The scope of IV room errors

There’s an interesting article in Pharmacy Practice News this month (In the IV Room, Robots Come to the Rescue). While the title of the article is a bit misleading – I think ‘rescue’ is a bit strong – it does contain quite a bit of good information.

The article discusses some of the technology being used at Brigham and Women’s Hospital (BWH) in Boston, and the University of California, San Francisco (UCSF) Mission Bay pharmacy. I’ve been in both pharmacies. BWH and UCSF both make extensive use of technology, but believe me when I say that they have very different approaches. Anyway, the article is worth a few minutes of your time.

Deep in the article, the author, Rajiv Leventhal spends a few paragraphs discussing the scope of the problem in the IV room, and some of the challenges of using robotics. Rajiv acknowledges that the iv room is a dangerous place for a host of reasons.

Regardless of the technology chosen, the need to automate IV compounding to at least some degree is hard to dispute, given the relatively high rate of errors that occur when technology is limited. In 1997, when many of the recent advances in robotics were not available, the error rate for IV compounding was 9%—or one mistake in every 11 medications coming out of the IV room.

As for the main cases [sic] of those errors, many factors have been identified, including sterility and other drug safety issues, according to a safety alert released last year by the Institute for Safe Medication Practices. The alert identified five core causes: 1) depreciating importance of the compounding and dispensing processes in pharmacy practice; 2) lack of knowledge and standardization around best practices; 3) training based on traditions handed down from one pharmacist to the next; 4) learned workplace tolerance of risk and routine practice deviations that persist; and 5) a reluctance to learn from the mistakes of others.

It seems intuitively obvious that the use of technologies like iv workflow management software, barcode scanning, gravimetrics, imaging, and even robotics can potentially decrease errors described in the article referenced above (Am J Health Syst Pharm1997;54[8]:904-912 ). However, of the causes identified in the second paragraph, only #2 can really be addressed with the use of technology alone. The rest of the items listed are symptoms of a deep-seeded problem growing in pharmacies today, and that is the failure to understand the need for our profession to provide patients with medications in the most efficient, safe, and economical way possible. Sounds ridiculous, I know, but it’s true nonetheless.

Most (all?) pharmacies I visit these days tout initiatives to improve patient care through increased ‘clinical activities’ of pharmacists, including electronic chart review, ADE follow-up, rounding with the medical team, monitoring and adjusting medications, antibiotic stewardship, and so on. However, I rarely, if ever hear directors talk about efforts to improve operations through streamlined processes, automation and technology, standardization, and heaven forbid, increased use of technicians and non-pharmacist personnel.

Examples of this can be found within open job listings at various healthcare systems. Recently I visited an acute care pharmacy with a large budget for several open ‘clinical pharmacist’s positions’ but no budget for improving operations or automating processes. In this particular case, a fraction of the money being allocated for open clinical pharmacist positions could be used to make significant improvements to the medication distribution process.

It’s an interesting dilemma for pharmacy directors. While spending tens of thousands of dollars on automation and technology to improve operations may not seem sexy, it goes without saying that a vast majority of care for a hospitalized patient involves getting the right drug at the right time. A majority of that falls to nursing staff, but the pharmacy owns a piece of the medication distribution/administration process. Nurses can’t administer medications if they’re not readily available, or wrong.

Regardless of what direction the profession wants to go, it is important that we understand that pharmacy is, at this time, tied to distribution. We must find ways to extricate ourselves from the medication distribution process first before we can begin to truly realize the benefits of pharmacists in patient care. Each time an error occurs for lack of focus, training, or sheer disinterest, the profession suffers. Preventable medication errors involving the pharmacy causes both the public and other healthcare practitioners to lose trust in our ability to get the job done. It’s difficult to recover from lack of trust. Think about it.

6 thoughts on “The scope of IV room errors”

  1. The law of large numbers governs here… If you can take people out of the process for the vast majority of the CSP does, you can impact the error rate introduced by human error.

    In order for “robots to fix it” (where “it” is “CSP compounding errors”) there need to be robots “you” can afford that can meet “your” needs for the timely dispensing of your overall CSP workload.

    Sure, we can watch the machinations of these devices at shows — and the machinery is in constant motion and some of those motions look pretty fast. However, if one ignores the entertaining view through the window and just watch/count the doses coming out, ready, labelled, etc. the issue becomes pretty obvious: these devices are slower than the process they replace (one competent technician manually compounding doses – MUCH slower than this same competent technician using non-robotic technologies like Repeater Pumps, etc).

    Is the right answer is to just put enough of these in place so that the total throughput meets your needs with some margin for one or more systems being down/maintained? Maybe.

    The universal desire is a system/black box that takes the people out of the equation, eliminates errors, and can easily crank out all needed CSP doses within real time constraints. If one of those is out there, I’ve never seen it/heard of it. There are many reasons why acceptance of IV Robots have been slow… it’s largely because *they* are.

    Perhaps – with unlimited funds and space and competent staff – it might be possible for “Robots to the Rescue” to truly defeat the “Law of Large numbers” with today’s IV robots.

    There are a lot of process improvements available in the IV Workflow tools that can help fight these error rates. Most/all are much less expensive and easier to deploy than IV Robots. Yet all are still working at being 20% deployed.

