27 December 2007


Scalpel had an interesting post today about treating a patient who was psychotic by any reasonable definition: hallucinating and with strong signs of potential violence, but at least partially cooperative with treatment. Scalpel and his nursing staff had to wrestle with the dilemma of whether they could administer anti-psychotic medications without his consent.

I always find this sort of thing interesting. In most cases, as long as you can document your clinical judgment that you believe that the patient was either unable to make an informed refusal of care, or presented an imminent danger to self or others, you're pretty safe doing whatever you need to in order to care for the patient. For ER docs, that often involves four-point leather restraints, sedation, occasionally medical procedures (i.e. gastric lavage, also known as pumping the stomach, or life-support such as endotracheal intubation), and often involuntarily detaining the person pending psychiatric hospitalization.

In our state, there is an interesting twist, though. To my knowledge, Washington is unique in that physicians do not have the authority to place patients on any sort of "72 hour hold." Not even psychiatrists. Instead, state law delegates this authority to trained officers of the court called "County Designated Mental Health Professionals," or CDMHPs.

It's an interesting system, and in my mind works well, though a bit kludge-y. I can basically do whatever I want to someone in my ER, provided that it is grounded in good medical reasoning. But if I think someone is or might be a threat, then I call in the CDMHP. They take over from there, and evaluate the patient and determine whether they meet the statutory definition of either an "imminent threat to self or others" or "gravely disabled." If so, they are detained, and the CDMHP finds a bed for them somewhere. If not, they are released, and both the CDMHP and I have immunity from liability should the patient go home and harm self or others.

But I get caught in the middle, in some cases, with a bit of a catch-22. I do not have the legal authority to detain them, so my clinical opinion is meaningless regarding whether the patient goes to the rubber room in the end. I must, however, do a medical evaluation and determine that there is no medical emergency or other non-psychiatric cause of the patient's behavior before I can call in the CDMHP. This usually involves some lab tests (a tox screen, for example) and maybe a CT scan of the brain, and the patient must be sober before the CDMHP can evaluate them. And they are busy. So there is always a lag time, sometimes significant, before the CDMHP evaluates the patient, and I need to control the patient during that time.

The best way to control an agitated, uncooperative, hallucinating patient is to administer an anti-psychotic med like Haldol, as Scalpel did. It's very effective, and much more pleasant than having them in leather four-points screaming and thrashing for hours on end. (Not to mention that physical restraints in such situations can be quite dangerous to the patient.)

Then, hours later, the CDMHP shows up and finds the patient polite and cooperative and not at all dangerous or disabled. The voices are gone and the patient no longer has the urge to kill. The CDMHPs ruefully shake their heads, agree that it sounded like he was pretty nutty when he first arrived, but, "Sorry, he just doesn't meet the criteria to be detained. Send him home."

So I have the choice of leaving the psych patients untreated for hours, possibly endangering patients and staff, or treating them early and precluding any more definitive treatment... argh. We do our best to work around it with short-acting sedatives and locked doors with heavy security presence. But when you have an irritating, obnoxious bipolar in the manic phase screaming sexual obscenities for hours on end, and your hands are tied, it's extremely frustrating.

It is, however, the happiest moment of my day when the CDMHP shows up, stands in the doorway for thirty seconds, then seeks me out and says, "Would you PLEASE get that guy some Haldol?"


  1. Have a video made prior to administration of medication.

  2. Wow. That system sounds like a pain in the butt.

  3. CDMHP is a really lousy system for families of loved ones who tend to go off their gourd, too.

    It creates an artificial division between conditions with a recognized biological cause (say, cognitive changes due to low blood glucose) and those where the biology has become secondary to the protected class of "mental health" subject to the CDMHP. So, when your father with the Lewy Body Dementia history and the florid psychosis is going apeshit because of an occult UTI, you get skittish doctors not wanting to admit them to the medical service until the CDMHP decides that they're not a candidate for a three-day psych hold. Or, said loved one manages to pull it together enough to sound oriented, meets with the CDMHP, gets cleared to go home, and then ratchets back to crazy again.

    My biggest objection to this is that a Masters-level professional can override even fellowship-trained attending physicians, including psychiatrists, and that doesn't feel right to me. There's also no effective mechanism to provide feedback about a CDMHP - if one is known to have a particularly high bar to declaring someone eligible for a hold, you'd better hope they're not the one that shows up, because the odds of being able to get the issue revisited in the next several days are incredibly low, especially in more rural counties - the system works very differently in King County than it does even in the close-in Boonies.

