There has been much press coverage of The Lancet Oncology’s paper this week on screening for ovarian cancer. Screening – looking for disease before a person has symptoms that suggest the disease – is often harder than it seems, thanks to the myriad problems it can create. That’s not to say that screening shouldn’t be considered or investigated as a way to try to add quality and years to life. But it is important to consider all the negatives as well as the positives when the screening is being tested – as the Lancet research is doing.

I have three concerns. First, this paper is an interim one – the study hasn’t finished, and as such, it isn’t possible to say whether or not screening for ovarian cancer will prevent deaths. However, it is a large, randomised trial, which is a good thing.

Second, the authors acknowledge that there is a degree of “over-diagnosis” going on: “44 per cent (22 of 45) of the primary ovarian cancers detected in the USS group were borderline.” (USS is ultrasound – one of the methods which was being compared as a tool to use for studying the effects of screening.) “Borderline ovarian tumours have 10-year survival rates in excess of 95 per cent… It could be argued that these cases would be best classified as false positives… Once borderline cancers are detected during screening, it is difficult not to operate given that borderline and stage I invasive ovarian cancers share common morphological features on ultrasound imaging.”

Third, we do not yet know about the full impact on health of patients who had to undergo more than one round of screening. These were cases where the first screening test was abnormal and further tests were run either right away or several weeks later. This may have an adverse impact on psychological wellbeing – some people may be made sick with worry.

None of this means that it might not be a worthwhile test to do. We have to wait for the full results of the trial.