I'm a nurse and I find the 7th point on the post especially relevant. I will add a disclaimer that I never worked in the ICU so I can't speak for what happens in that type of unit.
There is a serious issue with the flow of information in healthcare, (or at least in the U.S, I never worked elsewhere to know if it's any different). But If you find something during your shift which will be important to know later on, it will certainly be lost as soon as you are off for a few days, or even as soon as a new nurse comes on. To think of a somewhat crude example, if you find out that it is much easier to obtain a blood sample from the veins on the left arm of a patients vs the right, many nurses will still stick the right arm countless times hoping to get something.
And you can leave a chart note about things like that or speak about it during report, but for the most part few people will think "hm, I wonder what everybody else had to deal with." They are probably too busy handling a thousand different things happening all at once. And, even if that is not the case, from what I observed it's simply not part of how things are done. And very often patients will get (justifiably) angry, saying "I've been complaining of x thing for days!" or some version of that. I think it would be much better for both patients and healthcare staff alike if there was a greater emphasis placed on focusing on the series of successes and failures that happen over the course of someone's care, not just seeing it as a single shift or a single problem happening in some isolated point in time.
> I think it would be much better for both patients and healthcare staff alike if there was a greater emphasis placed on focusing on the series of successes and failures that happen over the course of someone's care, not just seeing it as a single shift or a single problem happening in some isolated point in time.
I once had a week as a patient at the Mayo Clinic in Scottsdale AZ. There were many remarkable aspects of care there versus the impossible mess out here in the other world.
But the single most significant aspect of care at Mayo Clinic is that the doctors and nurses and techs get to read your chart before seeing you.
That's it. You write something in the chart, it doesn't get tossed. It might not get parsed completely, but the essential info is there. And the staff does not get penalized for reading it.
(The other big reveal for me at Mayo was the sheer scale and throughput of the system. Healthcare at Mayo did not cost more than healthcare in my small town. It. Cost. The. Same.
It took six months to get in, I had a week, then it was someone else's turn. I presume that the high paying "celebrity" customers can get seen more regularly. So it's not perfect. But holy cow I wish it were easier for healthcare professionals to do their job.)
In general for US healthcare providers there is little relationship between price and quality. They have to meet certain quality standards in order to operate at all, but outsides few limited areas they don't get paid more for delivering higher quality care. So quality (or lack thereof) tends to come down to organizational culture and management.
For someone that hasn’t heard of they Mayo Clinic other than through webite articles describing medical conditions, are they that desired/high-quality?
I'm not sure if you live in the United States, but Mayo Clinic is probably a top 5 hospital system in the US. It is legendary. Just read the opening paragraph from Wiki: https://en.wikipedia.org/wiki/Mayo_Clinic
The Mayo Clinic (/ˈmeɪjoʊ/) is a nonprofit American academic medical center focused on integrated health care, education, and research.[6] It employs over 4,500 physicians and scientists, along with another 58,400 administrative and allied health staff, across three major campuses: Rochester, Minnesota; Jacksonville, Florida; and Phoenix/Scottsdale, Arizona.[7][8] The practice specializes in treating difficult cases through tertiary care and destination medicine. It is home to the top-15 ranked Mayo Clinic Alix School of Medicine in addition to many of the highest regarded residency education programs in the United States.[9][10][11] It spends over $660 million a year on research and has more than 3,000 full-time research personnel.[12][13]
A little deeper:
Mayo Clinic has ranked number one in the United States for seven consecutive years in U.S. News & World Report's Best Hospitals Honor Roll,[19] maintaining a position at or near the top for more than 35 years.
IIRC, the Mayo Clinic invented the medical checklist. Like, people actually read the checklist and make sure that they tick off all the boxes they are required to tick off.
That's why they are number one. Because they actually use checklists.
My wife was diagnosed Devic's disease, a rare disease with a grim prognosis. Almost every paper we could find on it had the name of a doctor that worked at the Mayo Clinic in Scottsdale. We lived in Arizona at the time so we went there and found that doctor. He corrected her diagnosis as MS, not Devic's. They both suck but Devic's is much worse. We paid out of pocket and getting the correct diagnosis was worth every penny.
