Proton Therapy Center KPI Analysis Exposes Where Time Is Lost Across the Treatment Pathway
Operational pressure in radiation oncology rarely comes from one dramatic failure. It accumulates in small losses: a planning handoff delayed by a day, a beam pause that eats into capacity, an appointment that starts a few minutes late, then another, then another. The KPI framework used at FHCC-PT turns those small losses into measurable clinical operations data, separating treatment planning time, beam uptime, schedule variance, slot utilization, and waiting-for-beam performance across calendar year 2024.
The center’s workflow map starts before a patient ever reaches the treatment room. CT simulation triggers a chain that moves through dosimetry, physician review, physics checks, and milling before a plan becomes a scheduled treatment. In practice, this means the bottleneck may sit in any one of several distinct layers rather than in the linear idea of “planning time” alone. By breaking the interval into component parts, the analysis captures where human review, technical verification, and fabrication steps consume business days before therapy can begin.
Planning Time Is Split Across Physicians, Dosimetry, Physics, and Milling
The treatment-planning figure shows a workflow designed for attribution rather than abstraction. Each interval is assigned to a specific operational domain so that planning delay can be traced to physician time, dosimetry time, physics time, or milling time. This is exactly how a modern radiation service should be audited: a late start is not simply a late start, but the visible endpoint of multiple upstream dependencies.
In calendar year 2024, planning time was presented both by month and by disease site, with the annual aggregate shown as a composite of those same components. Median total planning time and interquartile range were displayed alongside mean component times, giving a more realistic picture of throughput than a single average ever could. When the median and spread are both visible, the hidden cost of outlier cases becomes harder to ignore. Complex cases demand more physics review, more physician iteration, and sometimes more fabrication time, and those demands ripple outward into scheduling capacity.
The bedside consequence is straightforward. Every additional business day spent in planning is a day the patient remains between diagnosis and treatment, carrying both anxiety and potential clinical risk. In high-throughput radiation services, workflow latency is not administrative clutter; it is part of the treatment interval itself.
Beam Uptime, Pauses, and Lost Time Define Delivery Capacity
Delivery performance tells a different part of the story. Monthly beam metrics for 2024 tracked uptime by treatment room and at the aggregate site level, alongside the number of delivered treatment fractions. The same figure also quantified the total number of beam pauses each month and the cumulative time lost to those pauses.
Beam uptime is the operational heartbeat of a proton or radiation treatment unit. A system can be well staffed and still underperform if machine interruptions repeatedly fragment the day. Even brief pauses matter because they propagate through the schedule, compressing later slots and increasing the chance that a treatment runs beyond its planned window. In a setting where many patients are seen daily, minutes lost to pauses are not isolated technical nuisances; they become a recurring tax on the entire service line.
One sentence captures the operational reality plainly: beam pauses translate directly into lost treatment capacity.
The figure’s design makes the cumulative burden visible by plotting lost time downward, a useful visual reminder that every interruption subtracts from the available day. Monthly variability in uptime and pause frequency also suggests that operational stability is not uniform across the year. The service line therefore has to manage not only volume, but resilience.
Appointment Variance Shows How Delays Spread Across Treatment Rooms
Appointment schedule variance adds another layer of detail by showing how far actual appointment times drifted from schedule, month by month and by treatment room. The distribution was also categorized by delay intervals, which makes the pattern clinically relevant rather than merely statistical. A room with repeated late starts signals a local workflow issue, whether rooted in staffing, prior-slot overruns, machine turnover, or upstream readiness problems.
This kind of metric matters because schedule variance is not absorbed silently. Once a treatment day begins to slip, the next patient waits longer, the next team handoff becomes more rushed, and the rest of the session list inherits the delay. In a radiation oncology unit, punctuality is not cosmetic. It determines whether the day remains controllable or becomes progressively harder to recover.
The figure also ties variance to the number of treatment fractions, which helps distinguish a busy month from a disorganized one. High volume alone does not explain delay. Delay becomes a separate operational phenotype when it persists beyond what volume would reasonably predict.
Slot Utilization and Waiting-for-Beam Metrics Capture the Patient Experience of Delay
Slot utilization pushes the analysis from the machine to the appointment itself. The center measured the proportion of fractions that under-ran or over-ran the scheduled slot, then paired that with mean and 90th percentile durations for both under-run and over-run events. Waiting-for-beam performance was tracked by month and by clinical hours, again using both mean and P90 values.
The P90 measure is especially informative because it shows what happens at the long end of the distribution. Mean values can look acceptable while a subset of patients still experience large delays. In radiation treatment, those long-tail delays are the ones most likely to disrupt patient flow, extend chair time, and create a chain reaction for subsequent appointments. A service line focused only on averages can miss the patients and sessions that are most vulnerable to delay.
Waiting-for-beam is the most patient-facing metric in the set. It reflects the time a patient spends ready for therapy while the system prepares the beam, and it captures a real-world experience that is often felt more sharply than it is documented. Long waits are not just inconvenient. They can increase discomfort, strain transport and staffing workflows, and make adherence harder for patients already navigating repeated treatment visits.
Operational KPI Tracking Becomes a Clinical Quality Tool
What makes this KPI framework useful is the way it turns radiation operations into a measurable quality program rather than a vague efficiency effort. Planning intervals, beam uptime, appointment variance, slot utilization, and waiting-for-beam time all describe different failure modes, and each one can be acted on differently. A planning delay may call for revised review pathways. Repeated beam pauses may require engineering attention. Persistent waiting-for-beam delays may point to scheduling design rather than machine performance.
The structure also makes accountability more precise. When a center can separate dosimetry time from physician time and physics time, it can see whether delays are coming from cognitive review, technical verification, or fabrication. When machine uptime is separated from waiting-for-beam time, it can distinguish equipment instability from downstream scheduling inefficiency. That level of granularity is rare in routine clinical reporting, and it is exactly what allows operational medicine to move from intuition to correction.
The remaining challenge is not whether these metrics are useful; it is how consistently they can be sustained, compared across disease sites, and tied to patient throughput without sacrificing quality. Future operational reporting will need to show whether gains in one domain, such as shorter planning time, are preserved when beam interruptions and appointment drift remain in play. In a treatment environment governed by precision, the weakest link in the chain still defines the day.