Sub-optimizing the Patient Medication List

Hospitals frequently end up implementing suboptimal systems that actually result in a more dangerous patient safety environment. Healthcare is so inundated with poorly stratified safety guidelines and regulations that hospitals rush to initiate endless improvements in order to remain in compliance, without understanding the implementation challenges and without follow through. The results are initiatives that look great on paper at the executive level but that are completely dysfunctional at the level of patient care. Let’s look at an example.

LGR was under pressure to implement an electronic pre-admission Patient Medication List (PML) to reduce the incidence of medication error. A task force was put together to develop an IT-based process that would enable care providers seeing patients in the surgical pre-admission test center to enter the patient’s home medications into the computer; this record would then be printed as a separate sheet and placed into the patient’s paper chart behind the new PML tab for straightforward, easy access. Here’s the formal system:

1. Patient comes to the surgical pre-admission test center with their list of home medications.
2. The care provider evaluating the patient enters the medications into the new computerized PML template.
3. The PML is printed out and placed into the paper chart behind the new PML tab
4. The PML serves as an easy cross-reference during admission for any new medications added.
5. On discharge, the PML is referenced to ensure that the patient continues with their usual medications as appropriate.

Here’s what actually happened when the PML was launched (informal system). Computer entry of patient medications into the PML worked well enough in the pre-admit test center, provided that patients actually knew their detailed medication histories. Because the medications were now listed in the PML, the medication section of the history and physical form (H&P) was now simply noted with “see PML.” By the time the charts were assembled and the patients were on the floor recovering post-operatively, only 70% of patients seen in pre-admitting had copies of the PML in their chart. Further, only 25% of charts had the PML tab, resulting in the need for caregivers to search through the entire chart to find the PML, when it was present. With time, it became so frustrating to look through the chart for the PML that many caregivers stopped looking at all. They’re almost back to square one except that now there is no medication list in the 30% of the charts at all because the H&P says, “see PML.” You would think that caregivers could at least go online to look at the electronic version of the PML; however, access to the PML was restricted to a subset of users. So, while LGR is now in compliance with the PML guideline, appearing to be a leader in patient safety, here’s the suboptimal result that replaced the existing process:

1. In-chart documentation of pre-admission medications has dropped from 100% to 70%.
2. The PML process as developed works only 25% of the time! (PML in the chart & behind the PML tab).
3. There is now a double standard for documenting pre-admit medications since none of the surgical patients admitted directly or through the emergency department have a PML (not part of the roll-out).
4. Caregivers are so frustrated by the inconsistency that they largely forego looking for the PLM at all - it’s easier to ask the patients directly about their meds.

Food for thought:
1. As currently implemented at LGR, the PML has increased the potential for medication error by as much as 30% and it is now systematized with an IT interface. We frequently forget that while computer automation can improve safety, it is not the magic bullet. Computers are tools and not solutions: if the underlying process is not understood in its entirety, we will only automate what we already have, or make it worse.
2. The process is only understood if the people doing the work are involved! In addition to the IT folks we need to include caregivers doing the pre-operative assessment, the people involved in chart assembly and the caregivers taking care of patients during their admission.
3. Closing the implementation loop is critical. Too frequently, we launch an initiative and assume that the execution will follow to completion on its own as we run off to address the next guideline.

Sub-optimizing the Paging System

There are so many challenges in healthcare needing immediate attention that it can be overwhelming to know where even to start. What frequently happens is that a particular department or division will take it upon themselves to make an improvement that benefits their area, without recognizing the impact that this change has on a broader scale. When an improvement occurs that benefits one area to the detriment of “outside” process participants, this is called sub-optimization. Sub-optimization looks great as a formal process but in the informal process (i.e. the way it plays out in the front line of care) it quickly becomes unmasked. Not only does sub-optimization make the overall process worse but it also creates an immediate work-around. Here’s an example:

A large healthcare network reached a point in its growth where the computer-enabled paging system could no longer accommodate abbreviated callback numbers with only five digits. The decision was made to make a change in the system that would require users to enter the full ten-digit number into the callback field. Advanced warning of the change was made by email and on e-bulletins several weeks in advance to launch so that care providers wouldn’t be caught by surprise. So here’s the formal improvement:

1. Care provider goes onto the computer to page someone
2. On the paging screen, they enter the ten digit callback number
3. The person paged calls back using the ten-digit callback number
4. Communication as before!

