Skip to content

The Ben Geen case: Did a naive interpretation of a cluster of cases send an innocent nurse to prison until 2035?

In a paper called “Rarity of Respiratory Arrest,” Richard Gill writes:

Statistical analysis of monthly rates of events in around 20 hospitals and over a period of about 10 years shows that respiratory arrest, though about five times less frequent than cardio-respiratory arrest, is a common occurrence in the Emergency Department of a typical smaller UK hospital.

He has lots of detailed and commonsensical (but hardly routine) data analysis. Those of you who read my recent posts (here and here) on the World Cup might be interested in this, because it has lots of practical details of statistical analysis but of a different sort (and it’s a topic that’s more important than soccer). The analysis is interesting and clearly written.

But the subject seems pretty specialized, no? Why am I sharing with you an analysis of respiratory arrests (whatever that is) in emergency departments? The background is a possibly huge miscarriage of justice.

Here’s how Gill told it to me in an email:

I’m wondering if you can do anything to help Ben Geen – a British nurse sentenced for 30 years for a heap of crimes that were not committed by anyone. Show the world that in the good hands, statistics can actually be useful!

There are important statistical issues and they need to be made known to the public.

There is in fact an international epidemic of falsely accused health care serial killers.

In Netherlands: Lucia de Berk
In the UK: Colin Norris, Ben Geen
In Canada: Susan Nelles
In the US:

So I [Gill] got involved in the Ben Geen case (asked by defence lawyer to write an expert statistical report). This is it

I became convinced that most of what the media repeated (snippets of over the top info from the prosecution, out of context, misinterpreted) was lies, and actually the real evidence was overwhelmingly strong that Ben was completely, totally innocent.

Here’s the scientific side of the story. It’s connected to the law of small numbers (Poisson and super Poisson variation) and to the Baader-Meinhof effect (observer bias). And to the psycho-social dynamics in a present-day, dysfunctional (financially threatened, badly run) hospital.

In the UK Colin Norris and Ben Geen are in 30 year sentences and absolutely clearly, they are completely innocent. Since no one was murdered there never will be a confession by the true murderer. Because there were no murders there won’t ever be new evidence pointing in a different direction. There will never be a new fact so the system will never allow the cases to be reviewed. Since the medical profession was complicent in putting those guys away no medical doctor will ever say a word to compromise his esteemed colleagues.

What is going on? why this international epidemic of falsely accused “health care serial killers”?

Answer: in the UK: the scare which followed Shipman triggered increased paranoia in the National Health Service. Already stressed, overburdened, underfunded … managers, nurses, specialists all with different interests, under one roof in a hospital … different social classes, lack of communication

So here’s the ingredients for a Lucia / Ben / Colin:

(1) a dysfunctional hospital (chaos, stress, short-cuts being taken)
(2) a nurse who is different from the other nurses. Stands up in the crowd. Different sex or age or class. More than average intelligence. Big mouth, critical.
(3) something goes wrong. Someone dies and everyone is surprised. (Why surprised: because of wrong diagnosis, disinformation, ….)
(4) Something clicks in someone’s mind (a paranoid doctor) and the link is made between the scary nurse and the event
(5) Something else clicks in … we had a lot more cases like that recently (eg. the seasonal bump in respiratory arrests. 7 this month but usually 0, 1 or 2)
(6) The spectre of a serial killer has now taken possession of the minds of the first doctor who got alarmed and he or she rapidly spreads the virus to his close colleagues. They start looking at the other recent cases and letting their minds fall back to other odd things which happened in recent months and stuck in their minds. The scary nurse also stuck in their mind and they connect the two. They go trawling and soon they have 20 or 30 “incidents” which are now bothering them. They check each one for any sign of involvement of the scary nurse and if he’s involved the incident quickly takes on a very sinister look. On the other hand if he was on a week’s vacation then obviously everything must have been OK and the case is forgotten.
(7) Another conference, gather some dossiers – half a dozen very suspicious cases to report to the police to begin with. The process of “retelling” the medical history of these “star cases” has already started. Everyone who was involved and does know something about the screw-ups and mistakes says nothing about them but confirms the fears of the others. That’s a relief – there was a killer around, it wasn’t my prescription mistake or oversight of some complicating condition. The dossiers which will go to the police (and importantly, the layman’s summary, written by the coordinating doctor) does contain “truth” but not the *whole truth*. And there is lots of truth which is not even in hospital dossiers (culture of lying, of covering up for mistakes).
(8) The police are called it, the arrest, there is of course an announcement inside the hospital and there has to be an announcement to the press. Now of course the director of the hospital is in control – probably misinformed by his doctors, obviously having to show his “damage control” capacities and to minimize any bad PR for his hospital. The whole thing explodes out of control and the media feeding frenzy starts. Witch hunt, and then witch trial.

Then of course there is also the bad luck. The *syringe*, in Ben’s case, which clinches his guilt to anyone who nowadays does a quick Google search.

This is what Wendy Hesketh (a lawyer who is writing a book on the topic) wrote to me:

“I agree with your view on the “politics” behind incidences of death in the medical arena; that there is a culture endorsing collective lying”

“Inquries into medico-crime or medical malpractice in the UK see to have been commandeered for political purposes too: rather than investigate the scale of the actual problem at hand; or learn lessons on how to avoid it in future, the inquiries seem designed only to push through current health policy”

“The “Establishment” want the public to believe that, since the Shipman case, it is now easier to detect when a health professional kills (or sexually assaults) a patient. It’s good if the public think there will never be “another Shipman” and Ben Geen and Colin Norris being jailed for 30 years apiece sent out that message; as has the string of doctors convicted of sexual assault but statistics have shown that a GP would have to have a killing rate to rival Shipman’s in order to have any chance of coming to the attention of the criminal justice system. In fact, the case of Northumberland GP, Dr. David Moor, who openly admitted in the media to killing (sorry, “helping to die”) around 300 patients in the media (he wasn’t “caught”) reflects this. I argue in my book that it is not easier to detect a medico-killer now since Shipman, but it is much more difficult for an innocent person to defend themselves once accused of medico-murder.”

Indeed, the rate of serial killers in the UK’s National Health Service must be tiny and if there are good ones around they won’t even be noticed.
Yet is is so so easy in a failing health care organization for the suspicion to arise that there is one around. And once the chances are aligned and the triggering event has happened there is no going back. The thing snowballs. The “victim” has no chance.

