What overnight radiology coverage costs: 3 models

One overnight radiologist costs almost two. Nobody covers 365 nights alone, so once nights off, vacation, CME, and sick leave are counted, continuous in-house coverage takes 1.6 to 2.0 full-time radiologists at night-differential pay. That's why overnight radiology coverage comes down to a choice between three cost shapes. An employed nocturnist is a seven-figure fixed cost. Per-report teleradiology is a variable cost that falls to zero on a quiet night. The common default, holding studies for a morning batch, is an unpriced cost that lands on your emergency department in boarding hours and defensive admissions. This guide prices all three, including the one most vendors pretend doesn't exist, costs weekend and holiday call on the same arithmetic, and closes with a worksheet to fill in before you compare any two quotes.

How do hospitals cover radiology overnight?

Overnight radiology coverage runs as an employed nocturnist, as per-report teleradiology, or as a morning batch with no night reader at all. Almost every US facility runs one of the three or a blend, and each fails in its own place. The nocturnist breaks on the arithmetic of one human calendar. Per-report coverage breaks when a low headline rate hides recurring fees. The morning batch breaks in the emergency department, where the waiting happens.

Buyers usually compare the first two. At community hospitals and critical access facilities, the comparison that matters is the third against either of the others, because the third is the model already running.

Model 1: a dedicated nocturnist runs 1.6 to 2.0 FTE

A dedicated night radiologist is priced honestly only when you budget 1.6 to 2.0 FTE, because one person can't staff 365 nights. Start with base compensation, among the highest of any specialty in published salary surveys. Add the night differential, which industry surveys commonly place between 15 and 30 percent of base, or its equivalent in fewer shifts for the same pay. Add benefits, payroll taxes, and malpractice, commonly another 20 to 30 percent. Add recruiting: search fees, signing bonus, relocation, and months of vacancy in a shortage market, months during which you're buying outside coverage anyway.

That shortage has a literature behind it: a 2025 review of the US radiology workforce attributes it to rising imaging demand from an aging population set against limited residency positions and substantial retirements, with longer patient wait times and diagnostic delay as the downstream effects.

Then add the line most business cases omit: coverage for that radiologist's nights off and leave weeks. That gap is where the second 0.6 to 1.0 FTE comes from. Every figure in this section is an industry-typical employment benchmark; AstraRad's own prices are quoted per report, in writing. Summed on those benchmarks, continuous in-house nights are a seven-figure annual commitment at most community facilities.

The hire also narrows the coverage question without closing it. A single generalist reading everything overnight is the configuration the subspecialty literature ties to the highest discrepancy rates, so cardiac, breast, pediatric, and nuclear studies typically still route out. In a pediatric hospital series of second opinion interpretations by specialty radiologists, 168 of 773 outside reports carried a major disagreement, 21.7 percent, although these were cases already selected for reinterpretation, which lifts the rate well above what a blind sample of everything read would show. For the daytime version of this staffing math, see in-house radiologists vs outsourced teleradiology. Overnight, the numbers tilt harder: volume is lower, the differential is higher, and the single point of failure is one person's calendar.

Model 2: five fees hide under a low per-report rate

Per-report teleradiology matches the cost curve of the night itself, provided the headline rate is the whole rate. Overnight volume is low, spiky, and unpredictable, and per-report pricing is the only structure that tracks it. A quiet Tuesday costs nothing. A multi-vehicle trauma costs exactly the reports it produces.

Five line items routinely sit under the headline number. Monthly minimums commit you to a study count or dollar floor, and an overnight worklist spends much of the month below any floor. After-hours multipliers add a premium between defined hours, sometimes stacked with a weekend or holiday surcharge on top of a priority fee. Platform and interface fees bill for the connection itself every month; on a low-volume worklist the connection can out-cost the interpretations. Integration and onboarding appear as one-time DICOM, HL7, or VPN charges, occasionally multiplied per modality or per facility. Add-on surcharges attach to prior comparisons, extra series, a phone call on a critical finding, or a corrected report.

The test is short. Ask each vendor to price your last three months of actual overnight volume, by study type, with every line item shown. A vendor who will only quote a per-report figure is quoting the smallest number, not the price. Industry-typical per-report ranges by modality, and how priority tiers move them, are in the teleradiology cost guide, quoted there as market ranges.

