Radiology second opinion reads by subspecialists

A radiology second opinion read is a complete, independent re-interpretation of a study that already carries a signed radiologist report, performed by a subspecialist in that anatomy and delivered as a new signed report that names and addresses the original findings. At AstraRad it runs on the routine tier, 24 to 48 hours measured from last-image arrival to radiologist signature, and it is billed per report on the same card as a primary interpretation. There is no separate second opinion surcharge and no minimum order.

This page is for the imaging director, the surgical service, or the quality committee deciding whether a formal re-interpretation is worth ordering on outside work, and what arrives when it is.

What is a radiology second opinion read?

Four things in this market get called a second read, and they are different products with different buyers. Confusing them is how a quality committee ends up paying for peer review when what they wanted was a report they can act on.

Product	Who orders it	Prior report present	Result delivered to
Primary interpretation	The ordering facility	No	The patient's chart
Preliminary read	The facility, overnight	No	A local radiologist, who then finalizes
Overread	The facility, inside a workflow	Sometimes, often from a non-radiologist	The patient's chart
Second opinion	A clinician, facility, or quality program	Yes, from a radiologist	The patient's chart, alongside the original

A second opinion is the only one of the four where the reader is handed an existing radiologist report and asked to say whether it holds. That changes the work. The reader has the prior report, the comparison imaging, and the clinical question in view, and the report that comes back has to reconcile with what was said before: agree with it, extend it, or contradict it in writing, with the reasoning visible. A report that says something different from the original without saying so is worse than no second opinion at all, because the treating clinician now holds two documents and no way to choose between them.

AstraRad has never operated a preliminary tier, so the distinction between a draft and an interpretation of record does not arise here; STAT vs preliminary vs final reads sets out the operational and reimbursement mechanics of that split across the market. Second opinions are assigned from the same panel of board-certified subspecialists across ten subspecialties that reads everything else, described on our reading panel.

How often does a second read change the answer?

Often enough that the published series are uncomfortable reading, and the honest answer is that it depends heavily on what gets sent.

Two verifiable examples. A 2023 series from a tertiary center's abdominal imaging section re-interpreted outside general ultrasound studies and found a discrepancy between the first and second reads in 74 of 196 studies where the original report was available, with a potential management change in 26.0% of those studies (Abdominal Radiology, 2023). A 2025 systematic review of second-opinion reporting by subspecialist musculoskeletal radiologists pooled eight studies covering 11,186 re-interpreted examinations and found clinically significant discrepancy rates ranging from 1.4% to 27.9%, with the high end concentrated in musculoskeletal oncology (European Journal of Radiology, 2025).

Those are other institutions' published research findings on their own case mix, and they are industry literature, none of them an AstraRad performance figure. Read them for the shape of the effect and not as a number to expect. Two things drive the spread. The first is selection: studies nominated for a second opinion are already the hard ones, so the cohort is enriched for disagreement before anyone reads anything. The second is subspecialty distance. Where the first interpretation came from a generalist and the second from a fellowship-trained subspecialist in that exact anatomy, the gap widens, which is the mechanism the musculoskeletal review isolates.

Our own quality program measures a different thing and should not be set against those figures. Our protocol pulls 1 in 20 signed reports into a blind independent double-read, an unselected sample of our own primary work. The protocol sits in our published SLA.

When a second opinion is worth buying

The test is whether a decision hangs on the report. If nothing changes based on the answer, a re-interpretation buys a second document to reconcile and nothing else.

• Transfer of care in oncology. Staging, restaging, and treatment-response studies arriving from an outside facility, where the therapy plan is built directly on the measurements in the report.
• Pre-surgical planning. Complex spine, musculoskeletal tumor, and hepatobiliary cases where the operative approach follows the imaging description.
• An outside study arriving with the patient. A CD or a network transfer from another facility, read somewhere with no subspecialist in that anatomy.
• Quality programs sampling outside work. Facilities auditing interpretations produced by a prior vendor or a departed reader; a radiology report quality audit covers how to structure that sample so the result survives committee scrutiny.
• Advanced studies read by a generalist. Cardiac MRI, CT coronary angiography, and PET-CT are where subspecialty training moves the report most, which is why they carry their own service pages, including PET-CT teleradiology and MRI teleradiology reporting.

