Blog Viewer

I need to state up front that these are my opinions which are a bit off the beaten path. You are welcome to disagree with me, as I know many will.

This year contains the 50th anniversary of my graduation from pharmacy school, and my 50th year as a member of ASHP. During those 50 years, I have the opportunity to work within a completely floor-stock based medication delivery system (nursing was in charge of it all), individual in patient prescription systems (both with, and without a patient profile), the origins of unit-dose drug distribution (including cart exchange), the introduction of pharmacy-based IV admixtures, and the introduction and use of automated dispensing cabinet technology (ADC). These changes were introduced by pharmacy, primarily based on expectations of improved patient safety. These changes also required nurses to become increasingly dependent on pharmacy operations, a dependence nurses found (and continue to find) challenging.

Why should we care about what nurses want? In my opinion, the answer to that question is that we should care because we do not administer medications to patients. Nurses do. Even our clinical practices focus more on ensuring that patients are being ordered the right medication at the right dose than they focus on facilitating the availability of those medications for administration. Everything that we manage to do clinically to improve our patients' medication therapy is lost if the medication is unavailable to be administered when needed. 

I observe that our changes in process and technology for the distribution of medications have often resulted in added non-productive work for nurses. Even ADCs, which at least returned the availability of most commonly used drugs to the patient care areas have complicated medication administration by requiring medications to be stored in multiple places (the ADC doesn't hold everything), by failing to maintain adequate supplies in those cabinets, and by making unannounced changes in the list of items stored in a given cabinet, leaving the nurse to have to perform more work just to determine whether the medication they need is available somewhere on the unit.

Don't get me wrong; I would not propose for a second that we return to the days when we as a profession had no involvement at all in how medications were stored, acquired and administered and the medication error rate hovered around 20%. But my 50-year perspective tells me that all those changes we have made have been incomplete solutions, most of which have had to preserve some elements of older solutions in order to handle all the medications that needed to be delivered. Go into any acute care facility today who are operating ADCs on an "all meds model" and you will find most medications available in the ADC, some medications supplied as patient-specific supplies (back to the individual inpatient prescription model), and others still kept on the unit as uncontrolled floor stock.

There is an interesting video by Russ Ackoff that recounts how what was then "the phone company" created the telephony systems we enjoy today by first realizing that they had been spending 50 years "perfecting" current systems (e.g., the rotary dial telephone), then imagining what an ideal telephony system might look like (by first cataloguing all the problems with the current system that needed to be fixed), and then reimagining and reconstructing that telephony system into what we have today. I was 8 years old when the meetings described by Ackoff took place; what they eventually created from those meetings would have seemed like science fiction (or even magic) to that 8 year-old boy though, in the  Sunday comics, Dick Tracy had a 2-way wrist TV.

Similarly, we have had ADCs as our primary medication distribution technology for over 30 years and I will bet that even those who used the very first of these products could figure out how to use what we have today. We made a lot of things better; but we made some things worse. And the nursing process around our drug delivery systems remain chaotic. Studies of nursing workflow show that they spend most of their time walking, they are frequently interrupted while attempting to complete a patient care activity, and they spend a lot of time as hunter-seekers of the supplies and medications they need to render care to patients^1,2,3,4,5.

Interestingly, these technologies haven't really made our professional lives easier either. I had the opportunity to develop a model^6,7 from ADC transactional data around trips from the pharmacy to ADCs. I have run this model against many health-systems' data sets with the same result each time. The pharmacy is nearly constantly running items up to the ADC and the vast majority of those trips involve few medications and even fewer ADCs (the model doesn't measure the delivery of patient-specific supplies, though it could in some cases).

So if, like "ma Bell", we were to set out to reimagine medication delivery in the acute care setting, what problems would we have to solve?

Here's my list  - for the sake of simplicity, I presume that most medications are administered in the patient room - there are other use cases:

• All the medications for a patient at any particular time would be in one place, in the patient's room.
• Medications would be delivered to that location more or less just-in-time.
• There would be sufficient supplies of as-needed medications to ensure availability when needed.
• There would be storage for multiple-dose containers.
• The nurse could enter the patient's room, assess the patient, acquire and administer the medications, document administration, assess the impact of the medications (where appropriate) all without leaving the room.
• Humans would be out of the run-drugs-up-to-the-unit business.
• Humans would be out of the move-drugs-to-the-new-patient-location business (when patients were transferred).

At first blush it appears that this is likely not possible with the technologies we have today. We certainly couldn't accomplish this cost-effectively by growing headcount. Nor could we do this by placing ADCs as they exist now in every room. But there are some interesting technologies working today that might contribute to an automated solution:

• There are lockable, plug-in, self-identifying medication storage containers and "buses" onto which to place them that permit them to be opened and used and document their activity.
• The problems of robotic navigation around a facility have largely been solved; the current technologies still require a human to come take things out of the robot.
• There are known pass-through cabinets that could permit robotic technologies to place (or replace) medication supplies in patient rooms.  These "nurse servers" have failed in popularity primarily because the medications placed into them are not conveniently packaged to facilitate quick checking and use and once-a-day delivery virtually guarantees that the medication supply will be incomplete by the end of a day. But they could be outfitted with better technology.

I have had some wild imaginings regarding how such a system might be engineered. There are even patent submissions around a couple of them.  But, sadly, most of these ideas would be affordable only to large institutions. So I feel unlikely to be the one who eventually re-imagines, and delivers the medication delivery system that truly meets the needs of all the stakeholders.

• If this involved robots, what would they do and how would they work? What would they need to do that they currently cannot do?
• Might this involve better automating the medication dispensing process in the pharmacy?
• Could we remove the human entirely from the dispensing process?
• Might this involve completely different drug packaging purposefully designed for this process?
• What might the in-room storage look like and how might it operate?

I therefore propose this to my informatics colleagues in the hope you find or develop the necessary technologies to make something like this practical, and have the courage to drag our healthcare systems kicking and screaming into a brave new world that improves both pharmacy and nursing productivity. With rapid advances in artificial intelligence, quantum computing, robotics,  and the internet of things, it may just be possible. Flynn et al have articulated a goal for complete automation of pharmacy practice⁸. They have dared to dream. I encourage you to dream large.

As always, these opinions are my own, and are not necessarily those of ASHP.

Dennis A. Tribble, PharmD, FASHP

Ormond Beach, FL

tribbledennis@gmail.com

______________________________________________

¹ Keohane, C et al Quantifying Nursing Workflow in Medication Administration JONA 38(1);19-26 January 2008

² Garrett, S and Craig J Medication Administration and the Complexity of Nursing Workflow Clemson University(2009)

³ Cain C, Haque S Chapter 31 organizational Workflow and Its Impact on Work Quality Patient Safety and Quality: An Evidence-Based Handbook for Nurses Hughes, RG editor published by AHRQ, April 2008

⁴ Cornell P et al Transforming Nursing Workflow, Part 1: The Chaotic Nature of Nurse Activities JONA 40(9):366-373 September 2010

⁵ Cornell P et al Transforming Nursing Workflow, Part2: The Impact of Technology on Nurse Activities JONA 40(10):432-439 October 2010

⁶ Tribble D et al, Trips Model, Patent pending – Application # WO2022271384 A1

⁷ Tribble D et al, TRIP MANAGEMENT SYSTEM, Patent pending - US Application # 63/215,342

⁸ Flynn A et al The Autonomous Pharmacy Amer J Health-Syst Pharm 78(7):636-645 April 1, 2021