skip to Main Content

How Michigan Medicine unlocked 8% higher volume with 20% fewer chairs during COVID

At our winter 2021 Transform event, Michigan Rogel Cancer Center’s Suzanne Burke, RN, BSN, Clinical Nurse Supervisor, and Barbara Walters, MSA, PMP, Senior Project Manager, University of Michigan Rogel Cancer Center, discussed their successful transformation of infusion operations with LeanTaaS’ Akanksha Shukla, Senior Product Manager, iQueue for Infusion Centers.  Visit the Transform Infusion track page to hear the full talk.

Moderator: It is now my pleasure to welcome today’s speakers. Joining us is Suzanne Burke, who is a Clinical Nurse Supervisor at the University of Michigan Rogel Cancer Center. Suzanne has 20 years of experience as a hematology oncology nurse, caring for patients, including the last 10 years in supervision roles at Michigan Medicine in both cancer research and adult oncology infusion. She’s an active member of the Society of Clinical Research Associates, and the Oncology Nursing Society. In Suzanne’s current role as Clinical Nurse Supervisor for oncology infusion services, she’s responsible for the management and development of 68 nurses and overseeing care delivery operations for four infusion areas consisting of 44 chairs and beds, servicing an average of 130 patients daily. 

We’re also joined today by Barbara Walters, who is the Senior Project Manager at the University of Michigan Rogel Cancer Center. Barbara joined the center in 2016 to manage the university’s participation with the oncology care model through the Center for Medicare and Medicaid Innovation Program. Barbara is also responsible for implementing other quality of care programs and pathologies within the Cancer Center including improving care coordination, access to care, and utilization of Clinical Services. She is a certified Project Management Professional with over 25 years of experience in project implementation. Barbara collaborated with the clinical team and Michigan Medicine and LeanTaaS to implement iQueue for oncology infusion services in 2020. 

Joining us as well, from LeanTaaS, is Senior Product Manager Akanksha Shukla. Suzanne, Barbara, Akanksha, thank you so much for being here today. We are really looking forward to this presentation. So with that, I’ll turn the floor over to Akanksha to get started.

Shukla: Thank you again for the introduction. And thank you to all our audience for joining us today. So we will start our conversation with an overview of LeanTaaS in Michigan Medicine, then understand some of the common challenges faced at infusion centers, how that impacts the number of patients that are seen in the center and then dive into unlocking your center’s capacity. 

So LeanTaaS was founded with the aim to use advanced math to solve complex operational problems. Today LeanTaaS has three core products that are commercial on the market. We have iQueue for Operating Rooms, iQueue for Inpatient Beds, and iQueue for Infusion Centers.

iQueue for Infusion Centers is our original product. It’s been around the longest, and it currently powers over 9700 infusion chairs across the country at 480 different infusion centers belonging to 95 health systems. This represents a little over 15% of the infusion capacity in the United States that is currently being optimized by the iQueue for Infusion Centers product. The health systems we work with represent 13 of the top 20 and we are in 42 states, only eight to go.  

I’ll pass the baton to Barbara, who we have here on the call from Michigan Medicine. 

Walters: Thank you Akanksha, I’m happy to be here with all of you today. Just wanted to give you a little bit of our facts and figures from Michigan Medicine. We’re one of the largest healthcare complexes in Michigan in an academic medical center. Our research and patient care have rated us among the best in the nation in a broad range of adult and pediatric specialties. According to US News and World Report, we are ranked Number 11 in the nation and Number One in Michigan. Our health system has three hospitals and 125 clinics. We have over 1,000 licensed beds and see about 2.6 million outpatient clinic visits annually. Also very proud that we are part of the ANCC magnet recognition program. 

As far as our cancer center, the Rogel Cancer Center has 27 disease specialized oncology clinics, and over 300 faculty members providing over 140,000 outpatient visits annually. Our infusion clinics provide chemotherapy and clinical trials for over 80,000 patients annually. We’ve earned national recognition from oncology quality organizations, including National Comprehensive Cancer Network. We’re designated as a comprehensive cancer center. As with Michigan Medicine, the Rogel Cancer Center is consistently ranked as one of the best national cancer centers by US News and World Report.

