19 November 2008

Lost in Transition

I remember a bad hand-off once, long ago. It was the classic admitted patient, long forgotten two shifts after a bed had been ordered, but hanging out in the ER waiting for the assigned bed to be vacated and cleaned. It was a chest pain admission, a "low-risk rule-out," meaning that the patient was to get a blood tests to rule out a heart attack and then a stress test. Turns out the diagnosis was quite wrong: it was an aortic dissection, and when the patient crashed, there was chaos because nobody remembered why the patient was there and who was responsible for him.

In that case, there were other problems: the nurses in the ED had been quite content to ignore the patient while he slept. No vital signs were obtained, at one point the cardiac monitoring had been discontinued for the patient to go to the bathroom, and the schedule of blood tests designed to detect an evolving heart attack were not drawn.

The outcome was bad, and we as an organization learned a lot from it.

Doctor RW writes about this topic today, linking to an interesting article in Today's Hospitalist. RW's recommendation is good, but does not go far enough:

Hospitalist groups should meet with their emergency medicine colleagues regularly to discuss cases, offer feedback and improve professional relationships.
To be sure, this is necessary and I won't disagree with it. But the fact is this: if the patient is physically in your ED after "admission," they remain your responsibility, and as a department, you must have procedures in place to ensure the patient will continue to receive excellent care during the transition.

Key points which such procedures must address include:
  • There must be an ED physician who is the designated responsible provider and who is aware of the patient. This is pretty standard and is easily accomplished with most patient tracking system, be they simple grease boards or sophisticated EMRs.
  • Transfer of care to the hospitalist does not take place until the patient has either left the ED or until the hospitalist has physically seen the patient in the ED.
  • The ED doc must perform interval assessments of the "boarding" patients in the ED, regardless of whether the hospitalist has seen them. If they're in the ED, they're still your responsibility. Generally, stable patients don't take much attention, but sicker patients, ICU admits, etc will require this assessment. If nothing else, it can add to your critical care time!
  • Once the hospitalist has seen the "boarded" patient, they are the primary caregiver, and simple questions or non-urgent issues can be directed to them. But the ED doc must remain available for urgent issues and to keep tabs on the patient's condition.
  • Inpatient admitting orders should be written at the time the patient is designated as "admitted." The ED nursing staff need to follow these orders as if the patient were in their inpatient bed, especially if the patient will be boarding more than a couple of hours.
  • If you are in the enviable position of sending patients upstairs before the admitting doc has seen them, you need to write adequate holding orders.
For some reason, this last point has been controversial in EM. The AAEM, I think, particularly crusades against this practice. I cannot understand why. Sending patients upstairs to languish until the hospitalists see them is a clear extension of liability for the ER doc. Yeah, it would be great if all admits were seen within ten minutes of arrival to the floor, but that's not reality. There's a persistent idea that writing admitting or holding orders somehow increases the ER docs' liability and muddies the question of who bears responsibility for the inpatient.

My opinion is that by writing good holding orders, the ER docs improve patient care, help the hospitalists, and reduce everybody's risk. The key is that these orders do not need to be comprehensive, but they do need to be adequate. In my opinion, the minimum acceptable holding order set includes:
  • The name of the responsible admitting doc.
  • A clear statement that for problems, questions, or changes in condition, the admitting doc should be promptly notified.
  • A defined time during which your holding orders are valid (i.e. an expiration time for your orders, by which time the admitting doc needs to have seen the patient).
  • Any critical patient safety information (ie allergies, NPO, fall precautions, etc)
  • Any scheduled tests or treatments which will forseeably be needed before the patient may be seen (serial enzymes, nebs, pain meds, blood sugars, e.g.)
  • Parameters to notify the admitting doc (vitals, test results, etc)
When well done, this practice can improve patient care and safety and foster the sense of collaboration between the hospitalists and the ER docs. Standardization is your friend; we have a pre-printed "holding order" set which is very useful and help ensure nothing important is omitted. As hospital-based medicine is a team sport, it is also useful to have joint committees set up between the ER and hospitalist teams. Working together regularly outside of the clinical setting also helps foster a sense of collegiality, and to dispel the "us-vs-them" sentiment that is engendered in the trenches.

