A teleradiology RFP that gets useful answers
A teleradiology RFP gets useful answers when every question demands a number and its method: turnaround clocks, discrepancy math, licensure, fixed fees.
A vague teleradiology RFP wastes six weeks of committee time up front, then keeps costing you for the length of the contract, because the turnaround and quality figures you signed were never commitments. The fix costs nothing but discipline in the drafting: make every question ask for a number and the method behind it. Which two timestamps does the turnaround clock run between; what share of studies met each tier last month, on what denominator; what does the complete fee schedule look like applied to your own twelve-month volume. A vendor that measures its operation answers each of those in a paragraph. A vendor that doesn't sends adjectives back, and the adjectives tell you plenty.
This page is the RFP itself. If you're still building the shortlist, start with the buyer's guide and the vendor landscape. If you already have responses in hand, verifying vendor claims covers what to ask for as evidence, and contract terms covers the redline pass that follows.
Why do most teleradiology RFPs get useless answers?
An RFP built on open questions produces prose, and prose from five vendors is indistinguishable. The standard RFP asks each of them to describe: describe your turnaround times, describe your quality program, describe your integration capabilities. Every vendor can describe. The responses come back at similar length, in similar language, carrying the same adjectives, and the committee ends up scoring writing style. Price and the better sales meeting decide the rest. Closed questions with a required unit break this pattern. "What is your STAT turnaround?" invites a claim, while "what percentage of STAT studies received last calendar month were signed as final reports within 60 minutes of the last image arriving, and what were the median and 90th percentile?" invites a number or an excuse, and both are informative. A vendor that measures its operation has that figure in a dashboard and pastes it into the response in an afternoon. A vendor that has never measured it must build the measurement, decline the question, or improvise, and each of those shows up on the page in its own recognizable way. The document sorts vendors before you score a single line.
Three drafting habits carry the whole exercise. Name the unit in every question, because a question without a unit will be answered without one. Ask for the method behind every result, because a discrepancy rate means nothing without its sample and review process. And cap each section's response at one page, because length is where vagueness hides.
One honest limit: if your facility sends a few hundred studies a year in a single modality, skip the RFP. Two reference calls and a week of sample reports will teach you more than seven sections of paperwork.
Five numbers about your own volume come first
An RFP that doesn't describe your volume can't produce comparable pricing, so pull five figures from your RIS for the trailing twelve months before you write a single question.
- Studies per year by modality (CT, MRI, ultrasound, X-ray, mammography, NM/PET): the mix drives your effective price per study more than any negotiated discount.
- The split by priority tier, as your clinicians mark it today: this surfaces STAT inflation before it becomes a contract dispute.
- Hourly distribution across a typical week: overnight-heavy volume prices differently from a daytime overflow pattern.
- Subspecialty mix, including the studies you already send elsewhere: this decides whether a vendor's panel covers you at all.
- Peak month against trough month: the spread determines whether any volume minimum is survivable.
Publish all five inside the document. A quote priced against your real volume file is comparable line by line; a quote built on a generic rate card starts moving after signature.
Section 1: coverage and capacity
Coverage questions establish who is awake and reading, in what numbers, with how much slack.
- What hours are covered by radiologists actively reading, as opposed to on call, on each day of the week including holidays?
- How many board-certified radiologists read for your service, and how many are fellowship-trained in each subspecialty?
- What is your current unused monthly capacity in studies, and how is it measured?
- If our volume grew 30% within a quarter, what changes on your side and over what timeline?
- Have you absorbed a backlog for a client in the last twelve months? State the study count and the elapsed days to clear it.
A specific answer states unused monthly headroom against current staffed reading hours and how long a stated backlog takes to clear. Every hour on a credible coverage grid has a radiologist at a workstation behind it.
Section 2: turnaround with the clock defined
Turnaround questions fail at the definition of the clock more often than at the number.
- Between which two timestamps is turnaround measured? Name the events.
- For each tier, state last full month: median, 90th percentile, and the percentage of studies that met the tier.
- Does a study you reject, hold for a query, or reassign restart the clock?
- Are the figures for signed final reports or preliminary reads?
- Do you provide a monthly SLA report by tier, with study-level timestamps available on request?
Insist that the clock start at the arrival of the last image of the study and stop at the radiologist signing the final report. Any other pair of timestamps hides queue time or transmission time inside the gap. A preliminary read and a final signed report are different products, and preliminary versus final reads changes what a turnaround figure is worth. Our own SLA methodology documents the clock we measure against.