Shukla: Perfect, thank you for that, Barb. So the implementation of iQueue for the Rogel Cancer Center started at a very interesting time, it was right at the cusp of the COVID-19 pandemic in early 2020. The Rogel team really wanted to mitigate the operational and the financial impact of COVID-19, which the iQueue templates addressed by unlocking capacity within their existing resources. So I would love to have Suzanne speak more about that.

Burke: I agree, thank you. The Rogel Cancer Center adult infusion was experiencing a variety of issues prior to COVID and the COVID-19 pandemic pandemic only cause additional operational challenges. The infusion area had excessive overbooking during the especially highly desirable 10am to 2pm timeframe, which in turn caused lengthy patient wait time, because our pharmacy does struggle to keep up with the demand. And despite having long hours of operation, open from 7:30 until 8:30pm, we’ve had extremely underutilized morning and afternoon hours, despite having adequate nursing staffing during these times. 

As the COVID pandemic emerged, the need for social distancing in our infusion areas also caused us to lose very valuable treatment chairs. In addition, in June of 2020, we had to close one of our infusion units due to a preplanned renovation project. So we definitely had some pain points and challenges. 

Shukla: Thank you, Suzanne, for sharing that. So some of the Michigan Medicine challenges may sound familiar to some of the audience members here. Let’s take a step back and understand why a lot of infusion centers actually experienced similar challenges.

Alright, so if anybody has ever played Tetris before, then you are familiar with the fundamentals of the game, which is to fit blocks of various shapes and length into a fixed space without any wasted space in between them. So the complexity of the game is that you have to make a decision on where to place the block without knowing the full set of appointments that will end up in the space. Once a block is set, they cannot be moved. And there is no information on the next block that will need to go within that space. 

So you’re making the placement decision for every block, one block at a time, and with only the information that is immediately visible to you. This usually results in the blocks stacking up slowly, one on top of each other, to create a peak and leaving a lot of empty space in between them. And now if this sounds familiar to you outside of Tetris, it’s because placing appointments in an open schedule is essentially the same activity. 

This is a very complex activity in itself. And this is what your schedulers are doing mentally every time they schedule an appointment, all while responding to patient and provider requests, answering phone calls, and managing a whole bunch of other tasks. As you can imagine, this is not an effective way of scheduling. iQueue for Infusion Centers solves this problem more effectively by using optimization, predictive analytics, machine learning, and other mathematical tools to generate a scheduling template. This template predicts the volume of appointments for each day of the week with a high degree of precision, and also predicts the mix with a high degree of precision. So in other words, it predicts if 20 of your appointments on a typical Monday are going to be an hour long, 15 are going to be two, etc. Then it factors in the reality that scheduled durations may not be perfect. So it looks at the previous 1000 times someone was scheduled for a three hour appointment to learn the actual duration ranges possible for a three hour appointment at your infusion center. 

This determines whether the expected duration at time of scheduling is likely to be accurate or not accurate. After factoring in that error, it can then start to arrange the appointments. So now we’re back to the Tetris. It has factored in the volume, the mix, the accuracy of the mix, and expanded your blocks to the right length. It can then figure out the perfect way to lay the blocks out so that it matches your chair availability and the nurse workflow, giving you a more level loaded template that you see on the right. 

So what does this optimized template do for you? Your schedulers can focus on scheduling, with more plug and play templates built directly into your scheduling tool. Second, it unlocks capacity. All the white space you saw in the previous template is now filled with a patient. And right at the peak where you need capacity the most, you’ve got a few chairs available as a buffer. This helps you address patients who show up late or spend more time in-chair than expected, etc. The patients also have choices, notice appointments of different durations are spread throughout the day. It flattens the nurse workload for a more steady flow. And finally, it matches your nurses’ schedules. So nurses that show up first, leave first and the nurses that are on the late shift leave on time.

