UMEM Educational Pearls - By Quincy Tran

Show References

Background: SOFA score has been used as a predictor for poor outcomes in patients with sepsis.  However, the original SOFA score utilizes PaO2/FiO2 ratio to calculate the SOFA’s respiratory component.  When there are no ABG, thus no PaO2, we have to convert patients’ spO2 to PaO2, and the amount of oxygen support to FiO2 (for example, 2 liters of oxygen via nasal cannula = 0.27).  This is cumbersome.

Objective: This study assessed whether spO2 can be used instead of PaO2/FiO2 ratio for SOFA’s respiratory score.

Settings: 8 hospitals across Sweden and Canada

Patients: Adults with sepsis.  19396 patients were included for the derivation group while there were 10586 patients for the validation cohort.

Study Results:

• When PaO2 was not measured, assigning respiratory SOFA score of 1 for spO2 94% and respiratory SOFA score of 2 for spO2 < 90% had good discriminatory capability.
• The AUROC with SOFA score using spO2 threshold as above was 0.783 (0.767-0.798), which was slightly increased from the model using previous methods to calculate respiratory SOFA scores (0.781 [0.765-0.796]).

Discussion:

• For Emergency Medicine, using a cut-off threshold  for spO2 of 94% (respiratory SOFA of 1) and 90% (respiratory SOFA of 2) can simplify how to calculate the SOFA score.  
• It’s also simpler to do research when we collect SOFA score retrospectively.

Show References

Settings: International multicenter trials; 1:1 randomization, blinded assessment of outcomes.

Patients: adults with witnessed OHCA, regardless of initial rhythm.  Patients had more than 20 minutes of CPR.  Eligible patients were unconscious, not able to follow command, no verbal responses to painful stimuli.

Intervention: hypothermia to target of 33C for 28 hours, then rewarming at rate of 1/3C every hour until  37C.

Comparison: maintaining temperature at 37.5C or less.  Cooling if body temperature reached 37.8C to 37.5C

Outcome: primary outcome was Any cause mortality at 6 months; secondary outcome was poor functional outcome at 6 months (modified Rankin Scale 4-6).

Study Results:

1. 930 hypothermia, mortality 465/925 (50%, RR 1.04, 95%CI 0.94-1.14); 488/881 (55%) had mRS 4-6 (RR 1.0, 95%CI 0.92-1.09).

2. 931 normothermia, mortality 446/925 (48%); 479/866 (55%) had mRS 4-6.

Discussion Points:

• Hypothermia would lead to higher rates of arrhythmia-related hemodynamic instability.
• More studies reinforced that preventing fever is beneficial.
• ED clinicians will not have to rush to cool patients while awaiting for ICU beds (Yay).

Conclusion:

Normothermia in coma patients after OHCA did not lead to higher morality or worse neurologic outcomes.

Show References

Background:

Cardiac arrest from massive pulmonary embolism (PE) can be up to 90% (1). A recent systemic review evaluated the efficacy of Venoarterial-Extracorporeal Membrane Oxygenation (VA-ECMO) for PE-related cardiac arrest.

Results:

The authors screened 1115 articles and included 77 articles, including gray literature.  The authors performed a quantitative analysis of a total of 301 patients.

Overall, 183/301 (61%) patients survived to hospital discharge, a significant improvement from 90%.

Patients who were cannulated during chest compression were associated with 7x higher odds of death (OR, 6.84; 95% CI, 1.53–30.58; p = 0.01), compared to those who were cannulated after ROSC.  However, cannulation in the ED was not associated with improved outcomes, compared with other cannulation site.

No increased risk of death among patients who received tPA prior to VA-ECMO vs. those who did not (OR, 0.78; 95% CI, 0.39–1.54; p = 0.48).

Patients whose age > 65 years of age were associated with 3X risk of death, compared to those with age < 65 years (OR, 3.56; 95% CI, 1.29–9.87; p = 0.02).

Take-home points

Please consider “early” VA-ECMO for eligible patients who have cardiac arrest from massive PE.  However, it will take great convincing to push the PERT team to cannulate for VA-ECMO while the patient is still receiving chest compression.