This is similar to problems that would often happen in car manufacturing. The person assembling the car the standard way finds a problem, but the problem doesn't get addressed, or information isn't disseminated correctly, so the cars go out with problems. Toyota developed a methodology whereby such problems are addressed immediately and fixes were disseminated immediately, and would not send a car out otherwise. That kind of obsessive attention to detail and "crazy" focus on quality is what made them the top automaker. But most businesses are led by management that refuse to believe that being slow or focusing on quality first will result in more profits. And none of their lower-level workers are trained on how to spot and fix quality issues, nor are they told to care.
Hospital systems are the same way. Moronic, scared management that is fine with these kinds of problems as long as the dough keeps coming in, ignorant of the fact that more dough would roll in (in addition to better health outcomes, which of course is not their first priority) if they would just focus on quality.
This is my favourite ever episode of This American Life, about NUMMI, the shop floor level and individually fraternal miracle that was created by workers at Toyota and GM in a CA GM plant, until management shut them down:
This episode is incredible! I also listened to it a few years ago. It provided so much insight into (a) Japanese vs American manufactoring and (b) the impact of poor labour union relations. (Please do not read [b] as me being personally anti-labour union. Some of the revelations from union members in that podcast were shocking to me -- drinking and drugging while on the manuf line!)
Very cool point. In my ideal world a whole nursing unit or facility would be a self-correcting operation. There are problems that a nurse on the floor can fix but it takes up time. If those problems were prevented to begin with, it would be much easier. I would like a system where the nurse notices a problem and simply sends it up to a manager / supervisor who 1) finds a way to handle the immediate problem and 2) always writes up and enforces a new guideline to prevent it from happening again.
Good managers probably already do this, but healthcare has a very short supply of such people. It would be great if this type of improvement were the standard across the board. Let's say, for example, that you have latex and non-latex foley catheters mixed in the same bin in a supply closet. Your patients with latex allergies have gotten a latex catheter put in more than once and it now becomes a problem. Well, someone notices the issue, sends it up to someone above and now there is a new guideline to place the different catheters at least 3 feet apart, or something to that effect. It almost sounds silly, but people would be surprised how many of these mistakes happen over and over again due to equally silly reasons / lack of basic prevention.
Not lost, but I've had a lot of trouble with information being captured incorrectly.
Things like date are pretty commonly messed up. I've also had doctors and nurses put their own, incorrect, interpretation on information I've given them when they repeat it to others. When I say "my child wasn't eating and drinking normally and had half of what they normally do throughout the day", it's incorrect to say "the patient didn't eat or drink all day". That's the type of shit that can look really bad if it's recorded and looked at later. But it's like nobody cares if they record things correctly.
I've also had trouble with people not doing anything with important information. Like maybe you should slow down on the morphine and oxy if the patient is answering fewer basic questions correctly than when they came out of surgery. But it's OK if they can't tell you their own birth date - just give them more and later order a CT ro check for a stroke. Sorry guys, but it should be pretty obvious you're putting them into a opium stupor...
I've noticed recently that there are people out there who simply can't accurately listen to others and repeat back what they say. It's not about being stressed for time, the skill is just not there. You say ABC, they write CBD, and they have no idea it's not the same thing.
OpenNotes can help a little with this, but only if the patient or one of their caregivers has the time and ability to do a detailed review of every chart note.
This comes up every once in a while when discussing the crazy 24+ hour shifts that doctors in residency are often assigned. One argument in favor of keeping the hours is that continuity of care is by far the factor most strongly correlated with good patient outcomes. So the argument goes that a change in caregiver is more detrimental to the patient than continued care from one doctor even if that doctor is sleep deprived.
I am not knowledgeable or qualified enough to weigh in on this, but it's something I've heard cited by multiple friends in the field.
As a physician, shift length is honestly a red herring.
As much as I hated doing 24-28 hour shifts on inpatient services, continuity of care does matter and errors do occur in handover.
