7 Questions to Ask Before You Trust a Preliminary Read Services Provider
Turnaround claims, credentials, discrepancy rates, contract clarity. Seven questions that separate a real preliminary read services provider from a brochure.
Preliminary read services are bought under pressure — a rota gap, a resignation, a surge in overnight volume — and pressure is a poor environment for due diligence. Every provider quotes a fast turnaround and describes its readers as experienced. These seven questions are the ones that separate the providers who can evidence those claims from the ones who cannot. Full detail of how RAD365 structures this service is on the preliminary reads page.
⚠️ Important scope note
For human patients at USA facilities, RAD365 provides preliminary reads only. Your facility's own licensed radiologist always completes and signs the final interpretation. RAD365 does not provide final reads for human patients.
For veterinary and pet imaging there are no preliminary reads — veterinary studies are read as complete final reports by specialist veterinary radiology groups.
typical industry commitment for STAT/emergent reads
benchmark target ED-to-imaging turnaround (AAG Health)
average CT order-to-final-report interval in published QI research
common ED target range for STAT CT/MRI
1. What exactly are you delivering — and who signs the final report?
Start here, because everything else depends on the answer. A preliminary read is a rapid, documented, explicitly provisional interpretation intended to support an immediate clinical decision. A final report is the signed diagnostic document of record. For human patients at USA facilities, RAD365 delivers the preliminary read and your own licensed radiologist completes and signs the final interpretation — that responsibility never transfers. Veterinary imaging has no preliminary stage — those studies are read as complete final reports by specialist veterinary radiology groups.
Ask any provider to state their model in one sentence and then to point at the clause in the scope of work that says it. If the sentence and the clause do not match, you have found the first problem.
2. How do you measure turnaround — and can I see the distribution?
Every provider will quote a target. Reading-service industry standards typically commit to STAT and emergent studies within about one hour. The number is close to meaningless without two further details: when the clock starts and stops, and what the distribution looks like rather than the average.
Ask for median, 90th percentile and 95th percentile turnaround by modality and priority, broken out by hour of day. Averages hide the 3am outliers, and the outliers are the cases that generate incident reports. For operational context, AAG Health's radiology turnaround-time benchmarking puts a target ED-to-imaging turnaround at around 77 minutes, and emergency departments commonly work to a 60-to-120-minute range for STAT CT and MRI overall.
3. What is your discrepancy rate, and what happens after a major variance?
Discrepancy between a preliminary and a final interpretation occurs at some rate in every honest programme. What distinguishes providers is whether they measure it. Ask for the rate, how variances are categorised by clinical significance, who reviews them, how findings feed back to reading staff, and how quickly your facility is notified when a difference is clinically actionable.
A provider who says the rate is effectively zero is either not measuring or not telling you. This is also where an independent peer review and QA process earns its place — external review of a sample of reads is the only way a facility gains assurance that is not self-reported.
4. Who is actually on shift at 3am on a Sunday?
Coverage claims are the easiest thing on a website to write and the hardest to verify. The useful questions are concrete: how many readers are on shift simultaneously in your lowest-volume hour? What happens when two STAT cases arrive within five minutes of each other? What is the escalation path if the on-shift reader is unavailable? What subspecialty depth is genuinely present overnight for neuro, body, chest, MSK, paediatric and advanced MRI?
Depth shows up in the answers to those questions, and in the 3am section of the turnaround distribution you asked for in question two.
5. How will this actually connect to our PACS and reporting workflow?
A preliminary read service that returns reports into a separate portal creates work rather than removing it. Studies should route from your archive by rule — modality, priority, time window, body part — over a secure connection, typically through a DICOM gateway, and preliminary reports should return attached to the correct accession inside the reporting workflow your radiologists already use.
Ask what integration work is required on your side, who performs it, how identity matching is validated, and what the go-live test plan looks like. Two to six weeks is a normal implementation window; a provider promising same-week go-live on a complex environment is describing an untested one.
6. What does the contract say about the things that go wrong?
Read the scope of work for the unpleasant scenarios rather than the pleasant ones. It should define severity levels and turnaround targets with measurement methods, the critical-findings notification protocol and timescale, the discrepancy process, credentialing verification obligations, data protection and retention, SLA remedies, and the exit path including how your data is returned.
| Clause | Weak version | What to require instead |
|---|---|---|
| Turnaround | "Rapid STAT reporting" | Target per priority, clock definition, percentile reporting |
| Final report | Unstated | Facility's licensed radiologist signs all final human reports |
| Critical findings | "Promptly communicated" | Named channel, timescale, documented acknowledgement |
| Discrepancy | Not mentioned | Logged, categorised, trended, reported to the facility |
| Remedy | None | Defined credits plus root-cause report on miss |
| Exit | Silent | Notice period, data return format and timescale |
7. How will we prove this worked in ninety days?
Agree the measurement before go-live, not after. Baseline your current order-to-report intervals by modality and priority for the ninety days before the service starts, then compare the same series afterwards. Published emergency-radiology quality-improvement research has documented average CT order-to-final-report intervals of roughly 5.9 hours, with a median of 4.2 hours, before workflow redesign — which is the scale of delay that a well-implemented preliminary read service is built to compress.
