Today’s sterile compounding climate is a lesson on availability cascades

I’ve been slowly reading ‘Thinking, Fast and Slow’, a best-selling book by Daniel Kahneman, a Nobel Prize winning psychologist and economist known for his work on the psychology of judgment and decision-making. 

The book summarizes behavioral science research conducted by Kahneman. The recurring theme of the book is what the author defines as two modes of human thought, “System 1” and “System 2”. System 1 is fast, instinctive, and emotional. System 2, as you might guess, is slower, deliberate, and logical. The author spends a lot of time discussing how System 1 and System 2 impact choices and judgment, creating an incredible amount of bias in our decision making process. The basic thesis being that people really don’t have as much control over their decisions as they believe. According to Kahneman, humans constantly fall into predictable thinking traps, often due to the fast and emotional System 1. The concept is philosophically interesting but practically devastating.[1] 

One section of the book that I found particularly compelling, and applicable to pharmacy, was Chapter 13, ‘Availability, Emotion, and Risk’. The chapter goes into great detail about how bias impacts decisions, even when presented with facts to the contrary. According to Kahneman, these biases are warped by media coverage, especially for emotionally charged issues. This creates problems because the average person is driven more by emotion than reason. [2] 

This area of the book hit home for me because of my current involvement in a massive project to bring hospital pharmacy cleanrooms into compliance with new USP <797> and <800> guidelines. The entire project is the direct result of decision bias surrounding sterile compounding. 

If one were to go back and research the impetus for developing current sterile compounding guidelines, one would be hard pressed to find anything concrete. Current USP Chapters <797> and <800> are based more on “expert opinion” than science. If it were the other way around, one would find more references for specific requirements and fewer reader comments. The guidelines would read more like a journal article and less like best practice recommendations. However, that’s not what we find in these chapters. 

While there’s nothing wrong with expert opinion, per se, they should not override common sense. Expert opinions should be considered recommendations and guidance, not requirements and law. The problem with expert opinion is that it can be influenced by emotion, the same as an average person. Emotional decision making can lead to problems. According to Kahneman, “…biased reactions to risks are an important source of erratic and misplaced priorities in public policy. Lawmakers and regulators may be overly responsive to the irrational concerns of citizens, both because of political sensitivity and because they are prone to the same cognitive biases as other citizens”. 

Consider this, Kahneman describes something in his book called an availability cascade. An availability cascade is “a self-sustaining chain of events, which may start from media reports … and lead up to public panic and large-scale government action. …[the] emotional reaction becomes a story in itself, prompting additional coverage in the media, which in turn produces greater concern and involvement”. This describes perfectly the current state of sterile compounding regulation in pharmacy. 

Remember a little thing in 2012 that caused outrage aimed at pharmacies across the country? For those of you that don’t recall, Google New England Compounding Center (NECC).  In summary, people died as the result of fungal meningitis caused by contaminated steroid injections compounded by NECC. It was pretty bad. The incident lead to congressional hearings, more power being granted to the FDA, increased scrutiny by regulatory and licensing agencies, and ultimately, more pharmacy regulation. Pharmacy is still reeling from the impact that the NECC tragedy had on the profession. 

Here’s the thing, it was a horrible, terrible, no good tragedy, but it had nothing to do with lack of regulation. The folks at NECC simply chose to ignore best practice and place profit ahead of common sense and standard safety precautions. Lest we forget, USP <797> was already on the books in 2012. Would NECC have followed standard sterile compounding processes in place at the time, we wouldn’t be having this discussion today.  

A single event, the NECC tragedy, was championed into significant changes without objective analysis. Why? According to Kahneman, we have “a basic limitation in the ability of our mind to deal with small risks, we either ignore them altogether or give them far too much weight…..biased reactions to risks are an important source of erratic and misplaced priorities in public policy. Lawmakers and regulators may be overly responsive to the irrational concerns of citizens, both because of political sensitivity and because they are prone to the same cognitive biases as other citizens”. That’s a perfect description of the current sterile compounding climate. Unfortunately, there’s no going back. As one prominent sterile compounding consultant told me, that ship has sailed. The only thing we can do now is be more prudent with decisions moving forward. 

