The 90-Minute Window Between Your Scan and the Final Report
What happens inside the 60–120 minute window between a scan and the final radiologist sign-off — and why preliminary radiology interpretations exist to close it.
⚠️ Important scope note — please read before continuing
RAD365 provides offshore preliminary reads for human patients in the USA only — we do not provide final reads; your facility's own licensed radiologist completes every final interpretation. For veterinary and pet imaging there are no preliminary reads; veterinary studies are read as complete final reports by specialist veterinary radiology groups.
By Trisha Seal — September 14, 2026. RAD365 has run offshore radiology operations for 12+ years, building PACS-integrated preliminary read workflows inside each client's existing worklist. That experience is specifically in the preliminary and offshore workflow layer and in PACS integration — RAD365 does not sign final human reads, and every final interpretation is completed by the facility's own licensed radiologist.
The scan finishes at 11:42 p.m. The patient is back in the emergency department by 11:48. Somewhere between that moment and the final signed report, a clock is running that almost nobody outside imaging ever sees — and for a STAT study, independent benchmark reviews put the expected length of that clock at 60 to 120 minutes. Preliminary radiology interpretations exist because of what has to happen inside it.
This article walks through that window literally, minute by minute: what the study does, who touches it, where it waits, and what breaks when the roster is thin. It is not an argument for a service. It is a description of a process most clinicians rely on daily and few have ever seen laid out end to end.
The Clock Starts Before Anyone Looks at the Images
The moment the last series reconstructs, the study begins a journey that is mostly invisible from the bedside. It transfers to the archive. Priors are located — or are not. It lands on a worklist, where it takes a position determined by priority flag, modality, location and, at many facilities, by nothing more sophisticated than arrival order. Then it waits for a pair of eyes.
That wait is the part nobody budgets for. At 2 p.m. on a Tuesday, the gap between arrival and first read is small enough to be invisible. At 11:48 p.m. on a Saturday, the same study lands on a worklist being held by one reader who is already mid-dictation on something else, with four more studies behind it. Nothing has malfunctioned. The queue is simply longer than the capacity pointed at it.
Independent turnaround-time benchmark work published by AAG Health sets out the expectations most facilities operate against: STAT and emergency CT and MRI reads are generally expected within 60 to 120 minutes; routine CT and MRI runs from a few hours up to 24 hours; and outpatient non-emergent studies commonly run 24 to 72 hours. Those are not aspirational numbers. They are the working assumptions the ordering clinician has in their head when they decide whether to hold a patient, transfer them, or start treatment empirically.
Why the Window Keeps Closing
The 60 to 120 minute expectation has not changed much. What has changed is how hard it is to hold with in-house staffing alone, and the reason is a workforce arithmetic problem that two well-regarded sources describe with unusual clarity.
Radiologists are leaving faster than they used to
The American College of Radiology's 2026 ACR Bulletin workforce update, drawing on Harvey L. Neiman Health Policy Institute data, reports that radiologists are now leaving practice entirely at over twice the rate they were less than 10 years ago. The update also finds that subspecialty radiologists are 37% more likely to exit the workforce than generalists — which matters disproportionately, because the subspecialised reader is usually the one a department cannot replace within a hiring cycle.
In plain operational terms: the people who hold your nights, weekends and holidays are the hardest to replace, and they are leaving faster than they were a decade ago. A single resignation in a small group does not shrink the roster by a percentage point. It removes an entire shift block.
The supply side is not scheduled to catch up
The Harvey L. Neiman Health Policy Institute projects that the radiologist workforce will grow only 25.7% by 2055 if no new residency positions are added, while imaging utilisation demand is projected to rise 16.9% to 26.9% depending on modality. Read those two figures together and the conclusion is uncomfortable but simple: without workforce intervention, the current gap is on track to persist rather than close. Neiman HPI also projects that post-COVID attrition — up roughly 50% since 2020 — will leave the 2055 workforce about 3,116 radiologists smaller than the pre-pandemic trend line implied.
Residency expansion is the real fix, and it is worth advocating for. But it operates on a decade-plus timescale. The scan that landed at 11:42 tonight needs a different answer.
Expected STAT CT/MRI turnaround (AAG Health benchmark review)
Projected radiologist workforce growth by 2055 (Neiman HPI)
Projected rise in imaging utilisation demand by modality (Neiman HPI)
Rate radiologists now leave practice vs. under 10 years ago (ACR Bulletin 2026)
What Actually Happens in a Preliminary Read
This is the part that is usually described vaguely and shouldn't be. The workflow behind preliminary read services is specific, sequential, and deliberately bounded.
1. The study lands on the existing PACS worklist. No export, no portal, no second archive. Routing rules in the facility's own system decide which studies — by modality, priority flag, time window or location — become visible to the covering team. Technologists change nothing about how they work.
