Subspecialty radiology reads by study type

"Who signs our knee MRI at midnight, and what fellowship did they finish?" Buyers ask us that question in almost exactly those words, and it's the right question. Subspecialty radiology reads by study type mean each study is signed by a radiologist fellowship-trained in the body system it covers. Brain and stroke imaging goes to neuroradiology, joint MRI to MSK (musculoskeletal), staging CT to oncologic body imaging, cardiac CTA to cardiac imaging, mammography to breast imaging, and a child's imaging to a pediatric radiologist. AstraRad routes every study this way, on every shift, and each one comes back as a final signed report from a US based radiologist licensed in the state where your patient is located.

This page is the map. Study type on one side, the subspecialist who should sign it on the other, and the questions that show you whether a vendor routes by fellowship or just clears a queue.

What counts as a subspecialty radiology read?

A subspecialty radiology read requires fellowship training in the relevant body system plus a signature on the final report, and each half is worthless without the other. Board certification in diagnostic radiology is the floor, and it's only the floor: under 42 CFR 482.12(a)(7), a hospital may never make staff membership or privileges depend solely on certification, fellowship or membership in a specialty body. Fellowship is the year that teaches a physician to call subtle cartilage injury, perfusion mismatch or a pediatric growth plate with confidence.

The signature matters just as much. A subspecialist whose preliminary impression your own group has to over read the next morning has handed you a consult, and the billing, the liability and the discrepancy exposure all sit wherever the final signature lands. That's why the difference between STAT, preliminary and final reads matters more than any tier name in your contract. AstraRad signs finals only.

Map your own study mix before you shop

Your own RIS export is the shopping list: one month of studies, grouped by modality and body part, sorted by volume. In a typical month's export, five or six study types carry 70 to 80 percent of your volume. Mark the rows where the clinical question is subtle or where the referring physician acts on the report within the hour. That marked list is your requirement. Bring it to every vendor conversation and ask who signs your knee MRI at 11 p.m. on a Saturday, because that's the study you're buying.

One honest limit before you shop: if your facility needs a radiologist physically in the building, for fluoroscopy, biopsies or contrast coverage, teleradiology is the wrong tool for that part of your list, ours included.

One routing map, from study type to subspecialty

The table below maps the study types a community hospital or imaging center produces in a typical month to the subspecialty that should sign each one.

Spine shows why the indication drives the routing. A post operative cervical spine with hardware belongs with a neuroradiologist even when "spine" sits on the MSK side of a worklist, because the clinical question, myelopathy versus mechanical pain, decides who should read it. A vendor that routes purely on body part will send both spines to whoever reads spine that night.

Coverage thins overnight in five subspecialties

Pediatric, cardiac, breast, nuclear medicine and oncologic body imaging are the five areas where after hours coverage narrows at most vendors, and each thins for its own reason. Pediatric radiology is the smallest fellowship pool, and overnight pediatric volume is unpredictable. Cardiac CTA needs its own post processing and reporting conventions. Mammography depends on the prior comparison being in the reader's hands before the read starts, so ask how priors travel to the physician who signs.

Breast imaging also carries a federal floor the other modalities do not. 21 CFR 900.12 requires an interpreting physician to have read at least 960 mammographic examinations in the preceding 24 months to stay qualified. The same rule sets the reporting clock: the written report has to reach the referring provider within 30 days, with a lay language summary to the patient within seven calendar days when the assessment is suspicious or highly suggestive of malignancy. PET-CT is a separate training track with low overnight volume. Oncologic restaging is only as good as its measurement convention: when this quarter's target lesions are measured differently from last quarter's, the report can't answer the question the oncologist asked.

When a vendor says "we cover that," ask for the coverage window in clock hours. AstraRad staffs these five through the night, and the same turnaround tiers, STAT under one hour, urgent under four, routine under 24, apply to a 2 a.m. study and a 2 p.m. one alike. Each tier is measured from last-image arrival to radiologist signature, whatever hour the study lands.

Pediatric studies carry the highest mismatch cost

A child's abdominal CT is read by a pediatric radiologist, not by a body imager who reads adults all day. That sentence describes AstraRad's routing, and it should describe your vendor's.