But what if you are already scheduling really, really well. And maybe you already have a level loaded day? What happens then? Especially if even despite being a level loaded day, you find yourself experiencing when you have patients that need to be seen with short notice, you don’t have availability for them. Your schedules look full, but the center ends up with empty chairs during the day. Or maybe you want to grow volumes, but the days are completely booked. So let’s talk a little bit more about how you can increase your utilization to solve some of these issues. 

Now an intuitive solution to increasing utilization, may be to expand the available resources. So you add more chairs, maybe add more physical space or nurses. As you well know, this process can be costly, time consuming, and space is not always readily available, especially if you are in a bigger city. Another way to increase utilization is to schedule strategically, to unlock existing capacity. Doing this increases appointment volumes without adding physical space, nursing etc. Instead it is achieved by fully utilizing the existing resources within your center. So this approach brings together the true capacity of your infusion center and some machine learning techniques. So let’s go ahead and dive into that.

Let’s begin by understanding the true capacity of an infusion center. We can use one of the Rogel units as an example. First, let’s define the capacity. Here capacity is the available patient hours for treatment. The simplistic way to calculate capacity is to multiply the total number of chairs with a total number of operating hours in a day.

For the single Rogel unit, let’s say we have 32 chairs and it operates for 12.5 hours, which will give us about 400 total patient hours. Now this 400 total patient hours is a theoretical upper limit, since to achieve it we’ll have to have 32 patients in the treatment chair from the minute the center opens to the minute it closes, which is simply unrealistic. To correct for that, we consider operational factors like nurse staffing, pharmacy prep time, lab turnaround times, etc. With all these operational limiting factors, the center will have a ramp up in the morning and a ramp down in the evening. The trapezoids in the graph reflect that. For our example of the global unit, if the ramp up and ramp down is about two hours each, which can be typical for an infusion center, then our true capacity will be 312 patient hours, which is slightly reduced from the calculation. So now that we have an understanding of our center’s basic capacity, we can incorporate machine learning and predictive analytics to account for the other pieces.

So I’m sure you all agree that despite our best efforts, the operations of an infusion center will have unpredictability, and that can cause it to run very differently than what you had initially planned. One of these unpredictable factors tends to be same day changes to the scheduled appointment volumes, which can open up chairs during times when the center was scheduled to run out of chairs, or very quickly fill in previously available chairs. And it’s usually caused by patients who don’t make it to or need to cancel their appointments for various reasons. That can be not passing their labs or not feeling well or something completely unrelated, or patients or treatments that need to be added on to the schedule at the last minute. In either of these cases, there’s very little advance notice. And the scheduler would not be able to accurately or precisely predict that at time of scheduling to be able to plan around it. iqueue is able to analyze the center’s historical data, learn and predict the add on and no shows and same day cancel patterns, for a center for a day. 

Another factor in this unpredictability are patients who show up for their appointments, not at their scheduled time. This can cause a treatment to run late, creating a domino effect or take away a chair from a patient who was actually scheduled to use it at that time. While the infusion team may know from multiple experiences that specific patients for example, Mr. Smith, will likely show up late for his appointment, since he’s coming from a clinic visit with a provider who tends to run late or that another patient, let’s say Mrs. Jones will arrive exactly 10 minutes early. They cannot predict how many other patients of their total demographics will show up early, late or on time with hyper-precision. So patients showing up late is obviously impactful for your operations. 

But it is interesting to also consider that, and maybe not as intuitive, patients showing up early has a negative impact on your infusion operations as well. Perhaps there are some patients who think “hey, if I have to show up early, they’ll be able to squeeze me in,” but that behavior can often strain the pharmacy, and there may not be a chair available or a nurse available for that patient when they arrive, resulting in a wait time for that patient. 