Chance events are clustered!!! Pure chance gives little bunches of tightly clustered events with big gaps between them. When chances are changing (e.g. seasonal variation, changes in hospital policies, staffing, new personel with new habits when filling in woefully inadequate diagnosis forms) then the phenomenon is stronger still!

eg three airliners crashed within a couple of days this week!!!
How odd is a cluster of cases? Well by the law of *small* numbers (Poisson and even super-Poisson variation – Poisson means pure chance .. super-Poisson means pure chance but with the “chance per day” slowly varying in time) “short intervals between crashes are more likely than long ones”. (actually – very short, and very long, intervals, are both common. Pure chance means that accidents are *not* uniformly spread out in time. They are clustered. Big gap, cluster, biggish gap, smallish cluster… that’s pure randomness!!!)

Then there is the Baader-Meinhof phenomenon
[I replaced an earlier link for this which pointed to a flaky news site — AG]
“Baader-Meinhof is the phenomenon where one happens upon some obscure piece of information– often an unfamiliar word or name– and soon afterwards encounters the same subject again, often repeatedly. Anytime the phrase “That’s so weird, I just heard about that the other day” would be appropriate, the utterer is hip-deep in Baader-Meinhof.”

Another name for this is *observer bias*. You (a medical doctor having to fill in a diagnose for a patient in a standard form, which is totally inadequate for the complexity of medicine) saw one case which they had to give a rather unusual label to, and the next weeks that “unusual diagnosis” will suddenly come up several times.

Well, Professor Jane Hutton (Warwick university, UK) wrote all these things in her expert report for the appeal 6 years ago but the judge said that such kind of statistical evidence “is barely more than common sense” so refused the request for her to tell this common sense out loud in court.

OK, it’s me again. I haven’t looked at this case in any detail and so can’t add anything of substance to Gill’s analysis. But what I will say is, if Gill is correct, this example demonstrates both the dangers and the potential of statistics. The danger because it is statistical analysis that has been used to convict Geen (both in court and in the “court of public opinion” as measured by what can be found on Google). The potential because a careful statistical analysis reveals the problems with the case (again, I’m relying on Gill’s report here; that is, my comments here are conditional on Gill’s report being reasonable).

Just to be clear, I’m not not saying that statistical arguments cannot or should not be used to make public health decisions. Indeed, I was involved last year in a case in which the local public health department made a recommendation based on statistical evidence, and this recommendation was questioned, and I (at the request of the public health department, and for no compensation) wrote a brief concurrence saying why I did not agree with the critics. So I am not saying that any statistical argument can be shot down, or that the (inevitable) reliance of any argument on assumptions makes that argument suspect. What I am doing is passing along Richard Gill’s analysis in this particular case, where he has found it possible for people to draw conclusions from noise, to the extent of, in his view, sending an innocent person to jail for 30 years.


  1. […] Did a naive interpretation of a cluster of cases send an innocent nurse to prison until 2035? […]

  2. My father and sister are trying a criminal defense case this week and the prosecution brought in an expert genetic witness to match blood samples. The lab “expert” (there’s a legal definition of what an expert is that’s related to the ordinary use of the word and based on a judge’s decision [sort of like the definition of a strike in baseball]). The “expert” said there was a one in sextillion chance of a false positive. The “expert” spent a long time explaining to the jury that a sextillion (10^21) is a really big number. My father and sister aren’t statisticians, and this particular fact wasn’t even relevant to their case, so they just left it.

    A couple years ago, I went to a workshop on forensic linguistics at the Brooklyn College of Law, where I believe it was Michael J. Saks who pointed out that there’s a slightly higher than 1 in a gazillion (for some value of gazillion) chance of the lab making an honest mistake and mixing up labels on samples. Not to mention the chance of dishonest mistakes, given the number of medical lab fraud cases (more than 1). He also did a great job of laying out the prosecutor’s fallacy.

    From a statistical perspective, these odds are conditioned on the chain of evidence being sound and the lab not making a mistake.

    • It seems unlikely even that a proper yet naive calculation would give you 1 in sextillion. That sounds like the kind of thing one gets from assuming each base pair in a DNA sequence is just a random multinomial chosen from ATCG. A little like calculating the probability of this sentence being generated at random by pulling random letters and spaces out of the alphabet instead of say random phrases from books.

      Having someone destroy the “expert” calculation can be beneficial to a case even if the DNA evidence isn’t relevant, because an expert who appears incompetent then puts a lot more reasonable doubt on other aspects of the prosecution’s case. Of course, your relatives know a lot more about the case than I do, but I really hate to hear that kind of crap being still put into trial evidence.

      • In this case, they had neither the expertise nor the expert to rebut this. Not to mention it was all going to support an already established fact that was actually beneficial to their defense.

        I’ll have to ask them about the general point about rebutting experts. It’s a tricky business, though, as you have to get the right expert who will be able to convince the judge and the jury, where they’d be lucky if a single person had actually understood high school algebra (both sides tend to throw out engineers and scientists during jury selection because they’re too analytical). So it usually comes down to a charisma and rhetoric battle, not a scientific one.

        It may be a failed independence assumption — there’s one kicking around in most models.

        • K? O'Rourke says:

          Agree, convince the judge and the jury is all that is of importance and I would not want to predict that on the basis of sound statistical logic, even if we were really sure what that was.

        • Yes, I was recently up for jury duty, and was thrown out because I coincidentally knew the judge, but there were probably only 2 other people in the first 25 of the pool for that trial that had any numeracy to speak of. One was a financial planner, I think maybe one other was a pre-med college student or something. Most of the other people were things like painting contractors, a brewery foreman, a hairdresser, a secretary, a retail clerk or whatever. Things that don’t require mathematics beyond making change, estimating costs from materials lists, or scheduling people for shift work. Those aren’t trivial mathematical skills, but they’re usually learned intuitively and not relevant for evaluating formalized statistics in a courtroom.

          I’ve worked as a forensic analyst on cases involving engineering and construction defects and similar things. We’ve seen a lot of crap expert calculations in that field. Usually there’s some statistical argument that goes like “we hand-selected 3 or 4 locations to test, and found 100% of them had some defect, so 100% of all 1000 locations need to be repaired, our estimated bill is $35 Million”, what they don’t tell you is the methodology for hand-selecting those 4 locations (usually involves someone with some expertise in the physical problem looking around for areas where they think such problems are likely to be, especially if they can find physical indicators for that problem such as cracks, water stains, etc).

