# 1 Preamble I must prefix the events I am about to recount by a disclaimer: I am a man of science. A rationalist. I do not believe in ghosts, fate, signs, and the like. Yes, I have read Carl Jung, but no one will convince me that 'synchronicity' is anything but the mind primed to spot certain coincidences, like noticing, after buying a new car, how many of this model are on the road. What we see is what we get, and it works according to rules, which we may not fully know yet, but are figuring out. There is space for metaphysics in my philosophy, but it is, as its name ought to indicate, firmly outside that of actual, material, physics. Maybe this would all be easier if I were a bit more woo-woo, but here we are. This... essay? case study? confession, maybe - is a summary of my notes dating back from December **XXX**, which are attached alongside the present. My names is Charles Anderson. I am a GP with a small, independent practice in a small town on the South coast of England. Yes, there still are small independent practices left in the country, though they are mostly, like mine, in rural areas, serving an ageing population. Most of my days are spent caring for the town's elders, whiling away the time before I soon join them. I was given the chance to sell the practice, join a larger surgery, and the like, but I have always stubbornly resisted it. Now there is little point as I expect to retire in the next few years, and, well... I have grown to like the arrangement. Until about a decade ago, there were other physicians working at the practice. Now, after the last one retired, it is just me and my medical secretary Susan Walker. Susy, you will probably be the first person reading this, I am so sorry. None of this is your fault. I wish you the best. A large factor in my keeping my independent practice, against economic and practical reason, has been the data collection practices of the NHS. We have resisted the various IT system upgrades offered by our trust, initially out of habit, then quickly out of principle. Faced with a cohort of elderly, vulnerable patients with little technical knowledge, I have decided to opt them all out of NHS data sharing. This means that, to this day, we handle all patient information on the surgery's own computer. It's a beige machine like you do not see anymore. Sitting on Suzy's desk, with the sun coming in from the left, it has, over two decades, discoloured unevenly, ending up with very distinct shades of yellowish gray on each side. At some point, we replaced her original near-cubic CRT monitor with a black, flat screen one. This patchwork of colours earnt the machine the nickname of 'Harley Quinn', after the Batman character in Harlequin make-up. I know nothing about these things, this was all the coinage of Suzy's nephew, Jack, who has been keeping Harley in working condition in his unofficial capacity of the surgery's IT guy. It is a miracle we can keep working like this. Harley runs a version of Windows from an age when these were named after years, not version numbers. On it runs a 'legacy' version of NHS software, which handles all patient records, appointments, prescriptions and the like. This cannot connect to the data mothership, so, every week, we run an export procedure to send record updates to the trust. This gives me control as to what data is shared, and, health analytics be damned, I keep the surveillance system on a need-to-know basis. This arrangement has been a reason to keep Harley alive and working. We need the data entry system, and it only runs on the old Windows, which in turns only runs on vintage equipment. Jack has been changing hard drives, memory, even the whole motherboard once, in order to keep us working like this. This will not last forever: the trust has been moving everyone onto newer, ever more centralised systems, and the little loophole we're exploiting will soon close. Up until this morning, I expected this would be the point at which I'll retire and close the shop. So, in spite of her - its - advanced age, Harley does a lot of work for us. The downside of what Jack calls our 'operational security' is that a lot of things that, elsewhere, are automated, need to be done by hand. What they do in the click of a button takes us a minimum of five minutes. And this is on a good day. But it's not exactly like we are very busy, and overall the system works. Slightly more problematic is that, like many computers her age, I suppose, Harley suffers from intermittent freezes: you'd be filing in phlebotomy results, or renewing a prescription, and the computer just stops responding. Sometimes the mouse moves, sometimes everything is frozen. No error, no 'blue screen of death' (another of Jack's terms), just an unresponsive computer. The only solution is to restart the machine, which is, of course, another five minutes job. Still, it's not exactly a deal-breaker, just another of these IT niggles one learns to write off and live with. Harley's 'brain freezes', as we've taken to call them - after I veto'd 'senior moments', on account of our clientele, are at the core of the... phenomenon. I do not know how else to call it. Harley' freezes have never been such an issue that we'd ask Jack to look into them. Thinking back, it seems they have always occurred, regardless of the patches, upgrades and kludges of Jack's. They may be happening a bit more of late, but, since no one was keeping track before December **20XX**, it is hard to tell. The first freeze that really came onto my radar was in August **202X** . This is where the story starts. I was renewing a prescription for a patient, Ms. Sullivan. 84, relatively sprightly for her age, but with a chronic kidney disease that's her main cause for being on my radar. As you can imagine, most of my patients, being elderly, have some kind of chronic disease. This means I will be doing this kind of medication renewal on Harley every month or two at most. That's six to twelve times a year. This will become relevant as I explain, for the stats and the spreadsheets. For some reason, I was sorting out Sullivan's prescription as a bit of last minute admin, before catching a flight. Really, before catching a car to the catching of the flight - this is not important. What is important is that when Harley froze on the patient details screen, I didn't have the 7 minutes it took to restart the machine, then get to the same screen. Yes, I have timed it, several times actually. So when the taxi called for the second time, after ten minutes parked in front of the building, I asked Susan to finish the job. This is why this is important. Suzy, I say, Harley's frozen again, when I was renewing Sarah Sullivan's. I've rebooted it and logged in, could you finish it for me - it's the same as last month, no change, it just needs pushed through? Yes of course, Doc, now you rush off to your car, your man's been honking, couldn't you hear it? etc. Nothing unusual about this interaction. There are certain clinical acts, and their administrative counterparts, which must be carried by a physician - prescribing medications and interventions, ostensibly. Others can be done by less qualified staff. Your medical secretary can arrange for a urine sample. Technically they could also arrange the testing of a blood sample, but the *drawing* of the blood is an act that needs authorised by a MD... As you can imagine, there are many rules. The clinical aspect is entirely unobjectionable, it is the enforcement of this permission system on the administrative side that is the typical NHS IT soup you would expect. Actually we can only do this - I mean me, asking Suzy to sort out what results in a clinical act, but is in practice a data entry exercise - *because* we still use Harley. The rest of the county's health system has long been moved to these computer with the badge reader on the keyboard, the system knows who's allowed to do what, and, if they do, where and when. And a lot of the permissions that clinically, should be available from day one, need to be granted in the system, with the NHS requiring specific training to tick boxes and click buttons. I do not mean to be derisive, it is important that people doing this do this correctly. If we were working anywhere else, the only way for Suzy to put in Sullivan's renewal on my behalf would be for her to impersonate my identity, have *my* card in the keyboard as *she* is using it. Admittedly, this is technically what she was doing this day. This was my computer, and I logged in before running for the cab. So, this is the first confession, I suppose. On occasion, we violate the rules, because we can. Then again, I will argue that the rule is about who makes the decision, not refusing to trust a civilian to action a clinical decision from a physician. I made the clinical decision on Mrs Sullivan's medicine. Susan did the admin. Preventing people like her from doing things like this is not a matter of qualification, because the competence needed is, well, boxes, numbers in fields, buttons and the like. And actually, I know plenty of physicians who are much less competent at this kind of stuff than their staff are, in fact I'm probably one of them. So it's not a matter of competence, it is a matter of *trust*. Do we trust someone to not deviate, accidentally or maliciously, from a physician's instructions, essentially. We have a system where this trust is based on qualifications, completely unrelated to the moral and practical competence needed. What is the logic here? Someone's got this or that degree, presumably without cheating, they were taught about the ethics of clinical practice, so this make them more qualified, more trustworthy? Claptrap. The ethics of clinical practice are for the doctor making the decision. The requirement here, beyond the practical, is really moral character. I object to equating education with moral character. This to say that although both Susan and I use Harley every day, Suzy has little call to be looking at specific patients' records, because most of the time this is needed, it is in order to do something only I am allowed to do. This time