Critical Care Scenarios

Podcast: Play in new window | Download

Subscribe: Apple Podcasts | Spotify | Android | Pandora | iHeartRadio | TuneIn | RSS

We discuss the practical barriers to implementing the A-F ICU liberation bundle, with Kali Dayton, ACNP-BC (@daytonicu), host of the Walking Home from the ICU podcast, and consultant to ICUs working on these issues.

Learn more at the Intensive Care Academy!

Find us on Patreon here!

Buy your merch here!

Podcast: Play in new window | Download

Subscribe: Apple Podcasts | Spotify | Android | Pandora | iHeartRadio | TuneIn | RSS

We discuss transfusion reactions, risks, and management, including infection, consent, TRALI, TACO, and hemolytic reactions—with Dr. Joe Chaffin (@bloodbankguy), the “Blood Bank Guy” and transfusion medicine specialist.

Learn more at the Intensive Care Academy!

Find us on Patreon here!

Buy your merch here!

Takeaway lessons

1. The risk of transfusion-related infection (HIV, hepatitis B, and hepatitis C) is around 1 in 3 million.
2. Acute hemolytic transfusion reactions (usually due to clerical errors or unit mix-ups) occur about 1 in every 75 or 76 thousand transfusions. Mortality is only one per million or so, however.
3. Simple febrile transfusion reactions occur about 1/100-300 transfusions.
4. Transfusion is always slightly immunosuppressing, perhaps increasing risk of post-op infection, cancer recurrence, etc. This effect is real, but small and not easily quantified.
5. Urticarial reactions (hives) seem trivial to clinicians, but can be very frightening to patients, even causing them to refuse future transfusions.
6. 80% of hemolytic reactions initially present with only fever, perhaps some chills. There is no way to differentiate from non-hemolytic febrile reaction at this stage. While the odds favor a non-hemolytic reaction, if you presume this and continue your transfusions, you are relying on luck, and you will eventually be wrong, which would be an indefensible medical error.
7. Once a hemolytic reaction is obvious, you waited too long. The main determinant of mortality after hemolytic transfusion reaction is the volume of blood transfused.
8. Typical workup for a febrile, possible hemolytic reaction is to confirm the labels and clerical match, then return the blood to the blood bank, where they will check patient blood for hemolysis, direct Coomb’s, and usually repeating the ABO/Rh testing. This can cause a delay in transfusion and maybe loss of the unit of blood; by typical regulations, once blood is removed from the blood bank or portable cooler, it must be transfused within 4 hours or wasted.
9. The hallmark of ABO mismatch is severe intravascular hemolysis. Most other hemolytic reactions yield extravascular hemolysis, e.g. in the spleen. Cytokine storm will be be seen. Compared to the myoglobin released in rhabdomyolysis, the free hemoglobin released in intravascular hemolysis is not quite as nephrotoxic (the resulting AKI may be more related to shock than from direct toxicity).
10. Hemolysis is only destructive to the transfused blood, so anemia per se generally does not develop. One exception can occur in sickle cell patients, where transfusion can induce a “hyperhemolysis” phenomenon where native red cells are also hemolyzed.
11. Mortality from acute hemolytic reactions is fairly low in previously healthy patients. Patients already critically ill may not do as well.
12. TRALI is mostly diagnosed by consensus criteria. “Definitive” TRALI (there is no longer a less definite category) is defined as:
    • No evidence of lung injury prior to transfusion
    • Onset within 6 hours after end of transfusion
    • P/F ratio <300 or SaO2 <92% on room air
    • Radiographic evidence of bilateral infiltrates with no evidence of left atrial dysfunction
13. The challenge when hypoxia occurs after transfusion is usually to distinguish TRALI from TACO. The latter is mere volume overload; the former occurs when pre-existing inflammation primes neutrophils for activation in the lungs, whereupon factors in the transfused blood causes neutrophil activation as a second hit. The most common of these triggers is incompatible anti-HLA antibodies in the transfused blood.
14. TRALI is largely a clinical diagnosis. However, if a case of possible TRALI is reported, the donor will be investigated and potentially screened for anti-HLA antibodies (something usually not done without a suspicious case). Other products from that donor will also be recalled from the bank. Report your possible TRALI cases!
15. Now that female donors with previous pregnancies are excluded from donating plasma (without HLA screening), the old truism that plasma-rich products (e.g. FFP or platelets) are the highest risk for inducing TRALI is no longer true; the most common precipitant is PRBCs. Any product can induce TRALI, however, including HLA antibody-negative products.