          • Dan Wright says:

            Some of the problems with forensic science are highlighted in


            One approach that I think would improve things is covered in:


            It would be interesting to hear views about this report, from Gary Wells and colleagues, by folks here.

            Last time I was on jury selection for a case that involved eyewitness memory I got excused since I have done research (and been an expert) on that topic. Only other time I got called was for a murder case, but I was excused since I was from the same high school as the defendants.

          • Andrew says:

            I was on jury duty once and they didn’t throw me out. It was a slip-and-fall case. My statistical understanding helped me see through the ridiculous arguments of the person who was suing the city for tripping on a crack in the sidewalk. On the other hand, the other 5 people on the jury also had no problem seeing that the case was bogus. What a waste of a week. Afterwards I went up to the lawyer representing the city to express sympathy that she had to spend her time defending this sort of case and she replied that she gets a new one of these every week. What a horrible job!

            • Rahul says:

              What were these ridiculous arguments?

              • Andrew says:

                I don’t remember all the details but I do remember lots and lots of talk and then at some point we actually got to see a photo of the crack in the sidewalk and it was just an ordinary crack, nothing dangerous at all. The lawyer’s gimmick was that they had given the city a prior notification of this crack and thus they claimed the right to sue, but it turns out that there’s some consortium of slip-and-fall lawyers that, every year, officially notifies the city regarding something like 800,000 imperfections in the sidewalk. It’s basically a scam, which I assume pays off because every once in awhile, someone wins this sort of case.

              • No kidding. Our co-op’s insurance company payed off someone for a trip and fall that was a total scam. The tripper-and-faller had, perhaps not surprisingly, tripped and falled elsewhere and sued other people for it. Our insurance company said it was cheaper to just pay a few thousand bucks rather than fight it. It didn’t cost us any extra directly, but we all wind up paying indirectly, and the beneficiaries are the plaintiffs and their lawyers.

        • Phil says:

          I’ve only had to serve jury duty once. I was not the only PhD physicist on the jury.

    • Steve Sailer says:

      One summer I worked for Houston’s top criminal defense attorney Richard “Racehorse” Haynes. Back in the 1970s, he successfully defended the suspect in the first murder trial in Texas or maybe America in which the prosecution used genetic evidence. Haynes has an excellent brain for science, and he tore apart the credibility of the prosecution’s genetics witness. When I asked him about it, he surprised me by not reiterating his claims in the trial, but implicitly admitting the geneticist got the science right, but suggesting the cops had inadvertently framed his client by taking him to the scene of the crime and made him walk around it for an hour, leaving evidence for the genetics investigation.

  3. Anonymous says:

    So I followed the link trying to get further information. The additional info appears to be on “” which wants me to subscribe to read. Why put this behind a paywall? Don’t they want people to read it?

    • Richard Gill says:

      You can find Mark Heath’s report here:

      The new “campaign” website was set up only a couple of weeks ago and may have some teething problems.

      I would like to emphasize that I did my statistical analysis and wrote my report before studying the medical side of the Ben Geen case, though since then I have done my best to study that side of the case as well.

      Is the case about statistics? Well good statistics would have shown there is no case at all. Medically, there is (IMHO) no case at all, either. Jane Hutton said a lot of sensible things in her report but was not allowed to say those things in court. I think that that is a big scandal. In the Sally Clarke case, the president of the Royal Statistical Society presented written evidence on behalf of the defence. The judge did not allow that evidence to be presented in person to the jury, because “this is not rocket science”. That was also a terrible scandal. After a long, long battle, Sally was totally exhonerated and some medical persons disciplined. But still she was a broken woman and died probably of alcoholism.

    • Richard Gill says:

      The link does not bring you up against a pay-wall. It’s a file sharing service which you have to join up to before you can download stuff which people have put up there. It’s free.

  4. jonathan says:

    I don’t know that statistics is the main issue. I took a few minutes and read up on the Ben Geen case. The material I read is one-sided towards Ben Geen but what it paints is an argument about basic evidentiary matters getting at the question of whether the drug found on the needle in his pocket caused these deaths. Most of the material only cites contrary statements, but there was a trial and a defense so there’s no way to know.

    Statistics is part of the issue in considering if a crime was committed but again I know he had a trial and from the pro-Geen material I can see that was a major objection that winnowed out the number of charges.

    • jonathan says:

      I meant to say as well that one need only look at the Kercher case. Because she was British, you can read tons of material that proclaims the guilt of Knox and her then boyfriend in the British press. And by contrast, in a famous shaken-baby case, because the defendant was British, you could read tons of material exonerating “our Louise”.

    • Yes, even if Geen only committed one crime of injecting this drug improperly, it’s still a very serious crime. So what needs to come into evidence is some evidence that the drug was in fact in the dead person’s body and did in fact come from Ben Geen. Did that come out at trial? I don’t know. But this kind of circumstantial evidence by itself isn’t even close to “beyond a reasonable doubt”. Focusing instead on this one case, it’s plausible that Geen simply picked a syringe up off the floor in order to dispose of it properly, and that even if a crime were committed it wasn’t by Geen. The way that statistics enters is to give some kind of haze of legitimacy to this circumstantial evidence. If Geen was “caught” with this syringe once, and then some kind of analysis of past cases shows that he was abnormally frequently involved in similar outcomes in the past… then the circumstantial evidence becomes stronger in the mind of the jury. If those statistics are pure unadulterated BS generated by “experts” then… a serious miscarriage occurs.

      • Also, the statistics by Gill are relevant to whether crimes were even committed. If the rates of this respiratory arrest in other hospitals are essentially the same, and with similar fluctuations as in Geen’s hospital, then it’s implausible that there was any serial wrongdoing by Geen, as such wrongdoing would induce some other pattern different from random noise seen in other hospitals. Instead the analysis used in trial shows some cluster (which occurs frequently in other hospitals according to Gill) and was used as evidence that Geen was guilty.

        Let’s put it this way, since I know nothing much of the case except a few web pages and the Gill writeup, it is plausible to me that Geen may have done something bad in the one case involving the syringe, but the syringe evidence is by itself extremely flimsy. And since the “cluster” analysis is extremely flimsy, it offers no evidence to support the circumstantial evidence of the syringe. No jury should get “beyond a reasonable doubt” from such evidence.