was an exception, this to say that Susan was aware of Harley's freeze on Sarah Sullivan. Ms Sullivan's main issue was her kidneys. Chronic kidney disease can be managed, but it can also escalate, and be fatal. Like everything clinical, it depends. In the case of Sarah Sullivan, her bloodwork was showing improvement as recently as the month before the incident. In fact she's had no interaction with the practice between the good bloodwork and Harley's freeze. As we later found out, she was well enough to pick up the medication herself two days after the freeze. Five days later, though, she was dead. As I said, CKD is complex, and I'm no nephrologist, but when I said it can be managed and also be fatal, this is, tragically, often in the same patient. She picked up the medicine two days after Susan's renewal; as far as I was told, showing no symptoms. Two days after that, she was admitted to our local hospital with acute kidney failure, which took her life the following day. Of course, with my aging... I was about to write flock. I'm no shepherd, nor a minister, but I suppose my concern for my elderly patients has some of the pastoral. I mean for all my patients, but, as it happens mine are elderly. Well, most of them are. There's the exception of the odd new registration - some of which come to us because of the... peculiarity of our data practices, which Jack has be praising. But I'm a realist. Elderly people die. Famously, most of the money the NHS spends on a person is spent towards the end of their life. I remember someone jesting, on the radio, that, because of this, "National Death Service" would be better fitting. In practice, no pun intended, after decades of being a GP, I have grown inured to elderly patients dying. I do not mean to be callous here: as much as one develops an attachment to one's patients, we do not have the luxury to grieve them. Their outcomes are theirs, we do our best, but in the end, such is life, it is what it is, "them's the breaks". There are others who need our attention, and we move on. If there is a lesson to be learnt from a patient death, we address it - though this is the case of all clinical outcomes, not just death, this is just... Medicine. I digress. Clinically, the surprising thing with Ms S. was the speed of the escalation, particularly given her charts thus far, but, as I said, these things happen. I've described this freeze as the first one that came onto my radar, but it was really Suzy that noticed the coincidence - I am not one for so-called synchronicities, as I have said. But she could only notice it because she knew about the freeze in the first place. And her not normally handling patient records on Harley made me wonder how many such coincidences could have happened in the past without me noticing. Although, of course, not at the time - at the time, I squarely assumed the odds were that whilst Harley's freezes and patient deaths had each featured in my professional life with tragically predicable regularity, they had never coincided prior to Sarah Sullivan's. With her death, Suzy remarked on the coincidence, an amused what are the odds? and no more. Then, in November that same year, it was Suzy who was using the computer when it froze, again on a patient record. This time, we weren't breaking the rules: she was sending Mrs Patel for a scan, the machine froze with the patient record on screen. I heard a soft, breathy, very polite curse, and as I came out of the office Suzy was rebooting the machine. It was a Friday evening, everyone was tired, but Suzy, particularly, visibly so. I felt able to face the 7 minute reset with more equanimity than her, so I sent her home, telling her I'd do it alongside renewing Mrs Patel's medication. Which I did. In this order, without a freeze. Again, at this point, I had the impression this kind of sequence was a common occurrence. Because of the distribution of duties, who's got the moral authority to click the butons on this form, or that one, Suzy would lead the admin work (often right after a consultation), sorting out all she could, scheduling the next appointment, and ordering whatever labs. Then I would come in and sort out the precious 'privileged procedures'. Actually writing it down, it does look silly and needlessly complicated. Plenty of times, we've all had to tolerate Harley's freezes. That's how we ended up with our own term for it. This to say I am still, to this day, convinced Suzy and I must have experienced freezes on patient records, either separately or together, before Ms Sullivan. And Suzy would have noticed a pattern of deaths following in their wake. But maybe not, given how usual the freezes. Maybe it's the kerfuffle around Ms Sullivan that attracted Suzy's attention to the freeze, and the first coincidence that primed her to notice the second. Yes, I expect you will have guessed, Ms Patel, died shortly after Harley's freeze too. I could have started with this, but I wanted the reader - those readers who aren't Susan, who knows some, but far from all, of this story - to understand our state of mind, as these events unfolded. Again, I later pieced together the timeline: the referral for the scan went out, along with the medication review. She picked up the medication the very next day. She was fine that day, according to the pharmacist. I also talked to her son, who told me she started feeling run down that evening, then woke up the next day with a full on chest cold. The son, Amit - good kid, moved to the city, luckily was down that weekend - did his best to treat it at home, for two days, after which he brought his mother, by then hectic and delirious, to the ER. 36 hours after admission, that is, four days after the freeze, she died. She was the first victim of an influenza season that started early and claimed half a dozen of our patients that winter. The day after her death, the imaging department of the hospital in which she died sent her a letter giving her a date, three months hence, for her scan. It is grim when the right hand does not know what the left hand is doing, but these things run on different tracks, go through different processes. Only five days to get a date for a CAT scan is actually pretty good - in fact if they'd taken longer, they would eventually have found out the scan wasn't necessary. There's no bitter irony here, it is not like the scan would have saved her life, it was something completely unrelated, a sprained wrist. I took this second coincidence with a quizzical raised eyebrow, but Susan was positively thrown - freaked out, she said. Dead, like last time, what are the odds, she cried. Well, not like last time, I observed, there's been freezes in the meantime. She granted me that. And deaths since last time - she granted me that too. But still, what are the odds, first Sarah Sullivan, then Padma Patel - maybe we should look for alliterative names. Actually, I said, she was Sripadma Patel. I remember, we were sat in the waiting room, with the only light coming from the lamp on my office's desk, next door. When I gave Suzy the news she'd stood up, and had been pacing, increasingly agitatedly, as I chased her trying to explain the logic of the thing. At some point she walked to the waiting room and sat down, I joined her, and it didn't seem like a good idea to turn on the lights. Our waiting area has lighting better suited to an operating theatre. I remember looking at her profile as she, quite shaken, dabbed her eyes, in that half-light, and at that moment it struck me, very clearly, that on some level I... not exactly wasted a decade this young woman's life as establishing the conditions that allowed her to let it waste away. Inprisoned her, unwittingly, into the practice's, and my own, arc of decay. Suzy, I wish you I could have told you all this sooner, but I only realised that day in the waiting room, and after that, well... this is why I am writing the present. I'm also sorry for hiding from you some important matters concerning the practice, but one confession at a time. # 2 Suzy We hired Suzy Walker to work at reception, fresh out of university - some media studies BA or another, as were so popular back in the Blair years. She wasn't, like a lot of young people getting their first job in town, a local. Our little southern England seaside town is the next one over from a little southern England seaside city, which has a university, equally southern in its littoral Englishness. Not just the kind of university to *offer* BAs in media studies, for they all did back then, but the kind of university _famed_ for its BAs in media studies. This is why Suzy was studying there, and chose to move in town, rather than in the city, for reasons that were not immediately clear. It was apparent early enough that she had been studying, but never a 'student'. She much preferred, even pre-graduation, the quiet atmosphere of our town to that of city with a large student population - still, these cities have suburbs, free of students. She had options. Living on her own was non-negotiable, and she could only afford it here, or so she'd said. The truth was she could afford it *better* here. Getting to know her, I grew to think she had engineered her life so as never to bump into a peer unless she wanted to, even at the price of significant inconvenience. What I thought was significant inconvenience, until she told me how long she'd had the bike. We weren't as close then, so I only learnt that she owned and rode a motorbike in the first place, two, maybe three years after she'd started with us. She never rode it to the practice - she lived a street away, and rarely in town, since we're in the centre. But she would spend sunny weekends going on country rides, and even, but only once, and coming out of it vocally disinclined to repat the experience, attending a biker's meeting. Back when there were enough GPs at the practice to meaningfully gossip, Suzy's bike, and her, shocking, odd, *bikerness* became the talk of the practice. None of us had seen the machine, let alone her riding it, The Bike was a thing of legend. She never parked it in the street, and rented a