Skip to content

12 Must-Know EM Pharmacotherapy Articles of 2017


Welcome to the 4th annual installment of our must-know Emergency Medicine pharmacotherapy articles post, this time for 2017. We summarize some important EM pharmacotherapy articles from the last 12 months. We have tried to focus on articles that you may have missed, but are potentially high-impact for improving clinical practice in the ED. Without further ado, we present the 12 must-know EM pharmacotherapy articles of 2017.

1. IV Nitroglycerin Bolus for Acute Pulmonary Edema2. Are Vasopressors Useful for Toxin-Induced Cardiogenic Shock?3. Ketamine as a First-Line Treatment for Severe Agitation in the ED4. How to Administer Low-Dose IV Ketamine for Pain in the ED5. Hemodialysis and Survival in Intubated Salicylate-Poisoned Patients6. Importance of Second Antibiotic Doses in ED Sepsis Patients7. Uncomplicated Cellulitis? Consider Strep-only Coverage8. No Icatibant for ACE-I Induced Angioedema9. Don’t be Afraid of Physostigmine10. Risk of Bleeding After Antiplatelet or Oral Anticoagulant Overdose11. Is IV Acetaminophen Worth it in the ED?12. Don’t Dismiss Charcoal’s Potential Benefit in Overdose
Wilson S, Kwiatkowski G, Millis S, Purakal J, Mahajan A, Levy P. Use of nitroglycerin by bolus prevents intensive care unit admission in patients with acute hypertensive heart failure. Am J Emerg Med. 2017;35(1):126-131. [PubMed]

Nitroglycerin (NTG) is an important preload reducer in acute pulmonary edema, and even modestly reduces afterload with high doses. For pulmonary edema in the ED, NTG is often administered as a sublingual tablet and/or IV infusion. Starting the infusion at ≥ 80 mcg/min produces rapid effects in many patients, and can be titrated higher as needed. Combined with noninvasive positive pressure ventilation (and sometimes IV enalapril), patients often turn around quickly, from the precipice of intubation to comfortably lying in bed. But what about high-dose IV bolus NTG?

In a UMEM pearl, Dr. Rory Spiegel (@EMNerd_) highlighted 2 studies in which patients who received high doses of IV nitrates every 5 minute were intubated less often than patients who received a standard infusion.1,2

This new retrospective study by Wilson and colleagues evaluated the incidence of ICU admission and hospital LOS in acute pulmonary edema patients who received intermittent high-dose NTG bolus vs. standard NTG infusion vs. bolus + infusion. This was actually a follow-up to their previously published study.

What They Did

  • 395 patients: 182 received intermittent bolus, 182 received continuous infusion, 89 received bolus + infusion
    • The groups were relatively well-matched despite the retrospective nature
  • What was a high-dose bolus? 2 mg IV every 3-5 minutes

What They Found

  • ICU admission: 48% bolus, 69% infusion, 83% combo
  • Hospital Length of Stay: 3.7 days bolus, 4.7 days infusion, 5.0 days combo
  • 80% of patients in bolus group received only 1 dose
  • In both the infusion and combination groups, the median starting rate was 20 mcg/min with a maximum of 60 mcg/min
  • Incidence of hypotension was low (< 6%) and not different between the groups


  • Single center study means that ICU admission practices and hospital LOS (the 2 primary outcomes) are not necessarily similar or generalizable to other institutions.
  • The infusion rates in both the infusion and combination groups were quite low, and aren’t necessarily fair comparisons to a high-dose bolus.

Application to Clinical Practice

  • High-dose NTG seems to be effective at reducing ICU admissions in acute pulmonary edema, with a low incidence of hypotension.
  • This study does NOT prove high-dose bolus NTG is more effective than infusions. The infusion rates were too low to fairly compare the 2 methods.
  • High-dose NTG, whether by bolus, infusion, or bolus + infusion remains a cornerstone of early acute pulmonary edema management.

Originally posted December 27, 2016 on

Skoog C, Engebretsen K. Are vasopressors useful in toxin-induced cardiogenic shock? Clin Toxicol (Phila). 2017;55(4):285-304. [PubMed]

Toxin-induced cardiogenic shock is a life-threatening condition characterized by severe hypotension and ineffective tissue perfusion. Many drugs can lead to cardiogenic shock in overdose, for example beta blockers or calcium channel blockers. Given the poor prognosis of these cases AND theoretically-sound reasoning, vasoactive agents make sense as a therapeutic option. A detailed, comprehensive review, published in Clinical Toxicology, asks the question “Are vasopressors useful in toxin-induced cardiogenic shock?”