This is a pretty important topic. Change of shift is the most dangerous time in any ER, and the transfer of care is fraught with risk. It's curious, now that I think about it, that this has received so little attention in the evolving culture of patient safety and the Quality measures being developed. Look for this to gain prominence in coming years.


  1. Shadowfax,
    Disagree with the statement that "the fact is this: if the patient is physically in your ED after "admission," they remain your responsibility."
    ACEP created a policy a couple of years ago showing why this is NOT true. See this link:
    Have nurses call admitting docs for any issues that arise in the ED after a patient has been admitted. Show the policy to the admins at your hospital and recommend a change in policy to make the admitting docs responsible. That way the docs can't ignore the patients and "hope they go away."

  2. I hoped I was clear about this, but it's nuanced, so maybe reemphasizing it may help. I wrote:
    Once the hospitalist has seen the "boarded" patient, they are the primary caregiver, and simple questions or non-urgent issues can be directed to them. But the ED doc must remain available for urgent issues and to keep tabs on the patient's condition.

    I think we agree that once the baton is passed, the theoretical responsibility belongs to the admitting doc -- as a matter of policy, anyway. But the practical reality is that if the patient remains in the ED, and if they deteriorate, good luck evading liability is you disregard them and do not intervene to treat the deterioration.

    "He wasn't my patient any more," is not a defense which will hold water when the plaintiff's widow is alleging that "The ER doctor was right there and ignored us."

    So while the policy may be that the hospitalist is in charge and responsible once the patient is "admitted," you cannot avoid the reality that you are not "off the hook," while the patient resides in your unit.

  3. You were clear, but wrong. I'm with Dr. WhiteCoat and ACEP on this one.

  4. Riiight. Because ACEP policies have the binding force of law?

    Or something.

  5. No, because this time at least, they are right. Look, the a-hole lawyers are going to sue everyone on the chart anyway, and the idiot jury isn't going to be able to understand the situation regardless.

    I think there is a spectrum of responsibility for a given patient, and no hardline rules can apply. If I call a hospitalist to admit a patient and they code two minutes later, how responsible is he? Not much, I'd say. But if a patient has been in the ER for 18 hours already waiting on a room, I'd submit that the newly arrived graveyard shift ER doc doesn't own much responsibility either, regardless of the "assumption of care" passed off by the second or third ER doc to cover that room.

    Similarly, if there are 3 docs on duty and a patient codes in the waiting room prior to assessment but after sitting two hours in triage, which doc is responsible? All of them or none of them?

  6. I still don't agree with the premise that an admitting physician has to "see" a patient before being responsible for that patient. If a patient is admitted and transferred to the ICU, but the admitting physician doesn't "see" the patient until 24 hours later, is the ED physician still liable for the patient's care?
    Why does location of the patient matter in determining the responsibility for the patient. Are radiologists responsible for the patients when the patients are down getting xrays?
    You can practice however you want, but you're setting yourself, your group, and your hospital up for a lawsuit if you are providing inpatient care and haven't been credentialed for it.
    Attorney: The patient was admitted as an inpatient and you were still giving orders in a non-emergent situation? Are you credentialed for inpatient care? No? Were you even trained for inpatient care? So in other words, you're practicing outside the scope of your credentials and your training, doctor? Oh, and the hospital KNEW you were practicing outside the scope of your credentialing and training and it ALLOWED you to do so? Great. Oh, wait, you're not INSURED to provide inpatient hospital care? Looks like that $1 million judgment will be your responsibility.
    By the way, the ACEP policies aren't there to be the "force of law" - the policies are meant to help emergency physicians out of situations such as this one.

  7. Thank you. As a hospitalist, I work at two different hospitals, with two different EDs. At one, the attitude is, "Once I've called you, their yours." Catastrophes occur with this attitude. We try to see patients as fast as we can, but can't always get there. As to WhiteCoat, I say, "no way!" To have a nurse call me after YOU have seen the patient for questions BUT I HAVE NOT is a receipe for disaster. No offense, but I don't "ignore patients and hope they go away." We have crises on the floors/icus just like you, and get backed up, just like the ED. Please, don't get me wrong, I have worked in the ED, and realize it's a hard job, but passing off questions on a patient to a physician that hasn't even seen the patient yet is poor patient care.


Note: Only a member of this blog may post a comment.