garage somewhere she would not tell us. She had shown us, a picture, once, and very reluctantly. Of course by now I have seen both the bike and its more recent, reluctant replacement, and once ridden pillion - I remember Suzy teaching me the word - on the latter. I'd later learn that she had loved bikes as a girl like others love horses - and equally young. I remember her telling me it was learning about the influence of the bicycle on women's emancipation, aged 15, that sealed the deal - surely a motorised version meant that much more emancipation. Suzy had worked and saved through her sixth form to buy a bike the moment she started university. Living the next town over was a reason to use the bike, justify its purchase. The job we hired her for was a maternity cover, and we did not expect her to stay when the receptionist did not come back. Later I figured she'd come to the town so she could ride the bike, and that having graduated, but acquired little taste for the city, she'd no reason to work anywhere but the one place she wouldn't need to ride to. This way she'd save her riding for routes more varied than a commute. So she stayed in town, and stayed working here, with little career progression save a BSMSA qualification and membership so her secretarial duties could extend to the medical, when the practice lost its medical secretaries. We lost more and more staff, GPs retired one by one, and Susan was there through it all, doing more and more, always with her quiet, precise efficiency. Suzy, when Dr Raj retired and I told you I'd continue if you would, I didn't realise this was trapping you into the practice, out of my sake, although I knew you well enough, by then, that I should have figured it would have that result. But really, we should have folded when Sandeep retired, I could have done my last decade comfortably in a big surgery. I trapped myself into the practice, out of your sake. Either of us could end this arrangement, but because of the differential in power, it was impossible for you to end it. But really, maybe we were all comfortable with it. Neither of us really want to contemplate an alternative but you will have to find one, and since you would have had to sooner or later, it was my mistake not to give you your freedom, whether you wanted it or not, much sooner. So in that moment, I felt like I had more than the typical employer duty of care, and that, in the face of her distress, I owed Susy an explanation. Something to settle her mind. She's incredibly resilient, but there are moment when frustration and upset accumulate, cross a certain threshold, and she melts down. This was a high-risk situation in this regard, less because of the emotional impact of the deaths themselves than because of an unexplained coincidence. So, before I'd even sat down next to her, my mind had been working overtime on an explanation. Sniffling and sobbing, Susy says two, like that, what are the odds? I insisted on the existence of a rational explanation. Both of these two were old, at risk - they get scripts every month, so patients like them will always die within a month of their last prescription - that's a 25% chance they'll die a week after. And that's just for the scripts, some months we'll order labs or log in referral results. We'll have plenty of reasons of pulling up a patient file on Harley in the week preceding their death. Sure she says - already, the distraction of the argument, the challenge of constructing statistically plausible hypotheses and counterfactuals was pulling her back from the edge of a breakdown - we go through plenty of files on Harley every day. Some of these, statistically, will die in the week after. But the *freezes*? Twice in a row? Not in a row, I say, you told me there were freezes in between. Suddenly she wasn't so sure - not so sure that they happened on a patient file. Now it was my turn to pace the room. I think, retrospectively, that it was also my turn to be upset by a coincidence I couldn't explain. I'd been quite happy figuring out that these were patients whose records saw a lot of activity, but the fact of the co-occurences was throwing me. So I wasn't quite sure who I was trying to re-assure when I pointed out that both Sarah Sullivan and Padma Patel had been with the surgery forever, with increasingly complex chronic conditions, and correspondingly large patient records. Surely meaning their file would take more place in memory or something. The factors that made the patients more likely to die would also make their files bigger, and called to be pulled up by Harley... I mean pulled up by one of us, on Harley, more often. This only looked like an incredible coincidence if one presumed the freeze, the patient's record being displayed, and the patient death to be statistically independent. But in reality, their likelihood increased with the severity of their case, and the amount of care they received. Suzy was dubious, claiming this was the first time this happened. As far as she knew, I retorted - she's not aware of patients death as a matter of course, I am. But she notices coincidences, I don't. For all we know there's been three such cases every year on average for the past decade and we're only noticing now, because the circumstances of Harley's freeze with Sullivan had put it, prominently, on our radars, then her noticing this first coincidence primed us to notice the next one. Also, she said, freezes, in general, have increased over the past couple of years: there's three parts to the "coincidence" - she raised, then flexed the first two fingers of both hands - and the last one has only come in recently. Exactly I said - I was relieved to see her arguing for the rationality of the phenomenon rather than against it - there is a perfectly reasonable explanation. This was our working hypothesis: these coincidences were more frequent than we assumed, but our observation biais had kicked in. This explanation satisfied the both of us. So when it happened again in December **XX**, it didn't surprise me. Particularly as I'd called up the file of the patient in question - 78, male, overweight, lifelong smoker - as he'd just been admitted to hospital for pulmonary complications of this nasty strain of flu. He, too, died within the week. The puzzling thing is that December brought with it two more elderly flu fatalities in my flock, one of them with a long history in the surgery, and whose patient record must have come up on screen half a dozen times in the week preceding, without freeze. Ironically, having Harley freeze on these two too would have re-assured me: it cannot be a rare coincidence if there is a causal relationship. But these two pointed at chance still being a factor. I had not told Suzy about the third instance, because by the time he'd died, Harley had frozen on another file: the patient's profile was near-identical: already hospitalised by the time of the freeze, chronic illnesses and risk factor complicating a flu whose strain that we knew to kill even without them. What was different was the patient record: he had only been with the surgery for three years, when his daughter got him to move in with her family. Medically, his case was as complex, as high-risk as the others. Administratively - digitally, though, the typical three decades of patient records had been condensed into a single entry, a four page medical history with current and past treatments. From an IT perspective, the record was low-risk for freeze under the assumptions I had used to reassure Susan - and, as was growing on me even at the time, myself, really. So after he died, I immediately engineered circumstances that would minimise Suzy's work on patient records, because I didn't dare think how she would take more such coincidences. This was unsettling enough to me. And, in order to get to the bottom of this - back when I expected there to *be* a bottom to this in the first place, Suzy and I started logging freezes. It wasn't a hard sell - she did not know about the latest two deaths, as far as she knew the last "coincidence" was a month before, but she remembered it, and our conversation, obviously vividly. I pitched this as an exercise to establish what the odds, indeed, were. For this we need to get a sense of the frequency of freezes, look for any patterns. We had a notebook on Suzy's desk, where the both of us would, for each freeze, log their time and context. Over the first 6 weeks of the experiment, we experienced a bit more than thirty freezes, mostly observed by her, as, in spite of my keeping her away from patient files, she was doing most of the admin on Harley. Something we found out, but did not surprise either of us, was that freezes would only happen on our surgery records program. Web, spreadsheets, documents, export to the NHS data octopus... all worked fine. This was, we think, not just because we use the surgery management program more than any other: later I spent a weekend using Harley for general IT, and was able to leave the machine running without crashing for the 48 hours, when the typical median-time-before-freeze hovered around 6 hours. The software was a factor in the issue. I started, weekly, porting the entries in the freeze log book to the first spreadsheet in the attached data folder: `freeze_log.xls`. We could presume the computer to be started around 7.45 every morning, then restarted after every freeze. We were thus able to evaluate the median and standard deviation for the daily and weekly frequencies of freezes, as well as their distribution by context. In this instance, which screen the practice management software was showing when Harley Froze. Now, without an idea of the baseline, the way time using the program is distributed around various screens, there was no way of detecting which screens were particularly prone to it, but at a glance, the distribution appeared to match what I intuited of our patterns of use - at this stage, the sample was too small. With one exception, ironically: the patient notes screen in question. Enrolling Suzy in this data collection exercise put me in a tricky situation: I'd hidden