Date: April 24th, 2015

Guest Skeptic: Meghan Groth (@EMPharmGirl). Meghan is the emergency medicine pharmacy specialist at the University of Vermont Medical Center in Burlington, Vermont and an adjunct professor of pharmacy at the Albany College of Pharmacy and Health Sciences.

Case: A 58-year-old male presents to your emergency department complaining of a warm, painful, reddened area on his left thigh. His past medical history is only significant for generalized anxiety disorder and he has no known drug allergies. On exam, you find no evidence of an abscess, and you find his labs and vital signs are within normal limits. You confidently give him a diagnosis of uncomplicated cellulitis and need to determine an antibiotic regimen. You’d like to send him home with a five day course of cephalexin, but are thinking about giving him an intravenous (IV) dose of cefazolin before he leaves.

Background: In certain infectious disease conditions, such as meningitis or septic shock, we know that rapid administration of antimicrobials in adequate doses is associated with improvement in patient outcomes.

Septic shock is a time dependent emergency. It has been demonstrated for every hour delay in providing antibiotics to patients with septic shock mortality increased by almost 8%.

However, in patients who are well enough to be discharged home, this same relationship has not been demonstrated.

We all know that antibiotic associated diarrhea (AAD) is a common side effect of antibiotic therapy. The incidence in the literature has been reported to be between 5% and 39%. One of the most concerning types of AAD is Clostridium difficile (CDI). This has been on the increase.

There is such concern about C.diff that it has been one of the patient safety quality indicators in many organizations/jurisdictions.

There have been a number of factors that increase the risk of AAD/CDI including the type of antibiotic prescribed and the duration of use. While almost all antibiotics can cause AAD and CDI the cephalosporins, broad-spectrum penicillins and clindamycin are more often the cause.

Patient factors are also thought to be involved: age greater than 65 years, co-morbidities and having a history of AAD.

Clinical Question: What is the risk of getting antibiotic associated diarrhea with an IV dose of antibiotics prior to discharging patients home with an oral course?

Reference: Haran et al. Factors influencing the development of antibiotic associated diarrhea in ED patients discharged home: risk of administering IV antibiotics. Am J Emerg Med 2014

  • Population: Adult patients from three emergency departments (2 large urban and 1 small community site).
    • Excluded patients if they had diarrhea or Clostridium difficile in the previous 4 weeks, had received an antibiotic in the previous 4 weeks, or were admitted to the hospital.
  • Intervention: IV antibiotics as part of their emergency department and discharged home with a new prescription for antibiotics.
  • Comparison: Patients who were not given IV antibiotics as part of their emergency department care and discharged home with a new prescription for antibiotics.
  • Outcome:
    • Primary outcome was the development of AAD, which the authors defined as three or more loose stools per day for at least two days.
    • Secondary outcome was the development of CDI, which the authors defined as AAD that led to a diagnosis of CDI confirmed by a positive C. difficlie toxin A assay.

Authors Conclusions: Intravenous antibiotic therapy administered to ED patients before discharge was associated with higher rates of AAD and with 2 cases of CDI. Care should be taken when deciding to use broad-spectrum IV antibiotics to treat ED patients before discharge home.”

checklist-cartoonQuality Checklist for Observational Trials:

  1. Did the study address a clearly focused issue? Yes
  2. Did the authors use an appropriate method to answer their question? Yes, if looking for association not causation. 
  3. Was the cohort recruited in an acceptable way? Yes
  4. Was the exposure accurately measured to minimize bias? Yes
  5. Was the outcome accurately measured to minimize bias? Unsure
  6. Have the authors identified all important confounding factors? No
  7. Was the follow up of subjects complete enough? Yes
  8. How precise are the results/is the estimate of risk? Moderately
  9. Do you believe the results? Yes
  10. Can the results be applied to the local population? Yes
  11. Do the results of this study fit with other available evidence? Yes

Key Results:  The authors of this study were able to complete analysis on 247 patients. Most of these patients were generally healthy with no significant past medical history.

The most common infection being treated with antibiotics was a skin/soft tissue infection. The patients who received IV or oral only antibiotics were similar at baseline with respect to emergency severity index, % tachycardia or febrile, and medical histories.

