“Most ovarian cancer victims face delays in diagnosis that can kill”, “Ovarian cancer is being missed”, “Women with ovarian cancer dying because GPs fail to spot signs” – so shout the headlines this week, based on a new study published this week in the British Medical Journal.
But that’s not the whole story – let’s have a look at what the study’s really about.
It’s true that ovarian cancer is often diagnosed at a late stage, when survival is lower. So scientists are looking for ways to spot the disease earlier, and by doing that, help more women to beat the disease. But how?
We blogged a while ago about a big trial of screening for ovarian cancer, which is still ongoing. Although the early results look promising, they don’t provide enough evidence to conclude – yet – that screening would save lives, and do more good than harm. So for now, doctors can’t rely on screening tests to diagnose the disease earlier. They usually have to assess women based on their symptoms which as we’ll see, isn’t that easy.
Looking for a needle in a haystack
For a long time, doctors and scientists thought that ovarian cancer didn’t have any symptoms, earning the disease the infamous name ‘the silent killer’. Since then, further research has shown that the disease DOES in fact have symptoms – as Kat wrote last year. But because these symptoms are also very common among women who don’t have the disease, picking out the few people who actually need further tests is a bit like looking for a needle in a haystack.
When a GP thinks that a woman might have ovarian cancer, she’s referred to a specialist for more tests to find out whether or not she has the disease. But getting the balance right between making sure women who do have cancer are investigated promptly, and not putting healthy women through the stressful process unnecessarily, can be very difficult.
That’s why the National Institute for Clinical Excellence (NICE) produce guidelines for a number of different cancers, telling GPs which symptoms should be investigated for cancer, and how quickly. They can be used to help weigh up which women should have further investigations, and to help GPs refer the women who do have ovarian cancer, but not those who don’t.
These guidelines are partly based on what’s called the ‘positive predictive value’ of symptoms, which is simply the chance that someone with the symptom actually has the disease. So if a symptom had a predictive value of 5%, then out of every 100 women who had the symptom, 5 would eventually turn out to have the disease.
What’s the new research about?
The latest study, by a team of doctors at Bristol University, works out the predictive values of various different symptoms of ovarian cancer. The team have also done the same sort of thing for bowel, lung, prostate and brain cancers in the past.
In the new study, they looked at the medical records of more than 200 women with ovarian cancer, and more than 1,000 healthy women, to see what symptoms they had reported to their doctor. This is a more reliable method than interviewing patients after they’ve been diagnosed, because it’s much less prone to recall bias. There are still limitations in this method, though, because doctors sometimes record things in different ways and they may go into more detail if they suspect ovarian cancer.
They compared the symptoms reported by women who had ovarian cancer, and women who didn’t, and then some hefty number-crunching gave them the positive predictive values of symptoms associated with the disease, both alone and in combination with other symptoms.
All in all, there were seven symptoms associated with ovarian cancer:
- abdominal pain,
- change in urinary frequency,
- loss of appetite,
- abdominal distension,
- bleeding from the vagina after the menopause, and
- bleeding from the back passage.
And they found that the symptom with the highest predictive value for ovarian cancer was ‘abdominal distension’ – a persistent, progressive increase in the size of the tummy, which is different from the ‘off-and-on’ bloating that lots of women experience at some point. They found that the predictive value for this was 2.5 per cent – meaning that on average, 2.5 out of every hundred women with abdominal distension went on to be diagnosed with ovarian cancer.
What does it all mean?
The great news is that this new research gives a firmer foundation of evidence for doctors to use when they’re deciding which women to refer for further investigations. Interestingly, abdominal distension is not in the NICE referral guidelines at the moment, so including it in future editions of the guidelines could mean that more women are referred and diagnosed earlier.
Research like this is very important in developing reliable referral guidelines, as it allows experts to decide how likely people are to have cancer if they have certain symptoms, and so what to recommend to GPs.
So, far from criticising GPs for missing cancers, it’s important to remember that doctors need to decide who to refer for more testing based on experience and good quality evidence, like this, about what people’s symptoms really mean. And for ovarian cancer, this evidence has just got firmer.