You have to keep in mind that medicine between 12am and 6am is what we call “keep people alive.” 6am to 12pm after an overnight is for handover.
You’re not trying to diagnose a new illness overnight or make changes in management, your job is to deal with acute overnight concerns only. Furthermore, you’re supported by services such as RACE (an in hospital emergency response team) so you’re not dealing with critically ill patients alone. If you’re on a surgical service and need to go to the OR, staff/fellow + senior residents come in to help.
Acute care services where you’re seeing new/undifferentiated patients and need to be on your game, such as ER and radiology, tend to limit shifts to 8-12 hours.
> As a physician, shift length is honestly a red herring.
This is how the Stockholm syndrome feels. I manage a few T.A. in the university, and they barely can think after a 6 hours of teaching (two consecutive classrooms, with like half an hour of rest in each one for the students, and perhaps another informal half an hour in the middle). Sometimes they have to speak in the blackboard, sometime grade informal take home exercises, sometimes reply questions on the spot, and they get very tired. So we have a strict 6 hours per day rule. And if they make a mistake, nobody dies!
We had a case a few years ago in Argentina, when a child got an overdose of Potassium Chloride. The nurse was new in the hospital and in the previous hospital they had a different concentration, so she prepared a wrong dilution. [1]. Anyway, it's a problem that is common enough that the English NHS added it to a list of recommendations [page 8] https://www.england.nhs.uk/wp-content/uploads/2020/11/2018-N...
It looks like a random accident, but it's one of the silly mistakes that are more common when someone has worked 12 or 24 hours straight and has no checklists.
Not sure what this has to do with resident/physician work hours.
Nursing errors (i.e. administering the wrong dose) can certainly kill people. They also don’t / shouldn’t work 24 hours shifts (infrequently a nurse might work a double due to emergent staffing requirements, this is a systems issue though and not by design).
There are both technological (EMR and ordering systems) and human safeguards (nurses and pharmacy) protecting against “silly mistakes” by physicians.
Once again, resident physicians’ roles overnights are no where near as mission critical as a nurse.
You also identified a key point in why 24 hour resident call shifts are safe - we have checklists.
If I order the wrong med on the wrong patient on an overnight call shift this will be flagged by the nurse who’s checklist includes verifying order accuracy. This is especially true of medications that can have life threatening complications (e.g. insulin, potassium, hypertonic saline).
Please also note I’m only talking about places I’ve trained (US and Canada) where all of these systems exist. I cannot comment on other countries where the infrastructure is different, perhaps this is more of an issue in Argentina than it is here.
> By contract, [...], a resident has to serve eight hours a day, Monday through Friday, and do eight 24-hour shifts per month.
> “We work shifts of more than eight hours, which can reach 15 or more and with guards that are also on weekends. There are colleagues who work 40 hours straight,"
(The last one is a quote of one of the union leaders, so it may be a corner case.)
If it were so automatic and repetitive, it would be easy to pass the information to the next medic and have normal length shifts.
> If it were so automatic and repetitive, it would be easy to pass the information to the next medic and have normal length shifts.
Ward call for residents generally works like this:
I have an inpatient list of 15-20 patients I’m covering overnight, some of them I likely know as I’m often part of one of the relevant day teams (unless I’m flying in from another clinical service to help out).
I start by receiving handover from one of the day team members. We sit down together (or by phone) and go patient by patient on the list asking what their reason for admission is, any labs/results I need to follow up on from the day (e.g. patient A had a fever and a cough, we ordered a chest X-ray if it shows pneumonia start antibiotics), patient specific management plans (e.g. patient B may have a seizure overnight, he’s known for this and if it happens give drug Y.) and any patients that I specifically need to see (e.g. patient C was complaining of some belly pain this morning but has been fine the rest of the day, eyeball him in the evening and make sure nothing is brewing).
I then write these action items and notes down (either on paper or in an EMR patient list) for my shift and carry out the relevant actions from 5pm to ~10pm.
Between 5pm and ~10pm I’m following things up and seeing any patients I need to see. Depending on my service I may be taking ED/inpatient consults but that’s not the point here so I won’t get into that.