the last two patient freezes from her, and I knew form the start that, should one happen again, I would have to decide whether to tell her or not. Thankfully, none did over January, and the beginning of February; I went on to forget this conundrum, and got stuck into the statistics. By the time of the next occurrence, I was considering the exercise an endearing little pet project, having forgotten the unsettling implications that lead to its genesis. # 3 Tomazs February had been unexpectedly mild, hints of an early spring that would soon be dashed by another cold snap. I was meeting Nancy Fletcher for a check-up: she'd recently stared on HRT, with resounding success. This was a Thursday evening; this I know because Suzy is off on Thursdays afternoon, and I schedule Ms Fletcher's appointments, whenever feasible, at times when Suzy isn't in. The two of them were the same age, and had met at university. Nancy grew up in town, went to uni, then to London. When she came back with a husband and their first kid, then just an infant, they registered at the practice. In fact, the husband did, and it is only at Nancy's first appointment (a check up on her second pregnancy) that she bumped into Susan. I wouldn't say that sparks flew, exactly, but the emotional tone of the encounter wasn't the typical bumping-into-a-uni-friend one. In fact, 'bumping into a uni enemy' would be more accurate. Susan would later told me of a 'Queen Bee of undergrads' with her 'mean girls', which I understand to be a reference to a film, and their conducting a campaign of what we would now call bullying, upon Susy. Regardless of the reliability of this evidence, Susan was visibly rattled by her meeting a peer that had, in the space of less than a decade, grown into a comfortable, settled, successful by societal standards, life. Moving back to the hometown as they started a family, with a cushy job in marketing, working from home most days, with the occasional trip to the city. This wasn't the life Susan wanted, far from it, but this did not mean she enjoyed being reminded of its possibility. I thereafter arranged to minimise the chances of them meeting. This is not relevant to this story, though. At the end of our chat, having sent her prescription out, I took a detour I usually take with some of the parents amongst my patients. You see, when people have children, their GP tends to see them a fair bit as infants - vaccinations and the like; then still, albeit less, in their childhood. It is when the kids enter adolescence one must watch out. It is the parent who has the relationship with the clinician, and once their child enter their teenage years, most will have an automatic revulsion towards anything associated with their parents. This to say that we will only tend to see them for serious illnesses or accidents, and teens will drop off the practice's radar unless a modicum of effort is invested into checking in on them. So, when having the non-clinical - but ever so important - family small talk, whenever there's a teenaged child in the picture, I'll check the last news we've had of them, and if need be, arrange an appointment. As far as the Fletchers were concerned, I'd seen each of the younger daughters the previous autumn, but I wanted to enquire about their eldest, Tomasz. Chatting with Nancy, she tells me he's had offers from all of his choices of universities, and would be going to study Engineering in Manchester in October. Assuming of course that his marks on the A-level matched his school's expectations, but this was unlikely to be a problem. This always happens and this is why I have these chats: I had last seen Tom about a broken wrist, and in my head he was still barely fifteen. I told Nancy it would be a good idea to have a full check-up before he goes to university. Young adults, particularly the boys, can be... less than diligent in matters of health - not that the girls are particularly keen to sign up to a GP the moment they start university either. As Nancy was telling me she will tell him to call, and he will, being such a good kid, I pulled up his file on Harley. When I tried to scroll down to his history, the screen shifted by an inch, then froze. Nancy was merrily telling me about all of Tom's achievements. Sons will never found a bigger fan than their mother - especially if they're only one and/or the eldest, both the case here. I'm nodding and uh-uh-ing, as I wave the mouse and tap keys, to no avail. I tell Nancy Fletcher I would like to see Tom sooner rather than later, ideally within a week. Oh, I'm sure you will, she tells me. He will surely call tomorrow and arrange an appointment next week. I ask her if we could perhaps schedule it now, whilst we are talking about it. He is, after all, still a minor for another month - his date of birth was still visible on the frozen screen, its top just cut out by the edge of the display. Oh, she wouldn't want to meddle, to - she used a word that was new to me - 'overparent'. Quite understandable, I reassured her, I look forward to his call tomorrow. I restarted the computer and walked Ms Fletcher back. I logged this freeze directly in my spreadsheet. Susan had access to the notebook, but not to the digital record. Surely, a strapping young lad like Tom was unlikely to suffer an unexpected demise, but still. I'd keep this one... what is it Jack says? "on the DL". I wasn't too worried about this at the time. Harley freezing during my chat with Nancy Fletcher was, to me, an annoyance, and a very minor one at that: I'd done all I needed to do, administratively, it didn't even affect our appointment. Sure, I was keeping track of the freezes, tallying and tabulating, but I had nearly forgotten why. Although perhaps it would be more accurate to say I didn't think of myself as worried. But when on Friday Tom Fletcher failed to call the surgery, a certain sense of disquiet came over me. At the time, I was thinking this was more an irritation at the recklessness of youth, at the overconfidence of a loving mother in the conscientiousness of her son. But still, something must have been playing on my mind, because on Sunday, I took a walk in the park. I need to give context here: my taking walks is not rare. Neither is my taking them in the park, or on a Sunday. But I knew Tom Fletcher to be a keen footballer, playing winger positions for our local youth F.C. I know this, not from the Fletchers, but because when I started to practice here, I was briefly the medical officer for the club, and have stayed on drinks-at-the-pub terms with its coach, Tommy Williams. This is how I knew what a 'winger' was, and how Tomasz's left-footedness, would have Williams field him 'inverted', on the right side of his attack formation. I also knew that the club would play on the park's field, with practice on weekday evenings, and games on Sunday. So when I set off to the park on Sunday morning, although it did not occur to me I was going to see Tomasz, in hindsight, I had all the information to know he was going to be there, and some part of me must have engineered this encounter. Still, at the time, I considered it a surprise to see the under-18 squad playing lads from a nearby town, this early in the year. Still, not a bad occasion to check up on him. We - to the extent there was a we - were leading two-nil. Standing on the edge of the pitch on that sunny but crisp February morning, I thought to myself that eugenicists had it all wrong: miscegenation yields splendid specimens. Tall, blonde and Vitruvian-proportioned, Tomasz had a a healthy English boy look; furthermore, as his mother had reminded me three days before, he excelled academically, in addition to his evident athleticism. I do remember him as a smart kid, and so is Nancy, but as far as the physique went, this was all down to his father. Save perhaps for a roundness of jaw, and shorter brow, which made the kid pass as Kent-bred. Mirko, the husband Nancy bagged herself in London before coming back to roost, was a Polish IT something or other, moving in when Poland joined the EU. But what he got from his mother was her last name, at Mirko's own insistence. Smart man, this was bound to open a lot more doors. Healthy young lad either way. Although, I observed to myself, perhaps hogging the ball and playing a bit showy, but who could blame him. I had already turned away, in fact, I was halfway to the other side of the park when, even from there, I heard the commotion. My blood ran cold. I wouldn't say I ran back, but certainly marched as briskly as I could given my age. By the time I arrived, 911 had already been called, as no pulse could be found. I was out of breath and sweating watching Tommy pace, on the phone to the emergency services, whilst the game's medic gave CPR to Tomasz. His emergency care was clinically exemplary, with the field medic starting resuscitation *statim*, paramedics arriving quickly starting intubation, the full protocol. Let the record show that the system, in this instance, worked. If his heart had started again, I'm sure he would have turned out to suffer minimal ischemic injury, but I already knew it wouldn't. I was drenched in sweat, but felt colder inside than the February air. He was pronounced dead en route, although by the look in the paramedic's eyes, they knew they were loading in a corpse. This very afternoon, pretty much right after hanging up with the hospital, I went to the surgery, and moved Harley to my office desk. I moved the shiny Macbook that usually sits on it - closed most of the time, to Suzy's desk. Sat at it, thoughtfully, until darkness fell. There was a square of darker wood on the desktop where the processing unit had sat for twenty years. I opened the Macbook, ordered a wireless keyboard, mouse and a stand, next day delivery. Then I started the second spreadsheet. If an event has 1% chance of happening, every day, there's a 97% chance it will happen within a year. For 0.01%, over twenty years it's still more likely than not to happen. Once. This kind of stuff I knew how to work out algebraically: work out the chance of the event not happening by compounding the base