2. An offshore radiologist reads and dictates the preliminary. A named radiologist on the covering team opens the study inside the facility's PACS under a provisioned account, reads it against the clinical question, and dictates a preliminary interpretation that is timestamped, attributed and explicitly labelled non-final.
3. QA checks the report before release. The preliminary passes a quality check for completeness, clarity and protocol compliance before it returns to the record — not a second read, but a check that the document does the job it is supposed to do.
4. Critical findings are relayed per written protocol, with an audit trail. If the read surfaces a finding in an agreed critical category, escalation is by direct human contact along a named path with named fallbacks, inside a defined maximum interval, and the contact is logged. Leaving a result on a worklist is not escalation.
5. The facility's own licensed radiologist reviews and signs the final. On the next normal shift, the facility's radiologist opens the study, reads it in full with priors, and issues the signed final report. They are not countersigning someone else's work — the preliminary is context, not a draft. That final report is the interpretation of record and the one that is billed.
Step five is not a formality and it is not negotiable. RAD365 does not issue, sign or bill final interpretations for human patients under any arrangement. Where the wider question is about departmental capacity rather than the overnight window specifically, that conversation belongs with human radiology services.
Preliminary Read vs. Final Read
| Dimension | Preliminary read | Final read |
|---|---|---|
| Who performs it | A named radiologist on the covering team, working inside your PACS under a provisioned account | Your facility's own licensed radiologist |
| Billing status | Not separately billable to Medicare or insurers | The billable professional interpretation |
| Legal status | Documented, timestamped, explicitly non-final; context in the record | The diagnosis of record; carries medico-legal responsibility |
| Purpose | Answer the immediate clinical question inside the 60–120 minute window | Complete diagnostic interpretation with priors and full impression |
The two rows that matter most in a medical staff committee are billing and legal status, because together they explain why a preliminary-only arrangement changes nothing about who owns the diagnosis. Departments that instrument the relationship between the two reads — logging every discrepancy by modality and severity and routing it through their existing peer review and QA programme — end up with one of the more useful quality data sets they have, because it is one of the few places where two independent reads of the same study are compared as a matter of routine.
What the Window Looks Like When It Works
Back to 11:42 p.m. In a facility with a preliminary layer in place, the study reconstructs, routes, and appears on a worklist that has capacity pointed at it regardless of the hour. A radiologist reads it. QA checks it. If there is a bleed, someone picks up a phone and a documented conversation happens. A written preliminary lands in the record. The emergency physician makes the decision they needed to make, inside the window they expected to have.
In the morning, the facility's own radiologist opens the same study, reads it fully with priors, and signs the final. Nothing about their authority, their billing or their responsibility has changed. What changed is that the decision at midnight did not have to wait for them.
That is the whole proposition, and it is narrower than most vendor messaging suggests. It does not fix the workforce numbers — the ACR and Neiman HPI projections are what they are, and they are not improving on their own inside any realistic planning horizon. It closes one specific gap, in one specific window, in days rather than the many months a recruitment cycle takes.
Frequently Asked Questions About Preliminary Radiology Interpretations
Understanding Preliminary Radiology Interpretations
What is a preliminary radiology interpretation, exactly?
It is a rapid, documented first interpretation of an imaging study, issued by a radiologist and explicitly labelled as not final. Its purpose is narrow and practical: give the treating clinician an actionable answer inside the window where a decision has to be made, rather than making them wait for the final signed report. A preliminary radiology interpretation is timestamped, attributed to the named radiologist who issued it, and returned into the facility's own PACS or RIS alongside the study. It is not the diagnosis of record, it is not billed, and it does not replace anything. For human patients in the USA the facility's own licensed radiologist subsequently reads the study in full and signs the final report, which is the interpretation that carries diagnostic and legal weight.
How is a preliminary interpretation different from a wet read?
"Wet read" is older vocabulary from the film era, when a radiologist would look at a film still wet from processing and give a verbal impression on the spot. The phrase survives in conversation, but what it usually describes today is informal and often unwritten. A preliminary radiology interpretation in a modern workflow is a documented artefact: written, timestamped, attributed, stored in the imaging record, and governed by a turnaround protocol and a critical-finding escalation path agreed in advance. The practical difference is auditability. A wet read lives in someone's memory of a hallway conversation; a preliminary interpretation can be retrieved months later, compared against the final, and reviewed for agreement.
What is a preliminary radiology report used for in the first hour after a scan?