What They Did

The authors’ search identified 130 human case reports and 14 animal studies that met their inclusion criteria.

What They Found


Surprisingly (at least to me), human case report data showed vasopressors didn’t work more often than they did work. The largest case series (which was not included in the review article cases), of 48 diltiazem and verapamil overdose patients treated at one center over 25 years, reported good outcomes in almost all patients despite very high vasopressor doses in some cases.3 There was no comparison group in that case series and invasive monitoring was not reported, though no ischemic complications were noted.


In the majority of animal studies, vasopressor treatment failed to improve hemodynamic function and resulted in decreased survival.

Application to Clinical Practice

  1. Don’t throw out vasopressors as an option in these incredibly sick patients, but understand that other therapies, such as insulin, may be more effective.
  2. Expert consensus recommendations for management of calcium channel blocker poisoning still list norepinephrine and/or epinephrine among their first-line recommendations, though it was assigned 1D level of evidence (same as calcium and insulin).
  3. Human cases suggest that even though vasopressors are not often effective, they don’t seem to be harmful (unlike in the animal data). The largest series of 48 patients (not included in this review article), demonstrated effectiveness in most cases.

Originally posted February 13, 2017 on

Riddell J, Tran A, Bengiamin R, Hendey G, Armenian P. Ketamine as a first-line treatment for severely agitated emergency department patients. Am J Emerg Med. 2017;35(7):1000-1004. [PubMed]

Ketamine is steadily gaining traction as a treatment option for excited delirium and/or severe agitation in both the prehospital and ED settings. We published a summary of the available data back in 2015 on ALiEM. Last year in 2016, 2 prospective studies added important information to our understanding of the role of ketamine; 1 in the prehospital setting by Dr. Jon Cole’s group out of Minnesota4 and 1 in the ED from Dr. Geoffrey Isbister’s group in Australia.5 I was invited to write a commentary along with the Cole study, also published in 2016.6

Hot off the press in 2017 is another prospective study, this time from Dr. Jeff Riddell’s group in California (@Jeff__Riddell).

What They Did

  • Single-center, prospective, observational, convenience sample study examining agitation levels in acutely agitated adult ED patients after sedation with various parenteral medications (ketamine, midazolam, lorazepam, haloperidol, or a combination of lorazepam plus haloperidol)
  • Agitation levels were recorded at 0-, 5-, 10-, and 15-min after receiving sedation
  • Secondary outcomes were adverse events, repeat or rescue medication dosing, and changes in vital signs; it is important to note that secondary outcomes were collected via retrospective chart review so likely underrepresent the true incidence of adverse events

What They Found

  • 98 patients were enrolled across the 5 groups; 17 patients were needed in each group based on the sample size calculation – the haloperidol and combination groups did not meet the minimum number
  • Patients received medications via IV, IM, or intranasal (IN) routes
  • Mean doses are reported for each medication and for each route; for ketamine specifically, the mean dose was 0.87 mg/kg IV (n = 18) and 2.97 mg/kg IM (n = 6)
  • Based on agitation scores, more patients in the ketamine group were no longer agitated than the other medication groups at 5-, 10-, and 15-min after receiving medication (p = 0.001, p ≤ 0.001, p = 0.032)
  • 2 patients (8.7%) in the ketamine group were ultimately intubated; ketamine doses for these patients were not reported
  • Patients receiving ketamine had similar rates of redosing, changes in vital signs, and adverse events to the other groups

Application to Clinical Practice

  • Because of the observational nature, there was a wide variety of routes, medication choices, and doses used. So, it is difficult to pinpoint a single take-away. However, it is a ‘real-life’ study describing the practice for managing severe agitation in a busy ED. The groups were small, with 2 not reaching the minimum number needed from the power calculation.
  • It is encouraging that ketamine 1 mg/kg IV or 3 mg/kg IM seemed to work in most cases, with a much lower intubation rate than previous studies. These doses are consistent with those I suggested in the commentary, and lower than the 4-5 mg/kg IM previously studied.
  • That being said, repeat or rescue dosing of sedation meds occurred in almost 60% of patients in the ketamine group, statistically similar to the other groups
  • This is the first prospective study evaluating ketamine as a primary agent for severe agitation. A larger ED study is still needed, but it seems ketamine is continuing to demonstrate efficacy and safety for managing these challenging situations… determining the optimal dose is a crucial next step.

