Enough with the ED Rituals


Why are we still engaging in unnecessary treatments that have no evidence to back them up? 

Every day in EDs throughout the country, routine practices are taking place that have no proven patient benefit, or, even worse, have the potential to harm patients. These practices have been discussed in the literature and despite this, there seems to be little headway in reducing their use. Here are a couple of examples to consider.

The Ubiquitous Use of Oxygen
Have you ever seen a patient come into the ED via ambulance who was not on oxygen? They are reportable cases it would seem. But here’s something to consider. Oxygen, when used in the medical setting, is considered a drug. We often take the position with oxygen that “it can’t hurt,” so why not give it?


From a physiologic point-of-view we should remember that 95% of oxygen delivery to the cells is via the transfer of oxygen from hemoglobin. Giving more oxygen to attain supranormal levels is not helpful to the tissues when oxygen saturations are normal (96-98%), but supranormal levels of dissolved oxygen in the plasma may be harmful in a variety of circumstances.

Here’s an abstract that presents an overview of the literature. It points out the potential problems of supranormal oxygenation in a variety of conditions including myocardial infarction, CHF, COPD, stroke and post-resuscitative care of cardiac arrest. What is omitted is the use of hyperoxemia in infant resuscitation (see the abstract below), which has been condemned in all recent guidelines on infant resuscitation.

Cornet, A.D., et al, Crit Care 17(2):313, April 18, 2013
These Dutch authors comment on possible harm from routine use of supplemental oxygen in patients with a medical emergency. Several early studies reported high-flow oxygen in myocardial infarction patients was associated with reduced cardiac output and stroke volume with an increase in systemic vascular resistance and arterial blood pressure, largely due to vasoconstriction, with no evidence of a benefit of hyperoxia on myocardial ischemia. Clinical trials of supplemental oxygen in patients with cardiac emergencies are limited, but a 2010 Cochrane review reported increased mortality in MI patients receiving supplemental oxygen (RR 3.0). Experimental evidence also suggests an adverse hemodynamic effect of oxygen in patients with congestive heart failure. The adverse effects of supplemental oxygen are more widely appreciated in patients with chronic obstructive pulmonary disease (COPD), and a recent randomized trial noted decreased mortality in COPD patients receiving titrated rather than high-concentration oxygen. Hyperoxia in stroke has been noted to be associated with a decrease in cerebral blood flow and increased mortality, and use of supplemental oxygen for most ischemic stroke patients is not supported in American Stroke Association guidelines. Several large studies reported increased mortality and poor neurologic outcomes in hyperoxic patients after resuscitation from cardiac arrest. Although it is possible that vasoconstriction due to hyperoxia might be beneficial in patients with shock, no studies have demonstrated a benefit of supranormal oxygen levels in this setting. The authors advise caution with the routine use of supplemental oxygen for medical emergencies. While hypoxemia should be treated promptly, they recommend slow stepwise titration and feel that an oxygen saturation of 90-94% may be reasonable. 61 references (cornet@vumc.nl – no reprints)
Copyright 2014 by Emergency Medical Abstracts – All Rights Reserved 5/14 – #25


Biban, P., et al, Early Hum Dev 87(Suppl 1):S9, March 2011
These European authors summarized the most recent changes in neonatal resuscitation recommendations. At birth about 10% of newborns need some type of resuscitative maneuvers. The 2010 guidelines from the International Liaison Committee on Resuscitation (ILCOR) advise simultaneous measurement of pulse and respiration (along with pulse oximetry) once it is determined that a newborn is compromised. Resuscitation of term infants should begin with air, and a higher concentration of oxygen should be given only if the heart rate remains low despite effective ventilation, or if oxygenation (as guided by pulse oximetry) remains unacceptable. ILCOR recommends blended oxygen and air for preterm babies less than 32 weeks, but both hyperoxemia and hypoxemia are to be avoided. If an infant is apneic or gasping, or if the heart rate is less than 100/minute (bpm) after 30 seconds (sec) of initial resuscitation, positive pressure ventilation should be instituted. Suctioning of the oropharynx and nasopharynx before shoulder delivery in births with meconium-stained amniotic fluid is no longer recommended. Confirmation of endotracheal intubation should be performed with colorimetric exhaled CO2 detectors. Chest compressions are indicated in the newly-born infant with a heart rate below 60 bpm despite adequate ventilation for 30 sec. Adrenaline is considered only if an infant’s heart rate is below 60 bpm after adequate assisted ventilation for 30 sec and chest compressions for 30 sec. Therapeutic hypothermia is recommended for infants with suspected asphyxia. Physicians may consider not initiating resuscitation in situations associated with almost certain infant death and/or unacceptable morbidity; parents should be included in this decision. If there are no signs of life after 10 minutes of resuscitation, discontinuation is justified. 11 references 
Copyright 2011 by Emergency Medical Abstracts – All Rights Reserved 10/11 – #36
(paolo.biban@ospedaleuniverona.it – no reprints)