That early patient may sometimes take somebody else’s scheduled spot and cause the other patient to have a wait time. So analyzing that historical arrival pattern allows iQueue to learn and predict when a chair is scheduled to be used versus when it will actually be used.

Our third unpredictability factor is the actual length of the treatment cycle, which can cause a chair or nurse to be in use or taken for longer than planned, causing downstream delays or nurses or a chair to be available much sooner than planned to take the next treatment early. Cycles can be predictably longer because of known reactions or needs and then we know some can be shorter. However, most treatments the length can vary based on various factors like cycle schedule, patient health, and other variables that are difficult to predict. So regardless of known variances in treatment lengths, we also know that schedulers use a set expected duration for visits at time of scheduling to follow a very standardized workflow, that may also set up an accurate expectation of the duration of that appointment. So iQueue looks at and analyzes the likelihood of an appointment deviating from that schedule length and the variance in that deviation to recommend next appointment start that align most closely with a true chair availability. 

Alright, so the key to all these unpredictability factors is to figure out what the end result impact is. So knowing these different rates of occurrences, and the degree to which they impact operations is key to developing a machine learning model that maximizes utilization and unlocks previously unused capacity. For the Rogel Cancer Center, the iQueue template generation algorithm marries the center’s true capacity with the unpredictability machine learning outcomes to generate strategic template designs that actually push the chair utilization curves higher than realistic maximum that center can support, shown by the green line on the chart. We then run predictive models against these template designs to see the forecasted number of patients in chair at any given time throughout the day, which is shown in that blue line on the chart. So what is the result of this design and strategy? 

On this very specific day that you’re looking at, we can see that the iQueue template recommends scheduling more patients that need to be seated in chair, then there are chairs available for most of its key treatment hours. This is especially true during the late morning, early afternoon hours, when the machine learning forecasts more appointments being taken off the schedule then being added to the schedule on the date of treatment. The net loss of scheduled appointments during those hours will allow the center to maximize their resources without running out of chairs. The machine learning also forecasts a slight left shift in the number of patients in chair by the afternoon, which is being caused by the higher early to on time patient arrival ratios for that time of day. 

So the center is able to be at their maximum chair capacity, which is 43 chairs, quite a few times during the day or hover very close to it in the low 40s or high 30s during the treatment hours. So overall with this design, the template pushed the scheduled appointment volume to 150 anticipating an 8% shrink, so that is coming from those no shows same day add ons and cancels to complete a theoretical 130 appointments. And if you look closely, there’s still some chair buffer during those three key treating hours, which means that there is the continued possibility for the center to add on more volume for their future appointments. 

An important note here about this design and strategy, while we’ve used a 43 chair center for example today, this same data science and modeling principle can actually be applied to any sized center larger than this and much, much smaller. And now that we understand the methodology of strategically unlocking capacity with your existing infusion resources, I will pass the ball back to the Michigan team to speak about their experience of and the ROI from using this predictive template design. 

Burke: We’ve been able to meet our volume based despite having fewer chairs. Nurses are working more consistently throughout the day. With not having as many challenges around having these ebbs and flows where they’re working more at a peak time, as opposed to the ups and downs during the peak hours. So it’s more consistent during the day, we’ve been able to do our work a little bit differently on how we pull our patients back into our unit, which has been a very nice change. Overall we’re using our hours in the morning and in the evening, much more consistently. There’s less low census for nurses going home early, we’re able to have our pharmacy feel very consistent through the day, and we’ve made some real impact to our ability to kind of increase our patient load, despite having fewer chairs. 