Originally posted February 27, 2017 on

Motov S, Mai M, Pushkar I, et al. A prospective randomized, double-dummy trial comparing IV push low dose ketamine to short infusion of low dose ketamine for treatment of  pain in the ED. Am J Emerg Med. 2017;35(8):1095-1100. [PubMed]

Back in 2015, Dr. Sergey Motov’s (@painfreeED) group published a study demonstrating the efficacy of low-dose ketamine compared to morphine for analgesia in the ED. Here’s my quick analysis of that study as a UMEM pearl. The question, though, is how best to administer the 0.3 mg/kg IV ketamine dose while minimizing the risk of adverse effects.

Fortunately, Dr. Motov’s group has just published a follow-up study addressing that exact question.

What They Did

  • Prospective, randomized, double-blind, double-dummy trial comparing safety and analgesic efficacy of IV low-dose ketamine given as a push dose (over 5 minutes) versus given as a short infusion mixed in 100 mL 0.9% sodium chloride (over 15 minutes)
  • Pain scores, vital signs, and adverse effects were recorded at baseline, 5, 15, 30, 60, 90, and 120 minutes
  • Overall rates and specific severity levels of side effects were recorded in accordance with the Side Effects Rating Scale for Dissociative Anesthetics (SERSDA)
  • Patients < 46 kg or > 115 kg were excluded.

What They Found

  • 24 patients were enrolled in each group
  • At 5 min:
    • Median severity of feeling of unreality was 3.0 on SERSDA scale for the IV push group versus 0.0 for the short infusion group (p = 0.001).
    • Median sedation on RASS scale was greater in IV push group −2.0 versus 0.0 in short infusion group (p = 0.01).
  • Decrease in mean pain scores from baseline to 15 min was similar across groups.
  • No difference between groups for changes in vital signs or need for rescue medication.

Application to Clinical Practice

If you’re using low-dose ketamine for acute pain in the ED, or developing guidelines for its use, administer it in 100 mL over 15 minutes.

Note: A second study in the Journal of Emergency Medicine utilized this same approach with positive results.7

Originally posted March 6, 2017 on

McCabe D, Lu J. The association of hemodialysis and survival in intubated salicylate-poisoned patients. Am J Emerg Med. 2017;35(6):899-903. [PubMed]

Salicylate-poisoned patients can be incredibly complex and severely ill. Secondary to the significant acid-base abnormalities that can accompany salicylate poisoning, hemodialysis (HD) is sometimes required to facilitate removal and correct acid-base status. In addition, if intubation is needed, it is critical that ventilator settings match the patient’s high pre-intubation minute ventilation.

A new study from the Illinois Poison Center evaluated the relationship between salicylate level, intubation, HD, and mortality.

What They Did

  • Retrospective, observational study of poison center records.
  • Intubated patients with a salicylate level > 50 mg/dL were included.
  • Survival was compared to measured serum salicylate level and the administration of HD.

What They Found

  • Overall survival rate was 73% in the 56 cases reported
  • In patients that did not receive HD:
    • ASA level > 50 mg/dL had a 56% survival rate (14/25)
    • ASA level > 80 mg/dL had a 0% survival rate (0/9)
  • When patients received HD:
    • ASA level > 50 mg/dL had an 83.9% survival rate (26/31)
    • ASA level > 80 mg/dL had an 83.3% survival rate (15/18)

Application to Clinical Practice

  • Even taking into account the retrospective nature and incomplete data sets from voluntary reporting to poison centers, there is an association between hemodialysis and survival in mechanically-ventilated, salicylate-poisoned patients with levels > 50 mg/dL.
  • Early hemodialysis should be arranged in salicylate-poisoned patients who require intubation (or are headed in that direction).
  • While it may seem like an obvious treatment strategy, it is clear from these records that patients are not receiving early HD when it may be indicated.

Originally posted April 21, 2017 on

Leisman D, Huang V, Zhou Q, et al. Delayed Second Dose Antibiotics for Patients Admitted From the Emergency Department With Sepsis: Prevalence, Risk Factors, and Outcomes. Crit Care Med. 2017;45(6):956-965. [PubMed]

Most studies evaluating early antibiotic administration in sepsis patients focus on timing of the first dose. We highlight many of these studies in our recent review article on Appropriate Antibiotic Therapy in Emergency Medicine Clinics of North America. But, what about the second dose? A new study in Critical Care Medicine asks that question. Specifically, what is the frequency and magnitude of delays in second dose in patients admitted from the Emergency Department AND what are the risk factors for these delays?