        According to BBC news sites etc there were some patients who were found to have unusual drugs in their urine sample. I don’t know how they supposedly established the link to Geen though. The link seems to be entirely some kind of statistical matter (ie. that after the fact they were able to look back in history and find some cases where he was on duty, which of course is very likely spurious)

        • Richard Gill says:

          About the syringe: here is Ben’s story, corroborated by his (ex) girlfriend who is also a nurse, getting married to another guy this October. A very very nice girl, indeed. As are Ben’s parents. I did not yet meet Ben.

          By the way, she is convinced of his complete innocence and she’s (been) heartbroken. Because she’s a nurse and because of their break-up and up-coming marriage she doesn’t dare talk about the case. Nor does Ben’s mother – who is also a nurse – she would lose her job and her licence.
          The syringe (without a needle) which (on doctor’s orders) is used to inject muscle-relaxant into a drip of someone who’s going into an operation so you can get the pipes down their throat etc etc ended up after a more than average hectic Friday (understatement) in Emergency in his coat pocket. His girlfriend, doing the weekend washing (looking for dirty handkerchiefs?), found it and said he should take it back to dispose of properly. It was of course against the rules to have it. I guess it was half full or half empty or something… Now Ben had been getting funny looks from the doctors and other authorities the last few days, and expected to be in trouble. He had a big mouth, he was smart, he was a *guy* (the only one, probably), he had been in the army. His dad was a paratrooper, his mum a nurse. When he saw the police when he arrived at his work on Monday he panicked and he instinctively pressed the syringe empty. Not smart. Pure instinct.

          *On* his coat were found minute traces of everything a nurse gets to use in a hospital including anaesthetic. So, there was no “fatal lethal mixture” of anaesthetic and muscle-relaxant in the syringe. Even if the newspapers say there was. This is pure imagination of the journalists. No doubt encouraged by police and prosecution and hospital authorities.

          Incidentally the nurses were being told to do things by the doctors that really they weren’t allowed to do, and record keeping was also not quite up to scratch (more understatement). As other nurses have testified.

          So the syringe is not the clincher to his guilt but on the contrary a quite heartbreaking tragedy.

          But of course, this is just what I say … you can believe what you like.

          • Richard, I think the point that should be taken here, is that by itself, such a story, which undoubtedly came out in trial at some point and in some form, is by itself enough to put “reasonable doubt” into my mind and undoubtedly the mind of at least one juror about the importance of the syringe.

            However, combine it with expert testimony that there was an “implausible” “cluster” of cases associated with Geen, and … there is now much less reason to believe whatever Geen and his partner say.

            It sounds to me based on what you say, that this was the role of the “statistical” evidence… to give some kind of strong scientific sounding case which would make the syringe look highly suspicious rather than highly ordinary.

            • Richard Gill says:

              The jury was mostly asleep during the scientific presentations in court. The prosecution had powerful, highly experienced, mediagenic figures giving evidence. The journalists greedily wrote down what the prosecution had to say, and when a defence witness took the stand, marched out of the court in order to file their stories with their editors. Typically embellishing further what already were lies. The case was highly emotional, emotive. It wasn’t long after the Harold Shipman scare.

          • Rahul says:

            Some questions out of curiosity: Those in the medical professions, how often is it that you accidentally take home a syringe from work? Secondly, the times that you did would you just throw it in the trash (no needle, so no sharps) or take it all the way back to work for disposal?

            Also, how much drug exists in a typical (partially used) syringe and after three days in the pocket wouldn’t the liquid just leak out or evaporate?

            Wouldn’t it be scientific to test a few other random nurses’ coats to see if traces of the implicated anesthetic etc. are randomly to be found on such coats? That’d give us a control.

            • Richard Gill says:

              There was no needle in this syringe. If the syringe in the right way up the stuff that is in it when you take it home is still there three days later.

              Yes it would be scientific to look at other nurse’s coats. But the police are not going to waste time looking for evidence to disprove their hunch, are they?

              The defence is poor. They can’t commission expensive time-consuming scientific research. Moreover they are probably in the dark and as confused as anyone else. The defence lawyers are lawyers, not scientists…

              • Richard Gill says:

                The syringe went home in his nurse’s uniform (“scrubs”) by accident. He took the uniform home to get washed

                So he didn’t take it home deliberately in his outdoors coat. He brought it back to work on the Monday after the weekend in order to dispose of it properly.

                Of course, that is just what he says, and what his girlfriend says … You can say “they would say that, wouldn’t they”. I believe his girlfriend is honest and moreover has been heartbroken by this affair. I have exchanged a few emails with her and seen her social media activity. But then maybe I’m just a foolish old man …

              • Rahul says:


                No, your instincts might be right. Sometimes one must trust them. I have the handicap of only seeing the data and not the persons behind the story.

                In any case, was there previous reports of people taking syringes home accidentally and then bringing them back to work? What do nurses you’ve spoken with say? Is this a typical instinct? Have you asked any nurses what they’d do in similar situations?

              • Richard Gill says:

                Nobody should trust my instincts. Check the facts.

                Q: How often do nurses accidentally take stuff home which they shouldn’t?

                A: It happens all the time.

                Just ask some nurses. This was part of the case against Lucia de B… all the stuff she had in her medicine cabinet at home. OMG I have some paper and pens at home which belong to my boss! Lucia took an empty syringe home so that her daughter could show it at kindergarten “what my mum does”. The police found it and that was part of the case against her …

        • Richard Gill says:

          I believe that in just *two* patient’s urine samples, traces were found of medication which they were not supposed to be having, according to the medical records. However these traces represented therapeutic amounts of medication which logically *might* have been prescribed. Record keeping in that hospital was not quite up to scratch. Nurses were being instructed to administer medications or perform other tasks which doctors should have done.

    • Andrew says:


      I have no idea if statistics is the main issue but it was my impression that the existence of a surprising cluster of cases is what got all this going. The statistical issue addressed by Gill is that natural variation can lead to jumps in the data (“clusters”) that to a naive eye seem suspicious.

      • Richard Gill says:

        The cluster here is *manufactured* by the hospital doctors who already on the basis of one unexpected event, a nurse who people have been gossiping about, and the seasonal increase in a particular kind of event, go out on a trawling expedition to reclassify all kinds of innocent events if in any way they can be associated with the weird nurse’s presence.