  4. Washington State Nursing Practice Guide, p. 24 (pdf) shows a decision tree for determining appropriate action in situations with grey areas around nursing action/ patient advocacy.

    Don't know in which state Scalpel practices, but his attitude toward nurses to do whatever a physician orders is ludicrous. Nurses have full legal responsibility to do due diligence and to determine that every physician order falls within acceptable standards of care and practice, and that it also meets applicable regulatory requirements, not to mention employer-based approved policy and procedure.

    In the case of the patient situation described, it is not at all clear-cut that the patient was unable and incompetent to make decisions about accepting and comploying with medical care. Perhaps the actual circumstances were different than as described. This may be an artifact of the blogger's inclusion or lack thereof of patient assessment detail. There is also no mention of the nurse's rationale for declining to carry out the medication administration as originally ordered.

    That said, if the patient was deemed by the physician to be incompetent to make medical care decisions, and there was no available patient guardian or fmaily member to make proxy decisions, the nurse's refusal should have triggered the chain of decision making policy, which is required to exist and to be used by every accredited facility. (Refer to the leadership standard in the JCAHO regs.) Most likely, that would be the charge nurse, the nursing unit director or house supervisor, or a chief of emergency medicine physician - the idea being that at least one of those is always immediately available in house.

    If that still resulted in a decision which the treating physician believed was not acceptable, then the policy holds for yet another clinician in an authority position to be accessed in a timely manner - e.g. the chief of the medical staff and the chief nursing officer, or the administrator on call (AOC).

    When those charged via licensure fail to question prescriptions which do not appear to be congruent with evidence-based standards of care and practice and regulatory compliance, patient safety and outcomes are at risk.

    Just some food for thought to consider at a time when you aren't having to make fast decisions in the miasma of the time and resource-pressured ED.

    Interesting post.

  5. To anonymous: Patient privacy precludes videotaping patients, staff and physicians without their consent. There may be facilities which videotape, but I don't know of any that use videotape in direct patient care areas. Hallways, yes, but patient wards, rooms and cubbies, no. HIPAA, my friend. The areas which use video monitoring in patient rooms (intensive care units), to the best of my knowledge, do not capture any images for historical purposes. If I'm wrong, I'd like to have a link or two to learn otherwise.

    As the former CCF chief surgical resident learned, it's not kosher to snap cell phone pics of patients, either! Hot rod!

  6. AOC --

    While most of what you write is completely correct, I find that I cannot agree with you on the whole.

    First, far be it from me to defend Scalpel, but I think he made it clear that he views nurses as part of the team and preferred to talk it through. Some of his commenters were less charitable.

    Second, certainly it's impossible to accurately evaluate Scalpel's actions given the limited info, but the gestalt I got was that this guy was SCARY, which means, to me, that there was a reasonable potential for violence. That, IMO, goes a long way towards justifying over-riding the requirement for consent. This case is gray, which is what makes it interesting, but it also indicates that there may have been multiple acceptable ways to handle it. Especially in exigent circumstances, latitude should be given to the practitioner on the scene.

    Finally, while procedures do exist, as you admirably describe, to run things up the chain of command, there's also a spectrum of disagreement that determines whether to activate it. Different cases might be:

    1. "You want me to do that? Interesting."
    2. "I don't feel comfortable doing that. You do it."
    3. "I really don't feel comfortable doing that and we should get approval before proceeding."
    4. "I think what you are doing is wrong and you need to be stopped."

    I would contend that the procedures you describe are more appropriate for #3 and 4, but the case described was in #2 or even #1 and generally should not trigger that sort of review. Bear in mind that while these policies are ostensibly neutral reviews, for one clinician to invoke them on another is pretty confrontational, and most people are reluctant to do so in absence of very serious reservations regarding the proposed care. For less-serious concerns, retrospective review is reasonable, and even that bears a stigma of "being reported" which will reliably cause intra-staff friction.

    As I said in Scalpel's comments, I think it's reasonable and appropriate for nurses to speak out, and to say "no" when needful. It's an important safety check. You and I agree there. In this case, I do not think that escalation to the next level of review would be necessary.

    Thanks for your thoughtful response.

  7. Eric,

    The variability in the threshold to commit is frustrating. We see it too. I am surprised at the case you post, though. In my experience, those usually go to a gero-psych facility, even with a UTI, absent sepsis. Maybe you just don't have one out where you are?

