UMEM Educational Pearls - Critical Care

These 2 papers challenge management dogmas in critical care that have persisted despite low-quality/absent evidence.

In particular, one explores the dogma, “bicarbonate improves ventricular contractility in severe metabolic acidosis,” with the following points: 

-intracellular pH (which has a large impact on myocardial contractility) correlates poorly with blood gas pH

-many of the studies regarding bicarbonate in severe metabolic acidosis and hemodynamics are done on animal shock models

-two studies in patients with lactic acidosis showed increase in pH with bicarb administration without beneficial impact on hemodynamics (even in pts with pH < 7.1)

-bicarb administration is associated with hypernatremia, hypokalemia, and decreased ionized calcium levels

Show References

Extracorporeal cardiopulmonary resuscitation (ECPR) is a type of extracorporeal support following cardiac arrest available at a small, but growing number of ECMO centers around the world. After some initial promising results, more recent data have been mixed. There is a nice narrative review in JACEP Open recently which summarizes the most recent evidence. Implementation considerations and patient selection seemingly drive the variance seen in the studies reviewed.

To this point, a new article from Critical Care Medicine was just published looking at the outcomes of eCPR with respect to age using  5 years of ELSO patient data. Unsurprisingly, advancing age is associated with worse outcomes, with significantly reduced odds of survival above the age of 65.

Show References

The level of fitness/health a patient has entering the marathon of recovery from critical illness or trauma has a major impact on morbidity and mortality. Frailty is a measure of this fitness level. The clinical frailty scale can be used to assess your patients ability to survive critical illness. Age is a number. Frailty is more useful. 

Show References

Post-Intensive Care Syndrome (PICS) is an increasingly recognized phenomenon of impairment of physical, cognitive, and/or mental health after intensive care admission.  Even more recently, similar deficits in caregivers of patients admitted to the ICU, often called Post-Intensive Care Syndrome Family (PICS-F) is increasingly recognized.  A study recently published by Watland et al in Critical Care Medicine looking at reducing PICS-F through a “caregiver pathway” got me wondering if there's any literature out there about reducing PICS-F via interventions in the emergency department.  Patients' treatment course in the ED is a highly stressful and uncertain time for both the patient and family members, so it stands to reason this is an impactful period where intervention may help, and even in patients where their condition is too advanced for us to make a medical difference, our actions could have a positive impact on long term outcomes for the family members.

The short answer is no, to this author's knowledge and based on my review of the literature, there is no good evidence for reducing PICS-F by ED interventions (hint, hint: if anyone's looking for a good area to study…)  Based on evidence from the critical care realm, the following are probably reasonable approaches that would translate well to the ED:

1. Recognize, especially when you have a patient who likely has a very poor prognosis, that for our critical patients it is important to treat the family, as well as the patient.  
2. Update the family early and often.  Uncertainty is a key contributor to PICS-F.  
3. Consider developing a brochure for family of critically ill patients at your facility.  Basic information such as where to park, how to get into the hospital, where their loved one may go after the ED, where they can get food, what visiting hours are allowed, whom to contact with questions, etc seem exceptionally simple to us but are often early points of stress for family.  
4. Consider screening family members for PICS-F (probably better left to the ICU, but could be considered for longer ED stays or if patient prognosis is extremely poor).  There are multiple validated screening tools available.
5. Consider encouraging patient (if they are able) or family to keep a diary.  ICU diaries have been shown to decrease incidence of both PICS and PICS-F.  See also icu-diary.org
6. If feasible, consider follow up with family members at high risk of PICS-F.  Could be done as a joint venture between the ED and inpatient services or as a hospital-wide initiative.  
7. Engage ancillary services such as pastoral care, palliative care, integrative medicine, and others early and often to foster a multi-disciplinary approach.  Also, make sure to communicate well with your nursing team, who are at the bedside and often more in tune with family signs of future PICS-F.

Show References

In a fascinating perspective piece, Matt Bivens and colleagues explain that the combination of struggle and restraint leads to death not because of hypoxia, but because of acidosis.

The sequence is something like this: exertion or struggle results in an acidotic state -> restraint reduces respiratory ability, especially when held prone or weight is applied to back or chest -> acidosis worsens with the potential for cardiac arrhythmia and arrest.

In this setting, “I can’t breathe” does not mean that there is no air movement over the vocal cords but that respiration is impaired, much as it is in asthma or obstructive lung disease.