        The nurse is essentially convicted by a couple of hysterial or paranoid doctors. The required documentation is then handed over to the police including instructions how to interpret the medical dossiers. The media do the rest.

        This is *exactly* how it went with Lucia. We now know that four medical specialists have lied to the police and to the courts, covering up their own mistakes. Probably each one does this independently of the others, out of collegiality, and out of certainty that the weird nurse is indeed a killer (even if the case of their own patient might have an alternative explanation).

  5. question says:

    Richard writes:
    “We find that respiratory arrests in ED are about five times less frequent than cardiorespiratory
    arrests, which are of course extremely frequent. Respiratory arrest is certainly
    less common than cardio-respiratory, but certainly not rare at all, by any reasonable
    understanding of the meaning of the word “rare”.”

    But he does not plot this ratio directly. I agree that this data is very messy and if it is accepted that Cardio-respiratory arrest is not “rare” in emergency departments, then comparing these two incidences is probably the best we can do. Plots 1:3 are the ratios for each hospital, and plot 4 shows the distribution along with some filtering. We can see an example of a month where cases of respiratory arrest was 4x that of cario-resipiratory (March 2006 at Nottingham, with 4 and 1 cases respectively).

    For the figure 4 lower histogram I filtered the cases with ratio >1 for visibility. We can see these do occur (116 examples) but are rather rare. The number of occurrences of ratios equal to one and 1/2 are notable, this may be a problem with the data. And as noted by Gill, it is unclear whether zero values are measurements or missing data. The only other idea I have is to sum up the results from all the hospitals each month to get a lower/upper bounds. We can see that as many as 15 cases of respiratory arrest were reported in one month overall. However the missing data makes it difficult to say much:

    Good luck.

  6. Jane Hutton says:

    A response to Andrew’s point about the need for statistical arguments in court cases (and effectively an aside on jonathan says: July 29, 2014 at 11:59 am… one-sided). When I provided a report, I was absolutely clear to the lawyers that I had no opinion about the innocence or guilt of Ben Geen. My opinion was that the evidence I had been asked to consider was of very poor quality. In medicine, an article based on such poor quality data would have almost no chance of being published in a half-decent journal, and a grant application which proposed the methods used would not be funded.

    My longer term aim is to improve the collection of evidence by police and other authorities, and the understanding of judges of basic issues in data collection. As Sheila Bird says, “courts should not rely solely on anecdotal evidence when properly collected empirical evidence may differ from what witnesses claim…”. She was concerned about the quality of evidence related to prisoner deaths:
    BIRD SM. Fatal accident inquiries into 97 deaths in prison custody in Scotland
    (1999-2003, or during first five years of operation of Scotland’s only private
    prison): elapsed time to end of inquiry or written determination, issues and
    Howard Journal for Criminal Justice 2008 (published online 22 May ahead of print
    version); 47: 343 370.

    • K? O'Rourke says:

      > the understanding of judges of basic issues in data collection
      That would seem to be a necessary first step (lawyers unlikely that interested unless they think judges already get it)

      > I was absolutely clear to the lawyers that I had no opinion about the innocence or guilt of Ben Geen
      If you are going to actually fairly evaluate the evidence, that is (I believe) the best position to be in.

      But if I was on trial, I would want my lawyer to find an expert witness who understood the job was to find out how to best construe the _allowed findings_ as evidence of my innocence ;-)

  7. I recently read the book “Math on Trial: How Numbers Get Used and Abused in the Courtroom”, which show statistical errors in 10 serious cases. It’s written nicely, fluent to read, and they also give tips on how to avoid making such errors. I can really recommend that book! (There’s a German version as well).

  8. Richard Gill says:

    I’d also like to clarify that though right now I’m pretty convinced that Ben Geen is completely innocent (I see no reason whatsoever to believe anyone did anything wrong … except of course for all the usual medical blunders which go on all the time and which are never admitted to), the bigger issue to me is that even without not knowing a great deal about the case, I think it is absolutely clear that he did not get a fair trial. I even suspect that if the case had been handled properly in the hospital, when it first came up, it probably would not even have come to a trial at all. These two issues have everything to do with amateur opinions about whether or not a cluster of cases is unusual, and about an amateur approach to doing an investigation.

    In Horton General Hospital, the doctors first *knew* they had a serial killer on their hands. *Then* they went hunting for the cases in which he was involved. *Then* they pick the juiciest and give them, accompanied by their medical summaries about why case X, Y and Z are terribly suspicious, to the police.

    In science we know that after coming up with an exciting and daring scientific hypothesis, we should be looking hard for evidence to disprove it.

    Even good police investigators know that this is the good way to do a police investigation. You have your daring hunch … then you go all out to try to disprove it. Look for the kind of evidence which would prove you are wrong.

    However in medico-crime, the medical information is totally in control of the medics. The same people who make mistakes treating or diagnosing their own patients then suspect someone else of doing bad things, then hunt for information to prove they are right, then hand them over to Justice.

    Wendy Hesketh argues in her book “it is not easier to detect a medico-killer now since Shipman, but it is much more difficult for an innocent person to defend themselves once accused of medico-murder.” Moreover, from the Lucia case in the Netherlands, which in all main features is horrifically identical with Ben Geen’s case, we can learn how incredibly *easy* it is for a completley innocent person to get themselves accused of medico-murder. It just needs the a little chain of unlucky coincidences.

    How did Lucia get exhonerated? She was *lucky*. The sister-in-law of the chief paediatrician at the children’s hospital where it all started, who was herself a senior medical doctor, got suspicious when three years into investigations and trials, her sister-in-law (who had coordinated all the investigations and whose own patients were involved) for the very first time was called to give evidence. By now it was at a court of appeal. On the morning of the day she was due to give evidence she committed a little self-mutilation and urgently admitted to a psychiatric ward. She never did appear in court … Well that gave her sister-in-law something to think about and she began seeing all the events in a different light.

    Later there were more strokes of good luck. Who exactly was in the review committee which evaluated her case, which particular lawyer was assigned by the supreme court to prepare the submission for re-opening the case to the supreme court. It was the best guy possible. He saw that there was no legal “new fact” and commissioned new medical research which finally created what formally was indeed a “new fact”. In other equally obviously “wrong” convictions, some lazy lawyer handled such a submission and simply chucked it in the bin. No “new fact”.