Podcast: Play in new window | Download

Subscribe: Apple Podcasts | Spotify | Android | Pandora | iHeartRadio | TuneIn | RSS

We talk the nitty-gritty of assessing the right heart using echocardiography, with our friend Matt Siuba (@msiuba), intensivist at the Cleveland Clinic and master of zentensivism.

Learn more at the Intensive Care Academy!

Find us on Patreon here!

Buy your merch here!

Takeaway lessons

1. RV echo starts with evaluating three things: size, squeeze, and septal kinetics.
   • Size should be <2/3 the LV
   • Squeeze can be assessed in a variety of ways
   • The septum should not be bowing into the LV.
2. Dilation is an early and somewhat compensatory finding, and can be used as a screening test (the “D-dimer of RV dysfunction”). Septal changes are probably later and more of a sign of dysfunction (i.e. not compensatory).
3. Evaluating the RV’s ejection fraction is impractical due to its complex shape (without 3D echo or cardiac MRI or other advanced tools). So methods like TAPSE that reduce it to its longitudinal function become a more practical surrogate.
4. TAPSE is not an isolated marker of RV contractility, but a marker of the overall RV-PA unit. However, this is probably a feature, not a failure. You don’t really want to know how the RV is contracting in the abstract, but how it’s contracting in its current loading conditions. So TAPSE will vary by afterload and preload, but not artifactually—i.e. if the loading conditions change and TAPSE improves, then contractility is better in the current conditions.
5. s’ is similar to TAPSE, and similarly limited (mainly evaluating longitudinal function). It assesses velocity, not movement, which theoretically may represent something different (maybe a better marker of function?), although that difference is not very well studied; some studies do suggest that s’ may be more sensitive to changes after adding an inotrope, but who knows if that means anything. The most common cause for a big discrepancy between TAPSE and s’ is probably technical error, not a clinical distinction.
6. RVSP can be useful as a marker of afterload, but says nothing about the cause of RVSP—high left sided pressures vs high PVR—and also incorporates the RV function, so separating all this out can be difficult.
7. TAPSE/PASP (or TAPSE/RVSP) ratio might be a somewhat more accurate marker of RV/PA coupling, but not really clear if it’s clinically better than using the TAPSE alone, which is already a fair marker of RV/PA coupling. By measuring more things, it also introduces more room for technical error (usually underestimating RVSP), such as the need to estimate the TV gradient and the CVP. More tricuspid regurgitation will also tend to reduce the ratio, without necessarily indicating better RV function.
8. CVP estimates derived from the IVC are very unreliable in the critically ill. Many chronic PH patients have chronically distended IVCs regardless of their RAP. Using a transduced CVP is probably better. You can also just trend the TV gradient as a marker of its own and ignore the CVP component.
9. Shortening of the PA acceleration time (PAAT or PVAT) is a useful marker of pulmonary afterload. Notching of the waveform usually indicates a very high afterload, much more likely to be caused by pulmonary factors than high left heart pressures.
10. Fractional area change of the RV is another tool for approximating the LV “EF” which may work better in chronic dysfunction, where TAPSE may be misleadingly preserved. However, it requires a good view of the RV free wall, which is not always achievable.
11. Strain measurement has not yet penetrated point-of-care ultrasound machines reliably, but use is increasing. While still load-dependent, strain measurement is not angle dependent, which may make it helpful for right heart assessment.
12. In the less common clinical scenario of RV infarction/ischemia, most of the above still applies, yet the pulmonary afterload will not necessarily be elevated. In almost every other case, the problem driving RV failure is usually the afterload, hence reducing the afterload is usually the easiest treatment.
13. A proposed algorithm:
    1. Look for RV dilation
    2. Assess contractility using TAPSE and/or s’ (or other methods like eyeball gestalt, fractional area change, etc)
    3. Assess afterload using PA acceleration time and notching
    4. Compare contractility and afterload in context with the clinical scenario to understand the right heart’s function and conditions, with the understanding that your marker of contractility also incorporates the afterload to some extent.
14. Don’t forget that invasive monitoring, from CVP to a PA catheter, is always an option as well. CVP is “for free” and rarely wrong if you know how to interpret it, including the waveform, and in the sickest patients, a Swan can be quite helpful, particularly for monitoring; multiple advanced echo studies are not always possible or reliable, particularly with rotating shift staff.
15. If you have a Swan, wedge it. Otherwise it’s just a cardiac output monitor. Some fancy newer devices also allow measuring PA and RV pressures separately, which is an evolving science.