Saugstad, O.D., J Perinatol 30:S46, October 2010
The use of oxygen for newborn resuscitation was first proposed more than 200 years ago, and has been continued with little consideration of supporting evidence. This Norwegian author comments on the use of room air for newborn resuscitation. Animal studies provided the first evidence of the potential harm of using 100% oxygen for resuscitation. A multinational study involving more than 600 infants (the “Resuscitation of Newborn Infants with Room Air or Oxygen” or Resair 2 study), published in 1998, found that infants resuscitated with room air rather than 100% oxygen tended to have a lower neonatal mortality rate (odds ratio [OR] 0.69) as well as significantly higher Apgar scores at five minutes, and a subsequent follow-up study reported no neurodevelopmental differences in survivors after 18-24 months. Later experimental and clinical studies noted harms associated with resuscitation of newborns using 100% oxygen. Four meta-analyses and systematic reviews have concluded that newborn resuscitation with 21% rather than 100% oxygen significantly reduces newborn mortality, and newborn resuscitation protocols in some countries now specify use of a low-oxygen strategy. Updates of guidelines for newborn resuscitation have conservatively reflected this change in thinking, although the 2005 ILCOR, European and Cochrane Database guidelines proposed three different sets of recommendations. The authors suggest that, although more data are needed, the available evidence supports the use of 21% oxygen for the resuscitation of term or near-term infants. Emerging data suggest that a low-oxygen approach is also preferable for low birthweight infants and that an FiO2 of 0.3 appears to be sufficient for preterm infants (less than 35 weeks), with adjustment according to the response as reflected by the oxygen saturation and heart rate. 50 references (o.d.saugstad@medisin.uio.no – no reprints)
Copyright 2011 by Emergency Medical Abstracts – All Rights Reserved 3/11 – #24

Everybody Gets IV Fluids
When you can’t think of anything better to do for a patient, start an IV. Nobody will criticize you, it has become knee jerk for all manner of complaints – and in the rare case the patient actually needs IV fluids, it will be in place.
Unfortunately, I can’t locate the reference, but an old study did compare the use of IVs in England vs. the U.S. Turns out, no surprise, that we start a ton more IVs than our British colleagues.

And, as I’ve advocated many times in the past, clinicians should be aware of patient charges and see what $0.44 to $1.00 worth of salt water becomes on the bill. In a detail investigative report on charges for IV fluids by Nina Bernstein in the August 25th, 2013 issue of the New York Times, it was shown how IV fluids can result in big-time charges. She recounts multiple stories of patients’ bills for IV fluids after an epidemic of food poisoning in upstate New York (a place I intuit ED charges to be on the very low end). One sick anesthesiologist was billed $546 for six liters of saline. To my way of thinking, it was a bargain compared to prices I am used to here in metropolitan Los Angeles. And remember, this doesn’t cover the charge for administration of IV fluids—a totally different charge.


But back to the matter at hand, Do IVs harm patients? Not usually (except for the pain when it is started and having to be tethered to a hose and IV stand), but they often times don’t help when we think they should. For instance, aggressive IV fluids in patients with kidney stones don’t facilitate stone passage as noted in the abstract below.