Easy, straightforward and user-friendly, right?

Here’s how the launch went: When the new system was activated, frontline care providers quickly realized that they could no longer enter callback numbers to a significant number of phones because they didn’t have ten digit numbers! It turned out that in at least one of the network hospitals about one third of the phones are on internal phone lines with only five digits. Further, many of these phones are the ones that are next to the computers from which pages are sent and many are located in patient care areas where emergency contact may be critical.

Fortunately, frontline care providers are accustomed to “improvements” made without their input and quickly devised a work-around (informal process). The paging system is alpha-text capable, so almost immediately people started to write the five digit number in the text box, leaving the “improved” number entry blank. Of course, now everyone is largely ignoring the “improved” number entry site for all of the numbers, and they continue to abbreviate all numbers into five digits as before.

Sub-optimization occurred, at least in part, because the changes to the paging process were made without vital input from frontline caregivers who use the system every day. Without their input, the paging process was not visible in its entirety and the correction made was only from the vantage point of the IS/Communications department. The new and “improved” system is now worse than it was before. Not only has inconvenience been systematized, but it has also introduced an element of risk into the paging process for patients and staff when critical calls need to be made. Frontline care providers recognized this immediately and made the necessary work-around. Of course this work-around is also sub-optimal because it reinforces the likelihood that the wrong number will be called at another hospital (remember the initial reason for doing this change?)

So in the end, a lot of effort and resources went into what is now a suboptimal change because the right people weren’t at the table. It always takes far more effort to fix sub-optimization - it doesn't take long at all for the work-around to become very comfortable! I wonder if IS/Communications is even aware that a work-around is in place.

Formal vs. Informal Care Process: MRI

We touched on systems that we deal with in our daily lives outside of work. Now let’s look at an example of a formal vs. informal system in the hospital setting. (Reminder: the challenges that we face in healthcare are universal, and the purpose of this forum is not to separate out one institution from the next nor is it to finger point at individuals or departments. There will only be two references to hospitals, the larger hospital (>500 beds) as LGR, and the smaller hospital (<500>)).

Formal System: LGR provides anesthesia coverage for challenging MRI cases. This includes patients with severe anxiety and claustrophobia as well as patients with severe pain or co-morbidities that require extra attention. The formal system is set up as follows:

1. Anesthesia provides MRI coverage one morning per week on Thursdays.
2. The requesting service books the MRI case in advance with Radiology and with Surgical Scheduling so that the patient is officially on the anesthesia schedule
3. The outpatients are seen in the pre-admission test center on a day preceding their scheduled MRI; inpatients are evaluated the night before by the on-call anesthesia team.
4. The patients arrives at MRI about 30 minutes before the scan with complete paperwork
5. The anesthesiologist assigned to MRI reviews the chart, checks the anesthesia equipment set up by the anesthesia tech, and off they go.
6. The MRI cases are completed by 11:30AM and the anesthesiologist then proceeds to interventional radiology to provide similar afternoon anesthesia services.

Here is a description from the vantage point of the anesthesiologist of the system as it actually plays out at LGR. It’ not quite the same!