Show References

Background:

Cardiac arrest from massive pulmonary embolism (PE) can be up to 90% (1). A recent systemic review evaluated the efficacy of Venoarterial-Extracorporeal Membrane Oxygenation (VA-ECMO) for PE-related cardiac arrest.

Results:

The authors screened 1115 articles and included 77 articles, including gray literature.  The authors performed a quantitative analysis of a total of 301 patients.

Overall, 183/301 (61%) patients survived to hospital discharge, a significant improvement from 90%.

Patients who were cannulated during chest compression were associated with 7x higher odds of death (OR, 6.84; 95% CI, 1.53–30.58; p = 0.01), compared to those who were cannulated after ROSC.  However, cannulation in the ED was not associated with improved outcomes, compared with other cannulation sites.

No increased risk of death among patients who received tPA prior to VA-ECMO vs. those who did not (OR, 0.78; 95% CI, 0.39–1.54; p = 0.48).

Patients whose age > 65 years of age were associated with 3X risk of death, compared to those with age < 65 years (OR, 3.56; 95% CI, 1.29–9.87; p = 0.02).

Take-home points

Please consider “early” VA-ECMO for eligible patients who have cardiac arrest from massive PE.  However, it will take great convincing to push the PERT team to cannulate for VA-ECMO while the patient is still receiving chest compression.

Show References

Background:

The association between time intervals of ED antibiotic administration and outcome has been controversial.  While single studies showed there was increased mortality associated with delayed antibiotic administration (1-3). A meta-analysis of 13 studies and 33000 patients showed that there was no mortality difference between septic patients receiving immediate Abx (< 1 hour) vs. those receiving early abx (1-3 hours) (4).

Since delay in recognition of sepsis (defined as ED triage to Abx order) and delay in antibiotics delivery (Abx order to administration) contribute to total delay of Abx administration, a new retrospective study (3) attempted to investigate the contributions of either factor to hospital mortality.

Results:

The study used generalized linear mixed models and involved 24000 patients.

For All patients and outcome of hospital mortality:

Recognition delay (ED triage to Abx order): OR 2.7 (95% CI 1.5-4.7)*

Administration delay at 2-2.5 hours (Abx order to administration): OR 1.5 (1.1-2.0)

These results was associated with non-statistical significance in patients with septic shocks.

Conclusion:

Delayed recognition of sepsis was associated with higher hospital mortality.  Longer delay of abx administration was also associated with increased risk of hospital mortality.

Show References

A single center (Nebraska, USA), retrospective analysis investigated the prevalence of intubation during hospital stay for 105 patients who had COVID-19 between March 24 to May 5, 2020 (1).   40 patients underwent awake proning vs. 60 patients did not undergo awake proning.

After adjusting for either SOFA or APACHE scores, patients with awake proning were associated with lower Hazard Ratios of intubation for SOFA (HR 0.30, 95% CI 0.09-0.96, p=0.043) and APACHE (HR 0.30, 95%CI 0.1-0.91, p=0.034).

Discussion

While this US study seemed promising, another Brazilian study being published earlier in July 2020 showed no difference in the prevalence of intubation between COVID-19 patients with proning or without proning (2).

These 2 studies highlighted the nature of this disease: high practice variability, uncertainty of therapeutic modalities.  However, the complications from awake proning had been very low.

Conclusion:

Awake proning for hypoxic COVID-19 patients is a promising intervention but we will need more studies.  In the meanwhile, we can try this therapeutic modality as the risk is low.

Show References

Summary

Our group performed a meta-analysis to assess whether it is safe to infuse vasopressor through peripheral venous catheters.  We identified 9 studies with a total of 1835 patients.  The prevalence of complications among the pooled patient population was 9%.  Up to 96% of the complications was extravasation and almost no complications required any treatment.

A few studies reported safe infusion of norepinephrine up to 0.1 mcg/kg/min for up to 24 hours.

In exploratory meta-regression, catheter size 20 or larger was negatively associated with the rate of complications.

We also observed that studies that were published within the past 5 years reported significantly lower rate of complications from older studies.  This suggested that with careful planning and monitoring, it is safe to start vasopressor through peripheral IV.