Springhart, W.P., et al, J Endourol 20(10):713, October 2006
BACKGROUND: Aggressive hydration has been the traditional approach to the management of acute renal colic, but some recent studies have suggested that a minimal fluid strategy might be beneficial. Several studies found that use of desmopressin to decrease urine production prevents distention of the obstructed kidney and decreases pain.
METHODS: In this study, from Duke University, 43 adults (mean age, 41) with a clinical diagnosis of acute renal colic and CT confirmation of renal or ureteral calculi were treated with IV ketorolac and then randomized to “forced” IV hydration (500ml/hr for four hours) or minimal hydration (20ml/hour).
RESULTS: The mean stone diameter was 3.0mm and 3.4mm in the minimal hydration and forced hydration groups, respectively. In the total patient cohort, there were no significant differences between the two groups in median pain scores at baseline or during four hours of follow-up, change in pain scores, analgesic requirements, or the likelihood of spontaneous stone passage on follow-up (information available for 32 patients) (spontaneous stone passage rate, 22% in the minimal hydration group and 30% in the forced hydration group). In the subgroup of 30 patients with CT evidence of a stone in addition to evidence of secondary obstruction, there were no significant differences between the two groups in measures of pain or analgesic requirements, and the rate of spontaneous stone passage was 31% in the minimal hydration group and 29% in the forced hydration group.
CONCLUSIONS: Findings in this small study suggest that there is no advantage to aggressive IV hydration in patients with acute renal colic. 12 references (glenn.preminger@duke.edu)
Copyright 2007 by Emergency Medical Abstracts – All Rights Reserved 3/07 – #18

The American Academy of Pediatrics has been on emergency clinicians for having an inappropriately low threshold for starting IVs in mildly to moderately dehydrated children (the vast majority of the dehydrated kids we see) instead of using the oral “sippy” treatment (which has only a 5% failure rate).

Hom, J., et al, Ann Emerg Med 54(1):117, July 2009
METHODS: The authors, from the State University of New York in Brooklyn, reviewed 17 randomized or quasi-randomized controlled trials comparing oral and IV rehydration in 1,811 children (below the age of 18) with dehydration due to acute gastroenteritis for this paper from the “Evidence-Based Emergency Medicine” series.
RESULTS: In the inpatient setting, rates of failure to rehydrate or maintain hydration were 1.3% with IV rehydration vs. 4.9% with oral rehydration (number-needed-to-treat [NNT] with IV rather than oral rehydration to prevent one treatment failure, 36). The mean length of stay was 1.2 days less in children managed with oral rather than IV rehydration. There was no difference between the two strategies for other secondary outcomes, including weight gain at discharge, hypo- or hypernatremia, duration of dehydration, total fluid intake at 6 and 24 hours, or sodium intake. Paralytic ileus was more common in children managed with oral rehydration (risk difference 3%, NNT with IV rehydration to prevent one case 33), while phlebitis was more common in those managed with IV rehydration (risk difference 2%).
CONCLUSIONS: Given the advantages of oral rehydration for the management of mild to moderate dehydration in children with gastroenteritis (e.g., lower cost and ease of use), the authors feel it should be the method of first choice, even though an occasional patient will ultimately need to be switched to IV rehydration. 3 references
Copyright 2010 by Emergency Medical Abstracts – All Rights Reserved 1/10 – #20