At 10pm I do what’s called “tuck in rounds” and call up to the nursing station and ask if any of the nurses have issues they want me to address. Often this is something like morning labs that haven’t been ordered, laxative orders, etc. If there are any patients I’m worried about (uncommon on routine inpatient wards) I will pop my head in the room to make sure everything is alright. Cumulatively, the evening usually represents 1-2 hours of active work (again disregarding consults because that workflow is very different).
After that, and until the next morning, I am either asleep in a call room bed or at home. I will only be practicing medicine if there is an overnight issue that needs addressing (e.g. a patient is short of breath, their heart rate is elevated, decreased level of consciousness). These acute ward issues are beaten into every physician from the beginning of medical school and we follow very routine diagnostic workups (i.e. CBC, lytes, glucose, VBG), many of which are codified in algorithms such as ACLS.
If a patient is really unstable I call the RACE/code team (an in-house service to deal with unstable issues staffed by an ICU trainee, RT, and ICU nurse with advanced training) who assume care while I provide support and context as the home service/MRP resident.
This is a very safe system. It is really hard to kill an inpatient with a medical error in an acute setting.
Now let’s pretend I handed over to a night resident starting at 11pm. Two potential sources for error arise:
1. We would go over the same process of “running the list” and discussing patients, except now it’s second hand information I’m relating (versus my initial handover was from the primary team/MRP who knows the patient intimately). Broken telephone / forgotten action items becomes more likely.
2. An acute situation happens overnight and the 3rd shift person alerts the RACE service, except now the resident from the home team/MRP has never actually met the patient (you don’t go round and familiarize yourself with sleeping patients) and has no idea what they’ve been like all day except from what I’ve told them. This creates a huge problem because now they’re reading through the chart/notes to make sure this is a new symptom and not something I forgot to tell them about, they’re also reading the chart to see if there were any action items I addressed in my evening shift that didn’t merit handover but may be related to the acute concern. Whereas with the same resident on a 16-24 hour shift you have a much better understanding of the patients and their unique circumstances.
Many, many, many studies have shown medical errors happen a lot more due to handover than physician fatigue. You can argue that we should have better systems/IT in place to make handover safer, but we do not. Even places with systems like Epic/Cerner, it takes too much effort to maintain the handover list with accuracy and direct verbal communication remains the mainstay.
Furthermore, it’s important to keep in mind that dealing with ward issues between 12am and 7am is also pretty uncommon unless there is a late admission or someone that’s active, but that’s atypical. On-call is for emergency coverage not active medical practice.
Even outside the medical field, it seems like most humans are pretty bad about both writing down and consulting notes. Even worse for the notes written by another human. We really aren't particularly good at transferring knowledge / experience and it takes a lot of effort to do a good job of it, so most people don't even make much of an effort.
This really seems like a problem that still needs a lot more attention, especially in critical places like hospitals and really any long term crisis response situation where there is important knowledge gained over time with a (poorly handled) hand-off to successors.
I had some exposure to formalized incident management[1] at a previous job. There, I learned a few formalities and practices that seemed valuable, especially assigning a single coordinator to be responsible for continuity of information and coordination between many independent actors over a long period. The coordinator role had explicit hand off to their successor where the stated purpose was to transfer important working knowledge and prevent the kind of problems you (and the article) describe.
I liked the way you framed this as something universal. Is there any field where one can quickly reference knowledge from your peers to just as quickly solve a problem in practice? Maybe it's asking too much. Though I suppose it wouldn't be necessary to get everyone on board with such an idea if you have that single coordinator who everyone knows as the reference point. Although, if you think about it, even then that person would have to be available 24/7, which isn't feasible.
With patient documentation specifically, what I would really love to have is a simple search mechanism for patient notes. This still wouldn’t solve the problem of getting everyone to capture the right information. But assuming the information is there, and I'm having a real hard time sticking that right arm, I would love to be able to search for "arm", "blood draw," "stick" and see what pops up. I hope it's not something I missed entirely, but I have never used an EMR with such a feature.
