How accurate was the test?

Key points

• Around one in 100 people who had the blood test received a positive test result. 
• Around half of people who had a positive result were diagnosed with cancer.
• Around a third of people who had cancer got a positive test result. 
• Nearly everyone who did not have cancer got a negative test result. This means that the test gave a false alarm less than 0.5% of the time. 

One of the aims of the NHS-Galleri trial was to measure how accurate the blood test was at screening for cancer.

One way to measure the accuracy of a screening test is to look at whether people who have a ​‘positive’ test result are diagnosed with cancer.

Another way of measuring how well a screening test works is to look at how good it is at finding people with cancer and those who do not have cancer. This is important because a screening test needs to find cancer, and also to avoid false alarms in people who do not have cancer.

What the researchers looked at

To measure the accuracy of the blood test, the researchers looked at only people who had the blood test (test group). They compared how many people got a cancer signal detected (‘positive’) test result and how many people got a no cancer signal detected (‘negative’) test result. 

The researchers looked at several different measures of test accuracy.

What the results show

The results showed that less than 1% of people had a positive test result. Around half of people who had a positive result were diagnosed with cancer. 

This means that if 200 people had the test, two would get a positive test result. Of those two people, one would be diagnosed with cancer.

The results also showed that around a third of people who had cancer got a positive test result. Nearly everyone who did not have cancer got a negative test result. This means that the test gave a false alarm less than 0.5% of the time.

What the results mean

For an individual person, the important measures of test accuracy are how likely they are to have a positive test result and how likely that a positive test result means that they actually have cancer.

A small proportion of people in the trial who had the blood test got a positive test result. This is because cancer is not common in people who do not have symptoms. When someone got a positive test result, they were more likely than not to have cancer diagnosed.

Test accuracy can also be measured on a more general level by looking at how good a test is at finding people who have cancer and how good it is at finding people who do not have cancer.

The test was very unlikely to give a false alarm in people who did not have cancer. This helps to prevent people from having further tests that they do not need.

These results show how the test performed in the trial and give an indication of how it might perform in a wider population. These results are similar to what was found in earlier research with the test.