Jablonski, S., J Emerg Nurs 38(2):185, March 2012
IV rehydration is often the initial strategy used by providers for children with mild to moderate dehydration to gastroenteritis. This practice causes anxiety and discomfort in the child and might be psychologically harmful as well. This author, from Somerset Medical Center in Somerville, NJ, reviews the use of oral rehydration for mild to moderate dehydration in pediatric gastroenteritis. This practice has been shown to be as effective as IV rehydration, and is recommended by both the American Academy of Pediatrics and the World Health Organization. Because routine laboratory testing is not required for the child with mild to moderate dehydration, venipuncture is not necessary. Involvement of caretakers in the oral rehydration process reduces healthcare worker time requirements and educates caretakers in the potential for initiation of oral rehydration at home. It is recommended that oral rehydration solutions contain 75mEq/L of sodium and 75mmol/L of glucose, with a total osmolarity of 245mOsm/L. Parents should be advised to avoid using sports drinks, which contain inappropriate proportions of electrolytes. Correction of mild or moderate dehydration involves administration of 50ml/kg or 100ml/kg, respectively, as well as replacement of continuing losses, over a four-hour period. For administration of 100-150ml per hour, the author suggests starting with 5ml every one to two minutes over the initial ten minutes, followed by 10ml every five minutes over the next 20 minutes, and then 20ml every 10 to 15 minutes over the subsequent 30 minutes. Even children who have been vomiting can be rehydrated orally, but caregivers should wait 15 to 30 minutes after a vomiting episode before starting oral rehydration. Children can be started on an age-appropriate diet when rehydration is completed. 7 references stevejablonski@comcast.net – no reprints)
Copyright 2012 by Emergency Medical Abstracts – All Rights Reserved 8/12 – #36

In some circumstances, IVs can potentially harm patients. A study by Behnood Bikdeli, et al, entitled Intravenous Fluids in Acute Decompensated Heart Failure in the February 2, 2015 issue of the Journal of the American College of Cardiology: Heart Failure noted that in a survey of 131,430 hospitalizations for heart failure from 2009-2010 in the U.S., treated at 346 hospitals, 11% were treated with IV fluids for the first two days with a median volume of 1L (80% received normal saline). Patients treated with fluids had higher rates of subsequent critical care admissions (5.7%vs 3.8%), intubation (1.4% vs 1%), renal replacement therapy (0.6% vs 0.3%) and hospital death (3.3% vs 1.8%). As anticipated, hospital variation in the use of IV fluids was huge (from 0% to 71%, with a median of 12.5%). So in lots of pulmonary edema patient we’re putting fluids in one end and taking them out another – pretty nutty.

These two examples, unnecessary and potentially harmful oxygen use and use of IV fluids barely scratch the surface of routine practices that don’t benefit patients and may harm them. There are lots more.


Dr. Bukata is the Editor of Emergency Medical Abstracts.


  1. Keith Raymond, MD on

    As my schizophrenic colleague from Fight Club reported, the reason they give you oxygen on a crashing jet is: Oxygen makes you High. It is a distraction from the burning fuel and chaos of a busy EC. As for IV therapy, it is likely more about allowing Nurses to practice a skill that can easily slip away. Tongue and cheek aside, I still try oral rehydration, but its the parents that demand the IV. Both treatments are a customer satisfaction issue, and we must have those numbers for our evaluations. All sad, all true.

  2. Thank you,
    We need reminding of the “sometimes less is better” and the “power of observation and time”
    How do we resolve the issue of teaching “management” who wants the revenue of “procedures” and RVU boosting not to mention the pressure of “metrics” for shorter treatment times. ED MD’s “production” and job is controlled by it not “what is best” for patient in long run. Factory efficiency versus Practice of Medicine.

  3. Steve Cripe RN on

    Contrary to your assumption regarding starting IVs on every patient, the EDs I worked at even back as far as the 1980s we did not routinely do this. What we did do more times than not, is start an IV and just made it into a heparin lock. That way if the patient crashed in thhe ED or even upstairs later, they had a port to plug in intravenous fluids or push meds.

    • I agree with the saline lock comment. Having access when you need it and least expect to need it can be invaluable. So starting IV’s on MOST patients is a pretty common practice here. In my ED anyways, we don’t bolus patients Just for the patient satisfaction numbers. However, they do suffer because of this I’m sure. Without procedures patients tend to feel like they’re being ripped off, which isn’t the case. Also IV vs. oral intake becomes a problem in the emergency setting due to the NPO status requirements for procedures. That’s a topic for another article I’m sure. Also there’s the whole TKO practice.

  4. Andrew Loughlin on

    Your article would be better to state no oxygen in newborn resuscitation; resuscitation of all other infants will normally be carried out with oxygen.

Leave A Reply