AstraRad's structure removes all five. You pay per final signed report, with no minimums, no subscriptions, no platform fees, and no separate charge for nights, weekends, or holidays. You choose the turnaround tier each study needs: STAT under 1 hour, Urgent under 4 hours, Routine under 24.

Model 3: the morning batch bills the emergency department

Running no overnight coverage keeps the cost off the radiology budget and moves it, larger, onto departments that never attribute it back to imaging. A large share of rural and community facilities run exactly this model: studies acquired after late evening sit unread until the day team arrives, and emergency physicians describe the pattern openly in professional forums.

Walk the cost through a single unread study. A patient held until a report exists occupies an emergency department bed, a nurse, and a physician's attention, and boarding hours are among the most expensive hours a hospital sells. In a propensity-matched cohort of emergency patients with acute abdominal pain, a CT report available in under 30 minutes was associated with shorter emergency department and inpatient stays and lower hospitalization costs among patients triaged at higher severity.

When no interpretation exists, the safe decision is to admit. A defensive overnight admission costs more than any signed report a vendor has ever invoiced. Transfers get ordered on the ordering clinician's own look at the images, and some of those transfers, transport and receiving-facility charges included, would never have happened with a signed report in hand. The receiving facility that can't open or won't trust the unread study images the same anatomy again.

And a significant finding that sat unread for seven hours is a documented delay in a legal record, the one line item here with no upper bound. In a Finnish national review of 1,054 imaging-related patient injury claims, delayed diagnosis was the single most common cause, at 38.3 percent of claims. That is a Finnish dataset in a different liability environment, and the ranking of delay against other failure modes is still the point.

None of this appears on the radiology budget, which is how the morning batch keeps its reputation as the free option. Priced honestly, it competes with the nocturnist for the most expensive model on this page.

What does one night of radiology coverage cost?

One night costs whatever your model charges for the hours it holds open plus whatever it charges for the studies that arrive, and those two numbers agree at exactly one volume. A fixed-cost model buys covered hours: the price is identical whether ten studies arrive or eighty. A per-report model buys signed reports: the price moves with the count and with nothing else. Divide each into your own overnight volume and the comparison stops being a matter of opinion.

Two published anchors make the arithmetic real. On the variable side, NDX Imaging lists starting per-study interpretation rates on its 2026 US teleradiology price list, checked August 9, 2026: X-ray from $12, ultrasound from $28, mammogram from $32, CT from $40, MRI from $60, and PET-CT from $99. A high-acuity overnight mix that is mostly CT with some MRI blends to roughly $80 a study on those published starting figures. On the reimbursement side, what the interpretation earns back is set by the professional component in the Medicare Physician Fee Schedule, which is the number your finance office will net the coverage cost against. Those are a third party's published rates and a federal fee schedule, industry-typical figures and none of them an AstraRad price; AstraRad rates are quoted per client in a written card within one business day.

For the fixed side, take an illustrative fully loaded figure of $1.5 million a year for the 1.6 to 2.0 FTE this page has already costed. That is an illustrative industry-typical assumption for the worked example below, neither a published benchmark nor an AstraRad price. Spread across a 12-hour night, 365 nights a year, it covers 4,380 hours at about $342 an hour, roughly $4,110 a night. Substitute your own compensation figure and every cell in the table moves with it.

The table costs one night under each model at three overnight volumes, on the illustrative fixed cost above and the published third-party starting rates.

Read the crossover off the first two rows. At 10 studies the fixed model costs about five times the variable one on the same night. At 30 it is still ahead by a wide margin. Past roughly 50 high-acuity studies the fixed cost per report drops under the blended per-study rate, which is the same crossover named further down this page. Swap in a plain-film-heavy blend nearer $40 a study and the crossover moves beyond 100 studies a night, because cheap studies take far longer to amortize an expensive salary.

Row four is the row that looks cheapest on a per-study invoice and quietly is not. A preliminary read overnight pays a vendor for an interpretation your own radiologist repeats after 8 a.m., so the facility buys the same study twice and spends its scarcest daytime capacity on yesterday's work. The clinical and billing distinction between the two documents is set out in STAT, preliminary and final reads. AstraRad runs no preliminary tier at any hour, so nothing on the overnight worklist comes back for a second signature in the morning.