Some additional benefits that we’ve had – a LeanTaaS application continues to support us in our scheduling decisions. So it provides diagnostics that show us specific metrics, which allow us to make changes on a particular day. The metrics provide us details about where opportunities exist to schedule patients, or where there may be challenges due the way that we are scheduling if we’ve scheduled too many patients in one area. And so that allows us to work ahead to ensure a smooth schedule through the day. Nurses have more consistent work, like I previously spoke about, and they’re not experiencing those peaks and flows that we were in prior to implementation. And then the LeanTaaS team and the Michigan Medicine team have been able to meet every other week to continue to validate our operational metrics, and to discuss our successes and our challenges that continue. We collaborate consistently, in order to review our ever changing data. And we determine the best ways to optimize our templates. LeanTaaS provides ongoing support to the team throughout any issues that arise. Because of the success that we’ve experienced in the Rogel Cancer Center, we’re looking at expanding iQueue to our other infusion sites. We are working with IQueue to strategically open a few of our closed COVID chairs. And we are looking to schedule changes in real time to IQ so that cuddling decisions can be made a point of care. And we want to look at ways that we can smooth our schedule based off of our provider scheduled clinic days. And iQueue will continue to support us on our ongoing training of our cancer center schedulers. 

Shukla: Thank you, Suzanne. So I quickly want to touch on how iQueue for infusion centers is pretty uniquely positioned to incorporate advanced analytics and assist infusion centers with strategic templates designs. We’ve mostly talked about the Optimize box. But machine learning and optimization isn’t a one time thing. So as Suzanne mentioned, once the strategy and the templates design is in place, it is important for us to have access to forward looking tools to continually plan and know what is coming for approaching days and weeks. And to continually learn from the data and adapt to the changes in the center, the staff scheduling operating hours, etc. So using the sophisticated data science tools and having this type of information at your fingertips can certainly help you and your center operate at its fullest possible capacity.

Moderator: Wonderful, thank you Barbara, Suzanne, and Akanksha, for talking us through this today, it’s fascinating to hear about Rogel Cancer Center’s journey to improve capacity, so I appreciate you sharing your thoughts there. At this point, I’d love to open the floor up to any questions from our audience members. So you can submit any questions you have by typing them into the Q&A chat box you see on your webinar console. And if we don’t get through all the questions today, we’ll be sure to follow up with you directly after the webinar. We’ve already had a lot of great questions rolling in. The first is, how many patients does each nurse treat on a given shift? Eight hours or 10 hours depending on your shift. 

Burke: I would say that each nurse is about to is able to treat somewhere between four and nine patients in a 10 hour shift. Most people work 10 and 12 hour shifts on our unit. We do have some eight hours, but it’s just dependent on the treatment and the duration of the treatment. But anywhere between four and nine, I have some nurses who can treat as many as 13. 

Moderator: Wonderful. A follow up with one quick clarification as well, if you can repeat how many chairs are at the cancer center? 

Burke: Right now we’re operating with 44. We still have two chairs closed for social distancing in one of our units, but we’re strategically looking at how we can open those chairs in the near future. 

Moderator: Great, thank you for clarifying that. The next question is from an audience member who asked how actual wait times and treatment are times calculated. Where’s the data coming from?

Walters: So that’s a lot of the data we’re giving to iQueue or LeanTaaS daily. I know we’ll talk a little bit about the data we present to them before go-live that prepares the scheduling blocks, but we do give them data on a daily basis, we’re working to get that to be real time, but it is a batch job that runs nightly of any changes in our schedule, so any of the cancellations and no shows appear in there. But also every patient interaction is logged in Epic, the EMR that we’re utilizing, so we know when a patient checks in, we know when they’re roomed, or chaired in this case, we know when they’re vitals are taken, we know when the first bag was taken, we know when the last bag was taken, we know when they were discharged. So all of those time stamps help us identify that point of time from when a patient checks to when they’re seated and what their wait times are as well. 

Moderator: Wonderful, thank you for pointing that out. We have another question on the timing of the data as well. This audience member is asking, how long does it take to gather the historical data in building the machine learning model? 