Two situations where we will tell you it is not worth it. A technically inadequate study cannot be rescued by a better reader: if the acquisition missed the anatomy or the sequences are non-diagnostic, the patient needs a repeat exam and a second opinion only delays it. And a second opinion is not a mechanism for producing a preferred answer. The reader has no stake in which way it goes, and on the figures above most re-interpretations confirm the original report rather than overturn it.

How it differs from our internal double-read program

Both involve a second subspecialist reading a study that has already been read. Everything else differs, and the two are governed separately on purpose.

	Internal double-read	Second opinion
Who triggers it	AstraRad, on a 1 in 20 sample	The requesting facility, per study
What is being checked	Our own signed reports	An interpretation produced elsewhere
Reader's view of the prior	Blinded to it	Prior report and comparisons in view
Output	A discrepancy record	A signed report for the chart
Where it goes	Monthly quality report, discrepancy meeting	The patient's chart and the requesting clinician
Billing	Included in the program	Per report on the rate card

The double-read is blinded because its purpose is measurement, and a reader who has seen the first report cannot produce an independent one. The second opinion is deliberately unblinded because its purpose is reconciliation, and a reader who has not seen the first report cannot tell you where it went wrong. Same panel, same credentialing, opposite information design.

Discrepancies found in the internal program are categorized, traced to cause, closed with the radiologist who signed, and reported to the facility monthly. Discrepancies found on a second opinion belong to the facility that produced the original report, and we take no position on their program beyond what the new report states clinically.

What you send and what comes back

Send the full DICOM study, whatever priors you hold, the original signed report, and one sentence naming the clinical question. Studies route directly from your PACS with priors and order data attached, or upload through the portal with drag and drop, which needs no integration project and works on day one. Transfer is DICOM conformant and encrypted in transit and at rest, with an integrity hash recorded for every image; the control detail sits on our compliance and security posture page, where the handling is described as HIPAA aligned and performed as a business associate under a signed BAA.

What comes back is a structured report: clinical history, technique, comparison, findings, and impression, signed with the reading subspecialist's name and credentials, and carrying an explicit statement of how the interpretation relates to the prior report. Suspected critical findings are phoned to your designated contact within minutes of sign-off and documented on the report with the time and the person reached. Finalized reports are immutable, and any subsequent change arrives as a versioned addendum.

The signing radiologist holds a license in the state where the patient is located, which is the requirement that governs whether the report can be used, and teleradiology state licensing sets out how that works across state lines. Delivery is HL7 ORU into your RIS or EHR, FHIR, the portal, or a printable letter format for a patient or a referring office.

Timing: routine 24 to 48 hours, STAT under 4 hours, with 2-hour and 1-hour STAT available on request at a higher multiplier, all measured from the arrival of the last image to the radiologist's signature. How we measure turnaround is published in full, including what happens on a miss.

What does a second opinion read cost?

Per report, priced by modality and complexity across twelve study types, with the multiplier for each faster STAT speed printed on the rate card and itemized on every invoice line. There are no minimums, no ceilings, no subscription, no platform fee, and no second opinion premium: a re-interpretation of a chest CT is priced as a chest CT. Our pricing model sets out the four printed rules end to end.

Published market rates for teleradiology interpretation vary widely with modality and contract structure, and any figure you find quoted elsewhere is an industry-typical market number and not an AstraRad price. Ours goes in writing: request a volume quote and a complete per-report rate card arrives within one business day.

One billing caution worth raising with your coding team before volume is planned. The Medicare Claims Processing Manual, Chapter 13 pays the professional component to the physician who performs the interpretation, and an interpretation counts as one only once it includes a written report prepared for the patient's medical record. Whether a re-interpretation of an outside study is separately payable depends on the payer, the modifier, and the documentation, and that answer belongs to your billing staff.

Send ten studies you already hold reports for

The fastest way to evaluate a second opinion service is retrospective. Pick ten outside studies where you already have the original report and the eventual clinical outcome, send them through, and read the two reports side by side against what turned out to be true. That tells you more than any page describing the service can. Onboarding to first signed report runs 10 business days from countersignature, and second opinions are one of four lines described on our service lines.