Wait. What's that you say? You need an electronic order for that? OK, wait a second, let me find a computer … logging on … could I just get started without an official order? Not possible, eh? OK, here we go. It's in!
That's a typical communication in the ED. Latinphiles may recognize that the word “communication” has its origins in the Latin “communicare,” meaning to impart, share, or make common. A major component of ED communication is sharing and imparting treatment orders, and as we all know, we can do this in multiple ways — by pen, keystroke, and larynx. It is the latter on which I'd like to focus.
The spoken order, often sterilely called the “verbal order,” or VO, has become somewhat of an anathema at some U.S. hospitals. A recent systematic review described broad differences in verbal order policies across 40 acute care hospitals with many restricting the use of verbal orders and in some instances even simple face-to-face orders such as requesting a single dose of acetaminophen. (Jt Comm J Qual Patient Saf 2012;38:24.) As best as I can tell, the rationale for such restrictive policies is threefold: the availability of emerging technology (computerized order entry), a litany of anecdotal evidence on patient safety risk, and a strong suggestion from the Joint Commission as part of their national patient safety goals. What I'd like to know is, where is the evidence?
Wait, what's that? You need another order. Why? Because I am changing the nature of this discussion? Really? OK, let me log in. Here we go. Just a moment to jump into intergalactic hyperspace. Typing it in now. Order up!
Where was I? As it turns out, the evidence for restrictive verbal order policies is almost purely anecdotal. I'm certain you've heard some variation of this anecdote before. It proceeds along the lines of the broken telephone game where kids whisper a phrase to each other around a circle and see what comes out the other side. “I took my dog for a walk today, and then I gave him some food” morphs into “I took Michael for a walk today, and then I shaved him something good.” Or like an iPhone autocorrect embarrassment, you text about your “neighbor's child prodigy,” but your iPhone turns it into a text about your “neighbor's child prostitute.” Broken communication anecdotes in the healthcare arena go like this (examples courtesy of the Institute of Safe Medication Practices):
“An emergency room physician verbally ordered ‘morphine 2 mg IV,’ but the nurse heard ‘morphine 10 mg IV,’ and the patient received a 10 mg injection and developed respiratory arrest.”
“[A] physician called in an order for 15 mg of hydralazine to be given IV every TWO hours. The nurse, thinking that he had said ‘50 mg,’ administered an overdose to the patient who developed tachycardia and had a significant drop in blood pressure.” (Dynamics 2006;17:20.)
Anecdotes can be powerful (as can Narcan), and we all have experienced moments of misdirected communication in the workplace. The danger of the spoken order seems to make intuitive sense. But is there actual evidence in the literature supporting a patient safety benefit for restrictive verbal order policies?
Huh? Another order? Are you sure? This should definitely be covered by the previous order. Yes, OK, I know you want to protect patient safety, and your license, too. Yes, I know management is tracking this closely. No, you don't want to take a verbal on it? OK, then, but I have to admit that these interruptions are distracting. And I've been reading a lot about the danger of interruptions in the ED. A significant source of error, you know. Sorry, you don't want to hear about this without a separate order, so we'll drop it. Logging back in. Waiting. It's in.
Now, here we go, the actual evidence. It's quite sparse. A 2009 review on the topic, by Wakefield and Wakefield, summarizes it nicely: “The literature consists primarily of nonsystematic and anecdotal evidence of the relationship between verbal order utilization and actual or potential patient harm. To our knowledge, the only large-scale study of hospital verbal order policies is a 1990 report of a survey of nursing and pharmacy leaders' self-report of selected features of their hospitals' verbal order policies.” (Qual Saf Health Care 2009;18:165.)
The only study specifically looking at errors associated with verbal orders compared with handwritten or computer entry actually demonstrated a counterintuitive decrease in errors with verbal orders compared with other types. (Qual Saf Health Care 2009;18:165; Arch Pediatr Adolesc Med 1994;148:1322.)