    In these cases, there are no “new facts”. There is merely sensible evaluation of all the already known facts. This is not good enough to reopen a closed case even if for every sensible outsider it is clear that it is beyond reasonable doubt that the person in jail should not have been convicted and indeed, there is no reasonable argument that they are guilty.

    • Rahul says:

      I read your arxiv article & you seem to have effectively addressed the question “How odd is a cluster of cases?” In this case of respiratory arrests, not so odd at all, apparently.

      But the question that bugs me is “How odd is a cluster of 18 respiratory arrests, such that the same nurse has signed all 18 charts, has had a role in all 18 treatments?” Were there few enough nurses in A&E duty rosters to make this a statistically not-unlikely fluke?

      In the empirical data from other hospitals did you did into (anonymysed) nurse identities for the reported respiratory arrest events?

  9. Richard Gill says:

    Here is a webpage with links to original documents on the case, mainly from the defence side, but including the judge’s summing up at the last (appeal court) verdict.

    • Rahul says:

      Thanks for the link. Very interesting reading.

      Let’s assume Geen is totally innocent. But this doesn’t still seem like an innocent fluke of cases does it? Some malfeasance seems indicated. How else do you explain the Vecuronium and Midazolam found in the last dead guy’s body fluids? Two, not one, potentially deadly drugs that have no innocuous explanation how they ended up in the guy’s system?

      Or is there an alternative explanation?

      I can go with Geen didn’t kill them, but this doesn’t seem like a routine death from natural causes, does it?

      • DK says:

        I am with Rahul on this one. Serial killer or not, it’s not merely a random cluster. The story about syringe accidentally brought home that is then brought back to work “for proper disposal” and is then accidentally pushed empty at the sight of police is really, really fishy. No one would bother bringing such a thing back!

        • Richard Gill says:

          Nobody who was a serial killer and who knew that the authorities were suspicious of him would come to work with a syringe of poison in his pocket.

          Here is Ben’s story, corroborated by his (ex) girlfriend who is also a nurse, getting married to another guy this October. A very very nice girl, indeed. As are Ben’s parents. I did not yet meet Ben.

          By the way, she is convinced of his complete innocence and she’s (been) heartbroken. Because she’s a nurse and because of their break-up and up-coming marriage she doesn’t dare talk about the case. Nor does Ben’s mother – who is also a nurse – she would lose her job and her licence.

          The syringe (without a needle) which (on doctor’s orders) is used to inject muscle-relaxant into a drip of someone who’s going into an operation so you can get the pipes down their throat etc etc ended up after a more than average hectic Friday (understatement) in Emergency in his coat pocket. His girlfriend, doing the weekend washing (looking for dirty handkerchiefs?), found it and said he should take it back to dispose of properly. It was of course against the rules to have it. I guess it was half full or half empty or something… Now Ben had been getting funny looks from the doctors and other authorities the last few days, and expected to be in trouble. He had a big mouth, he was smart, he was a *guy* (the only one, probably), he had been in the army. His dad was a paratrooper, his mum a nurse. When he saw the police when he arrived at his work on Monday he panicked and he instinctively pressed the syringe empty. Not smart. Pure instinct.

          *On* his coat were found minute traces of everything a nurse gets to use in a hospital including anaesthetic. So, there was no “fatal lethal mixture” of anaesthetic and muscle-relaxant in the syringe. Even if the newspapers say there was. This is pure imagination of the journalists. No doubt encouraged by police and prosecution and hospital authorities.

          Incidentally the nurses were being told to do things by the doctors that really they weren’t allowed to do, and record keeping was also not quite up to scratch (more understatement). As other nurses have testified.

          So the syringe is not the clincher to his guilt but on the contrary a quite heartbreaking tragedy.

          Look: if Ben is innocent then his innocent story of the syringe is pretty plausible. If Ben is guilty then the prosecution’s allegations about the syringe are hardly believable, but OK a crazy killer … Therefore the likelihood ratio Prob(evidence | defence) / Prob(evidence | prosecution) = 1. The syringe is no evidence at all, either way. On the other hand this is one of the best known cognitive biases: attaching enormous weight on just one item of evidence (emotional, graphic, …) and disregarding all the others

          • thom says:

            I’m persuaded by many of your arguments but this point “Nobody who was a serial killer and who knew that the authorities were suspicious of him would come to work with a syringe of poison in his pocket” is just weak. Criminals including serial killers make stupid mistakes all the time – most of them get caught by luck or by a build up of such stupid errors.

        • Richard Gill says:

          No it’s not a random cluster. The cluster was *created* by the hospital.
          Step 1: you have a nurse who people are gossiping about
          Step 2: you have the seasonal increase in some kind of event
          Step 3: you have the surprising death when the nurse was around (surprising because nobody had looked in the patient’s dossier. eg terminal alcolism leading to liver failure. Nobody knew)

          Now the hospital doctors get together and go on a trawling expedition. During which they also essentially rewrite the medical histories (certainly: they reinterpret them) of many patients.

          In Ben’s case the media say that there were 18 respiratory arrests in 3 months and he was there every time. This was in the winter months when the typical number in that hospital BTW is 5 per month. Actually those 18 were respiratory arrests, respiratory declines, and even one or two cases of a patient just fainting. The hospital doctors did not go through every one of the 3000 admissions to A&E in those three months. No they only looked at admissions where something remotely connected to “respiratory” was present and where Ben was present.

          So indeed, the cluster is not due to chance. It was manufactured by the hospital out of nothing.

      • Richard Gill says:

        Vecuronium and Midazolam found in the last dead guy’s body fluids are easily explained. BTW were these in toxic concentrations??? No. Could they have been there because of past medication? Yes. cf. Lucia de Berk. Yes there might have been digoxin in baby Amber’s body but it was not enough to have anything to do with her death and if could have leaked out of various organs where it had been stored after death. She had been treated with it officially, in previous weeks, and who knows what medicines the family give to their baby when the nurses and doctors aren’t looking. Then there are all the medicines which are given accidentally because of faulty administration. Then there are all the medicines given “officially” but forgotten to be filled in in the forms.

        • Rahul says:

          Past medication? Vercuronium is hardly an OTC drug you could get at the corner pharmacy.

          Besides, have you looked at its physiological half life? 80 minutes. So far as I can see, it’d be hard for any Vercuronium in remote history to show up in his fluids. It had to be a relatively recent injection. But I’m no expert. I’d love to be corrected.