Section 3: quality measured on a stated sample
A discrepancy rate means exactly what its sample and review method make it mean. The ACR White Paper on Teleradiology Practice expects a teleradiology provider to participate in an established quality assurance program including formal peer review, with longitudinal follow-up, second-opinion access and a remediation process for low-performing radiologists, so a vendor should be able to describe all of that without being asked twice.
- What is your major discrepancy rate, over what period, on what denominator?
- What share of studies receives an independent second read, and is the second reader blinded to the first interpretation?
- Who performs peer review, and how are reviewers selected?
- What happens operationally after a major discrepancy is identified?
- Will you provide a quality report specific to our account, at what interval?
AstraRad's protocol double-reads 1 in 20 signed reports independently, with the second radiologist blinded to the first interpretation. If you want your own check on any vendor's answer, the report quality audit method is a workable starting point.
Section 4: roster, subspecialty and licensure
This section confirms that a fellowship-trained physician, licensed in the state where your patients are located, will read the study and can be reached about it. The licensure line is not a formality: 42 CFR 482.22(a)(4)(iii) requires the distant-site physician to hold a license issued or recognized by the state in which the hospital whose patients receive the service is located.
- Are all interpreting radiologists board-certified, and where do they read from?
- Confirm that radiologists interpreting our studies are licensed in the state where our patients are located.
- Which subspecialties are covered by fellowship-trained readers, at what hours?
- Will the same small group of radiologists read our account, or are studies distributed across an open pool?
- How does a treating clinician reach the interpreting radiologist, by what channel, within what committed time, at 3 a.m. as well as at 3 p.m.?
- What is your committed maximum time from identification of a critical finding to direct verbal contact with a licensed provider, and how is that contact documented?
The escalation question separates responses. A vendor that commits to turnaround minutes but won't commit to time-to-reach-a-radiologist has told you which of the two it measures. Our radiologist roster and licensing approach describe how we answer this section, and direct radiologist access covers the contract language.
Section 5: integration and onboarding
Integration questions establish what your IT team installs, maintains, and gets paged about.
- How do studies reach you: DICOM send from our PACS, cloud gateway, portal upload, other?
- How are results returned, and in what format?
- How do you prevent MRN and accession number collisions between institutions?
- How are prior studies made available to the interpreting radiologist, and is retrieval billable?
- What is installed on our side, and what does our IT team have to maintain?
- Give the onboarding timeline from countersignature to first signed report, broken into credentialing, interface build, and test reads.
Credentialing is usually the long pole in that timeline, and it has a shortcut worth asking about. Under 42 CFR 482.22(a)(4) a hospital may rely on the credentialing and privileging decisions made by a distant-site telemedicine entity instead of running its own full appraisal. That relief applies only where there is a written agreement and the vendor's own process meets the medical staff standards named in that paragraph. Ask each vendor whether it supports credentialing by proxy and what it needs from your medical staff office.
AstraRad accepts DICOM directly from your PACS or via portal upload, returns results over HL7 or FHIR, and installs nothing in your environment. The PACS integration guide covers what to expect from any vendor here.
Section 6: compliance, security and liability
This is the section where you collect documents: the BAA itself, the certificate of insurance, the audit log specification. The first of those is a precondition rather than a courtesy, because 45 CFR 164.308(b)(1) permits a covered entity to let a business associate create, receive, maintain or transmit electronic protected health information only once it has obtained satisfactory assurances that the information will be safeguarded. Separately, 45 CFR 164.404(b) fixes the outer limit for individual breach notification at 60 calendar days after discovery.
- Provide your BAA, your breach notification commitment, and your encryption posture in transit and at rest.
- Describe audit logging: what is captured, retained how long, and produced to us on what terms.
- Confirm HIPAA and, if relevant to your organization, GDPR posture, plus DICOM conformance and HL7 or FHIR support.
- State professional liability coverage limits and confirm coverage extends to every interpreting radiologist.
- Who owns the reports and image data, and can any of it be used for benchmarking or product development without our written consent?
Our compliance page is written to be answerable in this format.
Section 7: pricing in one fixed format
Free-form pricing is where comparability dies, so issue this table and require it back completed.
| Line item | Required response |
|---|---|
| Price per report, by modality | One figure per modality in your volume file |
| Priority multipliers | STAT, urgent, routine, stated as a factor or a separate rate |
| After-hours, weekend, holiday multipliers | Stated or explicitly confirmed as none |
| Minimum volume commitment | Study count and consequence of shortfall, or none |
| Platform, licensing, per-seat fees | Amount and frequency, or none |
| Implementation and interface fees | One-time and recurring, including cost of adding a facility or modality later |
| Prior study retrieval | Per study, or included |
| Report amendments and addenda | Per report, or included |
| Annual escalator | Percentage and cap |
| Total applied to our volume file | One number, so responses can be ranked |
Then add the sentence that does the work: no fee absent from this table may be charged during the term. The hidden fees and pricing models guides cover where the effective cost per study moves, and teleradiology cost covers the underlying economics. AstraRad prices per report, with no minimums, no subscriptions and no platform fees.