probability, to the power of the number of draws. This is how likely it is not to happen over a year, or twenty. The reciprocal to 1 give the probability of it happening. I had tinkered with this stuff in another corner of excel, but only at a very basic level, finding the duration at which coincidence have a 50, 80, 95, 99% chance of occurrence based on daily chance of occurrence of each event. The twenty years thing had re-assured me, somewhat - this could explain the latest, and the size of the patient file the previous three. But every week passing without any... I almost wrote false positive. Every week passing without a patient freeze - one that the patient survives, made the base rate for freezes on patient file lower, meaning the coincidence itself rarer. Thursday, if anything had re-assured me. And yet I took that walk in the park where Tomasz Fletcher was playing. I started building a more complex model, taking in account categories of patient with probability of freeze versus probability of death. I went looking up textbooks and web pages, brushed up on a lot of statistics and quantitative analysis. I do not think I had worked that hard since med school, but, like then, it felt very stimulating. Certainly stopped me thinking of *other things*. At four in the morning, I decided to call it a night, which I took in my office. Thankfully sleep came fast. I was awoken, tense and achy, by Susan coming in. I used to love a night on the examination plinths when I was a younger man. Barely cushioned, really just some perfunctory yield. In my thirties I had a phase of sleeping horrendously poorly, until, as a why-not-couldn't-hurt, I changed by bed for a hard futon. Its feel brought me back to nights on gurneys and plinths as an intern, grabbing whatever sleep one could during back-to-back shifts. My body recognised that firmness and I was asleep in minutes. But that Monday morning, I wasn't in my thirties anymore. I explain to Suzy that - yes, I'd heard about Tomasz, I was there, actually, and didn't take it well, I was going to close for the day. Yes, I've swapped Harley and the Mac, I'm also getting her keyboard mouse etc. I just needed her to stay long enough to accept the parcel and shift whatever crumbs pass for my appointment diary to another day, then she could go home and I'd pay her for the full day. These days it was already dusk when we closed shop, so maybe she'd appreciate a ride in the sun. Yes I was OK, just hadn't slept well. Yes, here. I worked late. I was distressed enough by the whole thing; I couldn't get the numbers for an odd-son chance of occurring under once per millenium when accounting for Tom, even with generous assumptions on the underlying probabilities. Before going home, I also asked Suzy to draw up a referral to get Mirko checked for Hypertrophic Cardiomyopathy. This came back positive the next month, which was very reassuring, although by that point I was already reassured when the coroner called me the following Thursday. Tomasz post-mortem tox screen had thrown up MDMA, amphetamine and cocaine, no less. For minors it is our ethical duty to pass this information on to the parent. He thought the family GP might make for a better bearer when it came to this news. I had no reason to be re-assured. Both congenital heart defect and drug use may explain Tom's death. They did not explain Harley predicting it. What is information did is allay me of a fear I didn't, had the time, realise I was nursing: that it was, instead causing them. The next occurrence was in April. Like over the winter, this was an elderly patient, and one already in hospital - one of those hospitalisation when it's odds on they will come out by the front door. Cancer. On the fifth day, Nigel Ferguson passed. Long file at the surgery a further re-assurance. And so it went. Margaret Walker (no relation) in May, also not a surprise. With her immune system, she'd been on borrowed time for the last three years. One in June, two in July, none in August. I don't recall the names off the top of my head, but it's all in the spreadsheet. Cancer, heart failure, kidney failure. Elderly patients. Oh, and plenty of other deaths without freezes. Well, not 'plenty'... some. And for at least one, I am sure I had the patient file on screen the week before. Still, no freeze on patient files except for those of the then-deceased, so whilst the clinical circumstances of the patients were re-assuring, what wasn't was that for ten months now, every single patient whose file was on screen when Harley froze died within 2-6 days. I wasn't keeping the record up to date, let alone the models, all the numbers tending ever more towards the unlikely. I accepted the phenomenon with a near mystical detachment, like the very joke was that the universe behaved rationally *except* for this one wink, at me, for my hubris of thing I could understand its mechanics. This to say I wasn't afraid. Nothing like I would grow up to be later. # 4 Emily The first freeze of September had been in keeping with the 'new normal'. The second one, a Sunday afternoon - I found myself working weekends a lot this year, seeing as I was now handling all the data entry, to Susan's constant protestations. In fairness, I did get her to take a lot ofy my plate I'm exchange. She has not only taken down, but revised and in part ghost-written a letter to the editor of the British Medical Journal. Not about this, nothing about this - primary care perspectives on insuline resistance in the elderly. Before Emily, I still pursued academic correspondance with the scholarly community. If this is what you call readers of the BMJ. Emily Sheppard was 41 years old, a patient for more than ten years. She'd had a brush with breast cancer two three years into our relationship, and we had grown quite close for the ordeal. Since I've had her - she's had me had her, really - on and off antidepressants, with some concerning moments. She wasn't depressed at the moment, as far as I knew. She certainly wasn't on meds for it, but she's supplemented for a thyroid issue, and I was renewing her 8-weekly when Harley froze. I was overcome by the same cold dread that came over me when Tom collapsed. I stood up, paced round my office, then went out to pace round the block. Came back, and the computer was still frozen. Well, who knew, maybe patience was the trick all along, but no joy. I restarted Harley, put the prescription in, then called Emily. Apologised to disturb her on a Sunday. Not at all she said, she was on a girls' weekend at a posh spa-and-estate, ostensibly having a great time. I wasn't convinced. I've been depressed, and whilst I've never in the strictest sense enjoyed a girls' weekend at a posh spa, I have socialised and presented a cheery demeanour. I explained I thought I'd called now since if any of my patients wouldn't mind it was her, I was renewing her script, and realised I never had an ECG from her - the NHS changed their guidelines for the over forty. I had no reason to suspect anything, but this was standard, and she must have heard about the Fletcher's son. When could she come in to the surgery for for it - we could do a full physical, it'd been a while. The lies came easily, and she agreed to come in on Tuesday morning. I did not sleep well that night, nor Monday. Nor, as it would turn out, Tuesday night, as she was a no-show that morning. Susan definitely picked up that something was wrong. I gave her the rest of the day. She said it was rubbish biking weather and she was worried about me. I told her I was fine. I pulled up Emily Sheppard's home address on Harley, and it is only Susan being still in the next room that stopped me talking to the computer. Asking where Emily was, and what happened. For after all, the computer knew. The computer had known on Sunday. No one home at Emily's. I went back to the practice and called the hospital. Emily Sheppard, it turns out, what in the passenger seat of her friend's car, on the A3, on Monday morning when a lorry driver, on the road for 36 hours nodded off, swayed across lanes and clipped their tiny Fiat with the back of his trailer. He went into the ditch at a speed too low to deploy the airbag, yet sufficient to break his neck. The car flipped, and the handbrake pierced Emily's femoral artery. She was pronounced dead at the scene by the paramedics. The friend had cuts and bruises, but was otherwise fine. Well, as fine as you one can be after half an hour strapped upside down in her seat watching her friend bleed to death. I do not know why Emily's death, specifically, sent me off the rails. To be sure, the prediction not being of a medical death was unsettling, particularly given its gruesome nature, but the Fletcher boy's own was equally unlikely. In fact, they were all impossible to explain: even if individually the probability of a coincidence on an elderly patient whose file is big, and pulled up often may be higher than it seem, their rarity only compounds when looking at eight such 'coincidences'. I suppose it was, first that I'd gotten used to a certain pattern, that had been broken, again, and that, as I realised, I was affected in proportion with how the death on its own would have affected me. An elderly patient with cancer, a little. Tomasz Fletcher and Emily Sheppard, the former for his youth, and my presence at the scene, the second for our relationship, the horror of the details, and the irony of my contriving a medical examination only for her to meet her fate on the road. Whatever the reason, even if it does not make sense, I took it poorly. My zen-like acceptance of forces beyond my understanding suddenly went and took a walk. I started sleeping less and drinking more. Suzy installed a meditation app on my phone. There were more in the months that followed, at the same rate, roughly. It's all in the spreadsheet. It kept going, and I kept going with it, as if on autopilot. Instead of equanimity, my acceptance was fatalistic - literally, I suppose. Fate was at work. Susan grew worried. Almost a full year passed, with some month heavier than others. Often the additional freezes would bear