Every day, an infusion center juggles its scarce resources, namely its infusion chairs, nurses and other staff, and it’s pharmacy resources. When managers and schedulers get it right, the day flows relatively smoothly and patient care is maximized. However, when unexpected variances occur in cycle time – the time patients are seated in infusion chairs – problems can get quickly out of hand.
When an appointment runs longer than expected and scheduled, it results in a chair being unavailable for the next patient, and requires a longer than necessary wait time that negatively impacts patient satisfaction. Repeated over several occurrences each day, it makes for long days for nurses, increased overtime costs and, in time, employee retention issues.
When an appointment runs shorter than expected and scheduled, it illustrates gaping holes in an infusion center’s efficiencies. Occurring just a few instances each day, they leave underutilized resources that could have been used to attend to more patients and their medical needs.
At the heart of scheduling challenges for an infusion center is chair utilization data, and leveraging insights from that data empowers schedulers and managers to maximize their resources and create a productive, satisfying environment for patients and staff alike.
Cycle time
Chair utilization data revolves around cycle times. Since processes and systems differ across infusion centers, for the purpose of this article, cycle time is defined as the time elapsed from when the patient sits down in an infusion chair to the completion of the last infusion and patient check out.
Each patient visit will produce a single outcome with respect to chair utilization. The appointment cycle time either fell within the expected range, went longer than the expected range, or went shorter than the expected range.
Measuring chair utilization
Measuring chair utilization is fairly easy to do, and several automated solutions allow for both the collection of data and data visualization reporting tools.
In the graph below, from LeanTaaS’ iQueue for Infusion Centers product suite, you can view a 31-chair infusion center’s chair utilization graph for a single day. The X-axis represents the time of day in 10-minute increments, while the Y-axis represents the total number of chairs in use at any given time. The blue line shows the center’s schedule for the day, and can be compared to the pink line, which illustrates the actual chair utilization for the day.
On this particular day, the infusion center had 39 percent of its appointments run longer than expected. From 7:00 am to about 11:00 am, the two lines, planned appointments and actual appointments, ran close to one another. However, from about 11:00 am onwards, you can determine appointments were running longer than expected and patients were remaining seated in chairs for longer than scheduled. In fact, at 12:40 pm, the center almost ran out of available chairs.
Fortunately, this infusion center was scheduled lower than full capacity for the day. By 3:00 pm though, 25 patients were in chairs when the center – and its staff – had expected to have only 13. Scenarios like this day, for this center, lead to nurses delaying lunches and other breaks, and often result in extended workdays and overtime.
The graph below, however, illustrates quite the opposite problem, a situation where cycle times are running shorter than expected. On this particular day, 20 percent of the center’s appointments ran shorter than expected.
In the morning, actual cycle times were matching up well with scheduled appointments. Then, at noon, results began to shift and it became evident appointments were finishing earlier than expected. The center overestimated the duration of its appointments, and the result was an underutilization of its resources which they could have used to fit in more patient appointments.
Diagnosing the root cause(s)
Just as medical professionals strive to treat the underlying condition rather than its symptoms, infusion center managers and schedulers must do the same. There are many reasons why cycle times might not match expected duration, and it’s often helpful to split root causes into four categories:
- Unexpected reasons appointments run long
- Systemic reasons appointments run long
- Unexpected reasons appointments run short
- Systemic reasons appointments run short
Unexpected reasons an appointment might run long include adverse reactions, difficult access, long turnaround times for medications, add-on orders for blood cells or hydration and more. Any one or more of them can have a significant impact on cycle times.
Systematic reasons cycle times might run long include the underestimation of prep time, and accounting for intake, access, assessment, and drug prep time. Additionally, another factor might be treatments categorized incorrectly, setting the infusion center up for a negative result before the appointment even begins.
Unexpected reasons appointments run short include such adverse reactions that render a patient too ill to continue the remainder of a multi-regimen treatment, as well as a change in treatment plans after a clinic consultation.
Finally, systematic reasons appointments run short often include similar and opposite systematic reasons they run long – prep times overestimated rather than underestimated, and incorrectly categorized treatments.
Diving deeper into the data
Deeper root cause analysis requires classifying appointments as having 1) a cycle time falling within the expected range, 2) a cycle time extending longer than the expected range, or 3) a cycle time shorter than the expected range. After those data sets have been collated, they then need to be analyzed over a variety of different factors, including treatment length, time of day, day of week, regimen provider, patient profile variables and so many more variables that might be positively correlated to cycle time variability. As analyses go deeper and deeper, identifying patterns and trends become more and more valuable.
In the graph below, again from iQueue for Infusion Centers, a data visualization has been created that groups an infusion center’s expected treatment duration into six categories – one to two hours, two to three hours, and so on up to appointments expected to run more than five hours. For each duration group, the X-axis represents the actual duration of the appointment and the Y-axis shows the number appointments that lasted that long for the given time period, which in this case is three weeks. The light blue segment reflects appointments that ran short, the medium blue reflects appointments that ran within range, and the darkest blue represents appointments that ran long.
Scanning the data visualization, it’s evident that 54 percent of the scheduled two-hour appointments during the time period ran long. With an insight like this, the center can re-examine all of its appointment types scheduled for two hours and determine if there is a subset of appointments that consistently run long. If so, the center can then evaluate whether scheduling that subset of appointments for three hours would more positively affect chair utilization and infusion center effectiveness.
Similar analyses can be conducted for other variables, including time of day. In the graph, below, the different shades of blue indicate cycle time completion; light blue represents appointments that ran short, medium blue represents appointments that fell within range, and dark blue indicates appointments that ran long.
The graph shows a high degree of variability, both long and short, at the beginning of the day, and the 9:00 am time slot shows 35 percent of those appointments ran long. With this information, managers can begin to investigate more thoroughly their specific guidelines around scheduling particular treatment types or patients during early morning hours.
You’ve read through the analysis of two variables, and diving deep on any one particular variable delivers valuable insights. However, true actionable insights are developed when an infusion center can programatically dive deep across the entire spectrum of possible variables. Doing so uncovers opportunities to improve the entire systemic process and not just one component or subsystem.
Leveraging predictive analytics
As introduced, there are a multitude of ways to examine cycle times to develop some of the root causes of variability. However, the variety of factors and variables demonstrate the difficulty of conducting analysis through visual inspection alone, especially in uncovering hidden data trends that are driving variability in cycle time.
Analytics solutions like iQueue for Infusion Centers leverage machine learning to better predict an infusion center’s cycle times by taking into account every one of the factors covered above and more, including those data gems that are so often included in free-text notes fields on a patient’s record. To learn more, watch the on-demand webinar “ Data Driven Management of Infusion Centers”.
Utilizing predictive analytics solutions to accurately determine estimated cycle times unlocks more precise strategies for chair utilization, and that ultimately empowers the center to deliver more efficient and more effective care for its patients. For more information about how to maximize the allocation of your center’s resources by better understanding your patients’ cycle times, please watch the webinar on demand, “The Long & the Short of It: Understanding the Time Your Patients Spend in the Infusion Center.”