Use of sedation in this setting reduces respiration even further, worsening acidosis and risking death. It’s not hypoxia that kills; it’s acidosis.

See the complete perspective here: https://www.nejm.org/doi/full/10.1056/NEJMp2407162.

Show References

High-Intensity NIPPV for Acute COPD Exacerbations?

• Noninvasive positive pressure ventilation (NIPPV) is frequently used in the management of critically ill patients with an acute COPD exacerbation, and is associated with decreased intubation rates and decreased in-hospital mortality.
• “Low” intensity NIPPV, where the inspiratory positive airway pressure (IPAP) is < 18 cm H2O, is generally used in clinical practice.
• “High” intensity NIPPV, where the IPAP ranges from 20-30 cm H2O has recently been shown to improve gas exchange, ventilatory function, and reduced elevated PaCO2 when compared to low-intensity NIPPV.
• The recently published HAPPEN trial was a randomized trial performed in 30 centers across China and investigated whether high-intensity NIPPV reduced the need for intubation compared with low-intensity NIPPV in patients with an acute COPD exacerbation and hypercapnia.
• In this trial of 300 patients, investigators found that high-intensity NIPPV significantly reduced the number of patients who met criteria for intubation compared with low-intensity NIPPV.
• Importantly, patients were included and randomized in the trial if they remained hypercapnic after initially receiving 6 hours of low-intensity NIPPV.

Show References

The Venous Excess Ultrasound (VExUS) exam integrates IVC, portal, hepatic, and renal vein findings to assess venous congestion and guide management, such as diuresis, in critically ill patients.

Technique:

1. IVC: Measure the IVC diameter. If <2 cm, significant congestion is unlikely, and further assessment is not well validated.
2. Hepatic & Portal Veins: Use a curvilinear probe with color Doppler in the RUQ. The hepatic vein flows away from the probe (blue), and the portal vein, with thicker walls, flows toward the probe (red).
3. Hepatic Vein Doppler: Apply pulse wave Doppler to the hepatic vein or a tributary. If the waveform is not clear, try a different vein.
4. Portal Vein Doppler: After evaluating the hepatic vein, place PW Doppler on the portal vein.

Tips:

• Start from the right upper quadrant, Doppler signals are often easier to obtain and interpret here.
• Delay learning renal vein assessment until comfortable with the other views.
• If the IVC is hard to see subcostally, try a transhepatic view and adjust probe orientation (rotation and fanning).

Interpretation:

• Hepatic Vein: A normal hepatic vein waveform reflects atrial contraction (a wave), atrial filling during ventricular systole (S wave), and atrial filling during early diastole (D wave). As congestion worsens, the proportion of atrial filling during ventricular systole (S wave) decreases and eventually reverses.
• Portal Vein: Normally shows continuous flow. With congestion, it becomes more pulsatile.

Sometimes when other clinical information is contradictory, having the extra data point of the VExUS exam can be extremely useful to determine the best plan for a patient. Practice looking for the portal/hepatic veins and getting the waveforms on patients with a CLEAR clinical picture of venous congestion, then practice on more difficult cases.

Show References

Background:

Ultrasound-guided subclavian central venous catheter (CVC) placement has become a preferred site due to low risk of infection and a low risk of complication.  Complications include arterial puncture, pneumothorax, chylothorax, and malposition of the catheter.  Ultrasound guidance can significantly reduce the risk of these complications aside from catheter malposition.   The most common sites of malposition are in the ipsilateral internal jugular vein or the contralateral brachiocephalic vein.  This study sought to evaluate the rate of catheter malposition between left-and right-sided subclavian vein catheter placement using ultrasound guidance with an infraclavicular approach.

Study:

• Randomized controlled trial, single center, 449 patients
• Excluded patients with pacemaker near the insertion site, infection, patients on anticoagulation, tricuspid valve vegetation, vein thrombus, ports, or a preexisting catheter.
• The primary outcome was the rate of catheter malposition.
• Malposition was defined as not being in the ipsilateral subclavian and brachiocephalic veins and superior vena cava.

Results:

• Catheter malposition occurred in 4.5% in the left-sided group and 13.8% in the right-sided group, OR 0.29 (0.14-0.61 p=0.001). 
• Malposition of the catheter into the ipsilateral internal jugular vein was more common than the contralateral brachiocephalic vein.