Scoring the teleradiology RFP responses
Weight the criteria before the first response arrives, and score on evidence alone. A defensible split: turnaround performance with a defined clock and a percentile at 25%, quality method and measured discrepancy rate at 20%, roster depth and licensure at 20%, escalation and critical results commitment at 15%, total cost applied to your real volume at 15%, and integration timeline at 5%. An unanswered question scores zero.
Three answers that should end the evaluation
Some responses disqualify a vendor regardless of price. A turnaround figure with no stated clock, after you asked for the timestamps, means the vendor either doesn't measure it or doesn't want it compared. A refusal to confirm state licensure for the interpreting radiologist is a billing and liability question, not a technicality. And a fee schedule that points to a separate rate sheet the vendor can amend makes every other number in the response provisional.
Send the document to three to five vendors, give them two weeks, and require the pricing table back in the format you issued. When the responses land, the differences are visible before anyone books a meeting: the adjectives pile up in one stack, the numbers in the other. The best question in a teleradiology RFP is the one a marketing department can't answer. If you'd like AstraRad's answers to all seven sections in writing, priced against your own volume file, request a volume quote; how we work describes what happens after signature.
Frequently asked questions
What should a teleradiology RFP include?
Seven sections: your own volume and modality mix, coverage and capacity, turnaround with the clock defined, quality measurement method, radiologist roster and licensure, integration and onboarding, and a pricing schedule in a fixed format. Every question names the unit it wants back: a percentage, a count of minutes, dollars per report. The document exists to make three vendor responses comparable line by line.
What questions should I ask a teleradiology vendor?
Ask which two timestamps the turnaround clock runs between, what percentage of studies met each tier last month, how the discrepancy rate is calculated and on what sample, whether the interpreting radiologist is board-certified and fellowship-trained in the relevant subspecialty, and how a treating clinician reaches that radiologist at 3 a.m. Then ask for a complete fee schedule, with a sentence saying no unlisted fee may be charged. An answer that arrives as adjectives is itself a data point.
How do I compare teleradiology pricing across vendors?
Force every vendor onto the same grid: price per report by modality and by priority tier, applied to your actual twelve-month volume, with every additional fee listed. Headline rates mislead, because after-hours multipliers, minimums, platform fees, prior-study retrieval charges and addendum fees move the effective cost per study more than the base rate does. Ask each vendor to price your real volume file.
How long should a teleradiology RFP process take?
Four to six weeks is realistic for a competitive process: one week to assemble your volume data and issue the document, two weeks for vendor responses, one week for reference calls and scoring, and one to two weeks for redlines. Onboarding runs separately after signature, and credentialing plus state licensure set its length; the interface build is usually the fast part.
How many vendors should I invite to a teleradiology RFP?
Three to five. Fewer than three gives you no pricing reference point, and more than five produces a scoring exercise nobody finishes. Include at least one vendor outside your usual shortlist, because the national names and the subspecialist practices answer the roster and escalation questions very differently.
What are red flags in a teleradiology RFP response?
Five recur. A turnaround number with no stated clock, a quality claim with no sample size or peer review method, and a refusal to confirm that interpreting radiologists are licensed in the state where your patients are located. Then a fee schedule that references a separate rate sheet the vendor can change, and an escalation answer with no committed time to reach the interpreting radiologist. Any one of these means the numbers in the response are provisional.
Related on AstraRad
- Resources
How to choose a teleradiology company: 2026 buyer's guide
Choose a teleradiology company on five written answers: SLA compliance, state licensing, subspecialty match, all-in pricing, and a measured QA rate.
- Resources
Turnaround time SLA language for teleradiology contracts
A turnaround time SLA binds only when the contract defines the start event, stop event, statistic, exclusions, and remedy. Clause language for all five.
- Resources
Teleradiology turnaround time by modality in 2026
Compare teleradiology turnaround by priority, with STAT under 4 hours and routine in 24 to 48 hours across twelve study types for vendor review.
Put a radiologist's name on your next read.
Tell us your modalities and monthly volume. A complete per-report rate card, with turnaround tiers and SLA terms in writing, lands in your inbox within one business day.