Think of the resources that have gone into cleanroom construction, training, and testing since the NECC incident. It’s staggering. Now think of what other pharmacy initiatives could have been implemented with the same resources. Based on risk alone, patient safety would have been better served by putting the same effort toward improving sterile compounding accuracy instead of sterility.


[1] Please read the book. It contains little tests and puzzles that will blow your mind. I thought I had solid command of my thinking. I thought my decisions were logical and well reasoned. I was wrong. 

[2] Paul Slovic, a professor of psychology at the University of Oregon and the president of Decision Research stated that people are “guided by emotion rather than by reason, easily swayed by trivial details, and inadequately sensitive to differences between low and negligibly low probabilities”.

A “no-mistakes sponge system” — bar-coded sponges in the OR

While not directly related to pharmacy, the SurgiCount Safety-Sponge System is kinda’ cool. The system uses low-tech barcode technology to prevent surgical sponges from being left behind in patients. Simple yet effective.

The system uses sterile bar-coded sponges and a computer tablet loaded with proprietary software to ensure that all sponges are tracked. After approximately 11 million surgical procedures over the last five years, which involved the use of more than 200 million sponges, the system boasts zero sponges left behind.” source: MLive

The lowly barcode gets a bum wrap sometimes. It’s far from perfect, but overall I believe the use of barcodes has improved safety throughout the healthcare environment. I remember just a few short years ago when barcode scanning was uncommon. Now, not so much. “Overall, 92.6% of hospitals have barcode-assisted medication administration (BCMA) systems to verify patient identity and electronically check doses administered by nurses (Table 2). Over the past 13 years, the percentage of hospitals having BCMA has increased from 1.5% in 2002.“(1) I get calls in the pharmacy from nurses refusing to administer medication because they’re unable to scan the medication barcode. We’ve come a long way.

(1) Pedersen C, Schneider P, Scheckelhoff D. ASHP national survey of pharmacy practice in hospital settings: Prescribing and transcribing—2016. American Journal of Health-System Pharmacy. 2017;74(17):1336-1352. doi:10.2146/ajhp170228.

Does tall man lettering work?

First of all, is it tall man, tall-man, or tallman? And why is it called “tall man lettering” when none of the letters are actually taller than the others? Heck if I know. Just more questions in a mountain of questions piling up around tallMAN lettering.

Pharmacy Practice News: “[The study] found that there hasn’t been a substantial drop in drug name mix-ups since use of tall man lettering became widespread around 2007… “We saw no reassuring trend of declining rates of errors,” said study author Chris Feudtner, MD, PhD, MPH, a pediatrician at the University of Pennsylvania’s Perelman School of Medicine, in Philadelphia…If tall man lettering were working, the researchers expected to see a significant decrease in these types of errors after 2007 when the JC began recommending that hospitals implement tall man lettering and other typographic drug safety measures. No such drop was seen.”

The entire use of TaLlMaN lettering has always seemed odd to me. I could never understand how it would keep anyone from grabbing the wrong medication. I mean seriously, who in their right mind would confuse SUMAtriptan with ZOLMitriptan, or ARIPiprazole with RABEprazole. Crud, they’re not even remotely close when one considers the alphabet. When searching for the drug within a CPOE system one types “sum…” or “zol…”, not “…triptan. C’mon, people!

One classic mix up is hydrOXYzine and hydrALAzine. They definitely have similar names, but the former is an antihistamine used to treat itching, while the latter lowers blood pressure by exerting a vasodilating effect through a direct relaxation of vascular smooth muscle, i.e. it’s a blood pressure medication. Why the heck would anyone want to use a blood pressure medication to treat itching? They wouldn’t.