West et al reviewed all nonstanding inpatient pharmacy orders over a three-month period at a single children's hospital. Orders were classified as verbal, handwritten, or computer-entered. All orders were entered into the computerized system (either directly or by transcription), and error rates were determined from a pre-existing incident reporting system. Nearly a quarter (4,221 of 18,262 total orders) were verbal, and these carried a nearly fourfold decrease in error risk (2.6 errors per 1000 orders) compared with handwritten orders (8.5 per 1000), which had high rates of transcription error. Computer-entry order errors (6.3 out of 1000) were most likely to be dosage-based. Sensitivity analyses suggested that this reduction in errors diminished with increasing order complexity. But for common prescriptions, like intravenous fluid, beta-agonist inhalers, and acetaminophen, error rates were low or nonexistent (no mistakes of 613 beta-agonist orders) across all order types. Now granted, this study is two decades old, and was conducted in the early days of computerized order entry, but their results nonetheless make one wonder why restrictive verbal order policies should exist for common and simple medications.
Is it possible that a restrictive policy on verbal orders has intuitive and anecdotal appeal, and administrators have overlooked the potential unintended consequences? Every action has an equal and opposite reaction, as Newton would say. And, regarding computerized order entry, as the Romans once said, caveat emptor (let the buyer beware).
What now? You didn't get the prior order? Are you quite sure? Let's look. OK, logging in. Uh-oh, did I enter it on the wrong topic? Oops, not again. Oh, man, I am sure that's really mucked up the one about importance of clarity of thinking in an ED. Here we go again. Bear with me. Man, that's dumb. Funny how it's so easy to make a simple mistake like that. I suppose entering the right order on the wrong patient doesn't really advance patient safety, now does it? If only I weren't so human!
OK, where was I? Assumptions. Right. Take the electronic health record. Many people assume that it is superior by its very nature. I think most of us would agree that ultimately it will be, but this is not a smooth zero to 60. Some significant decelerations are to be expected. Consider the existing literature on how the EHR affects care.
Yes, yes, I know. New topic, new order. No leeway given for logical stepwise progression of thought or for just getting quick and clear approval to do the right thing and documenting it later. Logging in. Loving life in intergalactic hyperspace.
The past decade has seen a paradigm shift in the form and function of the medical record, but much work remains to optimize its ability to provide enhanced care. Handel et al have noted that to date the electronic health record has largely been designed with back-end processes such as billing in mind rather than frontline processes like order entry. (Acad Emerg Med 2011;18:e45.) We notice this as fewer electronic safeguards for the potentially calamitous “wrong patient entry” than for capturing the required review of systems for billing a level four visit. Not surprisingly then, despite the common assumption that EHR implementation has improved clinical care, this has not been consistently proven, and this fact has been emphasized in recent editorials and reviews. (Acad Emerg Med 2011;18:e45; Health Aff (Millwood) 2011;30:464; Mayo Clin Proc 2011;86:373; J Am Med Inform Assoc 2011;19:460.) In fact, poorly designed computerized order entry systems have been associated with increased mortality, and some authors have asserted that EHRs need to be studied in a robust, randomized clinical trial fashion. (Pediatrics 2005;116:1506; J Am Med Inform Assoc 2011;18:173.) The assumption that computerized order entry is universally safer than verbal orders actually has no basis in available evidence, and may actually put patients at risk in myriad ways.
What? You need me in room 2? A patient is in status epilepticus? OK, someone get the Ativan. One milligram Ativan intravenous. What? You need a computerized order for that? You don't remember how to get the medication out of Pyxis without one? Something's not right here.
If you haven't guessed by now, I think the patient safety pendulum of verbal orders has swung out of control, and is approaching the “ad absurdum” meter. Certain types of orders, like texted orders and complicated phone orders, don't make sense, but in the ED we need a safe and efficient means of utilizing verbal orders, especially for simple medications with standardized dosing. And we need staff comfortable utilizing them in appropriate situations using proven safety checks like the read-back. Consider the evidence. A recent abstract presented by clinicians from Cincinnati Children's Hospital Medical Center at the Pediatric Academic Societies annual meeting reported a zero error rate (of 75) after instituting an order read-back protocol for verbal orders given during team rounds (versus 9% prior to the policy). So there's hope. Maybe your ED has this verbal order procedure all compos mentis, and if so, I'd like to hear how you succeeded in figuring it out.
Of course, of course, a new request. I'll write an order for it.
Comments about this article? Write to EMN firstname.lastname@example.org.