Take Home:

For infraclavicular ultrasound-guided subclavian CVC placement, consider using the left-side over the right if no contraindications for left-sided access exist.

Show Additional Information

Show References

It’s the age-old question. We’ve read studies comparing propofol vs. etomidate, ketofol vs. etomidate, and now a meta-analysis about ketamine vs. etomidate.  Etomidate is the staple induction agent for RSI, mostly used by Emergency Medicine, and to a degree in the Intensive Care Unit. However, the question about adrenal suppression was initiated in the early 2000s and researchers have been looking for other alternatives. This meta analysis attempted to look for another answer.

Settings: A meta-analysis of randomized controlled trials

Participants: 2384 patients who needed emergent intubation were included.

Outcome measurement: Peri-intubation instability

Study Results:

Compared with etomidate, ketamine was associated with higher risk of hemodynamic instability and moderate certainty (RR 1.29, 95% CI 1.07-1.57). 

Ketamine was associated with lower risk of adrenal suppression, again, with moderate uncertainty (RR 0.54, 95% CI 0.45-0.66).

Ketamine was not associated with differences and risk of first successful intubation nor mortality.

Discussion:

Most studies were single center and involved small-moderate sample size, ranging from 20 patients to 700 patients.

For adrenal suppression, there were only 3 studies and a total of 1280 patients, thus, the results are still not definitive.

For an academic exercise, the Number Needed to Harm for both hemodynamic instability and adrenal suppression are calculated here.

Number Needed to Harm for hemodynamic instability: 25.

Number needed to harm for adrendal suppression: 11.

Show References

In out of hospital cardiac arrest (OHCA), does it matter if you choose an intraosseous (IO) vs intravenous (IV) approach to getting access and giving meds?

No, according to a recent study by Couper et al, just published in NEJM.  No significant difference in any clinically meaningful outcome including survival, neurologically intact discharge, etc.  Technically the IV group had slightly higher rates of ROSC, which just met statistical significance, and to be fair that group did trend very slightly towards better outcomes in some categories, but really well within the range expected by statistical noise.  

Interestingly, the median time from EMS arrival to access being established was the same in both groups (12 minutes), which I think raises some face validity questions.  Furthermore, of course, previous trials have raised questions as to whether ACLS meds even work or impact outcomes anyways, so naturally if they don't, the method by which they are given isn't likely to matter either.

Bottom Line: This large, well conducted trial continues to support the notion that either an IV-focused, or IO-focused approach to access and medication delivery in OHCA is reasonable.  You and your prehospital colleagues can likely continue to make this decision based on personal comfort, local protocols, and patient/case circumstances.  At the very least, this continues to support the notion that if an IV is proving challenging, pursuing an IO instead is a very appropriate thing to do.

Show References

Intravascular Volume and the IVC

• Point-of-care ultrasound (POCUS) assessments of the inferior vena cava (IVC) are frequently used in the fluid resuscitation strategy for critically ill patients.
• Numerous factors determine the appearance of the IVC, including intraabdominal pressure, mean systemic filling pressure, central venous pressure, intrathoracic pressure, and right heart function.
• Given these multitude of factors, it is not surprising that literature has demonstrated that the IVC is not a reliable marker of fluid responsiveness.
• Rather, focus on the use of the IVC has shifted towards assessing fluid tolerance, venous congestion, and its use as a marker on when to stop fluid administration.
• POCUS assessment of the IVC is commonly performed in the long axis a few centimeters distal to the diaphragm.  
• Rola, et al. highlight that this location may be misleading and recommend  that a more appropriate assessment be a short axis scan through the entire intrahepatic segment of the IVC, while taking into account the intrapleural and intraabdominal pressures.

Show References

Getting reliable venous and arterial access is crucial when resuscitating critically ill patients. These lines can be difficult due to patient and situation specific variables. 

Micropuncture kits contain a 21-gauge echogenic needle, a stainless-steel hard shaft/soft-tip wire, and a 4 Fr or 5 Fr sheath and introducer. The micropuncture kit offers several advantages that can help overcome difficult situations:

• Small, Sharp Needle: Easier puncture of compressible vessels.
• Echogenic Design: Improved visibility under ultrasound.
• Smooth Tissue Penetration: Moves through tissue more easily than a typical 18-gauge needle.
• Flexible Wire Tip: The 0.018-inch wire is soft, lacks a J-loop, and navigates tight corners and calcifications better than a standard J-tip wire. This is especially useful when entering at a steep angle or accessing small vessels.