Alongside turnaround, track the proportion of STAT cases meeting target, the discrepancy rate, the critical-findings notification compliance rate, and the time your own radiologists spend on overnight coverage. If none of those series move, the arrangement is a handoff rather than an improvement.
A short evaluation sequence
- Define the gap. Which hours, modalities and priorities are genuinely uncovered today, and at what volume.
- Baseline your current performance. Ninety days of order-to-report intervals, by modality and priority.
- Score providers on evidence, not claims. Turnaround distributions, discrepancy data, shift rosters, credentialing process.
- Test the integration. Routing, identity matching, report return and escalation, before any live clinical volume.
- Pilot on a limited case mix. Then review the same metrics you baselined before expanding.
Facilities running both human and veterinary imaging should note that the model differs by population — the human/USA boundary above is fixed, while veterinary radiology services involve no preliminary stage at all — the study is read once and returned as a complete final report, with veterinary peer review and QA available alongside. The full human-side service picture sits under human radiology services.
Get a scoped preliminary reads proposal
Tell us your uncovered hours, modality mix and volume. We'll come back with turnaround targets, integration requirements and a clear scope — no obligation.
Talk to our team →Preliminary Read Services: Frequently Asked Questions
Scope & Compliance
What exactly is a preliminary read service, and how is it different from a final radiology report?
A preliminary read is a structured, documented first interpretation delivered rapidly — typically to support an immediate clinical decision such as whether to admit, transfer or treat. It flags critical and actionable findings and describes the study, but it is explicitly provisional. The final report is the definitive, signed diagnostic document that enters the permanent record and supports billing. For human patients at USA facilities, RAD365 provides the preliminary read and the facility's own licensed radiologist always completes and signs the final interpretation. Veterinary imaging has no preliminary stage — those studies are read as complete final reports by specialist veterinary radiology groups.
Who is legally responsible for the final interpretation when a hospital uses preliminary read services?
The facility's own licensed radiologist. In a preliminary-read model the responsibility for the signed final interpretation never transfers — it remains with the credentialed radiologist at, or contracted directly to, the facility. The preliminary provider's obligation is accuracy, documentation, defined turnaround and immediate escalation of critical findings. Any provider that blurs this line for human patients is describing a different service and should be evaluated as such.
What's the difference between human preliminary read services and veterinary reporting?
The scope boundary. For human patients at USA facilities, RAD365 provides preliminary reads only and the facility's licensed radiologist completes and signs the final report. For veterinary and pet imaging there are no preliminary reads; veterinary studies are read as complete final reports by specialist veterinary radiology groups, because veterinary practice operates under a different regulatory and credentialing framework. Any provider you evaluate should state which of these two models applies to your patient population, in writing, before contracting.
What happens if a preliminary read and the final read disagree?
Discrepancy is expected at some rate in any preliminary model and should be managed, not hidden. A serious provider runs a formal discrepancy process: every variance is logged, categorised by clinical significance, reviewed, trended and fed back to the reading staff, with immediate notification to the facility where the difference is clinically actionable. Ask for the discrepancy rate, how it is categorised, and what happens after a major variance. A provider that has never measured it has not been looking.
How should a facility structure its contract to avoid ambiguity about who signs the final report?
State it explicitly in the scope of work rather than relying on the service name. The contract should name the deliverable as a preliminary interpretation, state that the facility's licensed radiologist completes and signs the final report, define the critical-findings notification protocol and timescale, specify the discrepancy process, and set out documentation and retention obligations for both parties. Ambiguity in a service description becomes ambiguity in a governance review.
Speed & Turnaround
What turnaround time should a hospital expect from a preliminary read service for a STAT case?
Reading-service industry standards typically commit to STAT and emergent studies within about one hour, and many facilities negotiate tighter targets for specific case types such as suspected intracranial haemorrhage or stroke protocol. What matters as much as the number is the definition: when does the clock start — at study completion, at arrival in the provider's worklist, or at assignment — and when does it stop. Two providers quoting the same hour can differ by twenty minutes on definition alone.
How is preliminary read service turnaround typically measured, and what's a reasonable benchmark?
Measure from study completion at the modality to preliminary report availability at the facility, and report it as a distribution rather than an average — median plus 90th and 95th percentile. Averages conceal the outliers that actually harm patients. On the wider operational picture, AAG Health's radiology turnaround-time benchmarking puts a target ED-to-imaging turnaround at around 77 minutes, and emergency departments commonly target roughly 60 to 120 minutes for STAT CT and MRI overall.