Podcast: Play in new window | Download

Subscribe: Apple Podcasts | Spotify | Android | Pandora | iHeartRadio | TuneIn | RSS

We discuss the practicalities of using airway pressure release ventilation (APRV) with Dr. Rory Spiegel (@EMnerd_), emergency physician and intensivist at MedStar Washington Hospital Center (and EMNerd at Emcrit).

Find us on Patreon here!

Buy your merch here!

Takeaway lessons

1. The most immediate benefit of APRV is to help restore lungs to FRC (functional residual capacity). While this can be achieved with PEEP, most people don’t use enough PEEP. APRV proves a higher mean airway pressure while also reducing sedation requirements, and provides a physiologically automatic titration of “PEEP” based on lung compliance.
2. Phigh can usually be set to equal the previous plateau pressure on a conventional mode (assuming reasonably appropriate settings there). This helps match higher Phigh to a more poorly compliant lung and vice versa. The release volume that results should be checked to give a sense of the effects; it should be more or less in the range of normal tidal volumes, although usually lower than your previous tidal volume on a conventional mode due to the intentional air trapping. (If it’s not lower, your Tlow may be set too long, allowing too much release.)
3. Thigh can range from 1.5 seconds to infinity. Longer T-high is better for recruitment, shorter is better for ventilation. When initially flipping to APRV, shorter Thigh is usually needed; try to match the patient’s minute ventilation (from the prior mode) fairly closely, although usually you’ll need to accept a small loss of ventilation. A too-long initial T-high is a common error; patients this sick usually cannot tolerate more acidosis. Usually an initial T-high of 2 seconds is about right.
4. Tlow should be set to terminate when the expiratory flow drops to 75% of the peak expiratory flow (so if the peak was 100 L/s, terminate Tlow when it drops to 75 L/s). This was about the point in pig models where alveolar derecruitment began to occur. Peak and end expiratory flow can be checked on most modern vents, although it may not be easy to find in the screens. Usually the right T-low is around 0.3–0.6 seconds.
5. Occasional patients may need a shorter T-low than this for optimal recruitment. But few need longer; Tlow should rarely be lengthened, even as patients recruit. Although the amount of air trapping will usually increase as the lung becomes more compliant (e.g. the same T-low duration will terminate expiration at 85% instead of 75% of peak expiratory flow), this is usually fine; this is when you’ll start weaning and stretching your Thigh.
6. Plow should be set to zero in almost all cases, allowing the fastest expiration (higher Plow reduces the driving pressure and substantially reduces expiratory flow). In a few vents (older Puritan Bennett, older Servos), the machine may attempt to synchronize with patient efforts by allowing the Tlow to “kick out” and extend, creating large release volumes and loss of desired air trapping. Increasing the Plow may provide some safety margin in this case, although switching from APRV altogether is probably the best solution.
7. As the patient recruits on APRV, release volume should gradually increase despite a fixed Phigh, as the lung recruits. The expiratory flow curve will flatten and the compliance will increase. Thus, release volumes are initially small—”lung protective” in conventional thinking—and later will increase. This increase should be allowed, as it’s still associated with a normal/low driving pressure, since the “PEEP” gradually increases as trapping increases. A large release volume + low driving pressure is felt to be lung protective in APRV thinking.