no news for a long while, then later news of a suicide or car accident would follow. Once each, an electrocution and diving accident. The next turning point would be in the next August, after a quiet period, both for the practice and in terms of freezes. A Tuesday afternoon. We'd gone back from a lunch of chips on the pier with Susan, whom I suspected to have suggested such an out-of-character outing in an attempt to cheer me up. Patrick Donnegan called to cancel his appointment tomorrow, his pee is burning so bad, he's going to A&E, I told him thank you for letting me know. I went to update his file, and Harley froze. I put my elbows on the desk, slid my index fingers under the pads of my glasses, where my nose gets sore at times, sliding them, and the glasses up, and pushing my face in my palms. I was in such a foul mood that, I admit, I - not exactly thought, it was too brief for this, but felt a jolt of *serves-you-right-you-fool-you-ignored-the-signs-for-so-long*, before simultaneously feeling ashamed, and realising this was a UTI we were talking about, not cancer. Paddy Donnegan was on his merry way to the hospital, where undoubtedly they would diagnose a case all the more severe because he's left it so long, and start him on IV antibiotics. I spread my fingers and open my eyes. Harley's screen was a blueish-white blur of boxes and scripts. I let my glasses fall back on my nose. Right across from it, at the top of Donnegan's file were the words: 'Pennicillin allergy'. # 5 Patrick I rushed to my feet and slapped the phone off the table in an attempt to pick it up. I replaced it on the desktop, sat - and tried to calm - down, then dialled Donnegan. Asked him if he knew to make sure to tell them he was allergic to pennicillin. I'm allergic to pennicillin, he asks? The old fool. Unbelievable. Yes, well, come to think of it, I think it was his daughter who told me. Oh, she would know. Well, so should they at the hospital - I glanced sideways at Harley's screen, still showing Donnegan's notes, and got the sense it dimmed briefly, as if winking at me. I shook my head. They *should* know, I continued but juuust in case, you understand, etc. Thank you doc, good thinking. He asked what would have happened if they'd injected pennicillin. Oh, I told him, you'd immediately go into anaphylactic shock, with a host of unpleasant symptoms, the most concerning of which being your airways closing up and you'd find yourself unable to breathe. I delivered this in an unwitting cold monotone that immediately struck me as very concerning. But not to worry, I quickly continued much more cheerfully, because they'd notice and immediately push some norepinephrine. And, I pointedly did not continue, with your elevated BP, you'd most likely burst that brain aneurysm we found last year, stroke out, and in the most optimistic prognosis only enter a coma today, before dying Monday at the latest. I know, because my magic computer told me so. Donnegan's daughter flew him back from the costa del sol, concerned at his depression she was mistaking for dementia, but foremost concerned at his denial of either charge. Big shot London lawyer, Saoirse, woman of means; she moved him back into the family home, with thrice-weekly carer visit. This to say he was registered in our practice straight from abroad, quite possibly making the NHS unaware of his allergy save for the data export we did the week of his registration. If it happened. We have had our ups and down in the reliability or our protocol, and although it is better now, data does fall through the gaps. Predictably, I later confirmed this note was absent from the record held at the hospital. I thought it a possibility I had saved the man's life. I mean, of course, as medics, we save lives. But not like that. This moment transformed me. I shut down Harley, and walked out of the office with a spring in my step, and told Susan the chips had worked, I'd cheered up, so much so I fancied desert, and I took her for macaroons and Riesling at the Michelin-starred restaurant. Yes, it lost the star nearly five years ago, but it's still the same restaurant, and we still called it that. This buzz kept on for the following months, although it never topped the highs of this day. I was a changed man; I wasn't cursed - quite the contrary. Instead of seeing... it, as a curse, a sleight against my rationality, I realised I had been blessed with a medical oracle. There was nothing I could do for cases like Emily's, or Tom's. But some could be saved, and I would, from then on, endeavour to try, aided by my trusty sidekick clinical divination machine, Harley. In our domain, it is surgeons who are typically accused of developing a 'God complex', very rightly so in the case of some surgeons of my acquaintance, although others belie this stereotype. I wouldn't go quite as far in my case, but I certainly approached this part of my clinical care, not with a sense of divine agency, but that of being an agent of forces greater than our understanding. Harley, or... whatever was behind Harley's behaviour could only alert us. Could only alert *me*, for I had been excluding Suzy of her chance to 'witness Signs', 'receive Messages' or whatever you would call it within this framing. I suppose that this is what prophets, messiahs and gurus of all stripes also do. In such a competitive world, why share your hotline to the transcendent? The next patients I "saved" did nothing to dissuade me that I was the mortal agent of supernatural forces, presumed benevolent by the mere fact of their reaching out to me. In September, the first freeze was on a stage 4 cancer patient, so I did not intervene. Then at the end of the month, Harley froze on Mary Fitzgerald's file. Over the period where I had this sense of mission, much as I hate to use this word, there was a fun part to it. Trying to figure out what would kill the patient within a week, given information available at the time, and whether I was in a position to intervene. Ms Fitzgerald, sixties, widowed, was scheduled for a colonoscopy under general anaesthetic. I reached out to the surgical team, pointing them to a couple of well-evidenced papers on propofol deaths in the elderly patient, and they decided to do the procedure under local anaesthetic. I followed up the day after her outpatients appointment, and she said she had no issues, 'except a massive pain in the arse'. Still, better a pain in the arse than a cardiac arrest in the operating theatre. Another result as far as I was concerned. This continued through the autumn. There were more freezes now, perhaps 2-3 a month. Then again I was using Harley more. There were these I couldn't figure out - I would never hear of a death through the clinical system for these, and, after Emily Sheppard, I knew better than to follow up. There were others I could figure out but couldn't do anything about - cancers, most of them. But amongst all these, were a handful where I could make a difference. Patrick and Mary were the first. Later that autumn, I managed to convince Francesca Lovell to cancel an elective surgery - the most likely cause of foretold death in a middle-aged, healthy patient; increased the dosage of Peter Tardjian's blood thinners, presuming his clotting disorder would cause a lethal embolism; **THIR PATIENT - NITROGLYCERIN???**. Another four I couldn't save: cancer, heart failure, cancer again, and what turned out to be a suicide. I am bundling this one in the medical fatalities, as I was notified of the patient's admission to ICU, and her death there the same night of acute liver failure. Paracetamol job. Actually, this is just how I know about it. I am classifying Lucy Ferret as a medical death because her depressive disorder had been diagnosed. In medical discourse, the term 'preventable death' gets used a lot. Of course, on a long enough timeframe, it is an oxymoron. Death comes to us all, etc., it is merely a matter of how quickly. A better term, perhaps would be 'postponable', or maybe 'deferable' death. Terminal conditions involve a grim calculus, intervene now to push back the inevitable? but at for how long? and at what cost? I am thinking of the human cost here, but for the NHS, indeed any public health system, literal cost factors. **LOOK UP ADDITIONAL YEARS QOL**.. Still, these considerations are typically the remit of oncologists, cardiologists, neurologists, not family doctors like me. However, my limited agency in the matter, predicated as it was on correctly guessing, and meaningfully intervening in the space of days, meant there was little for me to calculate. If I could act, I would. Who wouldn't? I have little memory of the other freezes - these where I would not hear of a corresponding death. The spreadsheet showed three such instances occurred before the Petersons, and I see it also includes day and cause of death, which I must have filled retrospectively, in the following year, when I started cross-referencing with data from the Registry. But I am getting ahead of myself. # 5 The Petersons The Peterson were my neighbours. Well, not at the time, not anymore, but they used to be. they lived next to me for nearly twenty years. Ten years ago, their only daughter left home, so they downsized, although they stayed in town. I wouldn't say we were close, exactly - I do not know if there is anyone about whom I could say this anyway - but we were close enough to ask each other over for dinner every now and then. Not so close that we kept doing so once they moved away, though, but they remained my patients - certainly the parents did. Clinically, Cynthia was on my radar more than Jacob, on account of her rheumatoid arthritis. Olivia, the daughter, I hadn't seen since she left for university, and frankly, probably not since her GCSEs. I have already explained how these things go. Cynthia had been on and off so called 'disease-modifying' drugs for her RA - a term that seems chosen to manage expectations of therapeutic effect in both patient and clinicians - mainly methotrexate. She was on it at the moment, and I was renewing her prescription when the freeze happened. Now, with this