Indeed, seemingly simple functionality like you describe should be bare-minimum requirement for just about any kind of information system, and yet it's almost always omitted or implemented so terribly as to negate any practical benefit to users.
> a single coordinator to be responsible for continuity of information and coordination
Years ago I saw a talk by a VA (US veteran's health care) thoracic surgeon, who was trying to entice tech folk to address the following problem. A surgeon is both team manager and skilled technician. When heads down in the technician role, the management role suffers for lack of attention. Especially severely at the VA, which did randomized staffing of operating teams, so you don't get the "group mind" and practiced gap filling of team which stays together. So surgeons would say "do X", and being distracted, not notice the order was dropped, and then proceed assuming X had happened, with regrettable results. The VA surgeon envisioned a voice system which noted the request, waited, and then whispered a nudge in someone's ear "did we do X?".
I think this is a symptom of hospitals being understaffed, whether that's from a deliberate lack of hiring or an actual labor shortage. I feel like many of the problems in this aspect of healthcare could be solved if doctors, nurses, etc weren't run ragged with insanely long shifts and expected to care for a ton of patients.
I don't know how to respond to your post. I'll try: Money. Skilled people cost money, a lot of money. I disagree with both of these: <<deliberate lack of hiring or an actual labor shortage>> They are making do with the amount of money that is available. It would be wiser to focus on why healthcare is so expensive in the United States compared to other highly advanced countries -- France, Germany, Netherlands, Finland, Japan, etc.
I wonder if part of the issue is having to remember a bunch of random things about a bunch of different people so that you can apply the information at the relevant time. E.g. the left arm is easier to find a vein, but this information is only useful for 30 seconds a day during a blood draw, so it's hard for people to remember or even retrieve (how much notes do you have to read through to find this info?)
If the info were somehow magically there when needed, it would be used, right?
Your comment made me think about this for a bit. It is almost fun imagining something like seeing a short little warning floating up in the air above a patient's arm, saying "blood draws from here". It would be pretty darn cool. If this were possible, it would be worth seeing how much it helps with continuity of care.
To let my imagination run a little wilder, I tried to think of what a system like that would be in practice, and how it relates to the problems that we currently face with the systems we already have. As much as it would be interesting to have all that information available through some sort of AR, there are really three important things that I would like to see about a patient: code status, vital signs and how they get up from bed.
It's really crazy to me how even the simplest of stuff is buried in a chart or EMR. Most do show the patients code status easily, but quite often it is in a small little font beside not-quite-as-relevant stuff like their marital status and what type of insurance they have. Why isn't this in big bold red letters in every room and in every chart as soon as you open up a document? Even for vital signs you have to click through two or three different things to get the information you need (but thank goodness I get to see some stuff right away, like that ICD-10 code for unspecified follow up for dietary counseling!)
One thing I think a lot of people may also not realize is how little information a nurse often has to go off of when walking into a room. If I am answering a call light for a patient who is not one my assigned ones, and they are screaming that they need to go to the bathroom yesterday, and you see them with both feet planted on the floor ready to get up, you have a quick second to think about a few different things. 1) How alert is this patient? 2) How mobile are they, do we need two people in the room? 3) Is this someone with a massive diabetic ulcer who wasn't supposed to be putting any pressure on that heel at all and they are about to do just that? Of course, you can look at the whiteboard, but you better pray that it's updated haha.
So, going back to the AR stuff. If I could have a snapshot of all this information as soon as I walked into a room, it would be a life saver, especially for situations like the above.
I have great respect for what people like you do. I've lived all over Europe and a have some first hand experience with care in different countries. Universally, nurses are heroes and it's just a very hard and often thankless job. Not to mention under paid in many places. But there's a big difference in what I would label as institutional stupidity between different countries. Some countries feature a lot of mismanaged health care facilities where things are bureaucratic, slow, etc. For example, Germany is hopeless on this front. Being in the health care system here means an endless sequence of forms that need to be filled in with the exact same data points over and over again and over worked nurses dealing with all that crap on top of their normal job. It's beyond stupid. Nobody shares any data. And it's inefficient and dangerous for patients because no doctor or nurse can possibly have the full picture.