What They Did

  • Retrospective cohort of 828 consecutive patients at a single, tertiary, academic center admitted from the ED with sepsis or septic shock
  • Delay defined as first-to-second dose time greater than or equal to 25% of the recommended interval

What They Found

  • One-third of patients had a delay in 2nd antibiotic dose
  • Not surprisingly, antibiotics that should be given q6 or q8 hours had a higher incidence of delay
    • 72% with q6 hour antibiotics
    • 47% with q8 hour antibiotics
  • Delays were more frequent for inpatients boarding in the ED AND for patients receiving 3-hour bundle compliant initial care
  • Of note, there was an association with increased mortality in patients with delayed 2nd dose antibiotics

Application to Clinical Practice

I see this happen every day and there is no easy solution. I am not surprised by the finding that patients receiving initially compliant care subsequently having a delay in the second antibiotic dose. In the ED, we generally have 2 options: order a 1-time dose of initial antibiotics OR order scheduled antibiotics (eg, piperacillin-tazobactam 4.5 gm IV q6 hours). Herein lies the problem, we generally don’t have labs resulted when the antibiotics are initially ordered. When we order a 1-time dose, we have a higher risk of forgetting to order a second dose later. If we order scheduled antibiotics, we risk the patient having a yet undetermined degree of renal insufficiency (acute or acute-on-chronic). We rarely go back to adjust the initial antibiotic order even if the serum creatinine is markedly elevated. Both can lead to patient harm.

How do we address this issue?

This is 1 (of several) reasons I generally prefer cefepime as my gram-negative agent of choice over piperacillin-tazobactam since the normal dosing interval would be q8 or q12 hours instead of q6 (with pip-tazo). If ordering 1-time doses initially (which I prefer), built-in prompts from the electronic medical record to order subsequent doses after labs are resulted would be extremely helpful. Unfortunately, I have not yet encountered an EMR that can do this. Depending on how inpatient boarders are handled in your ED, there may be additional opportunities to create a process to avoid the delays.

In conclusion, we must be mindful that there is often delay in the second antibiotic dose and it may contribute to worse patient outcomes.

Originally posted May 19, 2017 on

Moran G, Krishnadasan A, Mower W, et al. Effect of Cephalexin Plus Trimethoprim-Sulfamethoxazole vs Cephalexin Alone on Clinical Cure of Uncomplicated Cellulitis: A Randomized Clinical Trial. JAMA. 2017;317(20):2088-2096. [PubMed]

Back in 2013, Dr. Pallin’s group in Boston published a study comparing clinical cure rates in uncomplicated cellulitis patients receiving cephalexin or cephalexin plus sulfamethoxazole-trimethoprim (SMX-TMP). I covered this study in a UMEM pearl, with the end result suggesting there was no difference in cure rate between the two treatment arms. Even in communities with high prevalence of MRSA, uncomplicated cellulitis cases without pus or abscess generally seem to be strep species. This was confirmed in the 2014 IDSA guidelines on SSTI in which they recommended streptococcal-only coverage for uncomplicated cases. A new study in JAMA reexamines this treatment strategy.

What They Did

  • Multicenter, double-blind, randomized superiority trial in 5 US EDs among outpatients older than 12 years with cellulitis and no wound, purulent drainage, or abscess (confirmed by bedside ultrasound)
  • 500 patients randomized to 2 treatment arms:
    • Cephalexin, 500 mg 4 times daily, plus SMX-TMP, 320 mg/1600 mg twice daily, for 7 days or
    • Cephalexin plus placebo for 7 days

What They Found

In the per-protocol population, clinical cure occurred in 83.5% participants in the cephalexin plus SMX-TMP group vs 85.5% in the cephalexin group (difference, −2.0%; 95% CI, −9.7% to 5.7%; P = .50).