Perhaps it would make more sense to simply put the drug class or use on the packaging. You know, hydroxyzine [antihistamine/itching] or hydralazine [vasodilator/blood pressure]. Better yet, let’s require prescribers to place an indication on all orders: hydroxyzine 25mg PO Q6H PRN ITCHING versus hydralazine 25mg PO Q6H FOR BLOOD PRESSURE. Might even be educational for some prescribers.(1)

How about we spend a little time creating smart EHR’s that know when something is amiss?(2) A system that won’t let the provider select a medication for an inappropriate indication without jumping through some hoops. Something like “You sure about this, bruh? Hydralazine isn’t typically used for itching. Were you trying to prescribe hydroxyzine?”(3)

Now combine smart prescribing practices like those above with safety measures in the pharmacy like barcode scanning for verification. Selecting the wrong medication in the pharmacy is always possible and happens for a host of reasons, regardless of t.a.l.l.m.a.n lettering. Barcode scanning is a pretty good way to help ensure that you have the prescribed medication in hand.

Overall, I’m not surprised that T-A-L-L-M-A-N lettering didn’t make much of a difference in the study. Even though it’s become a standard of practice, I don’t know that I’ve ever bought into it. My preference would be to use better technology with a little common sense.

If you’re interested in reading the article (BMJ Qual Saf 2016;25[4]:213-217; BMJ Qual Saf 2015 Dec 16. [Epub ahead of print]), it can be found here.



(1)    You might be surprised to find out how little some practitioners know about the medications they prescribe. I constantly appalled by the prescribing practices that I see in the acute care setting.

(2)    Something like AI or ML, perhaps. Hmm…

(3)    That’s kind of how the call goes when you have to let a prescriber know they may have inadvertently selected the wrong drug.

ISMP releases new medication safety best practices document

I quite literally stumbled across this the other day while doing research for another project. I heard that ISMP had updated their best practices document, but didn’t see an official announcement. It’s possible I just missed it.


The document contains some great new safety recommendations. All in all there are eleven best practices listed. Most of the recommendations are what I would consider minor, but there are a couple that I think are worth highlighting:

Dispensing vinCRIStine in a minibag instead of a syringe. This is one of those ideas that seems so simple, yet brilliant. When you read it, you instantly say to yourself “why didn’t I think of that?”. VinCRIStine is commonly dispenses in a syringe and given via short IV push. However, being dispensed in a syringe has led to the accidental administration of the drug via the intrathecal route. The result is devastating neurological damage, up to and including death. By simply putting the drug in a minibag, you effectively eliminate the possibility of it being administered intrathecally.

Performing independent verification of ingredients during sterile compounding. This includes a recommendation to use technology to “assist in the verification process (e.g., barcode scanning verification of ingredients, gravimetric verification, robotics, IV workflow software) to augment the manual processes.” I believe this is the first official document from an organization to include such a recommendation. Congratulations to ISMP for having the resolve to do this. ASHP needs to follow suite.

More information can be found here: 2016-2017 Targeted Medication Safety Best Practices for Hospitals [PDF]

More thoughts on USP <797> and pharmacy IV rooms [comment from reader]

A friend and colleague, Ray Vrabel, left a comment on my post from April 20th. I thought what Ray had to say was too good not to post. He raises some good points, which are worth more discussion.

Ray is a sharp guy, and he and I have had some good conversations over the past couple of years. He’s passionate about patient safety and pharmacy practice. While I don’t always agree with everything that Ray has to say, I definitely appreciate his thoughts and opinions.

Make sure to read more of Ray’s thoughts here.


Your post has got me wondering about a number of things: Area 51, The Kennedy Assassination, Obama’s birth certificate, and now USP797.

You raise the question which I have always wondered about: What was the problem that USP797 was attempting to solve? Was there documentation of significant problems associated with pharmacy-prepared IV admixtures by pharmacies who were following the ASHP Guidelines in place before USP797? Did anyone conduct a multi-hospital study to determine if hospitals following pre-USP797 ASHP Guidelines had any clinical problems associated with pharmacy IV admixtures? In other words, what is the science that drove the USP797 standards?

