY

The cholesterol test

I confess that it took me two weeks to get around to carrying out this test because I wasn't keen on having to prick my finger to extract blood. I don't bleed well - getting the stuff out of a stone is quicker.

The kit does make it easy, though. All you do is twist off the cap of a plastic lancet, push a trigger and something sharp and unseen shoots into the end of your finger. You then allow a drop of blood to fall on to the test area of a calibrated card, wait for three minutes, and your drop will have turned an unattractive shade of green. You match this to one of the colours on the card and can then read your cholesterol level. Frankly, the only difficult bit is this matching. Mine seemed closest to 5.9 mmmol/L, which suggests that my cholesterol level is slightly high and should be checked by a doctor. Yet in a recent reading taken during a BUPA examination it came out much lower and was declared by a doctor to be perfect. Speedy as the home-testing kit is, £11.99 seems a lot to spend on a test that may or may not be accurate, and in future I'd rather jump through the hoops needed to get an appointment at my local surgery than treat my health as a DIY mission. I'm not a doctor or a nurse and I'm happy to trust someone who is, even if I do have to read ancient Hello! magazines as I wait my turn.

PENNY WARK

Boots Test Kit cholesterol, £11.99

The prostate test

The logic of a DIY prostate-cancer-screening test kit seems impeccable. Prostate Specific Antigen (PSA) is a protein secreted into the blood by the prostate gland. An abnormally raised level of PSA in the blood can be an indication that you have prostate cancer.

Well, I'd love to be able to regale you with the results of my prostate self-test, but I can't. I paid £11.95 online to a Stockport-based supplier called SelfDiagnosis Ltd and waited a few days for my box to arrive in the post. When it did, I tried to follow the instructions - but they defeated me. Not only did I still not know if I had a high PSA level but I felt stupid, too.

Surely Dr Hammad Malik of Same Day Doctor, a private clinic with branches in London and Manchester, could decode the kit for me. And he did. “This is very confusing,” he said, “and it won't work.”

The little gizmo that you use to prick your finger to draw blood from was of a different sort to the one illustrated in the kit's instruction book. “The instructions show the type that we use, they're automatic,” said Dr Malik. “But what you've got is a needle you have to jab yourself with - and it's quite hard to jab yourself.”

The killer blow, though, was the vial of test solution not included in my box. Without it there was no way of getting a result.

Had it worked, the test would have flagged positive if I'd had a PSA level greater than 4ng/ml. However, PSA levels increase naturally with age, so a level greater than 1.4ng/ml would have been on the high side for me. And that might not have shown up on the test - I wouldn't have known that if I hadn't spoken to Dr Malik.

So what is my real prostate cancer risk? Dr Malik took out a large needle and drew some blood from my arm. The next day I received an e-mail - my PSA levels are normal for my age. It cost £150, along with the consultation, although I could have gone to my NHS GP. Later, Dr Malik phoned. “Men are becoming more and more aware of prostate problems. A lot of chaps in the know will take a PSA test every year.” I think I will too - but I won't try to do it myself.

Terri Duigenan, from Self Diagnosis Ltd, was aghast when I contacted her: “We tend to try out all the tests ourselves. My husband has done it and he didn't have a problem with it but we have since changed suppliers. I can't apologise enough. We would always advise that if you have any worries about your health, you go to see your GP.” She is also providing me with a refund.

ANDREW LAKE

SameDayDoctor.co.uk

The doctor's verdict

What's so tough about making a diagnosis? You have symptoms. You run a test. You get an answer. Simple. Deconstruct doctoring, make self-testing kits easily available and you end up with empowered patients and less hassled medics. So everyone's a winner, right?

Wrong. It's a self-test jungle out there. And without someone hacking through the undergrowth for you, there's a real danger that you could step on a snake.

For a start, do you really need that test at all? The PSA prostate check, say, sounds intuitively sensible. But you could write a book on its pros and cons - a situation neatly encapsulated by one authority stating: “The only certainty about the PSA is that it can cause harm.”

And if you really do need a test, how can you be sure you'll choose the right one? Check out your tiredness with a DIY blood-sugar reading and you might discover you haven't got diabetes. Fine. In which case you'll miss the diagnosis of depression or an underactive thyroid. After all, there are reasons why training to be a doctor takes years and a large chunk of taxpayers' money. One is that medics learn to tailor investigations so you don't end up down diagnostic cul-de-sacs.

But you've gone ahead anyway, and now you're staring at a result. What does it mean? Good question. Leaving aside obvious quality-control issues, result interpretation is a minefield. A negative result doesn't necessarily rule out an illness, nor does a positive one confirm it. So you may end up falsely reassured, unnecessarily anxious, or simply confused.

The bottom line is this: if you're worried enough to self-test then clearly you have a health issue. Your best bet is not to play doctor according to guesswork and what's on offer this week. Instead, see someone who can make sense of your problem, put it in context and apply some knowledge. It'll be worth the wait - and cheaper, too.

DR KEITH HOPCROFT