An MP has called for an independent investigation into the suspension of children's heart surgery at Leeds General Infirmary after Computer Weekly revealed claims that the decision had been based on a misuse of data.
Medical researchers say the data used by NHS medical director Bruce Keogh to justify the closure was incomplete and was quickly proved inaccurate.
It has also emerged that Leeds General Infirmary was asked to suspend the heart surgery unit without seeing the full data that purported to show a dramatic increase in deaths of children undergoing surgery at its specialist heart unit.
Computer Weekly has learned that Keogh pressed hospital managers and surgeons to stop children's heart operations immediately, on 28 March 2013, after showing them a paper graph he had extracted from a report of health data that insiders claim was incomplete.
Managers at Leeds General Infirmary were left to find the complete data themselves, according to a source close to staff attending the meeting. By the time they found the dataset, around 10 children had been moved to hospitals hundreds of miles away for urgent operations.
Questions in the Commons
Greg Mulholland, Liberal Democrat MP for Leeds North West (pictured), called for an investigation into Keogh's claim that the mortality data left him no choice but to suspend heart operations at the Leeds General Infirmary unit.
Greg Mulholland cited Computer Weekly's revelations in the House of Commons last week, quoting an earlier article concerning the purported misuse of data at Leeds General Infirmary: "Keogh, first by requesting unverified data, then by ignoring the data's obvious faults, then by breaching usual procedure in not seeking to clarify those faults, then apparently wilfully misinterpreting that data, and then using the weight of all these errors to tell the Leeds hospital to close its heart unit, may have committed a serious breach of protocols he had helped establish."
He then asked: "Why is the secretary of state for health still refusing to have an independent investigation into this matter?"
Mulholland was referred to health secretary Jeremy Hunt's last statement on the matter. Hunt had insisted Keogh made the right decision and said arguments over data quality should not stop services being closed when it indicated there may be risks to patients.
"Sir Bruce took the entirely appropriate decision to suspend children’s heart surgery while further investigations were made," Hunt said.
Hunt had blamed the scandal on Roger Boyle, an academic who took over from Keogh in 2007 as head of the National Institute for Cardiovascular Outcomes Research (Nicor), the unit of University College London that produced the data Keogh purportedly misused. Boyle subsequently resigned in April, after Hunt sacked him from a national review of heart surgery.
Mulholland's call for an independent investigation is based on claims that what happened at Leeds General Infirmary on 28 March 2013 was not as Keogh claimed when he explained his decision on the BBC Radio 4 Today programme.
Computer Weekly understands that on 27 March 2013 Keogh breached data safeguards to ask Boyle to send preliminary Nicor data he could use to single out Leeds General Infirmary.
David Cunningham is the Nicor senior strategist responsible for the data Keogh used. He said Boyle sent Keogh an anonymised copy of preliminary data on the day Nicor's expert committee received it for quality checking. Late that afternoon Keogh asked Boyle to send unanonymised preliminary data he then used to single out Leeds, said Cunningham. The data had been sent to experts on Nicor's quality committee for a routine check of its accuracy, but they were not permitted to share it until it had been verified. They very quickly saw it was inaccurate, said Cunningham.
Keogh travelled to Leeds General Infirmary the next morning with a copy of the inaccurate data. According to the Leeds source, at 8:00am Keogh laid a graph out on a table before Maggie Boyle, chief executive at Leeds teaching hospitals; Stacey Hunter, the general manager of the women and children division at Leeds General Infirmary; and associated cardiologists.
Compliance and doubt
Computer Weekly understands the graph came from the spreadsheet of data sent to Nicor's data quality committee and had been clearly marked as preliminary, although it is not clear if the copy of the graph shown to managers at Leeds General Infirmary was marked as such. Keogh had obtained a copy of just one page of the preliminary report that purported to show how children had died while undergoing heart surgery at Leeds General Infirmary at twice the national average rate.
Keogh told them the data meant they should suspend the heart unit. Hospital managers complied for the sake of safety. But they were doubtful because Keogh had not shown them any real data, said the source. What the graph told them also conflicted with what they knew about their own hospital - that there had been no dramatic increase in deaths of children undergoing heart operations.
The printed graph, later published on the website of the NHS Commissioning Board - since renamed NHS England - where Keogh works, was shorn of any reference to its preliminary nature.
Keogh took the graph away with him, said the source. Leeds General Infirmary hospital experts were forced to find the source data themselves, which they got from Nicor. They saw immediately there was something wrong with it, and confirmed the data was incomplete before the end of the day.
A spokeswoman for NHS England said Keogh was not ready to answer questions about the decision. Computer Weekly asked for answers about the story Greg Mulholland read out in Parliament.
Computer Weekly asked why Keogh had been unable to say how – with his experience as founder of Nicor and his familiarity with the procedures around the collection and verification of Nicor data – he did not see what other Nicor experts said was the data's obvious flaws. Nor was Keogh able to say why he had not waited for routine verification before acting on the data.
It is also unclear why Keogh sought to use death statistics that were about to be routinely verified by Nicor data experts. It remains to be explained why he would then have concealed all but a single, paper print-out from managers at the hospital – and persuade them to suspend operations – when the Nicor experts quickly verified the data was wrong. It is also unclear why Keogh blamed the hospital for supplying incomplete data when routine collection was still being completed.
Use of data in the NHS
A spokeswoman for Nicor would not clarify how much of its data collection and verification procedures had been influenced by Keogh himself. The medical director had been a pioneer in the development of heart surgery in work that had grown from the mid-1990s into Nicor itself.
Nicor was conducting its own investigation of what happened and would not comment until it was complete, she said.
The incident is the latest in a string of reports questioning the use of data that have stemmed from the Francis Inquiry into deaths of patients at Mid-Staffordshire NHS Foundation Trust. NHS England is increasingly using data to check on the performance of hospitals across the health service.