          Vercuronium has very specific usage in medical practice (anesthesiology / surgery) so I doubt it has some lower therapeutic dose that may have been accidentally / routinely administered by a lay nurse. Even if I assume this was so, I’d have to allow a combination of a doctor and a nurse who prescribe and administer the said drug and both forget / lie during the trial.

          A bit of a stretch I thought.

          • Richard Gill says:

            Interesting. I will look this one up.

            So how about the hypothesis: this was medication which was quite normal that the patient in question should have been given, but by a clerical error it hadn’t been registered.

            That did happen all the time, we know.

            It would be interesting to analyse urine samples of patients who experienced all kinds of events when Ben was *not* on duty to find out if there was medicine in them which shouldn’t have been there.

            And how do you know Ben was responsible for getting it into that patient’s body?

            He was never observed doing something with medications which he shouldn’t have been doing. Never. Yet he had been under suspicion (subject of gossip) for weeks.

            • Rahul says:

              Well, that was exactly my point though. I’m saying this isn’t a random cluster. Maybe Ben didn’t do it. But someone did. Malfeasance seems indicated.

              Regarding your hypothesis: I don’t think Vecuronium is given for much outside of actual surgical intubation (I could be wrong). So, yes, there’s a chance it was prescribed but not registered. But in that case why would the legit doctor / nurse pair lie. They could / would admit they had prescribed it. It’s not even an error.

              • Richard Gill says:

                I’m saying this is not a random cluster because it’s artificially created by the hospital in hindsight

                17 respiratory arrests in three months … weren’t 17 respiratory arrests. They were also respiratory depressions, even a couple of instances of a patient just fainting. Everything which by stretch of the imagination could be called a respiratory arrest and which happened when Ben was around was made into a respiratory arrest and was made suspicious.

                Why could/would they lie? In the Lucia case, four top medical specialists have all lied. They also all made mistakes which were not reported to the police or to the courts.

                This is called “medical collegiality”. Once a few doctors are convinced they have a serial killer on their hands, all the others try to help. Keep quiet about little things which they know which go in the other direction, believe everything their colleagues say. Once the police have been called in and the press is at the doorstep they are trapped in their lies (ie half-truths). No going back. That’s how it works. Again and again.

              • Couldn’t it be accident rather than malfeasance? Like some nurse came and administered the IV intended for a surgery patient to the wrong patient and this caused the respiratory arrest shortly after, but no one comes forward because they either never realized the mistake, or are afraid of the consequences (including having the “serial killer” investigation/witch-hunt turned on them)?

                If there were real malfeasance here, the prosecution sure seems to have been extremely lazy and reliant on this crappy statistics analysis rather than real investigative work, like perhaps trolling through the dispensary logs and looking for whether there were obvious differences in the character of these events when Geen was vs wasn’t involved (rather than simply taking a list of cases from the doctors which were pre-screened for Geen involvement).

                I honestly don’t have any idea whether Geen was guilty, mainly because I don’t have enough information, but it does seem clear to me that he got a crap trial, and that his defense lawyer dropped the ball on countering the “statistical” evidence (probably, based on the description here, most of this evidence could have been disallowed with a proper statistical analysis showing how flimsy and improper it was prior to trial).

              • Richard Gill says:

                I’m sure it can *all* be initially accident … an unfortunate chain of coincidences. What happens however is that at some point, inside the hospital, there is a kind of phase-transition. One doctor has become certain that there is a serial killer. Everyone has noticed that weird nurse. There were some “unexpected” incidents recently and he was there. Everyone is certain. Everyone has imperfect memory and the medical records are very very imperfect. They hurriedly meet together in conference, sort out some dossiers, and decide they have to inform the hospital management and report to the police. They remember more odd things, don’t remember ordinary stuff, don’t recall the innocent explanation for the (in retrospect) odd things they had noticed.

                From this point on, there is no going back. Doctors who might initially just have “forgotten” some extenuating circumstance, or just “forgot” to mention their own role in some incident (wrong prescription, wrong diagnosis, …) are not going to suddenly remember now. And they don’t withhold this information deliberately in order to mislead … no, they too are convinced (by what they heard from the other doctors) and their own impressions, that the nurse is a serial killer. So precisely for good motives: collegiality, and in order to stop a killer, … they totally suppress any qualms that they might have had, and information that might have pointed the other way. It’s called “collaborative story telling” in psychology. It’s just common group dynamics.

                This was exactly how it went in the case of Lucia de Berk. There is even strong evidence that a pair of top medical specialists doctors “corrected” medical records retrospectively, after an “unexpected” death, *before* going to the police, because they had convinced themselves that the baby had died of poison administered by that weird nurse and *therefore* the heart must have failed before the lungs, not the other way round.

                They were already convinced the nurse was a serial killer and when that baby died that night they were certain of it. So they indeed “corrected” medical records “in good faith”. They didn’t lie – lying is telling a falsehood with intention to deceive. They did not intend to deceive. They did not “know” that their “correction” was wrong.

              • Richard Gill says:

                Incidentally I don’t wish to argue that Ben Geen is innocent. I would wish anyone who is interested to inform themselves as well as possible and come to their own conclusion.

                I do wish to argue that he did not have a fair trial. I am pretty sure that if he had had a fair trial, he would have been acquitted. In fact, I am pretty sure that if the hospital had done the right thing – namely had an independent, confidential, thorough, medical investigation into the various cases *before* going to the police – the case would not even have gone to the police.

                Moreover I wish to argue that if the defence had called in a good statistician right at the start, he might have had a small chance to have been acquitted. However once the hospital and the police *know* they have caught a killer, and the media has taken this message on board, then actually no one has a chance. It was a public lynching. If one compares what is in the media (even the BBC!) to the true facts, this is rather clear.

              • Rahul says:


                I don’t agree: The hospital, once it even has a whiff of wrongdoing, is duty bound to approach the police.

                What you are proposing would be highly irregular i.e. the hospital conduct a confidential, internal investigation before approaching the cops.

                Note that the hospital need to be even close to certain of Geen’s guilt to justify them approaching the police. Proving guilt comes later but I think a hospital is eminently justified in contacting the police once a reasonable suspicion has been expressed.

