Margaret Mccartney:
We want IT systems fit for purpose – and to give patients the information they need to make choices
Technology in the NHS is both blessing and curse. We are reliant on it for communication, records, referrals and organisation. But despite the millions lavished on it, it is often so glitchy and difficult to use that it drives me to tears of frustration. Now an IT mistake is being blamed for an error that resulted in 450,000 women in England aged between 68 and 71 not being invited for breast screening over the past nine years.
Why didn’t women notice they were overdue? Why didn’t GPS realise? Well, I wouldn’t have noticed, either as patient or doctor. Screening programmes are enormous undertakings. Invitations and recalls aren’t organised through individual GP surgeries but centrally by the NHS. Given the need to audit and quality assure the programmes, this is vital.
But it is also confusing. Breast screening is organised geographically – in my area, practice by practice – and the first invitation arrives between the age of 50 and 53. How can you know an invitation is late when you didn’t know when it should arrive? As for GPS noticing: well. The pressures in general practice are no secret. The massive volume of correspondence is shared out between all the doctors, and even if I did notice it was more than three years since the last batch of results, it’s easy to assume that someone else picked it up.
Despite the enthusiasm for innovation and shiny gadgets in the NHS, there are still basic problems. We humans are often bad at tracking time. Wouldn’t it be nice to have a text message reminder that the follow-up colonoscopy recommended five years ago is due? Or a reminder that your annual blood test to monitor your medication falls in June? Millions have been spent setting up failed IT systems. If we spent a fraction of that on making electronic systems fit for purpose, we could have prioritised the simple things – like checking tests we expected to be done actually were.
Yet, even if the technology were perfect, are we giving women the information they need about the efficacy of screening?
Jeremy Hunt told Parliament that the “best estimate” was that “there may be between 135 and 270 women who have had their lives shortened as a result” of this error. I am not so sure. Breast cancer screening does reduce deaths from breast cancer. The 2012 Marmot review found that, for every 10,000 women invited, over 20 years, 681 breast cancers would be diagnosed, and 43 deaths from breast cancer would be prevented.
But there is a downside. The report found that, of those cancers, 129 were over-diagnosed. Over-diagnosis is a counter-intuitive concept. These are real cancers, but do not behave like a life-threatening disease. But because screening cannot tell us whether a cancer is life threatening or overdiagnosed, all are offered treatment. And even well-intended treatments do harm, through side effects of chemotherapy or surgery. So we need to know the “all cause mortality” – whether the total number of deaths is reduced because of breast screening. Hunt may have been quoting breast cancer deaths, not deaths overall – a different statistic.
Additionally, since three times as many women have an over-diagnosed cancer relative to the number of women who don’t die of breast cancer because of screening, the IT failure means that an estimated 405 to 810 women avoided being over-diagnosed and given treatment they couldn’t benefit from. Women need all the information to reckon with their own risk of being invited – or not.
Even more confusingly, this error seems to have involved some women invited because they were part of the Age Extension Study. Breast screening has usually involved women between 50-70. The Age Extension Study expanded the ages to 47-50 and 70-73. We can’t say whether screening in the over 69s is beneficial at all, but it seems absurd to tell women in that age group that not having screening will have been harmful. The truth is, we just do not know.
We keep hearing about personalised medicine and the benefits of technology – but this is nothing without the basics. We need IT systems designed around humans, and honest conversations where we hear about the risks as well as benefits.