Informal System:
0730: There is construction going on in one of the MRI bays at LGR – it’s very noisy from the use of Jack Hammers; everyone is yelling to be heard.
0730: The patient arrives, upon chart review it is determined that there is no laboratory data. The MRI scan requires the use of contrast, which requires a baseline creatinine level to be drawn. There is no record of any recent laboratory work in the computer (looking in two IT systems: inpatient and outpatient); the patient had been seen in the pre-admission clinic 2 days earlier and blood had been drawn only for a coagulation profile (PT/PTT/INR). This lab is not necessary for this study.
0740: The patient is approached by the nurse and the radiology tech about the lab issue. The patient informs them that he presented a form with the necessary lab data when he presented to pre-admit clinic. No such sheet in the chart. The patient is now angry that LGR could lose this information. He now will have to have labs sent before we can do the study. He wants to talk with his referring neurosurgeon in person.
0750: The resident and attending anesthesiologist meet with the patient, acknowledge his frustration and suggest that to save him an additional needle stick that they can draw the blood sample through the iv. They also review his history and come up with an anesthetic plan to keep him comfortable during the study.
0810: The nurse contacts one of physician assistants to personally transport the blood sample from the MRI area to the lab after it is drawn. She arrives, apologizes to the patient and informs him that she will go to pre-admit clinic to find out what the problem was with the lab process. The patient says that he would love to go with her and give them a piece of his mind!
0820: The patient is taken out of the waiting area to the MRI patient prep area and placed on a stretcher. The resident starts the iv and draws a blood sample for the PA to take to the lab as arranged.
0830: PA goes to the lab – even though the sample is labeled “super stat” the result will take about an hour to obtain.
0908: The second MRI case is an inpatient scheduled for 0930. The tech asks the attending anesthesiologist (attending) whether she should delay the transport of this patient from the floor to MRI prep given the delay. She informs him that transport is usually arranged the night before for scheduled inpatients. He suggests that they hold off on sending for this patient until 10AM – given the unpredictable nature of transit time this will provide plenty of leeway so that this second patient has arrived by the time they’re ready; transport usually takes at least 30 minutes.
0910: The tech calls transport and is informed that the patient had never been scheduled with them (had they been on time with the first patient this second patient would not have arrived!). The tech asks that the patient be sent for at 10AM. She then calls the patient’s floor to inform the nurse that the patient will be coming down at 10AM.
0914: The attending receives a text page from Angiography to please call. He speaks with the nurse-in-charge, who inquires whether they will be able to start their scheduled case with anesthesia at 1130 as booked. The attending informs him of this unlikelihood given the delay. Fortunately, their schedule is flexible and he asks that they call him when they begin the second case so that he can send for the Angio patient.
0928: The nurse looks online for the necessary lab result for the first patient and it is still not entered. She calls the lab for the result and after a minute of waiting is informed of the lab value: It’s a go.
0930: The patient is taken into the MRI scanner, monitors are placed and sedation is given.
0942: The patient requires more sedation than expected, so while the resident remains with the patient the attending goes to the OR pharmacy to retrieve additional medications (OR pharmacy is at the other end of the building).
0943: The attending encounters an outpatient who is lost and requests assistance in finding the nuclear medicine clinic. He has a sense of where it is but given the poor signage, he offers to take the patient there himself. The clinic is halfway back to MRI. The patient is very thankful.
0945: The attending heads back to the pharmacy. Along the way three elderly family members stop to ask him who the third member of the Three Stooges is: Mo, Curley and? Without hesitation he answers “Larry” – he knows this data cold! They all have a good laugh.
1000: The MRI scan finally begins. The patient is well sedated, however, the pulse oximeter that they rely on to measure the patient’s oxygenation malfunctions repeatedly, resulting in continued scan interruptions, going into the scanner, pulling the patient out of the tube, readjusting the pulse oximeter, sliding the patient back in the scanner, leaving the room. Finally they get an intermittent signal by securing the pulse oximeter to the patient’s finger with a piece of tape.
1001: The second patient arrives to MRI – apparently they sent for the patient immediately disregarding the call requesting a 10AM pick up. This patient will now spend at least an hour on the stretcher waiting for her scan.
1013: The attending goes out to see the second MRI patient. Although the case had been scheduled the day before and should have been seen by the anesthesia call team, there is no pre-procedural assessment, no consent. The nurse informs him that she has the necessary paperwork for him to complete this but that she cannot stamp them up with the patient identification because the patient’s black ID card did not come down with the patient’s chart. She has called the floor and asked them to send it down.
1015: He reviews the chart in the meantime and begin to fill in the pertinent information
1028: The black ID card is hand delivered to him by one of the radiology patient care assistants
1104: The first case is complete, They’re 1.5 hours delayed. The patient will be recovered in Day Surgery and this recovery area requires discharge orders to be completed by the service performing the surgery or procedure. Radiology has never incorporated this requirement into their care process, so the attending asks the nurse to page the neuroradiology fellow to please go to day surgery to complete the paperwork. The patient informs them that he and his wife have traveled by train to get into the city and that they have to catch the outbound train at 12PM. Given that the scan was supposed to be completed at 0900, he thought that they’d have plenty of time. He tells us he’s leaving in 30 minutes whether we like it or not.
1109: I sign out the patient to the recovery room nurse and we realize that the patient’s chart has been left behind in MRI. I offer to go get it.
1141: Next patient in the room for MRI. Pulse oximeter continues to malfunction in the same way, adding approximately 25 minutes of additional time to the study.
1304: Patient taken out of the scanner. In the meantime the attending is down in Angio getting the patient ready for the procedure there. The resident pages him to inform him that the recovery room is full and that they will be on hold in MRI with this second patient if she has to recover there as originally planned. He inquires whether they might bypass the recovery room given that the patient has already met discharge criteria (alertness etc.). Fortunately they are out of the main OR and can exercise this work around option. The attending calls the resident back and inform him that this would be fine.

