7 ms·
The system crashed while my coworker was running a code (aka doing CPR) in the ER last night. Healthcare IT is so bad at baseline that we are somewhat prepared
by davycro 2y ago
The system crashed while my coworker was running a code (aka doing CPR) in the ER last night. Healthcare IT is so bad at baseline that we are somewhat prepared for an outage while resuscitating a critical patient.
- smsm42 2y agoNow this is an unusual meeting of two meanings of "running a code".
- iamtheworstdev 2y agothere's a great meme out there that says something like: Everyone on my floor is coding! \n Software PMs: :-D \n Doctors: :-O
- jmcgough 2y agoWhen you're a software engineer turned doctor you get sent that by all of your friends xD
- dwatson92 2y agoExcuse my ignorance, but what systems are needed for CPR?
- pksebben 2y agoMy guess is the system that notifies the next caretaker in the chain that someone is currently receiving CPR. if it works, there's a lot more to be done to get the patient to stable.
- SamuelAdams 2y agoI used to work in healthcare IT. Running a code is not always only CPR. Different medications may be pushed (injected into the patient) to help stabilize them. These medications are recorded via a bar code and added to the patients chart in Epic. Epic is the source of truth for the current state of the patient. So if that is suddenly unavailable that is a big problem.
- dwatson92 2y agoMakes sense, thank you for the explanation.
- 627467 2y agoOkay,not having historical data avaliable to make decision on what to put into a patient is understandable - but maybe also print critical stuff per patient once a day? - but not being able to log an action in realtime should not be a critical problem.
- adolph 2y ago> maybe also print critical stuff per patient once a day? Yep, the business continuity boxes are basically minimally connected PDF archives of patient records "printed" multiple times a day.
- pas 2y agomaybe non-volatile e-paper, which can be updated easily if things are up, and if the system is down it still works as well as the printouts
- anonymous8888 2y agoupdatable e-paper is going to be very expensive
- rbanffy 2y agoA small printer connected to the scanner should do.
- pas 2y agoCompared to managing thousands of printers? And then the resulting printouts? Buying ink, changing the cartridges? Technologically it seems doable. Big enough order brings down the costs. https://soldered.com/product/soldered-inkplate-5-5-2%e2%80%b3-e-paper-board/ https://soldered.com/product/soldered-inkplate-5-5-2%e2%80%b... Of course the real backup plan should be designed based on the actual needs, perhaps the whole system needs an "offline mode" switch. I assume they already run things locally, in case the big cable seeker machine arrives in the neighborhood.
- monkmartinez 2y agoYou need to document everything during a code arrest. All interventions, vitals and other pertinent information must be logged for various reasons. Paper and pen work but they are very difficult to audit and/or keep track of. Electronic reporting is the standard and deviating from the standard is generally a recipe for a myriad of problems.
- devilbunny 2y agoWe chart all codes on paper first and then transfer to computer when it's done. There's a nurse whose entire job is to stay in one place and document times while the rest of us work. You don't make the documenter do anything else because it's a lot of work. And that's in the OR, where vitals are automatically captured. There just aren't enough computers to do real-time electronic documentation, and even if there were there wouldn't be enough space.
- monkmartinez 2y agoI chart codes on my EPCR, in the PT's house, almost everyday with one hand. Not joking about the one hand either. Its easier, faster, and more accurate than writing in my experience. We have a page solely dedicated to codes and the most common interventions. Got IO? I press a button and its documented with timestamp. Pushing EPI, button press with timestamp. Dropping an I-Gel or Intubating, button press... you get the idea. The details of the interventions can be documented later along with the narrative, but the bulk of the work was captured real-time. We can also sync with our monitors and show depth of compressions, rate of compressions and rhythms associated with the continuous chest compression style CPR we do for my agency. Going back to paper for codes would be ludicrous for my department. The data would be shit for a start. Hand writing is often shit and made worse under the stress of screaming bystanders. Depending on whether we achieved ROSC or not would increase the likelihood of losing paper in the shuffle
- czl 2y agoThe idea is to have the current system create a backup paper trail from which you practice resuming from for when computers go down. Nothing about current process for you need change only that you be familiar with falling back to paper backups when computers are down.
- davycro 2y agoYou can do CPR without a computer system, but changing systems in the middle of resuscitation where a delay of seconds can mean the difference between survival and death is absolutely not ideal. CPR in the hospital is a coordinated team response and if one person can’t do their job without a computer then the whole thing breaks down.
- SoftTalker 2y agoIf you're so close to death that you're depending on a few seconds give or take, you're in God's hands. I would not blame or credit anyone or any system for the outcome, either way.
- singleshot_ 2y agoI’m sure you meant “the physicians’ hands.”
- SoftTalker 2y agoNo. The physician will be running a standard ER code protocol, following a memorized flow chart.
- dreamcompiler 2y agoJudgement is always part of the process, but yeah running a routine code is pretty easy to train for. It's one of the easiest procedures in medicine. There are a small number of things that can go wrong that cause quick death, and for each a small number of ways to fix them. You can learn all that in a 150 hour EMT class.
- xyst 2y agoprobably the system used to pull and record medication uses in a hospital. It's been awhile, but "Pyxis" used to be the standard where I shadowed. Nurses hated it.
- supergirl 2y agoneed to play bee gees on windows media player
- peterleiser 2y agoThe second largest hospital group in Nashville experienced a ransomware attack about two months ago. Nurses told me they were using manual processes for three weeks.
- ethbr1 2y agoAscension?
- peterleiser 2y agoYes. And I was told by multiple nurses at St. Thomas Midtown that the hospital did not have manual procedures already in place. In their press release they refer to their hospitals as "ministries" [0], so apparently they practice faith-based cyber security (as in "we believe that we don't need backups") since it took over 3 weeks to recover. [0] https://about.ascension.org/cybersecurity-event https://about.ascension.org/cybersecurity-event
- bufferoverflow 2y agoIt takes a certain type of a criminal a55hole to attack hospitals and blackmail them. I would easily support life or death penalty for anyone attempting this cr@p.
- peterleiser 2y agoIn this case it was tracked to Russia.
- uptownJimmy 2y agoThat is absolutely one of the A-tier "certain type of a criminal a55hole".
- bart_spoon 2y agoMore than just Nashville, they have hospitals all over the country.
- SarahWSJ 2y agoHello, I'm a journalist looking to reach people impacted by the outage and wondering if you could kindly connect with your ER colleague. My email is sarah.needleman@wsj.com. Thanks!
- davycro 2y agoI sent them your contact info, pretty sure they will be asleep for the next few hours
- bookofjoe 2y agoSurprised and impressed at your using HN as a resource.
- yard2010 2y agoShe is living in the future. Way to go.
- unixhero 2y agoThe comments is the content. I have always said this.
- whatevertrevor 2y agoI mean if they're finding sources through the comment and then corroborating their stories via actual interviews, it's completely fine practice. As long as what's printed is corroborated and cross-referenced I don't see a problem. If they go and publish "According to hackernews user davycro ..." _then_ there's a problem.
- FireBeyond 2y agoAs a paramedic, there is very little about running a code that requires IT. You have the crash cart, so not even stuck trying to get meds out of the Pyxis. The biggest challenge is charting / scribing the encounter.
- JshWright 2y agolol, yep, that was my take on this... If you need a computer to run an ACLS algorithm, something has gone seriously wrong.
- dreamcompiler 2y agoEspecially out in the field where we have a lot more autonomy. If our iPads break we'll just use paper.