To use a micropuncture kit, gain vessel access with the needle and wire, railroad the sheath and introducer into the vessel, remove the wire, then remove the introducer. Now you have a 4 Fr or 5 Fr sheath in the vessel. This is typically used to introduce a normal central line wire. 

For arterial lines, you can place them directly over the wire without dilation. Keep in mind that the 4 Fr sheath (1.3 mm OD) and 5 Fr sheath (1.7 mm OD) are larger than a typical arterial line catheter (18g = 1.27 mm OD). If you dilate then you will cause hematoma.

Find out where your department stores micropuncture kits and get familiar with their components. While it adds an extra step to the procedure, it could make the difference between securing the line or not.

Show References

Background:

Septic shock is a severe and common critical illness that is managed in the emergency department.  Our current foundation of treatment includes IV fluids, empiric antibiotic coverage, vasopressor therapy, source control and corticosteroids for refractory shock.  The levels of nitric oxide (NO) and hydrogen sulfide (H2S) are elevated in sepsis and associated with worse outcomes.  Hydroxocobalamin is an inhibitor of NO activity and production and a scavenger of H2S [1,2].  Most of the current data is limited to observational studies looking at hydroxocobalamin in cardiac surgery related vasodilatory shock with few case series and reports for use in septic shock.  The available data has shown an improvement in hemodynamics and reduction in vasopressor requirements in various vasodilatory shock states [2].  Chromaturia and self-limited red skin discoloration are common side effects but current data has not shown significant adverse events [3,4].  Patel et al, performed a phase 2 single-center trial to evaluate use of high dose IV hydroxocobalamin in patients with septic shock. 

Study:

• Single-center, double-blind RCT, 20 patients (10 hydroxocobalamin, 10 placebo)
• Included patients >/= 18 years of age within 48 hours of admission with a diagnosis of septic shock (based on Sepsis 3 criteria) who were receiving norepinephrine (NE) of 0.10 mcg/kg/min for at least 15 minutes or an equivalent dose of alternative vasopressor.
• Notable exclusion criteria were patients with a history of urinary calcium oxalate crystals, active hemolysis or bleeding, impending death.
• Intervention group received a single dose of 5 grams of IV hydroxocobalamin administered over 15 minutes
• Primary outcome – Feasibility Study (*Initial primary outcome was reduction in vasopressor dose but was changed during the COVID-19 pandemic to a feasibility study*)
• Secondary outcomes – Change in H2S levels and NE dose from randomization to 30 minutes and 3 hours after IV hydroxocobalamin.

Results

• Achieved feasibility with enrollment goal, receiving intervention, no contamination and good follow up.
• For secondary outcomes the study group showed a statistically significant relative decrease in vasopressor dose compared to placebo at 30 minutes (-36% vs 4%, p < 0.001) and 3 hours after infusion (-28% vs 10%, p = 0.019). 
• Non-statistically significant reduction in H2S levels in the intervention group compared to placebo.
• Tertiary outcomes of hospital mortality, ICU mortality, ICU and vasopressor free days did not show any significant difference between the groups. (The study was not designed with the power to look for a difference in these outcomes).  

Take home

There is a low risk of serious adverse events from high dose hydroxocobalamin use [3,4].  For now, it may be reasonable to consider in cases of septic shock refractory to standard care but there isn’t enough data to support its regular use yet.

Show References

Title: Albumin Versus Balanced Crystalloid for the Early Resuscitation of Sepsis: An Open Parallel-Group Randomized Feasibility Trial— The ABC-Sepsis Trial

