skip to Main Content
Increase infusion volume with machine learning, for better infusion center scheduling

How AI technology supports marginal thinking for optimized infusion scheduling

Rick Schlieper

Manager of Product Implementation & Customer Success, iQueue for Infusion Centers

The following article on optimized infusion scheduling was originally published on Becker’s Hospital Review. 

Facing frequent reductions in resources and staffing, combined with the increasing needs of their patient population, infusion centers must grow their patient volumes despite limitations and make the best use of what they do have. However, scheduling challenges quickly arise as the schedule fills in advance, becoming a barrier to clinic growth and patient access. Marrying economic theory with AI-based technology can eliminate this barrier, supporting organizational growth and also a better patient experience.

Long-term infusion scheduling is a blank slate

Scheduling the very first patient on any given day is a relatively simple task. In fact, most infusion centers estimate the first 60-80% of the patients they schedule for the day are easy to fit in. The majority of scheduling decisions do not require overrides, shuffling appointments, or other efforts that likely come to mind when thinking about optimized infusion center scheduling.

Complications arise when scheduling infusion appointments for the near future

Roadblocks do arise and extra effort is required to schedule the final population of patients for the day, which accounts for 20-40% of appointments in total. The schedule is often full a week or two in advance of the desired treatment day. Next-day and same-day add-ons are especially difficult to fit in, even more so considering patients who must also see their provider or get labs during the same visit. Organizational goals of growing appointment volumes may seem unachievable in light of an already full schedule.   

So how can infusion leaders and staff more easily schedule this last set of appointments and grow their volumes as needed? One way is to borrow a concept of economic theory commonly referred to as marginal decision making – and deploy that concept to its full extent, using AI-based technology. 

Defining marginal decision making in economic terms

In economics, focusing a decision on the margin means thinking about the cost or benefit of the very next unit made or worked on, as opposed to the overall average cost. For example, if a person is working on doing 100 sit-ups and measuring the cost of doing them with time, and it takes five minutes to do the 100 sit-ups, then on average the cost is three seconds per sit-up. What likely happens, however, is that the person is able to do the first several sit-ups significantly faster than the last, meaning sit-ups #99 and #100 take longer than #1 and #2. Assuming the final two sit-ups ultimately take five seconds each, then even though the average time cost of each sit-up is three seconds, the marginal cost is five seconds.

Applying marginal decision making to infusion center scheduling

Consider a scheduling team working on scheduling 100 infusion appointments on a given day. This is naturally a process that happens over the course of several weeks, and as mentioned, the first 60-80 appointments to schedule are relatively easy due to the sizable lead time, available chairs, and free space in the center’s scheduling calendar. At this time the center does not yet feel constrained, and most centers end up having a chair utilization profile with gaps of free space within. Then, as more and more appointments are scheduled, the schedule feels increasingly full. While the scheduling team does its best, the level of effort needed to place each additional patient appointment increases. The result is some combination of the following:

  • More and more people-power is dedicated to the situation. There are escalations to leadership, calls made to patients to ask if they can move their appointments, revision of infusion center schedules (plus related provider and lab schedules), and overrides to the schedule, creating situations where the center is scheduled beyond its chair capacity.
  • All patients are not scheduled, because there doesn’t seem to be enough room. Patient lead times are extended, delaying treatment and creating a worse patient experience. Patient appointment volume may also be decanted to other units in the geographical area, resulting in loss of revenue.
  • The center feels unable to grow. It may have demand to increase volume, but it seems unable to accommodate this growth due to ongoing struggles to handle the current volume demands.

While a number of factors can contribute to solving these challenges, marginal decision making can help fill the gaps and improve infusion scheduling through a variety of methods. Infusion leaders, managers, and decision makers will likely find this too intensive to apply manually. A smart scheduling solution like iQueue for Infusion Centers, which takes into account past, present, and future schedule data to optimally support marginal decision making through AI, is an invaluable tool. 

Infusion scheduling pre-iQueue
Utilization profile of sample infusion center using open scheduling without the marginal effort required to schedule future appointments (pre-iQueue)
iQueue for Infusion Centers supports optimized infusion scheduling
Utilization profile of infusion center using pre-calculated optimized scheduling factoring in the need for future anticipated volume (optimized scheduling by iQueue)

Marginal decision making provides scheduling solutions

The right technology can utilize a number of best practices and scheduling tactics to reduce the severity of this scheduling crunch and even allow centers to increase the volume of patients treated. These focus on minimizing problems that will eventually increase the marginal effort of those last several appointments that must be scheduled. 

There are three main approaches scheduling teams can take into consideration when trying to solve these issues:

  • Where possible, focus on protecting peak demand times for patients who truly need them, and more fully utilize less desirable times of the day.
  • “Pre-play” the Tetris game of appointment time combinations to avoid gaps in the schedule, then offer those available appointment times to patients when scheduling.
  • Empower schedulers with tools that show the impact of their scheduling decisions as it relates to the utilization within the infusion center.

Technology like iQueue is designed to help infusion users easily follow the first approach by making the center’s common peak demand times clearly visible. It can implement and hone the latter two approaches by providing access to the center’s unique predictive scheduling information. Infusion centers who have implemented iQueue and pivoted to marginal decision making have achieved proven results including reduced patient scheduling lead times, increased appointment volumes, and fewer scheduling escalations and leadership interventions on scheduling decisions. These outcomes in turn result in improved access for patients, and increased infusion scheduler autonomy, independence, and job satisfaction. 

Read on to learn how leading infusion centers are already using healthcare AI to adopt marginal decision making and achieve these outcomes of optimized scheduling. 

Back To Top