Walters: If you have a data analyst, it doesn’t take much time at all. It’s not something that anyone can pull, because it’s in the weeds of your EMR, but LT does give us a straightforward record layout of what information they need, all the information is de-identified, so we’re not giving any PHI information. We actually just ran a test sample of maybe one month of data prior, made sure we could transfer that to them in a secure platform and that they were able to read it. We had a few issues we worked through on one conference call, and then we had those same specs that we ran it on. I know we were going to run it for 12 months, but then we ran it for 18 months worth of, just so we had a lot of historic data to share. Then what we pulled from historic data is very similar to now, it’s the same record layout we provide the daily feed from. 

Moderator: Thank you for elaborating on that. We have another question about turnover time between chairs. How does this technology account for that turnover time? 

Burke: We use durations for all of our treatments, and our durations include preparation time for patient care. So for this particular platform, we haven’t had any issues with not being able to turn the chairs over, since we build durations based off of having this prep and finish time. 

Moderator: Thank you for clarifying that. Another question, is this volume scalable? Can this translate between small, medium, and large facilities? 

Burke: We’re an interesting space, because we’re multiple infusion areas in different locations, but housed in the same building, so I used the iQueue for a small six chair unit while researching it, as well as my larger units, 32 chairs. I’ve found it works in both of those areas, especially with our nurse staffing. 

Moderator: The next two questions are asking about specific to the cancer center. The first is, do you use medical assistants in your infusion room? 

Burke: Yes, we do. They are responsible for getting the patients and bringing them back. They do a multitude of different things that help with the flow of our unit. 

Moderator: Wonderful, thank you so much. The related question is just what are your pharmacy hours? 

Burke: Our pharmacy hours are 6am to 8pm. And they don’t stay until the last patient has completed treatment, but they do stay until all the meds have been dispensed. 

Moderator: More on the data side, this audience member was asking what data points are being fed into the iQueue system, assuming that scheduling data has to be as close to the real world processes as possible? 

Walters: So that’s one of the issues that we have right now, I wouldn’t say it’s an issue, but what sure would make Suzanne and her team’s life much better is if we had a real time interface for the scheduling. So that’s a EDT SIU interface that we’re working to get, so that as soon as a patient gets scheduled, that’s going to send that. Right now, because we do that batch mode overnight,  Suzanne’s team is not able to see the same day cancellations or add ons in the iQueue system, they can see that in our Epic system. And I saw another question here, does this system integrate. So currently, not yet. But once we get those interfaces live, yes, but it really is a standalone system that complements what our team is using in looking at Epic.

Moderator: Wonderful, thank you for tackling both those questions. I appreciate it. Next question is from an audience member who says it sounded like your early and late appointments were underutilized because of patient preference, which would suggest that patients are now being slotted into appointments they maybe don’t prefer. How do they respond to that change? 

Burke: I wouldn’t say that they’re necessarily slotted into spots that they don’t prefer, I think the way that we’ve established is that we have less people wanting to come in the morning, who optimized our templates around the later morning and in the evening. So I think we really try to schedule our longer appointments to come earlier in the day. The ones that need to be here early because they are going to be here for six to eight hours, we try to put those scheduling blocks earlier in the day, and then have room for the people who do like to come, but knowing that we optimize around the time of the peak times that people like. It just helps to make it a little bit more flat across the top. But it is a challenge. I mean, patients do like to come when they want to come. But we are also trying to provide what’s available, because we will ensure shorter wait times and more throughput. 

Moderator: Wonderful, thank you so much for touching on that scheduling piece. The next question is, do you assign chairs to nurses, so their acuity is based on how the chair scheduling templates are built? 

Burke: We do not assign nurses to specific chairs. Nurses pull patients back off of the schedule, based off what they’re treating. So you know, if they are in the middle of having, you know, a first time Rituxan, they have the option to take the next patient in line who may be a bit of having instead of having the next Rituxan in line they may take the single agent therapy versus a multi regimen while they they move through. So they have a three to one ratio. And they basically pick their patients as their chair space opens and support each other quite well, to make sure that there’s no issues with acuity. 

Moderator: Another question about nurses here as well. How many nurses a day do you staff and do you have assistant staff to support the nursing team?  