  • Primary Outcome of Antibiotic Associated Diarrhea:
    • 45/247 (18%)
    • OR 2.73 (95% CI 1.38-5.43)
    • 7% IV group vs. 12.3% Oral group
    • Absolute difference of 13.4%
    • NNH=7
  • Secondary Outcome CDI:
    • 2/247 (1%)

The rate of AAD tended to increase with the duration of antibiotic therapy. Clindamycin, vancomycin, cephalosporins, penicillins, and macrolides were associated with the highest rates of AAD; quinolones and doxycycline had the lowest rates.

Of patients who developed AAD, 27.9% stopped taking the antibiotic and 16.3% had a subsequent healthcare visit to address the diarrhea symptoms.

Screen Shot 2015-04-25 at 3.11.12 PM

In this small, observational trial the authors sought to determine risk factors for the development of AAD. The authors found that about 18% of patients in their analysis developed AAD, which fits into the range of 5-39% previously described in the literature.

The primary outcome was patient-oriented, but didn’t necessarily take into account all potential confounders. It would have been nice to see a description of whether or not patients had any history of constipation, or were taking any medications that could have affected GI motility or the development of diarrhea/constipation (e.g. any sort of bowel regimen, iron supplements, opioids, naltrexone, metoclopramide, erythromycin, etc.).

Additionally, it’s unclear if this primary outcome was accurately measured to minimize bias, because of how this data was collected. Patients were contacted in a follow up survey 4 weeks after finishing antibiotic therapy and asked about development of AAD. This is subject to not only recall bias, but may be different depending on the patient’s perception of normal bowel movements.

Emergency Severity Index is endorsed by ACEP and the Emergency Nurses Association in the US, and used widely outside of the US as well. It has been shown to have an acceptable level of reliability, but may be subject to some variability in certain patient populations (e.g. pediatric patients) and a greater degree of variability observed in levels 3 through 5.

The tool has its limitations as it relates to describing the patients in this study, but represents the most feasible and widely recognized tool available for this population.

Despite its limitations, this study provides important information about potential risks of an intervention which has yet to demonstrate any benefit in this patient population (e.g. those well enough to go home with a course of oral antibiotics for their infection).

Additionally, if patients who receive IV antibiotics are more likely to develop AAD, and 28% of those patients stop taking the antibiotics early due the side effects, this could have implications not only for that patient but also the eventual development of antimicrobial resistance.

If IV antibiotics don’t result in a treatment benefit for these patients, are potentially harmful, and have greater cost implications, then why haven’t we abandoned this practice?

What is needed is a double blinded, randomized, placebo controlled trial of patients being treated for infections in the ED and discharged home. The patients could receive IV antibiotics or saline and all be discharged home on a new oral antibiotic prescription. Then follow them for the primary outcome of AAD. Have secondary outcomes of CDI, discontinuation rate and clinical cure.

Meghan Groth

Meghan Groth

Comment on authors conclusion compared to SGEM Conclusion: We agree that there was an association between IV antibiotics administration in the ED and increased rates of antibiotic associated diarrhea. Physicians should always think about the potential harm before providing any treatment.

SGEM Bottom Line: Administration of IV antibiotics in the emergency department is not without harms. In this small observational study, it looks like IV antibiotics are associated with an increased risk of AAD.

Case Resolution:  You decide to skip the IV dose of cefazolin and discharge your patient with a five day course of cephalexin for his cellulitis. You give the standard advice to return to the emergency department if they are getting worse (fever and/or increased redness/pain), not able to tolerate the oral antibiotic or are otherwise worried.

Clinical Application: For patients well enough to go home and a working gut an IV dose of antibiotics doesn’t have any benefit over oral therapy, and may pose an increased risk of adverse effects.

What do I tell my patient? We are going to start you on antibiotics for your infection. Here is a prescription for some pills that are easily absorbed into your system and highly effective. One of the most common side effects of antibiotics is diarrhea. Giving you an intravenous dose of antibiotics before you leave could put you at a higher risk of developing diarrhea. Additionally, intravenous antibiotics for this type of infection do not result in a faster cure. Please take all your antibiotics as prescribed. This will ensure your best chance of being cured of this infection.

Keener Kontest: Last weeks winner was Tyler Gombash from Fort Lauderdale, Florida. Tyler knew Murry et al published the first study on preconditioning ischemia to protect ischemic myocardium? The paper was called Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation 1986.

Listen to the podcast for this weeks keener question. If you know the answer send an email to TheSGEM@gmail.com with “keener” in the subject line. The first correct response will win a cool skeptical prize.

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.


EMA Conferences


Screen Shot 2014-11-15 at 7.07.05 PM