Think about the impact this informal system has on:
1. Levels of patient and care provider frustration
2. Efficiency and scheduling
3. Safety
4. Cost
5. Other?

This informal process has been in place for years and is so routine that it has become invisible!! Nobody thinks twice about this process – it’s just the way it is. Until we see it, dig into the core causes, and take a frontline system-based approach to correcting the issues, nothing changes; it’s the heroics of the people at the front who keep things running to the best of their abilities.

Systems Are Everywhere

Dear Friends,

What is a system?

Making systems visible: As I mentioned earlier, I had to completely re-learn my concept of what a system or a process is. I always thought of a system as something that was “out there” that I had to access in order to take care of my patients – the computer system, the blood bank, pharmacy, blood bank, the paging system. These are indeed all systems and what I slowly started to appreciate was that systems and processes vary in size and shape and that most importantly I was intimately involved in numerous systems myself. A system or process is a way that we do things in a repetitive manner to accomplish tasks at hand. Whether large or small, what’s amazing is how invisible they become once they are in place. Here’s an example: Every one of us has a specific way that we start our work day, every day, but how many of us recognize this and actually think about what we’re doing? How many times do you hit the snooze button before getting up? What’s the first thing you do when you get into the bathroom? You brush your teeth the same way every morning without thinking about it and if you want to challenge this idea, try brushing with your opposite hand or start the brushing at a different tooth location.

Formal vs. Informal systems: Another fundamental system concept that I had to acquaint myself with is the difference between formal and informal systems. The formal system is how the process is supposed to work; the informal system is how the process really works. The formal system is the one in the written manual, instructions and guidelines; the informal system is the one based on what actually works. It’s very rare that a formal and an informal system are the same. How many of us brush our teeth according to the instructions on the toothbrush box and on the tube of toothpaste, or the way the dentist tells us to? What’s important to note as well is that while the formal process is documented, the informal process usually isn’t – it gets passed on and modified by word of mouth. Two things happen to informal systems or processes as a result: each individual modifies them and they become invisible!

Why are these things important? What I realized is that just about everything I do has some sort of an informal process associated with it that has become invisible; in fact, I can run most of my daily routine on autopilot! This is all well and good if things are working well, but what if they aren’t? If I can no longer see the process because it’s become routine, how am I going to make an improvement that sticks? Chances are I’ll go back to the formal process and focus on making the changes there because it’s documented! But this isn’t going to do a great deal of good is it because the formal system isn’t the underlying process that’s driving the activity – it’s the informal process! The first order of business is to recognize that there is an informal process running the show and that is needs to be made visible in its entirety or nothing will change!