Application to Clinical Practice

  1. It is important to note that cephalosporins, such as cephalexin, cover skin strep species pretty well but do not provide coverage against MRSA. Conversely, SMX-TMP or doxycycline provide strong MRSA coverage with questionable skin strep species coverage. So, the real question here is do we need to be covering for MRSA in uncomplicated cellulitis cases? The available data (and the IDSA guidelines) would suggest ‘NO.’ The 2013 Pallin study enrolled patients in Boston, while this new JAMA study enrolled patients from a diverse group of patients from 5 EDs. I think the conclusion is the same: even in communities with high prevalence of MRSA, uncomplicated cellulitis cases without pus or abscess generally seem to be strep species.
  2. Adherence to the study protocol favored the double-coverage group (oddly), with almost twice as many patients in the cephalexin monotherapy cohort taking <75% of antimicrobial therapy, missing follow-up visits, or other protocol deviations. There is some data supporting q12 hour dosing of cephalexin; perhaps that strategy would increase adherence if proven to be equivalent to q6 hour dosing. Or, perhaps choosing a twice-a-day cephalosporin (eg, cefuroxime) for strep coverage is another option.
  3. For a more detailed statistical analysis, Dr. Ryan Radecki provides a great summary on his EM Lit of Note blog, including a description of the intention-to-treat subgroup in which the authors speculated there may be a difference between the treatments.
  4. Linezolid is now available generically in the U.S., which has decreased its price somewhat. It covers both strep and MRSA if the patient does, in fact, need coverage for both.
  5. As a related aside, let’s at least avoid one-time vancomycin doses for SSTI, as explored in a blog post by Dr. Zlatan Coralic on ALiEM.

* ‘Strep-only’ in the title refers to skin streptococcal species without the need for MRSA coverage. Cephalosporins, penicillins, and clindamycin are all options.


Originally posted May 26, 2017 on

Sinert R, Levy P, Bernstein J, et al. Randomized Trial of Icatibant for Angiotensin-Converting Enzyme Inhibitor-Induced Upper Airway Angioedema. J Allergy Clin Immunol Pract. 2017;5(5):1402-1409.e3. [PubMed]

When it comes to Angiotensin Converting Enzyme Inhibitor (ACE-I) induced angioedema, we don’t have a lot of therapeutic options. Traditionally, patients receive the standard allergic reaction medications including corticosteroids, histamine receptor blockers, and sometimes epinephrine. But, for true ACE-I induced angioedema, these therapies do not target the underlying cause and probably treat the clinician more than the patient. In severe cases with airway involvement, we long for a treatment that can reverse impending intubation (or worse).

Enter icatibant, a bradykinin B2 receptor antagonist, that theoretically does target the pathologic process.

A 2015 New England Journal of Medicine study8 generated quite a bit of publicity when the authors concluded from their phase II trial that icatibant shortened time to resolution of ACE-I induced angioedema.

The CAMEO study group just published their findings of the subsequent phase III study.

What They Did

  • 2-armed, randomized double-blind clinical trial conducted at 31 centers in 4 countries
  • Included: Adults on ACE-Is presenting within 12 hours of the onset of at least moderately severe angioedema
  • Randomized 1:1 to icatibant 30 mg or placebo (subcutaneously)
  • Primary efficacy endpoint: time to meeting discharge criteria after study drug administration, based on the severity of airway symptoms assessed hourly by a blinded physician using clinical ratings across 4 domains

What They Found

  • 121 subjects were randomized
  • Treatment administered a median of 7.8 hours from symptom onset
  • No difference in time to meeting discharge criteria between groups
  • No difference in time to onset of symptom relief or any other secondary end point

A separate European group in 2016 similarly found no clinical efficacy with icatibant for ACE-I induced angioedema.

Application to Clinical Practice

Don’t reach for icatibant in ACE-I induced angioedema cases. It doesn’t work and is very expensive.

Related Thoughts

  • Word on the street is that this study was rejected by NEJM, perhaps in part because it was a negative trial and contradicted the 2015 findings published in the journal.
  • This study was funded by Shire, the maker of icatibant. Props to the study investigators for publishing a negative trial funded by a pharmaceutical company.
  • When patients arrive in the ED with suspected ACE-I induced angioedema, please remember to update their allergy profiles. It is an oft-forgotten but critically important step.

Originally posted June 10, 2017 on

Arens A, Shah K, Al-Abri S, Olson K, Kearney T. Safety and effectiveness of physostigmine: a 10-year retrospective review<sup/>. Clin Toxicol (Phila). July 2017:1-7. [PubMed]

Physostigmine gets a bad rap, in my opinion. I remember back to my PGY-1 pharmacy residency at UMass when we had a teenage female present with AMS after being found in the woods (published in Clinical Toxicology).9 She was clearly anticholinergic and the suspected medication, by history, was olanzapine. Physostigmine transformed a delirious patient into 1 with normal mentation telling us exactly what happened. It was like watching pharmacology in action.

