Benjamin Geen was a 25-year-old staff nurse at Horton General Hospital in Banbury, Oxfordshire, when he was arrested on February 9, 2004. In April 2006 a jury at Oxford Crown Court found him guilty of two murders and of grievous bodily harm to other patients, all in the hospital's emergency department between December 2003 and February 2004. He was given a life sentence with a minimum term of at least 30 years. He has said throughout that he is innocent.
His is one of the British cases in which the proof of guilt rested heavily on a count of events: eighteen patients who stopped breathing or whose breathing failed in a period of about three months, all during shifts he worked. A small group of statisticians, doctors and lawyers has argued for more than a decade that the count was an artifact of how the cases were chosen and counted. The Court of Appeal rejected his appeal in 2009 and the Criminal Cases Review Commission has declined to send the case back. The campaign for review continues.
Little of Geen's life before 2004 is on record. He was a staff nurse in the emergency department at Horton General Hospital, a small hospital in the market town of Banbury, in north Oxfordshire. Jon Robins, writing in Byline Times in 2019, described him as a 25-year-old who was the only male nurse in the department and a lieutenant in the Territorial Army. Colleagues, Robins wrote, regarded him as "a bit gung-ho."
Robins also reported that Geen complained several times in the weeks before his arrest about understaffing in the department and was reprimanded by the chief nurse; the hospital denied the understaffing claim, according to a 2020 account in The Justice Gap. The same account recorded that emergency admissions had roughly doubled, from about 400 a month in 1999 to about 800 by February 2004, before falling sharply.
The prosecution case depended on how unusual the winter of 2003 and 2004 was in the emergency department. According to the statisticians who later studied the records, the hospital's own dossier was assembled over a single weekend in February 2004 by a team led by the department's head nurse. It listed about 30 incidents, all from December 2003 onward. The team then looked at the shifts Geen had worked and narrowed the list to 18 cases, which were passed to the police. The statisticians say the team did not examine records from periods when Geen was absent.
The bar chart prepared for that review combined three categories of event: cardio-respiratory arrests, respiratory arrests and a third category of collapse. Later figures obtained under freedom of information rules kept the categories apart. The authors of a 2022 paper in the journal Laws argue that this combination was misleading and that the mix of cases changed in the winter of 2003 to 2004, so that nearly all were classed as respiratory arrests. They say cardio-respiratory arrests are normally about five times as common as respiratory arrests in such a department, and that the ratio reversed in the weeks before the arrest. Richard Gill, one of the paper's authors, described what he saw as a large reclassification of events in the days after two collapses on February 5, 2004. These are the authors' arguments, and the court heard a different account.
On Thursday, February 5, 2004, two patients collapsed in the emergency department. Geen reported sick the next day, according to Robins. Over the following weekend, senior hospital staff and doctors brought in from the John Radcliffe Hospital in Oxford went through patient notes and identified 18 patients with unexplained collapses in a period beginning in December 2003. Geen had been on duty for each.
The Justice Gap reported in 2014 that a former medical director at the John Radcliffe, Dr Malcolm Benson, reviewed about 5,000 sets of patient notes. He is quoted as saying that it was the unexplained nature of the arrests, not their number, that convinced him crimes had been committed.
Geen was arrested on Monday, February 9, 2004.
Geen was tried at Oxford Crown Court before Mr Justice Crane. The trial began in February 2006 and, according to The Justice Gap and Byline Times, lasted about three months. He was charged with two counts of murder and with counts of grievous bodily harm. The prosecution said that in each case a patient had suffered a respiratory arrest or a failure of breathing that had no natural explanation, that Geen was the common factor, and that this "unusual pattern" pointed to deliberate action. Robins's 2016 article records the prosecution evidence that respiratory arrest in an emergency department was extremely rare; the defence disputed the reading of the medical tests.
The prosecution also put forward a motive. According to reports of the verdict, the court heard that Geen was a thrill-seeker who wanted to satisfy what was described as a "lust for excitement." Geen's solicitor said at the time that he maintained his innocence.
The defence asked to call two experts, Professor Jane Hutton, a medical statistician at the University of Warwick, and Professor David Denison, a clinical physiologist. The authors of the 2022 Laws paper write that the judge did not allow Hutton or Denison to give their evidence in person; the paper says the judge told the jury that the written evidence "consisted merely of common sense." The judge also told the jury to consider each of the 18 cases entirely on its own merits.
Medical News Today reported on April 18, 2006 that the jury had found Benjamin Geen guilty of causing the deaths of two patients and of harming 15 others, and that the judge wanted a psychological report before sentencing. According to the Laws paper, the verdicts were guilty on two counts of murder and on 15 counts of grievous bodily harm; a 16th count of grievous bodily harm was not considered proven. The paper says the jury found that he had deliberately harmed 17 patients and that in two cases the harm led to early deaths. The statutory wording of the grievous bodily harm counts is not on record.
On May 10, 2006, Mr Justice Crane sentenced Geen to life imprisonment, with a minimum term of at least 30 years, the figure given in contemporary reporting and in later accounts in The Justice Gap and The Spectator. A 2019 article in Byline Times gives a minimum of 38 years; that figure does not match the other accounts.