kind of patient record - middle aged, a chronic condition, but not a lethal one, otherwise healthy, I would normally let it slide. That is, I would assume the foretold demise to be out of my clinical bailiwick, put it all out of my mind, and move on. But I liked Cynthia. Also, as I mentioned, I was developing a sub-clinical God complex; so I racked my brain for medical hypothesis. Methotrexate is considered the gold standard for DMARDs, although this too is a term I find odd, on account of the gold standard, in monetary policy, having been abandoned in 1914. It has side effects, as all drugs do, but we know what they can be, and have decades of research informing a cost-benefit calculation. Most of these side effects will be minor - the usual carnival of gastric queasyness, general malaise, rashes. I am not dismissing these, they can be difficult for the patient, but even they will take nausea and itching over losing joint function. But, more rarely, the molecule will have adverse effects on the body's own life support systems, foremost hepatic and renal. The life threatening potential of these is magnified by the poor medical literacy of patients themselves: a pain on the left side of the chest will have many calling into A&E, but the very same will go on with their day if the pain is on the right side of their abdomen. I was thinking along these lines, particularly given the complex history of both Cynthia and her husband when it came to alcohol. I do not pass moral judgement on my patients for this. I have had, fought, and overcome a drinking problem myself, although these days it was creeping back on me, as I think I wrote above. But it is not as an individual I refrain to cast the first stone, but as a clinician. Managing a patient's alcohol or substance abuse, as a medic, simply works better if one withholds judgement. But it happened to be easier in this instance because of my past. Also, Cynthia was in perimenopause, this I knew, and taking over-the-counter supplements for it. A bit of research threw up potentially serious interactions, impacting liver function, between Methotrexate and *actea racemosa* - black cohosh - a plant present in many such supplements. Ironically the plant has long had use in folk medicine as an anti-inflammatory agent, particularly for joint pain, earning it the more rarely used nickname of 'rheumatism weed'. So these were the notes I had in front of me when I called Cynthia, under the pretence of catching up on her menopausal symptom, to warn her about drug interaction, re-iterating the need to avoid alcohol. I could not, it turned out, have been more mistaken. I always try to call on the patient's landline if I have the number. Susan often gently mocks this as my insisting on methods and tools of the past, but my rationale here is that I want my phone consultations to take place in a quiet, private place, which the patient's home can be presumed to be. I know how hard it is to talk to a GP, so patients will feel coerced having the sensitive conversation there-and-then, even if they happen to be in a crowded train carriage. Cynthia picked up on the fourth ring. I explain I thought it quicker to call than to schedule an appointment, and since I had renewed her methotrexate, I wanted to have a catch up with her on her RA, maybe consider HRT for her menopause. Nothing unusual here, this is how I often proceed. Arthritis symptoms are good, she says - actually now the law had been changed, would I consider switching her to cannabis? I explain it is not quite as simple as this, the NICE guidelines being rather narrow, given her clinical profile she would have to go private. She asks me if I am referring to the prescription, or the supply itself? Cynthia, I say, this is very unlike you, are you asking me to give you my blessing for you to buy weed for your rheumatism? Maybe I'd been thinking about the wrong rheumatism weed - cannabis is relatively harmless, in itself, but all bets are off when it comes to illegal products. Who knows what edulcorants and pesticides are in this stuff? Not exactly, she tells me, but Livvy's got her to try a bit, and she was amazed at the effects. Yes, I say, but is she speaking about the effects on arthritis - oh, these *as well*, she answered, and collapsed in giggles. I try not to get irritated and play along - she sounds very cheerful has Olivia got her to try a bit *that very evening*? ... and the previous two, she tells me. She's down from London. Interesting, I reply, extemporising, then explained about the unknowns of street cannabis, whatever adjuvants could be in it. Also, smoking when on methotrexate is hardly indicated. Not to worry, it is edible candy, and Olivia has got a 'great guy' that only sells 'the good shit', plus she was telling her how easy it is to grow, they were thinking they could finally start using the greenhouse again. Cynthia Peterson, what has become of you, I thought. The headmistress of the local secondary, and oddly choosing this title over the more modern 'headteacher', growing 'the good shit' in her back garden? Going from head to pothead? In spite of the '-mistress', she'd never been Miss Turnbull, but I wasn't expecting her to turn into Bob Marley. Look, I said, maybe we could speak about this in person, could she come in Monday? By then I may be able to figure out what was going to kill her in Olivia's good shit. She couldn't, Livvy was down because they were flying to Morocco on Saturday morning. 5am flight, its going to be a nightmare, but worth it. Half term, she explains, ticket prices skyrocket, it was that or an extra thousand pounds. Oh-kay, I say, well, speaking of Olivia, is she being followed? Oh, she has thousands on Instagram, she guffaws. Drunkards are ostensibly a nuisance, but stoners are annoying in more subtle ways. Very droll, I say, I was talking medically. Well, she doesn't know about any GeePees, but she definitely has a good herborist. More giggles, and I overhear said Olivia joining in. My phone call was the evening's entertainment for the local all-female reboot of Cheech and Chong. During this part of our exchange, I had been taking advantage of the long stretches of near-hysterical laughter to pull up Olivia Peterson's file. Looking at the dates, she would have been 16 at last contact. In October, so she'd had sat her GCSEs, but I wasn't far off. I scrolled down to look at earlier appointments, and Harley froze. I had an instant flashback of the freeze during my appointment with Nancy, my pulling up the child's file, Harley freezing, my walk in the park, Tomasz death. Although my couching it thus makes it sound sequential, but it all came back at the same time. At least this time it was during a phone appointment, not in person. When I returned my attention to Cynthia, she hadn't even noticed, but at least the giggle attack was over. Actually, speaking of herborists, was she still taking menopause supplements? She was. I explained about the black cohosh, it's in a lot of these, very dangerous with methotrexate, etc. Oh, she was tempted to ditch the methotrexate, now she had something *much better*. Further laughter. I think I pursued this because the evident conclusion only occurred to me: not the methotrexate, not anything medical. It was the holiday. They were both going to die on their holiday, or on the way. I wrapped up the conversation, telling her I will text her the name of the plant to watch out for, because I knew better than trusting her to take down *'Actaea racemosa*' and the older, but sometime still used '*Cimicifuga racemosa*', given her current state. I'd put the phone done, watching Harley reboot, pensive. I wasn't scared, or even distraught. This to tell you the new level of reality I was operating in. Fate wasn't going to kill my former neighbours; it was letting me know of the imminent *possibility* of their death, and giving me a privileged chance to preclude it. I pulled up Jacob Petersen patient record. Nothing much here beyond the typical middle-aged men complaints. Watching the blood pressure and the cholesterol, etc. I'd barely moved the mouse when Harley froze. As suspected. Whatever was going to - maybe - happen, it would kill the three of them. As I said, I normally keep away from non-medical deaths. One needs to draw a line. But the Peterson were... not *friends* exactly, but more than acquaintances. All I needed to do was prevent them from going to Morroco Saturday. I started pacing the length of my office. Many options, all unsatisfactory. The simplest would be to pinch their passports, hide them. Hardly an option. I could invoke a medical rationale, plenty of conditions will preclude flying, but none of the Petersons had anything ressembling them. I could sabotage their car, although this would be assuming they would drive to the airport, and they did not strike me as the type to leave their cherished Range Rover in long-term airport parking. Calling in a bomb threat could do the job, but even in the throes of my near-divine mission, this felt beyond the pale. I guess I didn't want to prevent *all* passengers in the South of England from going to Morroco, just Cynthia, Jacob and Olivia. Slip a gun, sword, bomb replica or something of that nature in their luggage? Again, reliable option, no collateral harm, but requiring a level of access I didn't have. If I were in a position to do this, I would rather have pinched one of their passports, this would lead to much less trouble. I started thinking about all the ways one could miss a flight. But how could I arrange roadworks, protests, or a traffic jam in general, at such short notice? By this point, I was back at my desk, I'd pulled out my pad and categorises interventions. There were these at the airport level - bomb threat, flying a drone, and the like. This wouldn't require access to the family, but this very lack of precision meant it would affect many more beyond them. Then there were these on the way to the airport. Booking all the cars from all the apps, sending them on the other side of town? Complex, fiddly, unreliable. Surely this would make them late, but maybe not