How does a facility verify a preliminary read service is actually reducing turnaround time, not just adding a step?
Baseline before you start. Capture ninety days of your current order-to-report intervals by modality and priority, then measure the same series after go-live. Published emergency-radiology quality-improvement research has documented average CT order-to-final-report intervals of roughly 5.9 hours with a median of 4.2 hours before workflow redesign — which is the scale of delay these services exist to compress. If your own measured interval does not move, the service is adding a handoff rather than removing a bottleneck.
How quickly can a new facility go live with a preliminary read services provider?
Typically two to six weeks. The sequence is connectivity and routing setup, protocol and priority mapping, template and report-format agreement, critical-findings escalation testing, a supervised pilot on a limited case mix, then full volume. Facilities with clean DICOM routing and an existing gateway move faster; those needing new interfaces or VPN work sit at the longer end.
Evaluating a Provider
Can preliminary read services be used for both overnight coverage and daytime overflow?
Yes, and most mature arrangements use both. Overnight coverage addresses the hours when no radiologist is on site. Daytime overflow absorbs surge — a trauma influx, a scanner catching up after downtime, a colleague on leave. The contract should let you route by time window, priority and modality rather than forcing an all-or-nothing volume commitment.
What subspecialty coverage should a preliminary read services provider offer?
At minimum, competent coverage of the emergency case mix: neuro (including stroke and haemorrhage protocols), body CT, chest, musculoskeletal trauma and abdominal imaging. Ask specifically how paediatric and advanced MRI cases are handled, since these are where thin providers rely on generalists. The right question is not whether subspecialty names appear on a website but who is actually on shift at 3am on a Sunday.
What credentials and licensure should preliminary read radiologists or reading staff hold?
Ask for documented qualifications, current registration in the jurisdiction where they practise, evidence of ongoing professional development, and the provider's internal QA and peer-review process. Then ask how credentialing is verified on an ongoing basis rather than at onboarding, and how the facility is notified when reading staff change. Credentialing is a continuing obligation, not a one-time file.
What questions reveal whether a preliminary read provider actually has 24/7 depth, not just a marketing claim?
How many readers are on shift simultaneously at 3am on a Sunday? What happens when two STAT cases arrive in the same five minutes? What is the escalation path if the on-shift reader is unavailable? Show me the last three months of turnaround distribution by hour of day. Depth reveals itself in the 3am percentile, not in the daytime average.
What volume or case types are typically NOT appropriate for a preliminary read service?
Cases requiring physical presence — image-guided procedures, fluoroscopy and interventional work. Complex multidisciplinary cases where the reading radiologist needs direct clinical dialogue. Studies requiring on-site protocolling decisions during acquisition. And any workflow where the facility cannot guarantee a licensed radiologist to complete the final human interpretation, since the preliminary model depends on that step existing.
Pricing & Integration
Does using preliminary read services change how a hospital's own radiologists are scheduled?
Usually for the better. Preliminary coverage removes the pressure to staff overnight on-site rotas for low-volume hours and lets the facility's radiologists concentrate their attention on final interpretation, complex cases and procedural work during the day. It does not remove them from the pathway — they still complete and sign every final human report — but it changes when and under what pressure that work happens.
How does a hospital integrate a preliminary read service with its existing PACS and reporting workflow?
Studies route from the archive to the provider by rule — modality, priority, time window, body part — over a secure connection, usually via a DICOM gateway. Preliminary reports return into the same reporting workflow the facility already uses, attached to the correct accession, so the final-signing radiologist sees the preliminary in context rather than in a separate portal. Routing, identity matching and report return should all be tested before go-live; this is the layer where poorly implemented services create rather than remove work.
Is preliminary read services pricing typically per-study or a flat monthly arrangement?
Both models exist. Per-study pricing suits variable or seasonal volume and keeps cost proportional to use; a flat monthly arrangement suits predictable volume and makes budgeting simpler. Whichever you choose, get the modality and priority differentials, minimum commitments, overage terms and SLA remedies written down. The headline rate is rarely the number that determines the annual figure.
Related Resources
- Preliminary reads — scope, turnaround targets and how the service runs
- How RAD365 structures preliminary read coverage — workflow and escalation detail
- Human radiology services — the wider service line
- Peer review and QA — independent quality assurance on reads
- Veterinary radiology services — complete final reports for pet imaging
- Veterinary peer review and QA — quality processes for veterinary reads
Written by Trisha Seal, RAD365 — 12 August 2026. RAD365 runs preliminary read and imaging-workflow operations for hospitals, imaging centres and veterinary practices. For human patients at USA facilities we provide preliminary reads only; the facility's own licensed radiologist completes and signs every final interpretation. Veterinary imaging has no preliminary stage — those studies are read as complete final reports by specialist veterinary radiology groups.