My home country the Netherlands is very different. My father had a stroke quite recently and spent some time in a very modern hospital where they are applying some of the latest insights for patient care. So, he was obviously hooked up to lots of equipment and intensely monitored. However, this hospital has separate rooms for all patients. Reason: it's best for the patients and helps them recover more quickly. Basically, more privacy for the patients and less restless nights. There are no TVs in these rooms. Instead patients are issued ipads with entertainment options and access to various things like indicating dietary preferences. Nurses carry ipads as well. Everything is digital. There are no paper charts in sight anywhere.
The rooms were modern, clean, and clearly optimized for making patient handling easy and straightforward. What struck me was the attention to detail and level of pragmatism in this. For example, my father's room had wall mounted hangers for folding chairs. These are for visitors. And when they are folded they are not in the way. The room had a whiteboard and a locked cabinet for medication and supplies. The doors are sliding. So, it's easy to move things in and out. Like beds, wheel chairs, equipment, trolleys. Etc. And so on. Just a really well designed and thought through design and architecture. Well managed and efficient.
BTW. This is not a private hospital: my country has a mandatory private insurance system: they can't reject people, people must be insured, and they can switch insurer. So, insurers mainly compete on quality care. Miserable patients and inefficient hospitals are bad for business and they are working to fix any issues there with hospitals. Which is why everyone, rich or poor, gets the same quality treatment in this hospital. It's way better than the private insurance I pay for in Germany. Way cheaper too. My German insurance is about 5x the price. I've been in a hospital here a few times and they can learn a thing or two about efficiency there.
I have heard similar complaints about UK and Canada. From a cost perspective, it makes sense to me. I also wonder: If you make people wait months, how many people skip/cancel the appointment? Probably many.
I have lived in two countries with very unfair healthcare systems. High income people get "health insurance" (whatever that term really means!) from their employer. They use it a LOT. Way too much. And their "health insurance" covers most of the cost. The number of times that I have seen high income people see a medical doctor for a runny nose (light head cold) stuns me. What an incredible waste of medical resources! As someone fortunate enough to have this "health insurance" at various times in my life, I am constantly saying "no" when doctors try to over-prescribe all manner of medicines. Obviously, they know my insurance will pay 100%!
The #1 duty of a public healthcare system absolutely must be "acute need". Everything else is second priority, else they go bankrupt. It's rough. I don't know a better solution.
Crazy idea: What if there was a kind of public auction system where people in the queue could set a price to sell their position? As long as it was fair and transparent, I might be OK with it.
I'd say that's universal across many countries. The flip side is that Dutch insurers do allow their patients to shop around for care. E.g. getting treated in Belgium or Germany for routine procedures is fairly common. Mostly these just are shortages of staff, equipment, etc. and being efficient sometimes also means that available care is fully utilized. Which just means people have to wait for non critical things sometimes.
That makes sense. When I worked shiftwork (8 hr days), you'd get good info from the people before you, but not about the people before them (after you), so you'd often have the same problem in that rhythm. But we would stay on shift for months on end, and obviously that helped with all the knowledge walking away all the time.
There is a serious issue with the flow of information in healthcare, (or at least in the U.S, I never worked elsewhere to know if it's any different). But If you find something during your shift which will be important to know later on, it will certainly be lost as soon as you are off for a few days, or even as soon as a new nurse comes on. To think of a somewhat crude example, if you find out that it is much easier to obtain a blood sample from the veins on the left arm of a patients vs the right, many nurses will still stick the right arm countless times hoping to get something.
And you can leave a chart note about things like that or speak about it during report, but for the most part few people will think "hm, I wonder what everybody else had to deal with." They are probably too busy handling a thousand different things happening all at once. And, even if that is not the case, from what I observed it's simply not part of how things are done. And very often patients will get (justifiably) angry, saying "I've been complaining of x thing for days!" or some version of that. I think it would be much better for both patients and healthcare staff alike if there was a greater emphasis placed on focusing on the series of successes and failures that happen over the course of someone's care, not just seeing it as a single shift or a single problem happening in some isolated point in time.