The second part of our look at a case of catastrophic intracranial hemorrhage, with a focus on goals of care, family interaction, prognostication, and other end-of-life aspects, with neurointensivist and palliative care physician Jess McFarlin (@JessMcFarlinMD).
Useful phrase: “Can I tell you what to expect during the dying time?”
Discuss the possibility of secretions, etc. Use glycopyrrolate.
Use opioids if you expect dyspnea, otherwise not always needed. Can try a pressure support trial on the ventilator to get a sense for tachypnea.
Let both family and the nurse know what to expect after extubation.
Other than the occasional incidence of troubling myoclonus with fentanyl, and restrictions on its use outside of the ICU in many centers, all opioids are probably equally good for end-of-life care. Consider hydromorphone in renal patients.
In general, stop tube feeds at the end of life, and stop trying to ensure full nutrition, but do offer food and drink for comfort. Dying tends to limit hunger and caloric needs anyhow. Stop IV fluids as well.
When families invoke a “miracle scenario,” reframe by asking what a miracle might look like for them, or raise the possibility that the miracle won’t be survival, but another outcome such as surviving until the rest of the family arrives, or being comfortable and pain-free during the dying process.
Use “I wish [it would work]” statements to express empathy and a shared perspective, while maintaining a fact-based reality. Stop there and don’t wade into details.
Turn miracles into concrete plans by establishing a time trial with a deadline, with clear markers for what success will look like.
Neurointensivist and palliative care physician Jess McFarlin (@JessMcFarlinMD) walks us through a case of catastrophic intracranial hemorrhage, with a focus on goals of care, family interaction, prognostication, and other end-of-life aspects.
We can undo most things except death, so in most cases, a short trial (perhaps 3 days) of fully aggressive care after an ICU admission is reasonable to help clarify the eventual prognosis. Set clear guideposts for when you’ll regroup to make more decisions about the direction of care.
The ICH score is a helpful guide for early prognostication in spontaneous intraparenchymal hemorrhage.
When prognosticating, express the range of outcomes in terms of three possibilities, as determined by the currently available data: the best case, the worst case, and the most likely case.
Useful questions for families:
If your [loved one] could hear this prognosis, what would he/she say is most important to him/her?
What’s most important to you at this stage? What are you most afraid of?(Many are more concerned about discomfort, pain, or “struggling” at the end of life than about the prospect of death itself.)
When you’re asked, “What would you do if it were your mom?” the question is not really about your mom; it’s a request for a recommendation about theirs.
Transition from open-ended questioning, and translate their values/goals into an actionable plan, by using alignment statements: “What I’m hearing you say/it sounds like…” leads to “Hearing that, may I make a recommendation?”
ICH score: mortality prediction score for spontaneous intraparenchymal hemorrhage
Vital Talk: training resources for the skills of executing palliative and end-of-life conversations
[this episode was reposted on the website 4/27/2020 due to a database reversion after an unfortunate system breach —ed.]
Neurologist and neurointensivist Gracia Mui shows us the workup, initial management, and escalation of care for a case of refractory status epilepticus.
First-time unprovoked seizures usually need no further workup except screening for an underlying trigger, such as a tox screen, basic chemistries, and imaging as appropriate.
Initial seizure therapy: wait around 5 minutes, then give 2–4 mg lorazepam. Repeat every minute or so until convulsions stop, up to 0.1 mg/kg total.
Give an anti-epileptic concurrently: fosphenytoin or phenytoin (15–20 mg/kg), levetiracetam, or valproic acid (40 mg/kg, up to 3000 mg) are all acceptable. If using levetiracetam (Keppra), give a real dose of 60 mg/kg (max of 4.5 g).
After loading with benzos and/or anti-epileptics, if convulsions stop and the patient remains unresponsive, consider the duration of the drug you used. If it’s wearing off (e.g. after about an hour for lorazepam) and they remain unresponsive, suspect non-convulsive status epilepticus.
Any patient not waking up needs an EEG. If not available, they may need empiric deep sedation and intubation until EEG can prove the absence of seizures.
If convulsions are absent, that’s good, as convulsive seizures are more harmful than non-convulsive, but not as good as obviating seizure activity on EEG.
Other than the practical, there is no upper limit for benzodiazepine dosing.
Once you’ve successfully achieved the desired EEG result (either burst suppression or simply the absence of seizure activity) using anesthetics, hold them for about a day, then lighten sedation to see if seizures recur. If so, re-deepen sedation (perhaps for twice as long), increase anti-epileptic agents, then try again.
The patient in status should routinely be screened for underlying triggers, including brain imaging and LP (remember autoimmune causes such as NMDA encephalitis). But about half the time, even in severe refractory status, no underlying cause will be identified.
Cardiothoracic critical care PA Brendan Riordan (@concernecus) shows us his initial approach to the patient in cardiogenic shock, including initiating mechanical support, managing ECMO (plus Impella), and eventual weaning and discontinuation of support.
Anticoagulation on VA ECMO can be titrated to bleeding risk, with a balance between bleeding and circuit longevity—the latter being more than an inconvenience, as changing the circuit in a patient fully dependent on the pump is fraught. Anti-Xa levels are more reliable than the PTT. In a patient with HIT, you may be able to treat through it with bivalirudin, as the heparin-bonded circuit usually cannot be switched out.
“Hypoxemia” on VA ECMO is either regional hypoxemia/North-South syndrome/harlequin syndrome, or oxygenator failure. Rule out the latter by checking a post-oxygenator ABG or just looking to ensure the outflow blood is bright red. Rule in the former by evaluating the ABG or SpO2 from the right upper extremity.
Preemptively placing an anterograde perfusion catheter in the femoral artery is not absolutely mandatory, but is probably simpler and perhaps safer than placing one reactively.
A PA catheter is more useful for weaning ECMO than during the period of full support.
Readiness for weaning is evaluated by recognition of improving cardiac pulsatility, followed by a trial of weaning down pump flow, and finally decannulation in the OR. Consider leaving the Impella if there are any lingering concerns.
Intensivist and passionate slayer of venous congestion Philippe Rola (@thinkingcc) shows us how to deresuscitate the septic patient, with guidance from his handy ultrasound.
Fluid overload is harmful and should be actively reduced, even in a patient in active shock; it will not harm them.
The VEXUS exam is a good method for stratifying fluid overload by severity, i.e. severe (and harmful) versus mild (and relatively benign).
The IVC, CVP, or hepatic vein doppler offer similar information, and are all effective means of assessing central venous pressure, the first and most important step in evaluating for venous congestion. The portal vein doppler offers the most additional diagnostic yield on top of this. Renal vascular doppler acts mostly as a “tiebreaker” when these other studies are equivocal.
No one study or datapoint tells the whole story in these patients. Gather data from as many sources as possible to form the clearest picture.
Our apologies for the section of missing audio and slightly below-par audio quality in this one.