It is used to make the next clinical decision rather than to close the case. In the first hour after a scan, the questions in front of the emergency physician or hospitalist are operational: does this patient need the operating room, a transfer, anticoagulation reversal, an intervention, or admission versus discharge. A preliminary report answers those questions with enough confidence to act. It does not attempt to be the complete diagnostic document — incidental findings, comparison with priors, subtle characterisation and the full structured impression belong to the final read. Treating the preliminary as a decision-support instrument rather than a draft final report is the mental model that keeps its scope clear for everyone using it.
Does a preliminary interpretation ever go into the permanent medical record on its own?
It enters the imaging record, but never as the standalone interpretation of record. Most facilities store the preliminary as a clearly labelled non-final report attached to the study, precisely so that the clinical reasoning behind an overnight decision is documented and traceable. What it does not do is stand in for the final. The permanent diagnostic entry is the final report signed by the facility's own licensed radiologist, and if the final differs materially from the preliminary, the facility's protocol should require the treating clinician to be notified so that any overnight decision can be revisited. The preliminary remains in the record as context and as an audit artefact, not as the conclusion.
Who is qualified to issue a preliminary radiology interpretation?
A qualified radiologist, working under an arrangement the facility has formally approved and documented. In academic settings a resident often issues the overnight preliminary under attending oversight. In offshore coverage models such as RAD365's, the preliminary is issued by a named radiologist on the covering team, working inside the facility's own PACS under provisioned accounts and a signed business associate agreement. What matters for compliance is not the geography but the documentation: who issued it, when, under what protocol, with what escalation path, and how the final signing radiologist is guaranteed to be the facility's own licensed physician. Ask any prospective partner to show all five in writing.
Accuracy, Turnaround and the Final Report
How accurate are preliminary radiology interpretations compared to the final signed report?
In well-run programmes, agreement between the preliminary and the final is high, and the disagreements that do occur are mostly minor — wording, degree of characterisation, incidental findings surfaced on fuller review with priors. That is the expected behaviour, not a defect, because the two reads are performed under deliberately different conditions. The preliminary is produced fast, against an urgent clinical question, often without the full prior imaging history. The final is produced with time, with priors, and with the whole record. The honest way to evaluate accuracy is not to trust a vendor's headline agreement rate but to measure your own: log every preliminary-to-final discrepancy by modality and severity and review the trend through your existing QA programme.
How often does a preliminary interpretation get revised at final read?
Some revision is normal and should be expected; the number that matters is not the raw revision rate but the rate of clinically significant revision — changes that would have altered the decision made on the preliminary. Minor edits to phrasing, added incidental findings and refined measurements are routine and tell you very little. A rising rate of significant discrepancy in one modality or one time window tells you something real about case mix, image quality, prior availability or protocol clarity. Facilities that track only an overall percentage usually cannot act on it. Facilities that stratify by modality and severity typically find the signal is concentrated in a small, fixable subset.
What turnaround time should a facility expect for a STAT preliminary read?
Independent turnaround-time benchmark reviews, including work published by AAG Health, put the general expectation for STAT and emergency CT and MRI reads at 60 to 120 minutes. Routine CT and MRI typically runs from a few hours up to 24 hours, and outpatient non-emergent studies commonly run 24 to 72 hours. A preliminary read layer exists specifically to protect the first of those windows when in-house staffing cannot reliably hold it — overnight, weekends, holidays and during surge. What you should insist on contractually is not a single headline number but written case definitions for STAT, urgent and routine, a target for each, and reporting that shows actual performance against those targets rather than an average across all study types.
What happens if a preliminary and final interpretation disagree?
The final governs — always — and the facility's protocol should trigger two things. First, prompt notification of the treating clinician where the difference is clinically material, so that any decision made on the preliminary can be reassessed while it is still actionable. Second, a logged discrepancy entry recording modality, severity and the nature of the difference, routed into the department's peer review and QA programme. That log is genuinely valuable data: it is one of the few routine situations in radiology where two independent interpretations of the same study are systematically compared. Programmes that treat discrepancies as a quality input rather than a blame event get better at both reads over time.
Is a preliminary interpretation the same thing as a "preliminary radiology report"?
Yes — the two phrases describe the same artefact and are used interchangeably in practice. "Preliminary radiology interpretation" emphasises the clinical act of reading the study; "preliminary radiology report" emphasises the written document that results from it. Contracts, accreditation documentation and policy manuals tend to use one or the other with no difference in meaning, scope, turnaround expectation or accountability. If you are comparing providers and finding that they appear to describe different services under the two terms, the divergence is almost always vocabulary rather than substance. Ask instead about the concrete specifics: who reads, what the turnaround protocol says, how criticals escalate, and who signs the final.
Workflow and PACS Integration
Does adding a preliminary-read layer change how our technologists use PACS?