I can imagine what you’re thinking. “What does this have to do with improving healthcare?” We need to start seeing systems and processes in our non-work environment to appreciate the intimate pervasiveness that they have on our daily lives. When we recognize that we perform most of our daily routines on autopilot (invisibly) then we can start to look at what we do on a daily basis at work. So here’s your homework: look at a particular aspect of your daily home routine and write down how you do it – it’ll open your eyes!
Next time we will start looking at healthcare.

With love and respect,
Rick

Learning from Adversity

Dear Friends,
I want to share with you an adverse event that I experienced almost nine years ago, which resulted in a major re-direct in my career and that served to define my approach to change.
On November 18, 1999 I was providing anesthetic care for a 37-year-old woman undergoing a total ankle replacement. I placed a popliteal fossa nerve block preoperatively with Bupivicaine and there was no deviation from the standard of care.
Within moments, the patient experienced a grand mal seizure and progressed to cardiac arrest. After approximately ten minutes of resuscitation the patient remained unresponsive. A fully prepped cardiac operating room was fortuitously available and our patient was rushed into the room, where she underwent a sternotomy for emergent cardiopulmonary bypass. The patient’s cardiac rhythm was restored and after being weaned off of the bypass machine she was taken intubated to the cardiac intensive care unit (ICU).
As is typical during medical emergencies, we were focused on the resuscitation with our emotions on hold. Only after the patient had been stabilized on bypass did the impact of what I had just done begin to sink in. I felt personally responsible for what had happened and compelled to communicate with the family. I thought I would be able to provide a factual account of the event to the husband but to my shock, the husband came at me with full emotional and physical force; fortunately the orthopedic surgeon intercepted him. I was now forced to confront my own emotional distress and I realized my complete lack of training in how to manage this situation. In an instant, the years of clinical training, my board certification and the respect of my colleagues as a competent anesthesiologist had become irrelevant and meaningless. I felt lost and alone.
The following day I was doing cases as though nothing had happened. No one mentioned the event as I performed my clinical duties with numb detachment.
In spite of multiple attempts to speak with the patient, there were three communication barriers: 1.) Risk management’s request that I leave communication with the patient to them; 2.) The ICU team’s desire not to be pulled into the aftermath; and 3.) The husband’s request that I keep my distance. When the patient was discharged home ten days after the event without any opportunity for me to communicate directly, a profound sense of responsibility broke through my fear and compelled me to write the patient a letter of apology with an invitation to open communication if and when she was ready.
Six months after the event, Linda Kenney called me in Seattle. This was Linda’s first opportunity to hear a factual account of the event and when Linda surprisingly offered me forgiveness, I felt an incredible emotional release. I had my life back and I could talk openly about what had happened. We met in person two years after the event, and as Linda described her frustration and anger at my institution’s refusal to communicate with her about the event, two things stood out for me: 1). That healthcare had strayed from compassionate care in a profound way; and 2). That the wall of silence that was there to protect against lawsuits was in fact a major contributor to them. I began to recognize that in addition to a gaping hole in the emotional support for patient’s and families that there was also very little support available for caregivers following adverse events. It was time to do something about this.
What followed was my commitment to be a voice for doing the right thing and to take courageous action in transforming healthcare, however remote the odds of success might appear. I supported Linda with the founding of Medically Induced Trauma Support Services (MITSS, see www.mitss.org), an organization whose mission is to support healing and to restore hope to patients, families and clinicians impacted by adverse medical events. At the same time, I took action within my hospital to create the Peer Support Service, a service that utilizes trains caregivers to provide emotional “first-aid” to colleagues involved in adverse medical events. Initially the climate was anything but receptive to these ideas, however, with commitment and perseverance what resulted was a gradual and steady recognition that the patient safety movement had to incorporate support into its improvement initiatives if it was to achieve sustainable change.