  8. @ Shadowfax:

    Interesting point of view. I read into your response that you perceive the clinical decision tree as being a chain of command. I view it as a depth of clinical resources upon which to draw and to get real time resolution of clinical questions.

    There are so many situational variables which weren't addressed in the original post, such as the nurse's experience and psychiatric nursing expertise, whether the nurse being requested to administer the meds was the patient's primary nurse or simply a covering nurse (that can make a significant difference in the willingness of nurses to administer certain meds), whether there had been an overt disagreement about the patient's competency assessment, his degree of volatility, the presence of family, the available security mechanisms, whether he was made a one to one (I am guessing yes, but have no idea who serves in that capacity at that particular facility - can be anyone from an unskilled patient observer to a security officer and lots of variability in between those two), and so on.

    But what I got out of the clinical scenario was that Scalpel felt that staffing was short - he specifically noted that no security staff was present, the patient had potential for acting out, but at no time did act out, and that he was medicating to prevent eescalation and not to de-escalate. That's tough, and given the understaffing situation, I would have (wearing my AOC hat and being a fly on the virtual wall) wanted someone in that department to alert the administrator to an actual unsafe patient situation. That should have been documented and addressed in real time. Was the nursing supervisor called to provide constant observation staff? Was the situation documented in either real time or retrospectively? Is this a chronic problem, and was the patient endangered by the staffing situation?

    I also wonder if there was some concern about adhering to established policy or procedure, and this also was not addressed in the scenario. In that light, I crafted a response which spoke to the available resources to address the more generic question of administering medications to a patient without full disclosure.

    Finally, Scalpel titles the post "nursing ethics", but this is a legal issue about the nurse's legal duty to the patient relative to informing him about the medications he is being given. What if, (in a true Murphy's Law scenario) the patient had a significant adverse reaction to the Haldol and hadn't consented to it? Haldol, in my experience, is considered a behavioral restraint and is subject to restraint use policy. Is that true everywhere?

    Any ED physician who prescribes and has a nurse refuse is within rights to access the charge nurse/unit director/house supervisor in order to get timely resolution or to determine that there is still an outstanding issues. That's not confrontational, and the reason those decision hierarchy tools and mechanisms are in place is to facilitate just these types of situations.

    Perhaps your experience about the use of this mechanism is different, and if so, it might be worth a call to your quality improvement people to see about getting that changed to make it more informal, uncoupled from any disciplinary repercussions and not tethered to peer review sanctions. It might be more helpful to view it as a sideways version of a rapid response team, but in this case for the physician or nurse to get resolution for a prescription. It might prevent some root cause analyses down the road.

    This might also be a terrific topic for physician and nurse joint CME/CE which is facility specific. It could go a long way to holding everyone accountable, and for making everyone aware of how to resolve problematic clinical conundrums without engendering interdisciplinary hostility and power plays. FWIW.

  9. Shadow:

    No geropsych options in-county, which leads to issues - in one case, someone was voluntary and transported by BLS to Highline's facility, then got violent. Highline opted to call in the CDMHP after the person in question threw a styrofoam cup of water at the charge nurse.

    The CDMHP opted not to do a three-day hold and transfer to Northwest (after making a snarky (and hilarious) comment about how the only person harmed by a cup of water would be the wicked witch of the west) , but Highline wouldn't keep the patient either.

    It turns out, had the patient left Highline and the CDMHP *had* decided to hold him, her actions in King County mental health court wouldn't hold water across the county line.

    The CDMHPs in our county are *loathe* to use the out-of-county geropsych resources, limited though they are, as a result.

    Again, though, I prefer the judgment of a geropsychiatrist over even the most knowledgable of psych RNs or LCSWs, which fill the role of CDMHP over here. What I find *MOST* obnoxious is that I can consent, as healthcare proxy, to having the poor guy shocked, poked, compressed, cut, sewn and folded in the name of healthcare without so much as a blink, but I can't say "Yes, it's OK for you to get his meds balanced".

    Without getting off on a rant, this is largely the work of "advocacy" organizations like NAMI. It's perpetuating the "special" class of disease that is mental illness, contrary to the rest of the culture's move towards medicalization.


  10. The case was mildly interesting, but similar episodes happen with this pattern:

    -doc orders med
    -nurse uncomfortable and declines
    -doc administers it
    -both gain experience from outcome
    -next time nurse administers it.

    Agree that "running up chain of command" is confrontational and wastes time.


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