It should not change their workflow in any meaningful way. In a correctly configured deployment, the technologist completes the study and pushes it to the archive exactly as they do today. Routing rules decide which studies, in which time window, appear on the covering team's worklist — that logic sits in the PACS or the router, not in the technologist's hands. The preliminary report returns into the same record the technologist and the clinical team already look at. If a proposed arrangement requires your technologists to learn a second interface, manually export studies, or push images to an external portal, that is a sign the integration has been designed around the provider's convenience rather than your department's.
Can preliminary interpretations be routed through our existing worklist without new software?
In most environments, yes. The standard approach is to work inside the PACS you already run: named user accounts provisioned under your own security model, access scoped to the studies in the agreed coverage window, and routing driven by existing rules on modality, time of day, priority flag or location. No new viewer for your staff, no parallel archive, no second worklist to monitor. Where additional infrastructure is genuinely needed it is usually a DICOM routing or gateway component on your side of the network rather than a new clinical application. The test to apply during scoping is simple: ask precisely what your team would have to install, log into, or learn. The correct answer is usually nothing.
How are critical findings escalated during a preliminary read?
By direct human contact, against a written protocol agreed before go-live, with the contact logged. The protocol should name the finding categories that trigger escalation, the contact path and its fallbacks for each time window, the maximum time allowed between identification and contact, and exactly what gets recorded. Leaving a result in a worklist is not escalation, and neither is an email. The reason to write this down in advance rather than rely on judgement is that critical findings arrive at the worst possible times — 3 a.m., during a shift change, when the named on-call has moved. A protocol with named fallbacks and an audit trail is what makes the escalation reliable rather than fortunate.
How many studies can one radiologist realistically read in a shift, and why does that matter for coverage gaps?
It varies enormously with modality and case complexity — a plain radiography list moves at a completely different pace from complex cross-sectional work with priors — so any single headline number is misleading. The operationally useful point is different: reading capacity per shift is finite and does not flex upward on demand. When overnight volume rises above what the rostered reader can hold inside a 60 to 120 minute STAT window, the queue lengthens and turnaround degrades, regardless of how efficient that radiologist is. That is why coverage planning should be based on your own volume by modality and hour of day across the past twelve months, not on a per-radiologist productivity figure borrowed from elsewhere.
Scope, Compliance and Getting Started
Is a preliminary interpretation billable to Medicare or insurance?
No. The preliminary interpretation is not separately billable to Medicare or to commercial insurers. Only the final signed interpretation — produced by the facility's own licensed radiologist — is the billable professional service. This is one of the clearest structural distinctions between a preliminary read layer and a full diagnostic remote reading arrangement, and it is worth stating explicitly inside the facility, because it removes the most common source of confusion in medical staff committee discussions. Under a preliminary-only model, nothing about the facility's billing, coding or professional fee arrangements changes. The revenue and the responsibility for the final interpretation remain exactly where they are today.
What compliance documentation should a facility expect around preliminary reads?
At minimum: an executed business associate agreement; a written scope document stating that the service is preliminary only and that the facility's own licensed radiologist signs every final; named user provisioning under the facility's own identity and access management, with access reviews; the written turnaround protocol with case definitions; the critical-finding escalation protocol with contacts and fallbacks; an audit trail of study access and report issuance; and a documented discrepancy logging and review process. A provider who can produce all seven as standard artefacts is running a compliance programme. One who assembles them on request during procurement is building them for you, and that is worth knowing before you sign.
Does this apply to veterinary imaging the same way?
No. There is no preliminary stage in veterinary and pet imaging at all. Veterinary studies are read as complete final reports by specialist veterinary radiology groups — there is no preliminary-then-final split, no interim non-final interpretation, and no second signing step. Everything described in this article about preliminary interpretations, coverage windows, turnaround protocols, discrepancy review and final signing applies exclusively to human patients at facilities in the USA. The two models get conflated surprisingly often in procurement conversations, so it is worth stating plainly: human imaging has a preliminary layer; veterinary imaging does not.
How does a facility start using an offshore preliminary-read layer?
Start narrow and instrument it. Define one coverage window and one modality where the gap is measurable today — for most hospitals that is overnight CT. Write case definitions for STAT, urgent and routine, and agree the critical-finding escalation path. Confirm PACS access, routing rules and prior availability with imaging IT. Execute the business associate agreement and provision named accounts under your own security model. Then run a short trial against success criteria set in advance: turnaround versus protocol, preliminary-to-final agreement, and how much of your own radiologists' morning it frees. Widen from there. Because nothing new is installed, the sequence usually takes days rather than the many months a recruitment cycle requires.
Scope preliminary read coverage for your own turnaround window
A short conversation with a RAD365 operations lead: your modality mix by hour, your STAT definitions, your escalation protocol, and what integration into your existing PACS involves. Preliminary reads only — your radiologists sign every final.
Request a coverage review →