We have effectively turned our pharmacy IV Rooms into GMP-like sterile manufacturing facilities. So now, if properly followed, we have hospital pharmacies preparing a very high quality product from a sterility standpoint. That’s a good thing, but we also have a number of unintended consequences:

(1) Most IV admixtures are now prepared by pharmacy technicians, but they are no longer being directly supervised by pharmacists because of the onerous garbing requirements, making it inconvenient for the pharmacist to move into and out of the IV room.

(2) While there is now a requirement that every pharmacy must follow UPS797 standards, we do not have a technician licensure/certification requirement for all technicians in all states.

(3) While USP797 has required the use of all types of environmental, operational, and testing products, there is no requirement for pharmacies to use barcode checking of the IV admixture ingredients (i.e., Label, bag, and additives). Why do we have excellent sterility requirements with no requirement for accuracy of IV additive preparation?

What’s wrong with this picture? We now have sterile IV admixtures, but we don’t have any standards to make sure that the IV admixture is made correctly (i.e., correct ingredients). I feel that barcode scanning during medication preparation (BCMP) should be the minimum standard for ALL IV admixtures in ALL pharmacies. For more on this, please see my LinkedIn post:

ISMP responds to deadly drug error in Oregon

Last week I wrote about the tragic death of a patient caused by a drug error (CSP error results in death of a patient). One day later on December 18, 2014, ISMP also addressed the error in the Acute Care edition of their biweekly ISMP Medication Safety Alert, i.e. one of their newsletter. I had hoped that ISMP was going to provide much greater detail and insight into the error, but that’s not the case. At least not at this point, anyway.

I had hoped to find out what occurred in the pharmacy to allow such a mistake to happen. Perhaps more details will come to light as time goes on. All we can do is wait.

With that said here are some things from ISMP worth noting:

To prevent inadvertent use, identify neuromuscular blockers available within your organization and where and how they are stored. Regularly review these storage areas, both inside and outside of the pharmacy, including agents that require refrigeration, to consider the potential for mix-ups.

Limiting access to these products is a strong deterrent to inadvertent use. Consider limiting the number of neuromuscular blockers on formulary, and segregate or even eliminate storage from active pharmacy stock when possible.

Restrict storage of paralyzing agents outside the pharmacy and operating room by sequestering them in refrigerated and nonrefrigerated locations.

ISMP recommends highly visible storage container for neuromuscular blockers (one example here:**

ISMP recommends affixing warning labels on vials and admixtures that clearly communicate the dangers of neuromuscular blockers.**

ISMP recommends the use of IV workflow technologies. “Now is the time for hospital leadership to support the acquisition of IV workflow technologies that utilize barcode scanning of products during pharmacy IV admixture preparation.” While the article lists only three systems, there are several on the market [see  In the Clean Room TOC for a current list of many of the available systems].

**I know that many think this is a good idea, but I’m not so sure that I’m one of them. On the surface, using highly visible storage containers and labels might seems like a good idea, but over time people become used to the idea and become blind to the differences. In addition, over the years the number of items that require alternate storage and labeling has grown, making differentiation “the norm”. It’s like the student that highlights everything in the textbook with five different colors. Eventually the entire book is highlighted, making the process meaningless to the reader.

CSP error results in death of a patient

A 65-year-old woman died at St. Charles Medical Center in Oregon after being given an infusion of rocuronium instead of fosphenytoin.

“The prescription was entered correctly into the electronic medical records system, and the pharmacy received the correct medication order, the AP reported. The IV bag was also labeled properly. After the pharmacy worker mistakenly filled Macpherson’s IV with rocuronium, a second employee did not catch the error while checking the vials of medication and the IV bag for the 65-year-old patient.” (via: Pharmacy Times).

As details of the tragedy continue to emerge, here’s what we know so far:

  • An infusion of fosphenytoin was ordered for the patient, presumably a piggyback.
  • Instead of fosphenytoin (anti-seizure med) the patient received rocuronium (a paralytic)
  • It is unclear at what point in the compounding process, if at all, the infusion was verified by a pharmacist. According to the article “a second employee did not catch the error while checking the vials of medication and the IV bag”. Not entirely sure what that means.
  • The infusion was hung
  • A fire alarm sounded
  • The nurse closed the patient’s door and didn’t check on her for 20 minutes. That was more than enough time for the drug to cause irreparable harm to the patient.