Geen's lawyers took the statistical case to the Court of Appeal in November 2009. The defence put forward a report by Professor Hutton, written at the request of his lawyers, which said that the information presented at trial was of inadequate quality and had no value in showing either that there was an unusual pattern or that any pattern was not the work of chance. Nigel Hawkes wrote in Full Fact's Straight Statistics in April 2010 that Hutton compared five incidents in December 2002, before Geen worked at the hospital, with six in December 2003. She described a risk she called diagnostic suspicion bias, in which unexpected events on a suspect's shifts are attributed to him without a search for other causes.
The court refused to hear Professor Hutton's evidence. Lady Justice Hallett, giving the judgment, held that the jury was entitled to conclude that the pattern resulted from deliberate action and not from chance. Hawkes's report records the appeal as rejected in November 2009. Druin Burch wrote in The Spectator in December 2024 that the Court of Appeal described the evidence against Geen as "overwhelming."
After the appeal, Geen's supporters turned to a wider circle of statisticians.
In 2016 Robins reported that the application to the Commission was backed by Professor David Hands, Professor Stephen Senn, Sheila Bird, Sir David Spiegelhalter and Professor Fenton. Spiegelhalter gave the example of three aircraft crashes in eight days, which he said had a 60 percent chance of occurring in some ten-year period, and warned against choosing historical data to fit a coincidence. Geen's barrister, Mark McDonald, said that "rarity" meant nothing without a statistical basis and that only Geen's patients' notes had been examined, not those of other nurses or other patients.
Richard Gill of Leiden University, who worked without fee as a defence expert on an initial review attempt in 2014 and 2015, called the proof circular: the hospital's medical experts had judged the events suspicious partly because Geen was present. In 2020 The Justice Gap reported that Gill, Fenton, Professor David Lagnado and Professor Martin Neil had prepared a paper using 13 years of admissions data, and that Gill described a large reclassification of events in the days after the two collapses. The paper was published in the journal Laws in August 2022. In it the authors say that Gill is personally convinced Geen is innocent.
The prosecution side of the dispute rests on the court record. At trial the evidence included a prosecution exhibit and the medical experts' opinion that the unexplained collapses were exceedingly rare. In 2014 The Justice Gap reported that the Criminal Cases Review Commission had at one stage described "a cogent and compelling body of evidence" of guilt, in particular that exhibit and the timing of some patients' deterioration. The 2022 paper notes that experts often gave conflicting explanations for the same collapses.
Geen's family has also pointed to medical evidence. In 2016 the family said it planned to submit a report from a medical expert, Professor Vincent Marks, who it described as an expert in the area, to the effect that the two patients who died did so from natural causes. The prosecution case at trial was the opposite, and the jury's verdicts found that both deaths resulted from deliberate harm. The Court of Appeal in 2009 did not disturb those findings.
The Criminal Cases Review Commission, which can refer a conviction back to the Court of Appeal, had Geen's case for years. Robins reported in March 2016 that the family had been dealing with it for about two years, and Byline Times recorded in March 2019 that the case had been with it for almost six years.
In October 2015 the Commission decided not to refer the case. Geen's lawyers, Stephen Knight and Mark McDonald, applied for judicial review. On March 23, 2016, the High Court granted permission; Mr Justice Gilbart held it was arguable that there were failings in the Commission's consideration of the application. One Pump Court, the chambers acting for him, reported the decision on April 15, 2016. On April 14, 2016 the Commission reallocated the case with immediate effect, which Robins reported ahead of the usual wait of about eight months. The 2022 paper states that the earlier rejection was challenged in court and withdrawn, and that the Commission had said the application presented only statistical evidence and did not deal again with the medical evidence from the trial.
In April 2020 the Commission was expected to decide within weeks whether to refer the case. The date and wording of the decisions after 2016 are not on record. By 2022 the Laws paper stated that all attempts to reopen the case had so far been unsuccessful, and in December 2024 Burch wrote that three applications to the Commission had been rejected. No referral to the Court of Appeal is on record.
In March 2019 Geen was reported to be held at HMP Long Lartin, in Worcestershire, and in April 2020 at HMP Frankland, in County Durham. He was serving a life sentence with a minimum term of at least 30 years. His family has campaigned on his behalf for more than ten years, according to The Justice Gap.
The verdicts of April 2006 and the sentence of May 10, 2006 stand. The Court of Appeal rejected his appeal in November 2009, and no referral by the Commission is on record. Geen has always denied the charges. Statisticians and doctors who have examined the figures say the cluster was not unusual and that diagnostic suspicion bias shaped the case; the Court of Appeal held that the jury was entitled to find the pattern the result of deliberate action. The count of events, the classification of them and the hospital's weekend review remain the central points of the argument. The figures for December 2002 and December 2003 differ between accounts: five and six in one, four and four in another, five and seven in a third.
December 2003 to February 2004: the deaths and collapses the prosecution charged take place at Horton General Hospital, Banbury.
February 5, 2004: two patients collapse in the emergency department.
February 9, 2004: Geen is arrested.
February 2006: trial begins at Oxford Crown Court before Mr Justice Crane.
April 18, 2006: reports of the guilty verdicts.
May 10, 2006: life sentence with a minimum term of at least 30 years.
November 2009: Court of Appeal rejects the appeal.
October 2015: the Criminal Cases Review Commission decides not to refer the case.
March 23, 2016: High Court grants permission for judicial review of that decision.
April 14, 2016: the Commission takes the case up again.
April 2020: a Commission decision is reported as expected.
August 2022: statisticians' paper on the case published in Laws.
December 2024: three applications to the Commission are reported as rejected.