Settings: 15 ED in the United Kingdom. This study is a feasibility study but it looked at mortality as a primary outcome.
Participants: 
•    Patients with Sepsis, with their National Early Warning Score (NEWS) ? 5 (These patients have estimated mortality of 20%). IV fluid resuscitation needs to be within 1 hour of assessment.
•    300 Patients were randomized to receive balanced crystalloids or 5% human albumin solution (HAS) only, within 6 hours of randomization.
Outcome measurement: 30-day mortality, Hospital length of stay (HLOS)
Study Results:
•    The median time for receiving IV fluid from randomization was 41 minutes (HAS) vs. 36 minutes (crystalloids).
•    Total volume of IV fluid per Kg  in first 6 hours 14.5 ml/kg  (HAS) vs. 18.8 ml/kg (crystalloids).
•    Other interventions (vasopressor, Renal replacement therapy, invasive ventilation) were similar.
•    Complications (AKI, pulmonary edema, allergy) were lower for Crystalloids group
•    Median hospital LOS = 6 days for both groups.
•    90-day mortality: 31 (21.1%) (HAS) vs. 22 (14.8%) (Crystalloids), OR 1.54 (95% 0.8-2.8)
Discussion:
•    Total volumes for resuscitation in the first 6 hours was 750 ml (HAS) and 1250 ml (crystalloids). This signified a trend toward lower total volume of resuscitation (remember that 30 ml/kg recommendation)
•    The 2024 guidelines from Chest (REF 2) suggested that: “In Critically ill adult patients (excluding patients with thermal injuries and ARDS), intravenous albumin is not suggested for first line volume replacement or to increase serum albumin levels. Therefore, we should not give patients (except for cirrhosis or spontaneous bacterial peritonitis) albumin just to reduce the volume of fluid.
•    The authors suggested that even a definitive trial in the future will not be able to demonstrate a significant benefit of using 5% albumin.
Conclusion: 
There is lower mortality (numerical but not statistically) among the group with balanced crystalloids.

Show References

Some points from this narrative review:

• much of the existing literature is based on animal models or small human studies
• successful ventilations per compression pause (“synchronous" ventilation, 30:2, without advanced airway) is unsurprisingly important for neurologically intact survival
• no clear difference in outcomes between “synchronous” vs. “asynchronous” (insufflation without pause in CPR) ventilation
• RR below 6 breaths per min were associated with decreased ROSC, whereas faster RR were not associated with worse outcomes (however, be cautious of breathstacking in pts with asthma/COPD)
• chest rise can be detected with TVs as low as 180 mL which is likely not sufficient for CPR
• the benefit of larger tidal volumes (improved oxygenation, less hypercapnia) may outweigh the perceived costs (gastric insufflation, impact on venous return/CO)

Take home pearls:

• use 2-person BVM to ensure adequate TVs and aim for more than just minimal chest rise
• err on the side of moderately larger TVs rather than smaller and moderately faster RR rather than slower (but be cautious in pts with asthma/COPD)

Show References

I wanted to send out two websites curated in part by UMEM current and past faculty/residents/fellows which have a wealth of critical care lectures and resources:

• The Maryland Critical Care Project is managed by the current UMMC CCM fellows and has one of the best collections of CC lectures out there with some big names in CC (it is also founded by several UMEM faculty/alumni). Check out their core content series.
• STCMTCC.com is a website curated by STC CC faculty with a wealth of trauma-oriented education (and STC-specific guidelines for our current residents).  Check out the extracorporeal organ support section.

Disclosure: *I am one of the webmasters for the STCMTCC, but have no affiliation with MCCP other than as an enthusiastic reader.

Airway management in the pre-hospital setting is a matter of much controversy, and overall I will defer to my EMS colleagues, but several previous studies have failed to show a benefit to endotracheal intubation in the field as opposed to alternate approaches like a supraglottic airway.  Another nod in this direction has recently come out, with Battaglini et al performing a post-hoc analysis of one of the larger studies in the history of cardiac arrest, TTM-2, looking specifically at outcomes stratified by pre-hospital airway management strategy.  

Do patients who undergo endotracheal intubation in the field do better than those who get a supraglottic airway?

No, they don't.  TTM-2 included 1900 patients, of whom 1702 had enough data to be included in this re-analysis.  28% got supraglottic airways, and 72% got endotracheal intubation.  The groups were reasonably well matched on most characteristics, and if anything most well-known prognostic factors favored the endotracheal intubation group (very slightly).  It should be noted that several outcome metrics, including modified Rankin scale, did show slight signs of benefit for the endotracheal intubation group, even sometimes in a statistically significant fashion, but fell out when a multi-regression analysis, which was the primary endpoint, was done.  

Bottom Line: In pre-hospital cardiac arrest, there remains limited data to support the notion that endotracheal intubation results in better outcomes than supraglottic airway placement.  You should defer to your local protocols and continue to work with your paramedics and EMS directors as evidence continues to evolve.  For now, I don't think there's sufficient data to suggest that a given patient should be intubated vs undergoing supraglottic airway placement, and it is probably best to defer to the judgement, training, and protocols of your folks on scene.

