Reduce Radiology Turnaround Time Without More Hires

What TAT measures, and what it gets measured against

Turnaround time is the span between an order (or an exam completion, depending on where a practice starts the clock) and a signed final report. That parenthetical carries more weight than it looks like it does: two groups quoting the same TAT figure may not be measuring the same interval at all. However they define it, ops teams watch it closely. In RBMA's quality-measures survey, about 67% of practices named TAT as a metric they actively track, ahead of every other measure. RBMA / Radiology Business

The number only means something next to the benchmark for its setting, and those benchmarks sit far apart:

• ED / stat reads: roughly 30–60 minutes from exam completion to preliminary or final report
• Inpatient: roughly 4–8 hours
• Outpatient / routine: under 24 hours

Blend those into one practice-wide average and you get a number that describes nobody. A group clearing stat reads in 40 minutes while routine outpatient studies drift past 30 hours has a specific problem in a specific place, and the roll-up figure is the one thing guaranteed not to show it.

Why the hiring reflex misfires

When TAT slips, someone will say the practice needs more radiologists. The workforce data makes that a shaky first move. U.S. imaging volume grew about 31% between 2017 and 2024 while the radiologist workforce grew about 24%. That gap is real, but average exams read per radiologist per day barely moved across the same years, which is not the shape a straightforward capacity shortage takes. Sort radiologists into quartiles by how much their workload shifted and it comes into focus: the busiest quartile went from 57 to 74 exams a day, while the quartile that got lighter fell from 79 to 54. JACR / ScienceDirect

So the pressure is real but lopsided, and that should change the diagnosis. A genuine headcount shortage drags TAT down more or less evenly. What the data shows instead is particular readers and particular workflows absorbing the friction while per-radiologist reading capacity holds roughly steady. That is the signature of an operations problem, and operations problems are indifferent to headcount: a sixth radiologist reads no faster than the first five when the queue is stalled ahead of all of them.

Where the time goes

Radiologists reading slowly is seldom what is on the clock. The delay accumulates around them, in a short list of places that repeat from practice to practice:

• Priors that aren't there yet. The comparison study lives in another PACS, or was never imported. The radiologist stops mid-worklist to go find it, and every affected study pays that detour.
• Orders with holes in them. A blank clinical-history field leaves the reader placing a call or working from a guess. Quality-improvement audits keep finding this one alive at sites that trained staff specifically to catch it. PMC
• Flat worklists. With no automatic re-prioritization, an urgent exam waits behind routine ones until a human notices and reshuffles by hand.
• Everything that isn't reading. One widely cited time-motion study clocked radiologists at roughly 43.8% of the working day on non-interpretive tasks — protocolling, triage, consults — against 36.4% on interpretation itself. Hours spent there are hours the queue doesn't move. JACR — Vancouver Workload Utilization Study00220-2/abstract)

There is a fifth cause that's easier to miss, because nobody involved is idle while it happens. Slow image transfer, processing bottlenecks, and transcription lag turn up repeatedly among the largest contributors to prolonged TAT, and they sit entirely outside interpretation. The study simply isn't where it needs to be yet.

What to fix first

Start with measurement, because it costs nothing and it aims everything after it. Break TAT out by setting and modality until you can name which segment is dragging. Then work the pre-read path itself, roughly in this order:

1. Confirm priors are retrieved and attached before the study reaches the worklist, so the gap is never discovered mid-read.
2. Check orders for completeness at intake — clinical history, laterality, requested protocol — so nothing needs a callback once it's in the queue.
3. Replace static queues with prioritization that responds to acuity and TAT deadlines rather than arrival order.
4. Move administrative and non-interpretive steps off the radiologist's desk wherever they don't turn on clinical judgment.

None of these require a clinical decision, which is exactly why they can be systematized. They're completeness and sequencing checks that happen to be tedious, high-volume, and easy to skip under pressure.

Where Corena fits

Corena is a radiology operations agent. It works the process around a study and never the diagnosis. Before a study reaches a radiologist, it runs read-ready QC against the gaps above — missing priors, incomplete orders, protocol mismatches — that the literature ties to TAT loss. When it finds one, it recommends a fix and a named person on your team reviews and approves it. Corena does not read, interpret, or diagnose. The scope is deliberately narrow: clear the pre-read delay so time on the worklist goes to reading instead of troubleshooting.

If your TAT problem looks more like a jammed intake process than a shortage of readers, that's worth confirming before the next req gets budgeted. Request a demo to see how Corena's read-ready checks map onto your current worklist.