8. Driving pressure on ARPV can be checked on most vents by performing an expiratory (not inspiratory) hold to determine the effective “PEEP.”
9. Patients can breathe spontaneously on APRV and be comfortable, but this is mostly determined by lung recruitment and how close they are to FRC. When the lungs are still tightly closed, spontaneously breathing will not be either comfortable or safe, so when initially flipped to APRV, patients should NOT be breathing; they will look uncomfortable, require very high minute ventilation, and generate high pressures. (There is also great discomfort here due to the hypercarbia usually unavoidably present.) Use a shorter Thigh and ventilate using the vent releases in this period, while using deep sedation and/or paralysis to suppress breathing.
10. As patients stabilize and recruit, the minute ventilation needed to maintain pCO2 will drop as ventilation becomes more efficient. When MV and the CO2 approach normal physiologic ranges, sedation can be lightened and patients allowed to breathe. Ultimately, severe ARDS patients on APRV require less total sedation and need for paralysis than in other modes.
11. Weaning occurs as thus: CO2 will gradually fall and release volumes naturally increase as the lungs recruit. Eventually they become hypocapnic, so Thigh must be increased to reduce the minute ventilation. As MV reaches normal, stretching the Thigh further causes hypercapnia, so patients should now be allowed to start breathing spontaneously to make up the difference in MV. Breathing should look comfortable, with a benign clinical appearance and gentle inspiratory flows (not sharp peaks); if not, recruitment may not yet be optimal and it may not be time for spontaneous breathing.
12. Rory does not drop the Phigh during the weaning period, although many teach this; he finds it often causes derecruitment. He adjusts Phigh only in response to the perceived disease state; for example, it may need to be weaned as the disease improves and the initial Phigh may start to cause overdistention. He rarely touches it until the patient is ready for breathing. Once the patient is breathing spontaneously, this provides a good feedback tool to adjust Phigh; if you drop Phigh and spontaneous breathing looks worse (like a failed SBT – lower volumes, high rates), you derecruited them and Phigh should go back up. Spontaneous effort is a more sensitive and faster method of feedback than monitoring the release volumes alone.
13. Permissive hypercarbia is okay. But severe hypercarbia before starting APRV is a marker of advanced underlying disease and lung injury which may make it difficult to tolerate APRV, and persistent hypercarbia on APRV is a marker of failure—the lung is not recruiting, and the mode is probably not totally safe as a result (persistent acidosis + persistently high driving pressures and risk of overdistention of ventilated lung).
14. Hypotension is not necessarily a contraindication to APRV. Cardiac output is best when the lungs are at FRC, neither over- or under-distended. However, it’s true that the lungs “overdistend instantly, but recruit over time,” so until the lung recruits, intrathoracic pressure may be elevated, and delicate patients (eg hypovolemic) may not tolerate this well.
15. Pneumothorax should not be a contraindication to APRV. The more recruited the lungs, the less strain on each individual lung unit given the same overall driving pressure.
16. Using APRV is a skill that requires practice. However, it also helps create a general mindset of approaching the lung physiologically with the goal of restoring FRC, as well as appreciating the value of using minute ventilation as a marker of recruitment; these tools probably benefit patients even in other modes. With this approach, APRV can often be avoided, and used mainly as a rescue modality.