A disclosure in the United Kingdom has sparked a heated debate about the health impacts of an errant algorithm
Nearly half a million elderly women in the United Kingdom missed mammography exams because of a scheduling error caused by one incorrect computer algorithm, and several hundred of those women may have died early as a result.
Last week, the U.K. Health Minister Jeremy Hunt announced that an independent inquiry had been launched to determine how a “computer algorithm failure” stretching back to 2009 caused some 450,000 patients in England between the ages of 68 to 71 to not be invited for their final breast cancer screenings.
The errant algorithm was in the National Health System’s (NHS) breast cancer screening scheduling software, and remained undiscovered for nine years.
“Tragically, there are likely to be some people in this group who would have been alive today if the failure had not happened,” Hunt went on to tell Parliament. He added that based on statistical modeling, the number who may have died prematurely as a result was estimated to be between 135 and 270 women.
Next of kin of those women who missed their final exams and subsequently died from breast cancer would be contacted, Hunt said. The government would offer a formal apology, and determine whether the families are eligible for compensation.
Hunt’s announcement spread across the British press immediately, with the word “scandal” featured conspicuously in stories. “CONDEMNED TO DEATH… BY AN NHS COMPUTER,” read one tabloid headline. The estimate of 270 premature deaths also began to morph from a statistical probability into an established “fact” in many stories. Stories featured angry family members who had lost a loved one to breast cancer, and called for those responsible to be held accountable.
Further fueling the public outrage were unconfirmed reports that the error, which apparently was found during an update to the screening algorithm, was uncovered early last year if not before. The error was at first thought to affect only a small number of women. It wasn’t until early this year that its full scope was finally understood by health officials at the NHS and Public Health England, which is responsible for breast cancer screening in England. Government ministers were informed of the situation in March, but the public wasn’t told until this month.
The government says there are some 309,000 woman still alive who missed their final screening. Those women under the age 72 will receive an appointment
for a makeup mammogram within the next six months or so, while women over 72 can call a special government hotline to see whether they would benefit from one.
However, the hotline, run by the services company Serco, is manned by staff with little more than an hour of medical training. Hunt defended their lack of training, saying that the hotline had to be set up quickly, and anyway, callers had other resources to turn to help answer their questions. More than 14,000 calls have been placed to the hotline to date.
Making the whole situation even more of a mess is that Hunt’s statement of between 135 and 270 possible early deaths has been vigorously challenged by a host of cancer experts. They argue the statistical basis for Hunt’s early death claims are dubious, with some saying that as many as 800 women who weren’t screened may actually be better off for it, because they avoided harm from unnecessary cancer treatments caused by over-diagnosis.
The experts also questioned the need for makeup mammograms since the efficacy of breast cancer screening for those above 70 is unclear. Unsurprisingly, many of the families of those who died from breast cancer, as well as women who weren’t called for their final screenings, were not convinced.
Exactly how the “algorithm failure” came about is clouded in controversy and contradictions. Right now, the NHS, PHE, and the software company Hitachi Consulting which maintains the software are pointing fingers at each other. However, it may eventually turn out that there is no fault to be assigned.
At least for now, the government is blaming “administrative incompetence” for causing the mammogram invitation omissions. The investigation into the how, why, and when it occurred will hopefully shed more insight into the matter when a report is released in about six months.
Interestingly, a few months ago, two prominent British computer scientists, Harold Thimbleby and Martyn Thomas, publicly cautioned that buggy and low quality NHS computer systems could be responsible for up to 900 patient deaths a year. They called for an in-depth study into the harm almost assuredly happening to patients by NHS’s use of inadequate and outdated IT systems.
The NHS responded to the warning raised by Thimbleby and Thomas by saying it did not recognize their figures. It further huffed that “patient safety is our priority, and our £4.2 billion investment in technology will help eliminate avoidable harm.”
No doubt now the NHS will revise its investment figure to include the £100 million that many lawyers feel the government will need to pay out in compensation for this latest case of avoidable harm.
Source: IEEE Spectrum Computing