Burke: We have assistant staff to support the nursing staff ,about 20 I think almost 30 nurses a day, and we do have a system staff as well that helped. We have about 17-20 medical assistants. But we are in different locations, all our infusion areas in the same building but they may be in a separate location, which changes your staffing a little bit. 

Moderator: That makes sense. Thank you again. The next question is, did you do any work on pre-visit activities for example getting labs a day ahead? 

Burke: We encourage patients to get their labs drawn the day ahead. We know that they will have decreased wait times ust in general, because those are some of the things we cannot control in our infusion area, is how long our labs take to process our labs. So part of our messaging is to encourage patients to go to a local laboratory to get their labs drawn prior to their treatment. Not everyone does that. 

Walters: I have to say that, through all of this, I would say anything that’s because behavior modification changes have been the biggest challenges. during implementation. It’s not been making changes to our schedule templates and epic, it’s not been working with the LeanTaaS team, they’re great. It’s really convincing patients that it’s okay to uncouple your office visit from your infusion visit. Or our staff when they’re scheduling, they’re so used to asking the patient, when would you like to come in? And we’re training them to say, “I have a 10 and a two available, what would you like?” So that shift in behavior changes is really what’s been the most challenging.

Moderator: Great, thank you for adding that. And I think it’s an important point to know. We have a follow up question here as well about the staffing ratios. Is a three to one ratio a standard your facility has come up with? Or is there a best practice guideline that you can share? 

Burke: Ours is actually one where we are benchmarked with other larger academic institutions as well as taking a look internally at our multi-drug regimens and the actual complexity of the treatment that we give in our infusion area. So I do have some best practice guidelines, and I could potentially share those with you if that’s of interest. 

Moderator: Wonderful, thank you so much. We have another question about the clinic as well. In terms of your hours, can you share, what the hours are, and if you are all open seven days a week? 

Burke: So we are in the Rogel Cancer Center. We are Monday through Friday, 7:30 in the morning till 8:30pm. Our Saturday hours are from 7:30 until 6pm. And then we are not open in the Rogel Cancer Center. But we do have infusion operations on Sunday. They’re located on site but ina  different building in the Michigan Medicine complex, and their hours are 7:30 to 4:30. 

Moderator:  One more follow up question as well about the staffing ratio we just mentioned, because that ratio applies to clinical trial infusions as well. 

Burke: No. It depends on the complexity of the clinical trial Phase One and Two complex clinical trials, we have a specialized unit for those. And typically the ratio and those units are one to one or one to two, depending on the complexity of the trial and the amount of nursing procedures or procedures in general that will need to occur and the monitoring. But it just depends on a Phase Three or even a Phase One as a single agent with no procedures may be seen in our general infusion area. And those continue to be a three to one ratio. 

Moderator: We have another question about the historical data piece here, which is what information was required to be uploaded as historical data? 

Walters: So if you think about pretty much every touch point, for a patient regarding infusion, that’s what we uploaded. So date of service, day of the week, the time, what the regimen was, the day it was scheduled. Was it the same day add-on? Was at a couple of visits. And then those timestamps that I mentioned earlier? What time if the patient had a one o’clock appointment, did they check in at 11:30? What time was the first drug given? What time was the bag switch? So it sounds very complicated and it is that’s why if you have somebody who’s an informatics specialist, all of that data is in Epic, or probably any EMR. But all of that kind of helps build these scheduling blocks for us.

Moderator: Got it. That’s helpful to know. We had another question here as well, about the scheduling duration. This audience member was curious about the unit of time currently scheduled for each treatment plan. For example, whether it’s a multi drug regimen or a single? 

Burke: I do, we use an infusion duration tool that was created by, actually by nursing. And it calculates based off of the treatment regimen, the monitoring period and the amount of time for administration of the drugs. So for example, a first time Rituxan would be scheduled for 420 minutes. But a subsequent infusion of Rituxan would be 360 traditional rate or 180 at the rapid rate. And our tool holds all this information and allows the scheduler to search the drug. And then to find that coinciding duration for that specific drug. 