Limitation

most of the included studies were observational. No studies had enough power to statistically analyze any variables that could predict complications.

Bottom line: we should start vasopressor as soon as indicated, if we have good, reliable IV access.

Show References

A retrospective study analyzed data from 757 patients with spontaneous intraparenchymal hemorrhage.

Within the first 6 hours of admission, patients who had systolic blood pressure reduction between 40 – 60 mm Hg (OR 1.9, 95% CI 1.1-3.5) or reduction ≥ 60 mm Hg (OR 1.9, 95%CI 1.01-3.8) were associated with almost double likelihood of poor discharge functional outcome (defined as modified Rankin Scale 3-6).

Additionally, large systolic blood pressure reduction ≥ 60 mm Hg in patients with large hematoma (≥ 30.47 ml) was associated with higher likelihood of very poor functional outcome (mRS 5-6).

Take home points: while more studies are still needed to confirm these observations, perhaps we may not want to drop blood pressure in patients with spontaneous intraparenchymal hemorrhage too much and too fast.

Show References

Show References

Patient

·         A North America multicenter study involving 821 asymptomatic patients who had exposure to Covid-19-positive patients.  The study was double-blind, placebo-controlled randomized trial.

Intervention

·         Within 4 days of exposure, participants were randomized to receive hydroxychloroquine.  Dose of hydroxychloroquine was 800 mg once then 600 mg in 6-8 hours then 600 mg daily for 4 more days.

·         There were 414 patients in this arm. Median age 41 years [IQR 33-51]

Comparison:

·         Placebo treatment.  There were 407 patients in this arm. Median age 40years [IQR 32-50]

Outcome:

·         Incidence of either laboratory-confirmed Covid-19 or Covid-19 symptoms within 14 days.

Results:

·         49 (11.8%) patients with treatment had Covid-19 findings (positive tests or symptoms)

·         58 (14.3%) patients with placebo had Covid-19 findings (p=0.35). 

·         The absolute difference was -2.4%.  The number need to treat (NNT) to prevent one infection is 42 patients.  Number needed to harm is 50 patients.

·         Symptoms were fatigue (49.5%), cough (44.9%), sore throat (40.2%) myalgia (37.4%), fever (34.6%), anosmia (23.4%), shortness of breath (18.7%).

Conclusion:

Hydroxychloroquine prophylaxis did not prevent post-exposure Covid-19 infection.

Show References

Title: Randomized Trial of Three Anticonvulsant Medications for Status Epilepticus

Settings:

• 57 US hospitals: 26 sites for adults only, 18 sites enrolling only children, 13 sites enroll both.

Patients:

• 384 patients whose ages were 2 years and older. 
• Patients who continued to have generalized seizure for at least 5 minutes after “accepted” cumulative dose of benzodiazepines.

Intervention:

• levetiracetam at a dose of 60 mg per kilogram (maximum, 4500 mg),
• Fosphenytoin at a dose of 20 mg PE per kilogram (maximum, 1500 mg PE),
• valproate at a dose of 40 mg per kilogram (maximum 3000 mg)

Comparison:

• Patients > 32 kg total body weight:  diazepam of 10 mg; Lorazepam 4mg Intravenously; midazolam 10 mg intravenously or intramuscularly.
• Patients < 32 kg total body weight: diazepam at a dose of 0.3 mg per kilogram (administered intravenously or rectally), lorazepam at a dose of 0.1 mg per kilogram (administered intravenously), or midazolam at a dose of 0.3 mg of per kilogram (administered intramuscularly) or 0.2 mg per kilogram (administered intravenously)

Outcome: absence of clinical seizure at 60 minutes after infusion of medication.

Study Results:

• Rates of cessation of status epilepticus were similar in all 3 groups: 47% of levetiracetam vs. 45% Fosphenytoin vs. 46% for valproate.
• Fosphenytoin was associated with non-significantly higher rate of hypotension (3.2%) vs other drugs.
• Levetiracetam was associated with non-significantly higher rate of death (4.7%) vs. other drugs.
• All three medication was associated with similar rate of intubation within 60 minutes of drug infusion.