Use in TCA Overdose

Physostigmine used to be part of some ‘coma cocktail(s),’ until 2 patients developed asystole in the setting of TCA overdose.10 There was also asystole reported in a 15-year-old11 and seizures reported in other cases.12 We understand that the pathophysiology of TCA poisoning is complex, and certainly much more than just anticholinergic effects are contributing. Unfortunately, we somewhat overreacted and stopped using physostigmine regularly even when true anticholinergic poisoning was staring at us. The safety of physostigmine use for seizures or cardiotoxicity in the setting of TCA toxicity is difficult to predict and thus not recommended.13

Use in Anticholinergic Poisoning

Clearly beneficial: Physostigmine controlled agitation and reversed delirium in 96% and 87% of patients, respectively.14 Benzodiazepines controlled agitation in only 24% of patients and were ineffective in reversing delirium.

Physostigmine has also been associated with less ICU admissions.15 In 45 patients receiving the antidote, most were discharged directly from the ED, only 31% required multiple doses, and none required the second dose more than 6.5 hours after the first.

A New 2017 Study

The California Poison Control System published a 10-year retrospective cohort study of 191 hospitalized patients who received physostigmine to reverse an anticholinergic toxidrome.

What They Found

Exposure (n) Response to Physostigmine
Non-diphenhydramine antihistamines (14), antipsychotics (4), TCAs (3) 21/21 (100%)
Anticholinergic Plants (67) 46/67 (68.7%)
Diphenhydramine (56) 43/56 (64.2%)
Combination products (10) 8/10 (80%)
Other/unknown (37) 22/37 (59.4%)
  • 142 patients (74.3%) were treated with physostigmine alone
  • 36 patients (18.8%) were discharged directly from the ED (57.6% of patients were admitted to the ICU)
  • 182 patients (95.3%) had no documented adverse effects
    • 4 patients experienced emesis
    • 2 patients experienced QTc prolongation
    • 2 patients experienced seizures
    • 1 patient died

Application to Clinical Practice

  • Physostigmine appeared effective in the majority of cases.
  • Most patients still were admitted to the ICU (contradicting the findings of the 2010 Rosenbaum study).15
  • Adverse effects were not reported in most cases, though adverse effects are under-reported in poison center data and should not be used, alone, to determine the safety of an intervention.
  • The one fatality (massive diphenhydramine – 950 mg) occurred 6 hours after physostigmine administration and was likely unrelated.

Take Home Points

  1. We should stop fearing physostigmine and use it in cases of anticholinergic poisoning (as long as TCAs are not the cause and there is no intraventricular conduction defects or AV block on ECG)
  2. Stock physostigmine in your ED to ensure timely administration

Further Reading

For further reading, please check out the handout from my 2017 American Academy of Emergency Medicine Scientific Assembly talk on Rational Use of Physostigmine and Flumazenil in the ED.

Originally posted July 21, 2017 on

Levine M, Beuhler M, Pizon A, et al. Assessing Bleeding Risk in Patients With Intentional Overdoses of Novel Antiplatelet and Anticoagulant Medications. Ann Emerg Med. October 2017. [PubMed]

What is the risk of bleeding after an acute (or acute-on-chronic) overdose of the newer oral antiplatelet and anticoagulant agents? A new study in the Annals of Emergency Medicine set out to answer this question.

What They Did

  • Retrospective study of poison control center calls involving patients with intentional ingestions of novel antiplatelet or anticoagulant medications (acute or acute-on-chronic)
  • Data pooled from 7 poison control centers in 4 states
  • Primary outcome measurement: presence of bleeding during the index call or during the follow-up period (bleeding was defined in the methods)
  • The narrative sections of each chart were reviewed in addition to the electronically-collected, dropdown-style data points (this adds important information not always available from the poison center database review alone)

What They Found

  • 322 cases over 10 years
    • 260 (80.7%) clopidogrel
    • 40 (12.4%) rivaroxaban
    • 9 (2.8%) dabigatran
    • 5 (1.6%) prasugrel
    • 3 (0.93%) ticagrelor
    • 2 (0.62%) ticlopidine – no longer on the U.S. market
  • Bleeding developed in 16 (4.9%) cases (7 clopidogrel, 6 rivaroxaban, 2 dabigatran, 1 apixaban)
    • Major bleeding: 9 cases, minor: 3 cases, trivial: 4 cases