Walters: It’s really critical for our schedulers, because one of the things that was difficult for us to get used to, I think, in the beginning was we used to schedule to a resource, meaning a chair or a bed, and now we schedule to a time duration. So it’s very critical that they’re looking at that scheduling duration tool so that they know this is a three hour duration. This is an eight hour, this is a six hour, if it’s a six hour, we know we can’t schedule after one o’clock, because we’ll run out of time. So it’s really a critical tool. It’s a little cumbersome, but it really works great. And the other valuable information that Suzanne and her nurses have put in there is. We do have offsites. So it indicates to us, to the scheduling team. Does this regimen require a bed? Does it require a private room with a bathroom? Can this be done at an offsite location? So it’s really a great all around tool that they’ve put together for our scheduling team. 

Moderator: Absolutely sounds incredibly helpful. So thank you both for weighing in on that question. Another audience member is asking what percentage of patients who need to see a doctor prior to their infusion visit? 

Burke: I’m going to tell you, 50%, to continue to see their doctor prior to their infusion visit. We were trying to uncouple more appointments so that we can spread it just helps with being able to, you know, smooth your schedule over the week but 50% is typically our mark that we just sit at right now.

Walters: I’ll clarify that. That’s 50% who do but where we sit at right now, is that the need? Probably not. So we still continue to work with our physicians to help change that behavior. But I need my patient to be infusioned the day that I’m in clinic. Because we are a research facility, so many of our physicians are, you know, in labs, and not always in clinic the same day. So we’re still trying to break that habit. There’s room for opportunity.

Moderator: Absolutely appreciate you both are weighing in on that question. The next is from an audience member who is curious if you provide treatments to non oncology patients. 

Burke: We tend the majority of the patients that we treat our cancer center or cancer patients, but we do if they need assistance. So we do have offices that are more dedicated to not oncology. 

Moderator: Thank you. The next question is, what percentage of your appointments are saying to add on? And did that change after bringing in iQueue? 

Burke: Our percentage of same day add-ons is low. I would say that it’s less than 5% of our day to day. Not sure that I’ve seen a dramatic change and same day add on iQueue.

Walter: I will say, Suzanne, I can elaborate on this a little bit. One of the things I became familiar with was working with Suzanne’s routine is that we had what was called a handoff list. And the handoff list was when a regular checkout scheduler could not find a time on the scheduling tool for a patient and so we would put those patients on a handoff list. And then Suzanne and her team would look for places where they could place those patients. That handoff list has dwindled to not very many patients. Maybe a few patients a week. Was that correct, Suzanne? I think that’s a direct impact of iQueue. 

Moderator: We have time for two more questions. The first is, does your pharmacy premix common meds? 

Burke: They will only premix meds for orders that are signed. So they have a very extensive list of what they will make ahead, so there’s a make ahead list, but it will all be very dependent on whether the orders are signed. 

Moderator: Got it. The last question we have, is how long does the implementation process take and how does that process work? Can you walk us through that a little? 

Barbara: So we started working with LeanTaas before our contract was signed a little bit. So our contract was signed in 2020. We had weekly meetings regarding the data and training that was needed for schedulers and this new concept of scheduling to duration rather than a chair or bed. But we went live with scheduling, so this was a hard concept for me to wrap my head around in the beginning. In August of 2020, we started using the new scheduling templates, with the blocks being in resource blocks rather than a chair or a bed, for dates of service in September. So August was really when we went live with this. It was a pretty quick implementation considering. 

Moderator: Suzanne, Barbara, Akanksha, thanks again for your time and thoughtful remarks on Rogel Cancer Center. I’d also like to thank LeanTaaS for sponsoring this webinar. Thank you all so much for joining us and I hope you have a wonderful rest of your day.

Back To Top