Discussion:

• The median time interval from start to cessation of status epilepticus appeared to be shorter for valproate but there was no formal analysis yet,
• Valproate (7.0 minutes) vs. levetiracetam (11.7 minutes) vs. Fosphenytoin (11.7 minutes)

Conclusion:

• Three medications, Fosphenytoin, levetiracetam, valproate were equally effective.

Show References

Settings: Multicenter randomized controlled trial

Patients: 710 patients

Intervention: 345 patients.  no sedative but only boluses of morphine as clinically indicated (Sedation group)

Comparison: 356 patients.  light sedation with daily interruption (Nonsedation group)

Outcome: all-cause mortality at 90 days after randomization

Study Results:

42.4% of nonsedation group died vs 37% of sedation group (95% confidence interval [CI], −2.2 to 12.2; P = 0.65). 

Number of ventilator-free days for nonsedation group was 27 days vs. 26 for sedation group. 

Discussion:

This study did not agree with previous studies that lighter sedation was associated with shorter length of stay on mechanical ventilation , ICU or hospital.  The authors attributed to the findings that RASS score was not significantly different between the 2 groups.

Conclusion:

Critically ill adult patients receiving mechanical ventilation, there was no difference in 90-day mortality between patients receiving light sedation or no sedation.

Show References

Settings: multicenter, double-blind, phase 3 trial (apparently vitamin D worked in phase 2 trials).

• Patients:
  • 1059 patients were enrolled within 12 hours of ICU admission.  The patients had to have risk factors warranted ICU admisions (pneumonia, sepsis, mechanical ventilation, shock, pancreatitis, etc.).
  • Vitamin D deficiency was defined as plasma level < 20 ng/ml
• Intervention:
  • 531 patients received a single oral dose of 540,000 IU of vitamin D3 within 2 hours after randomization
• Comparison
  • 528 patients received placebo
• Outcome
  • 90-day all-cause mortality

Study Results:

• Total SOFA score was similar in both groups (5.6 vs. 5.4).               
• On day 3, mean plasma vitamin D was higher (47 ng/ml) in treatment group vs 11 ng/ml in placebo group
• 90-day all cause mortality was similar.  Treatment group was 23.5% vs. 20.6% for placebo (95% CI, −2.1 to 7.9; P = 0.26).
• Vitamin D-related adverse events were similar in both groups.

Discussion:

• This trial enrolled patients early in their critical illness compared to phase 2 trial which enrolled patients after 3 days in the ICU.
• This phase 3 trial also enrolled mostly medical-related illness, whereas 75% of patients in phase 2 had either surgical or neurology-related illnesses.

Conclusion:

Early administration of high dose vitamin D did not improve 90-day all cause mortality.

Show References

Settings

• Patients: mechanical ventilation in the ICU. Randomization of 1000 patients.
• Intervention: conservative oxygen therapy, if spO2 reached 97%, then FiO2 was lowered to 0.21
• Comparison: no specific limits for FiO2 or SpO2.
• Outcome: number of ventilator-free days at 28 days after randomization.

Study Results:

• 484 conservative-oxygen group vs  481 to the usual oxygen group
• Comparing to the conservative-oxygen group had:
• more time at FiO2 21 (29 hours vs. 28 hours),
• less time with SpO2 > 97% (27 hours vs. 49 hours)
• Similar ventilator-free days: 21 days vs. 22 days.

Discussion:

This study’s results differed from previous single center study (Girardis JAMA 2016) or meta analysis (Chu DK, Lancer 2018), which showed mortality benefit in patients with conservative oxygen (Girardis & Chu) and more ventilator-free days (Girardis).

Conclusion: Conservative oxygen did not significantly affect the ventilator free days of mechanically ventilated patients.

Show References

Rationale: Data regarding temperature management in patients suffered from cardiac arrest with nonshockable rhythm was inconclusive.

Objective: whether moderate hypothermia at 33C, compared with normothermia at 37C would improve neurologic outcome in patients with coma after cardiac arrest with nonshockable rhythm.