Show References

Question

The incidence of opioid-overdose-related deaths has clearly increased in the past decade, with recent estimates of up to 17% of OHCA being opioid-related in 2023. [1,2] The use of naloxone for opiate reversal in overdose is well-established, with reasonable inference but no formal proof that its use could help in opioid-associated out of hospital cardiac arrest (OA-OHCA). [3] The August publication of two trials [4,5] retrospectively examining naloxone administration in OHCA offers some perspectives…

• Patients receiving naloxone for OHCA are:
  • More often be younger, with fewer comorbidities, but more often unwitnessed than their non-naloxoned counterparts
  • More likely to have opioid OD as a presumed etiology

and

• Naloxone administration is associated with:
  • Increased rates/odds of ROSC and survival to hospital discharge, whether OD is suspected or not
  • And “early” naloxone (given prior to EMS IV/IO access) is associated with increased rates of DC with good neuro outcome in PEA compared to receipt after IV/IO access or none at all

[View “Visual Diagnosis” for slightly more detail on the referenced studies.]

Bottom Line: While prospective trials are absolutely needed to offer more definitive evidence regarding the use of empiric naloxone in nontraumatic OHCA, the rising incidence of OA-OHCA in the U.S. and current findings are convincing enough to encourage early naloxone administration, especially in populations with higher incidence of opioid use.

U.S. Mortality due to Opioid Overdose (CDC data)

Show Answer

Show References

Attachments

• download-66d58c5aeeeff.gif (929 Kb)
• Study Info-66d5907c6d792.png (78 Kb)

Hepatorenal Syndrome

• Emergency physicians evaluate patients with cirrhosis and ascites daily.
• Patients with cirrhosis are particularly susceptible to acute kidney injury (AKI), which is associated with a significant increase in hospital mortality.
• Hepatorenal syndrome (HRS) is a specific type of renal dysfunction in patients with cirrhosis and ascites.
• The previous classification of HRS (Type 1, Type 2) has now been replaced by HRS-AKI, HRS-AKD, and HRS-CKD.
• The diagnostic criteria for HRS-AKI include:
  • Increase in creatinine 0.3 mg/dL within 48 hrs or 50% from baseline value within the prior 7 days
  • Lack of improvement in creatinine or urine output within 24 hrs of adequate volume resuscitation
  • Absence of an alternative explanation for AKI
• Management of HRS-AKI centers on accurate volume assessment, timely administration of a splanchnic vasoconstrictor (norepinephrine), administration of 20-25% albumin, and avoidance of additional nephrotoxins.

Show References

Euglycemic DKA (eDKA) is a medical emergency requiring prompt attention. It is caused by an imbalance of insulin and glucagon leading to ketone accumulation (1-3). In addition to typical risk factors for DKA, those for eDKA include SGLT-2 inhibitor use and pregnancy, with 30% of DKA cases in pregnancy presenting euglycemic (4, 5).

eDKA presents with an anion gap metabolic acidosis, ketosis/ketonuria, & blood glucose less than 250 mg/dL.

Diagnosis requires ruling out other causes of anion gap metabolic acidosis, including toxic ingestions.

The cornerstone of eDKA management is ensuring enough dextrose to allow needed insulin administration to reverse ketone accumulation.

Pitfalls

• Not giving enough insulin to reverse ketosis due to concern about low blood sugars
• Not giving enough dextrose to support sufficient insulin dosing
• Not uptitrating insulin for refractory acidosis caused by eDKA

Pearls

• Start insulin with at least 0.05 units/kg/hour along with IV dextrose (3,5,7,9)
• Start IV dextrose at 5-10 g/hr (9). This will be 100-200 mL/hr of a 5% dextrose solution (dextrose should be added to either normal or ½ normal saline to avoid causing hyponatremia!)
  • Dextrose concentrations: D5 = 50 g/L || D10 = 100 g/L || D20 = 200 g/L
• Euglycemic DKA may present WITHOUT ketonuria if the patient is on an SGLT-2 inhibitor (7,8) – send a beta hydroxybutyrate!
• eDKA is most common in the first two months of SGLT-2 inhibitor use, but can happen at any time (6)

Show References