Some Thoughts

  • This article addresses an important question that poison center records may help us answer, in part. The authors do a nice job of explaining some of the limitations of the study. Dr. Robert Hoffman (@bobhoffmd) wrote a piece back in 2007 that further helps interpret these data.16
  • Because of voluntary reporting, 322 cases likely underrepresents the total number of overdoses during the study period. The denominator is unknown, which makes the true incidence of bleeding difficult to estimate.
  • The authors rightly point out that though the study period covered 10 years, it ended December 31, 2014. In the 2 years following, it is likely that the newer anticoagulants were prescribed (and overdosed on) even more, which may have added important insight toward the study conclusions.
  • Data points that could add important insight:
    • How many patients were on concurrent aspirin therapy?
    • What co-morbidities existed that increased risk of bleeding?
    • How many patients who overdosed were therapeutically on the other agent (eg, overdosed on rivaroxaban but are therapeutically on clopidogrel)?
    • What was the cause of bleeding and was it unrelated to the overdose?
  • Without some laboratory confirmation that an overdose took place, we may be incorrectly estimating the prevalence of bleeding after an overdose.
  • Follow-up calls were made at least 4 hours after the initial call. This probably increased the capture rate for the outcomes being assessed (eg, bleeding). However, the bleeding risk may be high for up to 24 hours (or even longer) for some of these agents. Rivaroxaban, for example, is known to have a prolonged absorption time.
  • How many patients received charcoal? There is emerging evidence that charcoal may be effective in reducing rivaroxaban absorption even if given 8 hours after a dose.17

Application to Clinical Practice

It is certainly helpful to have an estimate of the potential risk of bleeding (~5%) after an antiplatelet or novel oral anticoagulant overdose. Even with this data, it is difficult to plan patient dispositions without knowing the denominator of all overdoses and without laboratory confirmation demonstrating at least an estimated supratherapeutic level at some point within 24 hours of the ingestion.

Originally posted October 21, 2017 on

Furyk J, Levas D, Close B, et al. Intravenous versus oral paracetamol for acute pain in adults in the emergency department setting: a prospective, double-blind, double-dummy, randomised controlled trial. Emerg Med J. December 2017. [PubMed]

In the U.S., intravenous acetaminophen (IV APAP) has made its way into some EDs as part of a multi-modal pain management approach, in part due to a perception of more rapid and effective pain control compared to the oral/rectal routes. In an effort to reduce opioid use, IV APAP seemingly has the potential to increase the number of non-opioid options such as ketorolac, lidocaine, and ketamine. Indeed, in some post-surgery literature, IV APAP has demonstrated a reduction in opioid use. The utility of IV APAP (compared to PO/PR) in the ED is unclear as there is a paucity of head-to-head efficacy studies comparing IV vs PO therapy… until now.

What They Did

  • Prospective, single-center (ED and short stay unit in Australian academic hospital), randomized, double-blind, double-dummy trial comparing safety and analgesic efficacy of adjunct IV APAP 1 gm versus PO APAP 1 gm in moderate to severe pain
  • Primary outcome: Pain scores via the Visual Analogue Scale (VAS) at 30-minutes after administration
  • Secondary outcomes: adverse events, repeat or rescue opioid medication dosing, patient satisfaction, and length of stay
  • Exclusions: < 18 years old, unable to tolerate PO administration, chronic pain, and/or history of opiate dependence

What They Found

  • 87 patients included: 47 IV group, 40 PO group (44 per group needed for power)
  • Majority of pain was abdominal/GI or musculoskeletal
  • Decrease in mean pain scores from baseline to 30 min was similar across groups
  • Repeat or rescue dosing of opioids occurred in almost 86.2% of patients in the IV group, statistically similar to PO group
  • No difference in efficacy of analgesia across 4-hour study period, length of stay, patient satisfaction, need for rescue analgesia, or side effects

Application to Clinical Practice

  • All patients were given an unknown amount of IV opioid prior to VAS assessment, and pain scores upon presentation to the ED prior to opioid administration were not reported, lessening the ability to compare the two populations. This is the biggest study limitation, in our opinion. If we want to use an opioid-alternative, then use an opioid-alternative. Don’t start treatment with an opioid and then randomize patients to IV or PO APAP. We also don’t know what dose of opioids were used prior to APAP.
  • A note on costs: at our institution, the acquisition cost for APAP 1 gm PO is about $0.12, compared to $30 for the same dose IV (~250 times more expensive). Patients may be charged much more than that. Because of the cost, many institutions have removed IV APAP from formulary or placed restrictions on its use. In the U.S. generic IV APAP is not available.
  • The PO group did not reach the pre-specified number needed for adequate power. However, if extrapolated, the increased cost would not be offset by the benefit in pain control or less opioid use.