Outcome: survival with favorable 90-day neurologic outcome (Cerebral Performance Category scale 1-2/5)

SummaryThere was higher percentage of patients achieving CPC 1-2 in the hypothermia group (10.2%) vs normothermia group (5.7%, Hazard Ratio 4.5, 95% CI 0.1-8.9, p=0.04)

This randomized multicenter trial involved 581 patients with cardiac arrest and nonshockable rhythm.  Hypothermia group included 284 patients vs. 297 in the normothermia group.  Median GCS at enrollment = 3.

Majority of patients was cooled with the use of a basic external cooling device: 37% for hypothermia and 50.8% for normothermia group.

There was higher percentage of patients achieving CPC 1-2 in the hypothermia group (10.2%) vs normothermia group (5.7%, Hazard Ratio 4.5, 95% CI 0.1-8.9, p=0.04)

Limitation:

A. The study used strict enrollment criteria:

1. CPR initiation within 10 minutes;
2. CPR to ROSC within 60 minutes;
3. epinephrine or norepinephrine infusion at < 1 ug/kg/min;
4. No Child-Pugh class C liver cirrhosis

B. normothermia group had higher proportion of patients with temperature at 38C.

C. Hypothermia group underwent temperature management of 56 hours vs. 48 hours for normothermia patients.

Take home points:

In a selected group of patients with cardiac arrest and nonshockable rhythm, moderate hypothermia at 33C may improve neurologic outcome.

Show References

A new study confirmed the previously-known antibiotics to be associated with Torsades de pointes and QT prolongation (Macrolides, Linezolid, Imipenem and Fluoroquinolones). However, this study found new association between amikacin and Torsades de pointes/QT prolongation.

Methods

The authors queried the United States FDA Adverse Event Reporting System (FAERS) from 01/01/2015 to 12/31/2017 for reports of Torsade de points/QT prolongation (TdP/QT).

Reporting Odd Ratio (ROR) was calculated as the ratio of the odds of reporting TdP/QTP versus all other ADRs for a given drug, compared with these reporting odds for all other drugs present in FAERS

Results

FAERS contained 2,042,801 reports from January 1, 2015 to December 31, 2017. There were 3,960 TdP/QTP reports from the study period (0.19%).

Macrolides               ROR 14 (95% CI 11.8-17.38)

Linezolid                  ROR 12 (95% CI 8.5-18)

Amikacin                 ROR 11.8 (5.57-24.97)

Imipenem-cilastatin ROR 6.6 (3.13-13.9)

Fluoroquinolones   ROR 5.68 (95% CI 4.78-6.76)

Limitations:

These adverse events are voluntary reports

There might be other confounded by concomitant drugs such as ondansetron, azole anti-fungals, antipsychotics.

Show References

Torsades de pointes and QT prolongation Associated with Antibiotics

Methods

The authors queried the United States FDA Adverse Event Reporting System (FAERS) from 01/01/2015 to 12/31/2017 for reports of Torsade de points/QT prolongation (TdP/QT).

Reporting Odd Ratio (ROR) was calculated as the ratio of the odds of reporting TdP/QTP versus all other ADRs for a given drug, compared with these reporting odds for all other drugs present in FAERS

Results

FAERS contained 2,042,801 reports from January 1, 2015 to December 31, 2017. There were 3,960 TdP/QTP reports from the study period (0.19%).

Macrolides               ROR 14 (95% CI 11.8-17.38)

Linezolid                  ROR 12 (95% CI 8.5-18)

Amikacin                 ROR 11.8 (5.57-24.97)

Imipenem-cilastatin ROR 6.6 (3.13-13.9)

Fluoroquinolones   ROR 5.68 (95% CI 4.78-6.76)

Limitations:

These adverse events are voluntary reports

There might be other confounded by concomitant drugs such as ondansetron, azole anti-fungals, antipsychotics.

Bottom Line:

This study confimed the previously-known antibiotics to be associated with Torsades de pointes and QT prolongation (Macrolides, Linezolid, Imipenem and Fluoroquinolones). However, this study  found new association between amikacin and Torsades de pointes/QT prolongation.

Show References