    When does the impact of the very real CSP compounding error rate rise to “do nothing is not an answer”?


  2. Thanks for your thoughts, Dennis. Sometimes I feel like you’re trying to argue a point *against* my point. However, I feel that you and I are on the same page. If you’ve read my site, you know that I agree with pretty much everything you wrote.

    I too believe that robots need some work. In my mind, they’re good in very specific niches, and if you have that niche you’ll be happy.

    I too believe that process improvements available through iv workflow tools can go a long way in fighting error rates.

    I too am baffled as to why the adoption rate of iv workflow tools is so low.

    I too wonder when the impact of CSP compounding errors will get the attention they deserve.

    Possibly the only place that you and I may not see eye-to-eye is with my thoughts on the need for more automation and technology in the pharmacy. I’m a true believer. Do I think automation/technology is always the right answer? Of course not. Sometimes automation isn’t the answer, at all. With that said, I do feel that available automation and technology is often overlooked and underutilized.

    Again, thanks for your thoughts. Always appreciate you taking the time.

  3. There was an innovation named VEINROM developed by bio-medical students from University of Iowa which surely could change the way intravenous drug administration errors could be obliterated

  4. Thanks for the information, Anurag, but I’m not sure that’s the solution. The VEINROM system is specific to anesthesia, and while it might seem to make sense, the need to incorporate a different injection manifold to every pump on a nursing unit would be cumbersome, costly, and labor intensive. Not to mention that the number of drugs used in a pharmacy and at the bedside is significantly larger than those used in the OR.

  5. When I graduated from the University of Michigan school of pharmacy over 45 years ago, the concept of “clinical pharmacy” was just coming into being. To better prepare us for the “new world of pharmacy”, the pharmacy school hired a clinical pharmacist out of the UCSF School of Pharmacy program who had also completed a clinical residency program there. He taught us the basics of clinical pharmacy during the last year of our five year BS program. After several years of practice, I went back to school at the PharmD/Residency program at the University of Kentucky to obtain additional clinically-oriented training. So, like you, I come from a program (i.e., UCSF, University of Kentucky) with a strong clinical component, but that doesn’t mean the only thing that is important are “clinical activities”.

    The basic responsibility of a pharmacist is to get the right drug to the patient. For many years, “right drug” referred mostly to distributing the correct drug to the patient, whether in inpatient or outpatient setting. Clinical Pharmacy brought new meaning to the term “right drug”, since it required pharmacists to make sure that what was ordered for the patient was appropriate from a therapeutic use standpoint and that there was no interference of any medication with the other conditions of the patient or other medications being taken.

    Jerry, as you have alluded to, I think many pharmacists, including those in leadership positions, only think about the newer “right drug” responsibility for pharmacists (i.e., the clinical role), while at the same time neglecting our distribution “right drug” responsibilities. The truth is that regardless of the scope of clinical pharmacy services provided to a patient, if the patient never receives the correct drug, then all of that effort is in vain. Our drug distribution “right drug” responsibilities should always come first. If the drug distribution system is broken, there is little value in wasting money on a clinical program. It doesn’t matter if a pharmacist can correctly dose an critical drug for a patient with renal and/or hepatic insufficiencies if the pharmacy system is unable to properly prepare and deliver the correct drug for administration by the nurse. Drug distribution is a pharmacy core responsibility. Doing that well, allows other pharmacy programs and services to be more valuable.

  6. Why is the adoption rate of iv workflow tools is so low?

    We (pharmacists) are the problem. Do you think a management/industrial engineer in charge of an IV production line in a pharmaceutical company would be slow to adopt technology and automation to minimize the impact of human error? NO WAY!!!

    Some pharmacists, on the other hand, do not appear to appreciate the changes in volume and complexity of the IV workflow processes that have occurred over the past 20-30 years. They must think that “their pharmacy staff” are not capable of making a fatal human-slip error, like the one that occurred in Bend, Oregon on December 1st, 2014. Just because a process worked 30 years ago, does not mean it will work today. In some regards, they appear to be oblivious to what is happening in their own IV rooms.

    In some regards, this reminds me of the Boiling Frog story which is a metaphor for the inability or unwillingness of people to react to or be aware of threats that occur gradually. Another expression that could be used to describe the views of some toward adoption of IV workflow safety systems is that they “…can’t see the forest for the trees”. So much time has been been spent complying with USP requirements, that attention to the IV workflow processes have been overlooked. The end result is that we have patient-specific IV products labeled with a barcode that can be scanned by the nurse at the bedside, even though the preparation of those same CSPs are done withOUT the benefit of barcode scanning of the ingredients to prevent human-slip errors during their preparation. This focus on preparation sterility in the absence of the use of IV workflow safety technologies leaves me with only one way to describe many CSPs: “Clean, but deadly”.

    If serious injury or death results from the failure to use IV workflow safety technologies, then like the frog thrown into a boiling pot of water, there will be an immediate reaction to fix the problem (i.e., the implementation of barcode scanning IV workflow technologies, change in pharmacy management). Why are so many pharmacies waiting for that to happen? An ounce of prevention is worth a pound of cure…

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.