late enough to miss the flight. Would reporting a gas leak create road-works in time? Even if it did, this would be in town, and the detour wouldn't be too long - wouldn't be long *enough*. I'd pulled up Google's maps. Blocking the A-road going to the airport guaranteed a missed flight, and, unlike the bomb threat, disrupting road traffic for a couple of hours in the early morning was a risk I was willing to take. So, what? Fell a tree from the nearby forest? Clandestinely, in the dead of night? Ludicrous. This left interventions at the point of departure. An emergency, medical or domestic, would ground them nicely. I had plenty of options, pharmacologically, to induce a temporary malaise. But I couldn't imagine any opportunity to effect this, well, *poisoning*. Transient poisoning, but poisoning to be sure. I felt ashamed that it was the practical difficulty rather than the ethical reprehensibility that held me back from such options. By elimination, this left domestic sabotage. I crumpled the sheet of paper with my matrix of options, and tossed it in the wastebasket. I had put my coat on when I reconsidered, fished it out of the bin, and ran it through the shredder. I suppose a part of my was still aware that in spite of being mandated by Destiny itself, in the mortal world, my behaviour would, at best, *raise questions*. This was the end of the day, but I was aware I would have cancelled the rest of the afternoon and left to *case* the Peterson home even if it hadn't, and some part of me was already raising questions regarding my priorities. After some minor trespassing, I went home satisfied of a solution, and slept better than I had since Tom's heart failure. Which was fortuitous as I would have a short night of sleep that Friday. The following Monday, Suzy popped her head into my office door. I was never going to believe this, she started; do go on, I replied. I knew that plane that had crashed in Morroco this weekend? She'd just had Cynthia Peterson on the phone, the three of them were meant to be on that plane. Oh? I hadn't - I leaned sideways to put Harley's screen between my face and her eyes. Yes, but they missed it, they got up at stupid o' clock Saturday morning, and when Olivia took her shower upstairs, it all backed up their downstairs toilet. Clogged septic drain, they had to get a guy in. Was there much damage, I asked. She said I was fixating on entirely the wrong thing - what were the odds. Having regained my composure, I came back in her line of sight. I had worked, ever closer, with Susan for longer than many couples have been married. I am sure most anyone else would have picked up that I was hiding something, but Suzy is not very good at reading people, particularly when her attention is on something else. This is astounding indeed - are they all OK? They can't believe their luck, they've had a terrible 48 hours, but it turned out an amazing stroke of luck, they're fine. I did not ask about the costs again, and frankly I failed to enquire after - though I failed to enquire about much more. If the trespassing was minor, the property damage wasn't, although they were insured, and would be the firsts to deem whatever cost, discomfort and inconvenience well worth the trade-off. Is that why she was calling? Is she telling everyone? Oh no, Suzy responded, but since they cancelled their holiday, she wanted to book Livvy in tomorrow. Well, I huffed, another positive to this story, eh. Saved by a septic tank, what are the odds, she asked. Well, it is like these people who were meant to be at the World Trade Center on Spetember 11 2001, but missed it. We probably all dodge death a couple of time a decades, we just notice for disasters I said. She was unconvinced. I pointed out she lived on her own, like me, and how many times did she slip a bit in the bathroom, or trip on furniture, catch herself and thought that was a close call? Her face changed. Plus obviously, we notice the time we dodge death, because we can't the time we don't. I felt like I had won an award, been told ahead of the ceremony, and having to keep it all under wraps in front of another in competition for it. Certainly, I blame my emotional state for what I said next. There really was no need for anything but feigning amazement, and getting Susan out of the room. I said "It's like Harley's freezes". What did I mean, had there been more? I realised my mistake in the moment before she started speaking. I tried to back-pedal: no, well, yes, the usual. I was talking about last year. This time, she saw through me immediately. There *had* been more, wasn't there she asked; I was hiding something she said, it was obvious. I was, admittedly, quite flustered. Yes, okay, but I didn't want to worry her about it, because it was all old, sick patients - Maggie Walker, Ferguson. Harry Ferguson, the kid who drowned himself in the river, she exclaimed. No, no, *Nigel* Ferguson - cancer, remember I said. I did not know about Harry - twenty-two, I later found out the coroner ruled it misadventure, because of the blood alcohol level, but for some there are nights where the boundary between that and suicide is not a hard one. Harry Ferguson had had a previous bout of depression, and a first bout will more likely than not recur. Harley froze when I pulled up his file to send him a reminder to book an appointment, I had not seen him in nearly three years. Needless to say, I had not heard back. --- Was I OK, I looked pale all of a sudden? Susy was staring at me with an uncomfortable intensity; I realised then that she wasn't very good only at intuiting people's emotional state. But when deliberately paying attention to it, she applied to the endeavour the same fierce intelligence she did everything else. I seriously considered coming clean, although if I had, I do not know that what was to follow would have been precluded, in fact perhaps the opposite, knowing Suzy. It occurred to me I had been in much less slippery terrain when she had come in, all excited. I guessed, I sais, the Petersons, and, to think about it, maybe Harley's freezes, yes, it was getting to me too, I found it a bit creepy... Ah-ah! she exclaimed, somehow victorious. She knew she would 'turn me over', there was more weird stuff happening in the world than existed in my philosophy, Horatio. We will have me on the crystals in no time, she laughed, then, all suspicion seemingly allayed, told me of Olivia Peterson's appointment time and a couple more administrative items and pretty much got up and left. 'The crystals' was a private joke between us, from a time a particularly woo-woo patient told her she didn't need us, or any NHS care, because she was 'on the crystals now'. Going on the crystals was our term for falling into pseudoscience, an arc that a worryingly increasing number of our patients were following. Susan, hanging up from a frustrating conversation, once referred to a patient as 'fully crystallised'. This incident harshed my buzz, I thought of it as almost getting caught; getting caught not just hiding things from Susan, but really, having my - what? 'destiny-informed saviour'? - endeavour exposed. In my notes at the time, I had started to think of the mechanics of the universe as deterministic - hence the forecasting I was tapping into, but not pre-ordained - hence the possibility of my successful intervention. Constructing an ontology allowing my access to future information through a superannuated computer was, as this description ostensibly implies, much more difficult. I fell back to some hand-wavy notion as consciousness being fundamental to the universe, not emerging from matter, and the extinguishing of consciousness sending ripples through causality, affecting the material world. My hard boundary between metaphysics and physics notwithstanding. This would also explain ghosts, premonitions, and a whole raft of ideas not so long ago would have deemed, to use a technical term batshit. This to say it wasn't just the flu season that lead to what was to follow: I had started to consider the phenomenon, and its implications, rationally. Or perhaps rationalise them. And it is this newfound position of rationalism after a mire of confusion that got me to approach that year's flu season like I did. # 6 Flu Season In late November, of an evening renewing prescriptions in bulk, there was another freeze. Elderly female, diabetes, dementia, living in a care home. By this point, freezes hardly registered emotionally. This profile was typical of the deaths we registered - being notified in advance didn't make a difference. Still, I noticed that she wasn't vaccinated against the flu. The over-65 get free flu jabs every year, it is just a matter of getting them in. When this year's immunological vintage comes out, we reach out to patients to try and get them to schedule a jab at their local pharmacy. For retirement homes, we arrange someone to come in and take care of all their guests in one afternoon. This is now someone from the pharmacy, or an NHS community nurse: I have stopped doing the visits myself after the last one, ironically, earned me a rhinovirus that laid me low for nearly a week. I had heard rumblings of a bad strain this year. I couldn't be sure it would be the flu killing her, but at least, this I could try to do something about; in any case her death wouldn't exactly be a tragedy. Perhaps even a relief to the family, I even thought somewhat cynically. When Harley had come back up, I added some Tamiflu, for luck as it were. In these cases, medication pick-up and administration is handled by staff in the home. I emailed them a note explaining that given the current public health context, I'd taken the liberty of proactively prescribing antivirals, that they should administer should any flu-like symptoms manifest. Better safe than sorry, etc. This was maybe the whole of ten minutes, including a little wander to the water dispenser whilst Harley was restarting. It didn't feel like I was involved in a supernatural event, just a bit of new, additional admin, for a new, experimental treatment, that wasn't sure to work, but couldn't do any harm.