Take Home Point

  • For the most common presentations of pain in the ED, IV APAP was not more effective than PO APAP, with the caveat that patients received IV opioids prior to APAP.

Further Reading

Dr. Ryan Radecki shared his critique of the study on his EM Lit of Note blog.

Originally posted October 21, 2017 on

Activated charcoal is a funny intervention. There is plenty of data, with various drugs, demonstrating decreased absorption after its administration (much of it in healthy volunteers). However, some argue there is a lack of quality evidence demonstrating benefit in clinically meaningful outcomes, such as mortality. In fact, the most recent Position Paper on Single-Dose Activated Charcoal (from 2005) suggests charcoal “should not be administered routinely in the management of poisoned patients.” This may be an oversimplification and I think Drs. Lotte Hoegberg and Anne-Bolette Gude astutely address this point in Goldfrank’s Toxicologic Emergencies:

“These opinions are unfortunately biased by the fact that most ‘routinely’ poisoned patients have low-risk exposures and do well with minimal intervention. Despite little scientific basis or support from clinical trials, less severely poisoned patients might benefit from activated charcoal in terms of reduced need for life support, monitoring, and antidotes.” (Isbister 2011)

And, there is data demonstrating a lower risk of seizures (venlafaxine, Kumar 2011), QT prolongation (citalopram, Friberg 2006), and delirium (promethazine, Page 2009) after charcoal administration.

Conventional teaching is that the most benefit will be gained if charcoal is administered within the first hour after an overdose. This is important because we know that if charcoal is not administered in the prehospital setting, patients arriving to the ED are unlikely to receive it within that time frame.

2 recent articles may change our thinking in when and how to use charcoal.

Study 1Chiew et al. 2017

Chiew A, Isbister G, Kirby K, Page C, Chan B, Buckley N. Massive paracetamol overdose: an observational study of the effect of activated charcoal and increased acetylcysteine dose (ATOM-2). Clin Toxicol (Phila). 2017;55(10):1055-1065. [PubMed]

In massive acetaminophen overdose, charcoal reduced hepatotoxicity

What They Did

  • 200 patients with massive acetaminophen overdose (≥40 gm)
  • Outcomes
    • Paracetamol (acetaminophen) ratio [defined as the ratio of the first paracetamol concentration taken 4-16 h post-ingestion to the standard (150 mg/L at 4 h) nomogram line at that time]
    • Hepatotoxicity (ALT >1000 U/L)

What They Found

  • Patients receiving activated charcoal within 4 h of ingestion had significantly lower paracetamol ratio versus those who did not: 1.4 (n = 33, IQR: 1.1–1.6) vs. 2.2 (n = 140, IQR: 1.5–3.0) (p < .0001)
  • Patients receiving activated charcoal had lower rates of hepatotoxicity [unadjusted OR: 0.12 (95% CI: <0.001–0.91); adjusted for time to acetylcysteine OR: 0.20 (95%CI: 0.002–1.74)]
  • Charcoal was just 1 intervention evaluated in this study, with increased acetylcysteine doses being another

Bonus Article: Ollier et al. 2017

Ollier E, Hodin S, Lanoiselée J, et al. Effect of Activated Charcoal on Rivaroxaban Complex Absorption. Clin Pharmacokinet. 2017;56(7):793-801. [PubMed]

Charcoal reduced rivaroxaban absorption up to 8 hours after administration

What They Did

  • 12 healthy volunteers were given a 40 mg dose of rivaroxaban
  • Activated charcoal was given at 2, 5, or 8 h post-dose

What They Found

In this pharmacokinetic study, charcoal reduced the area under the rivaroxaban concentration-time curve by 43% at 2 h post-dose, by 31% at 5 h post-dose, and by 29% at 8 h post-dose

Application to Clinical Practice

  • Each individual case should be evaluated for the potential benefit/harm of activated charcoal
  • Just because it’s been more than 1 hour since ingestion, charcoal may still reduce absorption and provide clinical benefit
  • We now have evidence that charcoal within 4 hours actually reduced hepatotoxicity from massive acetaminophen overdose, when combined with acetylcysteine

Previous Annual Summaries

Bryan D. Hayes, PharmD, FAACT, FASHP

Bryan D. Hayes, PharmD, FAACT, FASHP

Leadership Team, ALiEM
Creator and Lead Editor, Capsules series, ALiEMU
Attending Pharmacist, EM and Toxicology, MGH
Assistant Professor of EM, Harvard Medical School