How accurate are at-home stool tests for bowel cancer screening?

A test posted to your house, done in your own bathroom, is a very appealing alternative to bowel preparation and a hospital appointment. The fair question is whether you are giving something up by choosing it.

The US Preventive Services Task Force publishes the actual numbers, so you can see the trade-off rather than guess at it.

Two measures matter. Sensitivity is how often a test correctly picks up people who do have the condition. Specificity is how often it correctly clears people who do not.

For detecting colorectal cancer, USPSTF reports:

• FIT: sensitivity 0.74, specificity 0.94
• High-sensitivity gFOBT: sensitivity 0.50 to 0.75, specificity 0.96 to 0.98
• Stool DNA-FIT: sensitivity 0.93, specificity 0.84

A sensitivity of 0.74 means roughly three in four cancers present at the time of testing would be picked up. That is genuinely useful and it is not perfect.

Notice the trade in the stool DNA test: it catches more cancers, and its specificity is lower, which means more people without cancer will get a positive result and go on to colonoscopy.

The figures for finding advanced adenomas — growths that have not yet become cancer but are on that path — are much weaker:

• FIT: 0.23
• High-sensitivity gFOBT: 0.06 to 0.17
• Stool DNA-FIT: 0.43

This is the real difference between stool testing and colonoscopy. Colonoscopy can find and remove growths before they ever become cancer. Stool tests are primarily looking for signs that something is already bleeding.

Repeating the test on schedule is what makes stool-based screening work. One round is not the programme.

Because each single round misses things, USPSTF pairs each test with an interval:

• FIT: every year
• High-sensitivity gFOBT: every year
• Stool DNA-FIT: every one to three years

Screening every year with an imperfect test catches a great deal over a decade. Screening once with the same test does not.

USPSTF is unambiguous here: positive results on stool-based screening tests require follow-up with colonoscopy for the screening benefits to be achieved.

That sentence deserves reading twice, because a positive stool test that is not followed up is close to no screening at all. A positive result is not a diagnosis. Bleeding in the bowel has plenty of causes that are not cancer, including haemorrhoids. But the only way to find out which one applies to you is to look.

If you get a positive result, book the colonoscopy. If getting that appointment is difficult, say so to your practice rather than waiting quietly.

USPSTF recommends colorectal cancer screening with a grade A for adults aged 50 to 75, and a grade B for adults aged 45 to 49. If you have a family history of bowel cancer, inflammatory bowel disease, or a previous adenoma, your situation sits outside average-risk screening and should be planned with your clinician.

The best screening test, in the end, is one that actually gets done on time. For many people that is the one that arrives in the post.

A test kit that sits in a drawer contributes nothing, and one that is done carelessly can give a misleading answer. The instructions on sample collection and on posting the kit promptly exist because the chemistry depends on them.

Two practical points come up often. Read the leaflet about medicines and foods before you collect, because instructions vary between kits. And send the sample back quickly rather than letting it sit at home for a week.

If the process defeats you, say so. Practices would far rather send a replacement kit than record a patient as not screened.

It is also worth naming the embarrassment, since it stops a lot of people. Handling a stool sample is unpleasant and mildly undignified, and everyone who does it feels that way. It is also one of the few genuinely proven ways to reduce your chance of dying from a common cancer, which is a reasonable trade for ten awkward minutes in your own bathroom.