Pediatric volume at a general facility is low, and that's exactly why the mismatch survives. A community hospital might send a handful of pediatric MRIs a month, so no single misread recurs often enough to trigger a pattern review, and the roster says pediatric radiology is covered, so nobody asks who was on shift when your scanner finished.

Meanwhile the studies themselves are less forgiving than adult work. Growth plates, ossification centers, thymus and marrow signal all look alarming against adult reference ranges, and a confident adult read on a child is a recurring failure mode. Pediatric CT protocols run at deliberately low dose, which means lower signal to noise, and a physician used to adult image quality will hedge or overcall artifact where a pediatric radiologist reads cleanly. Non accidental trauma raises the stakes further: classic metaphyseal lesions, posterior rib fractures and injuries of differing ages have to be recognized and then reported in language that supports the clinical and legal process that follows. Congenital variants that would be incidental in an adult can be the diagnosis in a child.

The size of the gap has been measured. When pediatric subspecialists at one children's hospital reinterpreted 773 outside studies, a series in the American Journal of Roentgenology found disagreement in 41.8 percent of reports and a major disagreement in 21.7 percent. Those studies were referred for second opinion, so they were never a random sample. Treat the figure as the ceiling on what mismatched routing can cost. Low volume, high consequence, and a gap that only shows at night: that's why pediatric routing is the first thing to verify in any subspecialty arrangement.

So skip the roster question. Ask for the hours a pediatric trained radiologist is on shift, matched against the hours your facility scans children.

Audit the subspecialty radiology reads you already receive

The signature blocks on your own signed reports tell you who has been reading each study type, and you can audit them without any vendor's cooperation. Pull one month of final signed reports. Read the stated fellowship on each, compare it against the body part and indication of the study, and count the mismatches by study type. An aggregate match rate hides the failure that matters: when 96 percent of a month's studies match but every pediatric MRI sits in the other 4 percent, the headline number tells you nothing useful. Repeat the count monthly and the trend becomes a contract conversation.

On AstraRad's side, assignment is run by human schedulers and the radiologists on shift. Your studies arrive by DICOM from your PACS or by portal upload, get grouped by body system and clinical indication, and land with a subspecialist whose fellowship matches, who holds a license in the state where your patient is located, and who is awake and on shift when your study arrives. About 1 in 20 signed reports is pulled for an independent double read as a standing quality sample, major discrepancies run under 0.3 percent of signed reports, and every one is reviewed and closed with the physician who signed. The day to day routing and QA process is on how we work and the SLA page, and the radiologist page covers the fellowship composition of the panel.

Six questions, asked separately for each study type

For every marked row on your volume list, ask which subspecialty signs that study type and whether the report comes back final. Then ask for the coverage hours of that subspecialty in clock hours. Ask for the subspecialty match rate on that specific study type over the last quarter. Ask how double reads are sampled and how a major discrepancy is defined and reported back to your facility. Ask how one of your physicians reaches the radiologist who signed a report, by phone, at 3 a.m. And ask which state licenses cover your patients for that study type and how teleradiology state licensing is verified, because 42 CFR 482.22 requires a distant site telemedicine physician reading for a Medicare hospital to hold a license issued or recognized by the state in which that hospital is located.

A vendor that answers all six for each study type is routing by fellowship. A vendor that answers them once for the whole contract is running a queue.

Go deeper on the modality that dominates your list

Each service page below covers turnaround, report content and integration for one modality group. If your marked volume is mostly CT, including overnight ED CT, start with STAT CT reads. For MRI across neuro and MSK, read MRI teleradiology. Breast imaging is covered in telemammography, PET-CT and nuclear studies in PET-CT and nuclear medicine, and plain film and ultrasound in X-ray and ultrasound overflow. If your gap is a shift, start with nighthawk radiology or weekend and holiday coverage. If you're still comparing vendors, the selection guide covers the contract mechanics.

Carry one idea into every vendor meeting. The roster is the promise. The schedule is the product. Subspecialty radiology reads by study type only mean something when the schedule backs the roster on the shift your study lands. Pricing at AstraRad is per report with no minimums, so request a volume quote, send your one month volume export with the marked rows, and we'll price it row by row. More vendor-vetting guides live in the teleradiology resource library.