A missed opportunity for safety – why scanning a limited formulary in the i.v. room is a mistake

By | January 28, 2015

Although some hospitals have chosen to limit use of these systems [IV workflow technology] for focused areas like admixture of chemotherapy or high-alert drugs, there’s no telling when someone might accidentally introduce a high-alert drug when preparing other drug classes that wouldn’t ordinarily be scanned. Therefore, to be maximally effective, the system must be utilized for all compounded admixtures”. (ISMP)

A couple of weeks ago I wrote about the need to use bar-code scanning technology during compounded sterile product (CSPs) preparation. In my mind it’s a no-brainer. The i.v. room is a dangerous place, and no amount of “double checking” is going to change that.

The efficacy of i.v. medication administration hinges on the integrity of dose preparation and labeling in the pharmacy. Get it wrong in the pharmacy and you get it wrong at the bedside. Which is why I’m perplexed by pharmacies that have made a conscious choice to limit the use of i.v. room automation and technology to a small percentage of their formulary. As I continue to visit hospitals and look at the use of semi-automated i.v. workflow management systems and robotics, I am struck by how few pharmacies use these systems for all CSP production.

There’s a false belief that using these systems for high-alert medications, chemotherapy, and a randomly selected number of “dangerous medications” will somehow prevent all errors from occurring. As well-intentioned as this approach may be it fails to get to the crux of the problem, i.e. choosing the wrong item during sterile compounding.

Much effort has gone into increasing awareness around high-alert medications in acute care pharmacies. Drugs such as neuromuscular blockers, chemotherapy, concentrated electrolyte vials, and so on now contain a litany of warning labels attached to shelves, bins, and medication bags. Take a look at the ISMP high-alert medications list for acute care settings. The list incorporates a sizable chunk of medications used to treat acute illness, i.e. the medications used when people are in the hospital.

Today it’s easier to spot items that are free of warning labels than it is to determine which are more dangerous than another. Medications have been de-alphabetized, relocated, double and triple checked, and put under lock and key in an attempt to limit errors. It is unclear to me exactly what effect this has had on error reduction on the whole, but I can assure you that it makes medication distribution frustratingly more difficult. Which is why I don’t understand the failure of pharmacies to use readily available technology to limit sterile compounding errors. It’s even more perplexing when I see a pharmacy using an i.v. workflow management system – DoseEdge, BD Cato, Verification, and so on – on only a limited formulary.

There are those that will say certain medications present low risk for error, and even when an error occurs will likely not cause harm. That’s true in many cases. All errors are regrettable, but few are clinically significant. Only 2% of errors that occur during CSP production have been shown to be clinically relevant.[1]

With that said, what happens when someone grabs a potentially deadly medication in place of an item that has been identified as low risk? That’s exactly what happened in the recent tragedy in Oregon. [2] The ordered medication, fosphenytoin is not considered a high-alert medication, and as such doesn’t require the use of special precautions in most facilities. The drug that was mistakenly used, rocuronium is on every high-alert medication list known to man. The items look nothing alike, as can clearly be seen in the images below. Nonetheless, the error occurred and a patient died.

Rocuronium and Fosphenytoin Sides

Rocuronium and FosphenytoinTops

Perhaps the error would have been avoided if made in reverse, i.e. order for rocuronium instead of fosphenytoin. I wager that the pharmacy involved in the error has special safeguards in place when preparing a neuromuscular infusion for a patient. Every pharmacy I’ve ever worked in has special procedures in place for preparing CSPs containing a neuromuscular blocker. Why? Because a mistake with a neuromuscular blocker can be deadly. Regardless of whether or not the reverse would have prevented the mistake, a mistake was made. Scanning the medication vial against the patient label during CSP production would have prevented the mistake, or at the very least alerted the preparer of the mistake.

It’s time for pharmacies to stop limiting the items they scan during CSP preparation. Scan it all.

————————————————

  1. Flynn, EA, Pearson, RE, Barker, KN. “Observational study of accuracy in compounding IV admixtures at five hospitals.” Am J Health-Syst Pharm. 1997 Apr 15; 54: 904–912
  2. Wrong drug put in IV bag led to fatal Bend hospital error. 2015. Available at: http://www.ktvz.com/news/st-charles-to-issue-results-of-fatal-drugerror-probe/30115104. Accessed January 4, 2015.

2 thoughts on “A missed opportunity for safety – why scanning a limited formulary in the i.v. room is a mistake

  1. Ray Vrabel

    The introduction of technology in hospitals has always confused me. When it comes to the introduction of barcode scanning, we appear have done it in the opposite order of what should have been done based upon potential risks to patients.

    In my opinion, barcode scanning in the IV room should have come first, followed by barcode scanning for other non-sterile compounding activities, followed by bedside barcode scanning (BCMA). Instead, the first department to adopt barcode scanning is the hospital gift shop! That happened over 20 years ago. BCMA has been slowly rolling out over the past 10 years, and now finally we’re getting to the IV room.

    Now that pharmacies are doing barcode scanning in the IV room, many are only doing it for high-risk drugs. As you pointed out, this will DO NOTHING to prevent the accidental use (simple human error) of a high risk drug (e.g., rocuronium) for the compounding of a low risk IV admixture (e.g., fosphenytoin)!!! The only logical conclusion that can be reached is that to prevent errors associated with high risk drugs, then you MUST use barcode scanning for ALL IV admixtures, otherwise the accidental use of high risk drugs will never be prevented.

    It may “feel good” for the pharmacy director to say “…we use technology to prevent errors when compounding for our high risk IV admixtures”, but who are they fooling? Tell that to the family impacted by the accidental use of a high risk drug during a routine, non-high risk, IV additive. Product segregation, special warning stickers, and other physical methods can be used to warn against the accidental access to high risk drugs, but only barcode scanning will PREVENT their accidental use.

    Then, there is the elephant in the IV room. Could it be that the failure to use more sophisticated medication error prevention technology in the IV room is because of COST? Some automated IV Workflow solutions are sold on a “Cost per IV” model, which means that the larger percentage of the total IV additive workload that uses this technology, the more it will cost the pharmacy. Could it be that pharmacies are limiting their use of these technologies for only high risk drugs, as a way “justify” their operational costs? I hope not!

    Pharmacy needs to start getting their priorities in order. There are three things that should be considered before cost, workflow efficiency, and impact on staff: Patient Safety, Patient Safety, and finally Patient Safety!!! Follow these principles and you will “Do the right thing!”

  2. Jerry Fahrni Post author

    Spot on, Ray. It’s always interesting to me that hospitals, and pharmacies, will use patient safety as a reason to get something done, but rarely put it ahead of cost. I’ve dealt with it many times when I was still a practicing IT pharmacist.

    I’ve also experienced the less is more approach secondary to cost per dose. However, I understand that the cost per dose model is going away. Not sure if it was from customer pressure or some other market force, but I think we’ve seen the end of it.

    Thanks for stopping by.

Leave a Reply

Your email address will not be published. Required fields are marked *