It is unclear what process was used to make the infusion, or what safety safeguards were in place. The real shame here is that there are any number of available technologies that could have prevented the error. Any of the semi-automated workflow management systems on the market today would have worked. Bar code scanning, gravimetrics, perhaps image assisted verification, etc. Take your pick.

According to an article from The Bulletin “To help prevent similar mistakes from happening, the hospital’s pharmacy has begun placing orange stickers on IV bags containing paralytic agents that indicate what’s in them. [The patient’s] IV bag had a blue sticker indicating it was a neuromuscular agent, which Boileau [Dr. Michel Boileau, St. Charles’ chief clinical officer], said both fosphenytoin and rocuronium are.” Not exactly sure how using orange stickers instead of blue is going to do much. Seems kind of silly. I think I’d start looking at something a little more aggressive. I’d also rethink my classification of both fosphenytoin and rocuronim as “neuromuscular agents”. I think I’d call fosphenytoin a hydantoin anticonvulsant and rocuronium a nondepolarizing neuromuscular blocker. They’re clearly not the same class of drug.

It will be interesting to see how organizations like ISMP and ASHP respond to this latest error.

Medical errors remain a problem despite years of effort

I recently read an article at Senator Bernie Sanders website about preventable medication errors. The article lists preventable medical errors in hospitals as the third leading cause of death in the U.S. behind only heart disease and cancer.

The article goes on to say that “the Journal of Patient Safety recently published a study which concluded that as many as 440,000 people die each year from preventable medical errors in hospitals. Tens of thousands also die from preventable mistakes outside hospitals, such as deaths from missed diagnoses or because of injuries from medications.” I’m not exactly sure what article they’re referencing here as they didn’t provide a link or additional information, but I assume they’re referring to the article by James in September 2013. Just a guess, I could be wrong.1 Regardless of the actual reference, the bottom line is that the number of patients that die from preventable medical errors is high.
Continue reading Medical errors remain a problem despite years of effort

Monitoring Pharmaceutical Products in Twitter [article]

Pharmacovigilance of TwitterThere’s an interesting article in the April 2014 edition of Drug Safety that looks at English Twitter posts from November 2012 through May 2013 to see if there is any correlation between adverse event (AEs) reporting via Twitter and more “official” channels.

The authors collected public Tweets, which were subsequently stored for analysis using Amazon Web Services. See how they did that? Nothing magical about it. They used readily available information and a commercially available storage source.

Through the use of some human ingenuity, a “tree-based dictionary-matching algorithm”, and some manual labor, the authors collected 6.9 million Tweets, of which 61,402 were examined, ultimately leading to 4,401 AEs identified; referred to as Proto-AEs by the authors. During the same period 1,400 events were reported by consumers to the FDA.

While not perfect, and most certainly limited, I think the results were surprising, encouraging, and disappointing all at once.

Surprising because of the number of Proto-AEs found in the Twitter stream. People are savvy. “There was evidence that patients intend to passively report AEs in social media, as evidenced by hashtags and mentions such as #accutaneprobz and @EssureProblems. Even within 140 characters, some tweets demonstrate an understanding of basic concepts of causation in drug safety, such as alleviation of the AE after discontinuation of the drug.”

Encouraging because being able to mine social media streams like Twitter could open up an entirely new avenue of real-time AE tracking; we all know that AEs are under reported, which leads to a lack of information for pharmacists and other healthcare professionals.

Disappointing because of the limited number of AEs reported to the FDA. I used to see AEs in the hospital that were never reported. I’m as guilty as many for not reporting AEs.

More work needs to be done in this area before we can begin to rely on data mined from social media, but they again it’s probably as reliable as information collected elsewhere.

The article is open access and the full version is online for free, so there’s no excuse not to read it.