“They didn’t know it was impossible so they went ahead and did it” – Mark Twain

I have had the fortunate opportunity to do presentations on Peer Support at a national and international level over the past several years. These presentations have provided me with continued exposure to the latest developments in patient safety and quality. I have observed the increasing complexity with which the challenges are being addressed and the myriads of methodologies that now exist to solve our safety and quality woes. While there have been successes, what is now becoming apparent is a frustrating absence of sustained change and improvement. Things really aren’t getting a whole lot better and I started to become acutely aware that much of what hospitals are advertising as major improvements in presentations and publications is not visible at the level of patient care. Why?
I started to develop an interest in systems and process improvement and scrutinized improvement initiatives that were ongoing at my institution as well as others. I knew that there was some important connection between the work that I was doing with MITSS and in Peer Support and systems improvement but the link remained elusive for quite some time. Out of synchronous circumstances I happened to take a 10-day training offered by David Dibble titled “The New Agreements in the Workplace.” The core concepts for achieving sustainable change were simple and comprised of four things:
1. Find your purpose
2. Love, grow and serve your people
3. Be a systems-thinker, and
4. Practice a little every day.
I progressed through the initial few days of the training thinking that I already knew the curriculum, and I was happily validating my foundation of expertise. However, it started to dawn on me that I was anything but expert and that a lot of what I thought I knew was based on a wonderful set of incorrect assumptions. To my great surprise, I discovered that I was systems illiterate! What I was viewing as systems (IT systems, pharmacy, infection control, medical records etc.) were in fact macro systems far beyond the level at which change initiatives had to occur. I chose to participate in the remainder of the training with an open mind and as I continued to learn I suddenly saw the connection between peer support and systems thinking as critical elements to the transformation of healthcare.
We spend an enormous amount of time and effort in healthcare training our people in a fragmented manner about team building or process improvement. On the team training side, methodologies such as Good to Great or Crucial Conversations are sold as the magic bullets to safety or quality improvement. On the process side, we are introduced with great zeal and zest to complex methodologies such as Six Sigma, Lean Sigma, Kaizen and TQM. While all of these approaches are wonderful toolsets in the abstract, they produce very little sustained success, much to the frustration of their creators, to the consultants and to the client organizations. When we look at how they’re being implemented and by whom, it becomes very apparent why this is so.
There exists a universal systems principle that states that everything is connected. We have become so good at fragmenting healthcare into sub-subspecialties that we have in large part forgotten this key principle when we take care of patients. Everyone thinks that his or her focus is most important and the result is that nobody sees any connection in what they do to anything else, least of all the patient. This attitude transfers to the majority of improvement initiatives currently underway. It’s either team building and communication or process improvement. Then we take quality improvement away from the people doing the work at the frontline and relegate quality improvement to the “experts,” our quality departments and our patient safety offices. The simple reality is that it takes a compassionate and supportive work environment to create a foundation for teamwork, a methodology for applying systems improvement that is simple and accessible, and quality improvement has to be returned to the frontline people with the support of leadership. This is my purpose.
Chances are that most hospitals, consultants and process improvement gurus will posit that they incorporate all of these elements into their programs. What I believe will become uncomfortably clear as this blog develops is that the vast majority of quality improvement is top-down driven, largely disregarding the importance of frontline input, and that we are skimming the surface of the quality and safety challenge rather than going down into the infrastructure that is unglamorous and dysfunctional to the core. Healthcare is held together by the everyday heroic work arounds of the caregivers, who do everything in their creative power to see that patients get the care that they deserve. Everything is connected folks and until we start addressing the broken platform that supports the shiny, high tech side of care, we are going to continue on our stagnant path.
This blog is going to educate with stories from the frontline of care. This will not be a gripe site but rather one that shares the truth with compassion and with the intent of improving healthcare. This is not about finger pointing or about making any institution right or wrong – every hospital in America has the same challenges and what we need is the courage and the commitment to recognize this and to step up to do the right thing. Off we go!
With love and respect,
Rick