What a mismatched slot length really costs
If you own the practice, the first question is not whether the afternoon is running to plan, but what the scanner time earns. Depreciation, leasing, service contracts and staff run on regardless of whether an MRI is occupied or standing idle for a quarter of an hour.
The economic core: contribution margin per scanner hour
Section titled “The economic core: contribution margin per scanner hour”A slot length that does not match the actual occupied time hits you from both sides:
- Planned too short: the exam overruns and the backlog moves into the following appointments. By the end of the day staff stay longer, or a slot is dropped.
- Planned too long: scanner time sits unused between exams. The cost of those minutes is incurred anyway, with no revenue against it.
Both can be converted into contribution margin per scanner hour as soon as costs, revenue and the time actually occupied sit side by side. On top of that comes the revenue mix: how budget-bound, extrabudgetary and GOÄ services are distributed across the quarter helps determine what a scanner hour earns.
Why an average is not enough for this
Section titled “Why an average is not enough for this”Exam durations are rarely normally distributed. A few very long exams pull the mean upwards without most cases taking that long. Planning with that mean plans correctly for neither case.
More reliable is the duration as a range – in the form “80 % of exams of this type fall between X and Y”, where X and Y come from your own data. Part of that is how certain the range is: an exam type with few cases in the period under review supports less certain statements than one that occurs daily.
Only with this spread does the calculation become honest. The contribution margin of an exam type is then not a single number but a band – and decisions about slot lengths, opening hours or case mix can be aligned with it rather than with gut feeling.
Where the variation comes from
Section titled “Where the variation comes from”Part of the spread is explainable and therefore plannable:
- Protocol and scanner. The same exam can take different amounts of time on different scanners, yet is often planned with the same slot length.
- Exam-related attributes (pseudonymised). Contrast administration, limited ability to be positioned, claustrophobia, inpatient or outpatient.
- Day dynamics. Late arrivals, no-shows, urgent cases fitted in, and backlogs that accumulate through the day.
Whatever cannot be explained by these remains residual uncertainty. It belongs in the report, not averaged away.
What can be measured
Section titled “What can be measured”The timestamps in the modalities’ DICOM headers record when image acquisitions actually took place. The RIS knows appointments, exam types and status timestamps; billing knows the codes and the payer group. Combined and pseudonymised, they show how long each exam type really takes on each scanner, how much that varies, and what an hour on that scanner earns.
Waiting time: the guardrail, not the goal
Section titled “Waiting time: the guardrail, not the goal”Aligning slot lengths and case mix with economics needs a counterweight on the side of patient care. That is why lead times and waiting times are tracked as a guardrail: not as the primary goal of planning, but as a check that an economic improvement does not come at the expense of patient care.
Delays will never disappear entirely anyway; urgent cases and the unexpected are part of everyday work. But the share that stems from planning itself can be made visible, and the places where reductions could start show up in the same data.
More on cost and revenue per scanner hour, scan durations and waiting times.