              • Richard Gill says:

                Well Rahul then the police should do what I said ought to be done. Instead of just accepting the whole stack of paper and going into prosecution mode, uncritically. Notice: the hospital doesn’t only informs the police it also informs hospital staff and it puts out a press release. There’s now no going back. The media grab it and there’s no need for the police to do any work at all. Judge and jury will do the right thing …

              • Richard Gill says:

                The hospital can put Ben Geen on “non active” (suspect him) while they are having external independent medics do an independent scientific investigation into the medical situation. They don’t have to instantly notify the police the moment some paranoid doctor smells the whiff of a serial killer.

              • Richard Gill says:

                sorry I meant “suspend him”

              • Richard Gill says:

                Note: the hospital *did* conduct a confidential, internal investigation before approaching the cops. It did not call in any independent external experts. It was very hurried. The same doctors did the investigation as were responsible for treating those patients. Ben Geen was convicted by a small bunch of doctors who acted simultaneously as prosecution, judge and jury. No defence. Lucia de Berk: exactly the same.

              • Richard Gill says:

                From the police point of view (a) the doctors ought also to have been suspects, (b) also all other nurses should have been suspects, and (c) they should not have believed that there even had been crimes committed, without having an independent external multidisciplinary team evaluate each individual case separately, with full access to complete medical dossiers. In the Lucia case it took 9 years before this was done. When it was done, the conclusion was pretty clear, and was not difficult to come to.

  10. […] The Andrew Gelman Blog Statistical modelling, Causal Inference & Social Science. […]

  11. Richard Gill says:

    I suspect that the reason that in the US there appear to be relative few *innocent* nurses convicted of health care serial crime is the death sentence. Together with the fact that only an idiot would keep on insisting they are innocent. Certainly if you want to avoid the death sentence.

    Seriously, this should be studied scientifically, by a good statistician. So far I only found three such US nurses

    But then I also suspect that there are more psychopaths per head of the population in the US than elsewhere … did anyone study the statistics of serial murder worldwide? Is my impression merely another of those cognitive biases?

    • Rahul says:

      You mean they plea bargain out?

      • Richard Gill says:

        (a) they plea bargain out
        (b) if they don’t they get the death penalty and once they’re dead no one fights for their innocence any more
        (c) if you do not admit to the crime you have been charged with, and since you are going to be convicted anyway, you had better just admit guilt. You’ll get shorter sentence, better treatment in jail, you’ll be allowed parole and many other prerogatives. It’s not plea bargaining (ie not arranged by your lawyers). It’s aggressive police interrogation techniques, brain-washing, … you could almost say, torture. Sleep deprivation, good cop – bad cop routines, Stockholm syndrome. It almost always works.

        We know this because of the many instances where suspect admitted to crimes went to jail and it was later *proven* that others did it. Sometimes we have the police interrogation tapes to show how it was done.

        • Richard Gill says:

          I should mention that in the NL a suspect cannot insist on their lawyer being present, and that interrogations are not routinely taped.

          The police have said that if they were obliged to do that (a) they wouldn’t be able to get so many criminals into jail (b) there were not enough numbers to put on all the tapes. Huh? They use finite number theory. There only exist finitely many integers, and its less than the number of police interrogations.

          I’m being serious here.

        • Rahul says:

          Regarding your (b), do you have any specific examples in mind, of innocent (at least in your perception) Americans that were executed for medical crimes?

          • Richard Gill says:

            No I don’t have any US cases in mind! I’m just wondering! Maybe there are none. In the US doctors and hospitals are terrified of being sued for malpractice. In Europe that hardly happens. This must be connected to quite a huge difference in culture concerning whether or not medical errors get admitted in public.

            Dutch medical specialists don’t have to pay huge insurance premiums to protect themselves against malpractice suits because there aren’t any because if there were the patients wouldn’t win. They do not make mistakes. Every year there are 2000 deaths in NL (this is four times the rate of deaths in road accidents) because of avoidable and known medical errors in hospitals but only a tiny percentage are admitted.

          • Not for medical crimes, but I think the number of murder/rape convictions from the 60’s and 70’s etc that were later found by DNA evidence to be improper is far from zero. I don’t know how many actual executions there have been.


            As an aside, there’s a nice account of a famous case, the only case of the supreme court issuing a contempt of court citation, where a convicted black man was allowed to be removed from jail and lynched without any significant effort on the part of the police to prevent the lynching, prior to his supreme court appeal. The book is called “Contempt Of Court” I believe. It describes the rape/murder trial, and then the contempt trial against the local law enforcement who “allowed” the lynching (all the guards were sent home, the keys to the jail were left easily accessible, doors unlocked etc I think this was more or less “standard practice” at a certain time and certain places).

  12. Richard Gill says:

    I have explained how to get set to jail for serial murder without killing anyone here:

    The PhD thesis by John Field is interesting

    Note his reference to many works by Beatrice Yorker and by Catherine Ramsland. These two ladies combine academic and media careers by capitalizing on the public’s love of scare stories. They coined and promoted the term “HCSK” (health care serial killer), sounds pretty sick and horrific, right? It is interesting to look at their check-lists for presence of a health care serial killer

    1) the nurse is called “the angel of death” by their colleagues
    2) the nurse is often around when bad things happen
    3) the nurse has faked credentials
    4) – 9) and so on and so forth

    Now Lucia, in the Netherlands, satisfied every one of the items on this check-list. Yet she killed no-one and indeed she and her colleagues fought heroically to save lives which unfortunately were curtailed by medical errors of the medical specialists treating those babies … those medical errors which never found their way into police depositions, were never mentioned in the courts, never mentioned in the media.

    The key publication in this field is Yorker, B. C., Kizer, K. W., Lampe, P., Forrest, A. R. W., Lannan, J. M., & Russell,
    D. A. (2006). Serial Murder by Healthcare Professionals.
    Journal of Forensic Sciences, 51(6), 1362-1371.

    Most if not all of the authors have gained enormously, professionally and financially, from being apostles of the “epidemic of HCSK” gospel.

  13. Richard Gill says:

    PS you will notice that John Field author of

    *knows* that Lucia de Berk is guilty and trots out all the prosecution and media lies about her in his thesis. I wonder if he will correct these errors? The thesis is dated October 2007 one year after publication of a book and a great deal of media attention to the case of Lucia de Berk. I would say that in 2006, anyone with any sense in their head, and minimally informed, already could be sure that that case was a horrific miscarriage of justice.

Leave a Reply