Five line items the FTE budget leaves out

Five costs sit outside the salary line and land on the same budget anyway: recruitment, the sign-on package, the shift differential, vacancy backfill, and the emergency department cost of running no coverage at all. A business case built on base compensation alone understates continuous overnight coverage by a wide margin, and it is the standard reason an approved staffing plan overruns in year one.

• Recruitment. Search fees on a radiologist placement are quoted as a share of first-year compensation, in a market where the vacancy is already long. Every month the post sits open is a month of bought outside coverage stacked on top of the search fee.
• Sign-on and relocation. A night post in a shortage specialty rarely fills at list. Signing bonus, relocation, loan assistance, and a retention schedule are all cash in years one and two, and the retention schedule is what you forfeit if the hire leaves at month 14.
• Shift differential. The premium for working nights, commonly 15 to 30 percent of base in industry surveys, or its equivalent in fewer shifts for the same pay. It is charged on every night worked and it compounds through every merit cycle.
• Vacancy backfill. Nights off, vacation, CME, illness, and jury duty are most of one person's calendar. This is where the second 0.6 to 1.0 FTE comes from, it is usually bought as locum days or per-report coverage, and it is the reason the outsourced line on the budget never reaches zero.
• The cost of no coverage. The morning batch priced above is what a facility falls back to whenever the post is empty. Boarding hours, defensive admissions, avoidable transfers, repeat imaging at the receiving site, and a documented delay in a legal record are all real money, and none of them appear on the radiology budget.

Vendor quotes carry their own hidden set, and the checklist for pulling every conditional fee out of a proposal before signature is in how to choose a teleradiology company. Ask for the fully loaded number on both sides, or the comparison is a salary against a rate card, which are two different kinds of object.

Where in-house overnight radiology coverage starts to win

In-house overnight radiology coverage starts to win above roughly 50 steady studies a night; below about 10, on industry-typical staffing math, it never does. The break-even moves with modality mix as much as with count, because forty plain films and ten CTs with MRI are different economic propositions. Under ten studies a night, a nocturnist idles at full cost. Between ten and twenty-five, per-report coverage with a defined STAT tier wins comfortably. From twenty-five to fifty, model both, because the second FTE and the subspecialty leakage usually erase the apparent salary advantage. Above fifty steady, high-acuity studies, an employed generalist becomes defensible for the general worklist, and that steady trauma-center pattern is the one situation where AstraRad is the wrong first call.

Three corrections push the practical crossover higher than simple division suggests. Overnight volume spikes, and a fixed cost runs through the trough as well as the peak. The nocturnist only earns their keep on studies inside their own fellowship. And the hire can't cover their own nights off, so a per-report partner stays on the contract regardless, which makes the honest comparison in-house plus teleradiology against teleradiology alone. Weekend and holiday call runs this same arithmetic at even lower utilization, and it is costed in its own right below.

What does weekend and holiday radiology call cost?

Weekend and holiday call costs more per covered hour than a weeknight and delivers fewer studies for the money, which is why it deserves its own budget line. A Saturday and a Sunday are 48 hours, so 52 weekends are 2,496 hours a year before a single holiday. Add the 11 days the Office of Personnel Management publishes as federal holidays and close to a third of the calendar year sits outside ordinary weekday staffing, which is the share most radiology budgets still treat as an exception.

Three premiums stack on that third of the year, all of them industry-typical employment costs and none of them an AstraRad charge:

• Weekend call pay. Groups pay weekend call either as a stipend per day covered or as protected weekdays off. Both are cash: the stipend directly, the days off through the weekday reading capacity they remove.
• Holiday premium. Time and a half, double time, or a floating day in lieu is standard employment practice on the 11 federal holidays, while emergency department volume on those days does not fall to match.
• Differential on the same hours. A Saturday night carries the night differential and the weekend premium together, which makes it the single most expensive hour on the schedule.

Weekend volume also behaves differently from overnight volume, which is why the two do not collapse into one budget line. Saturday daytime at a community hospital runs well above a Tuesday at 3 a.m. and well below a Tuesday at 11 a.m., so a weekend day is too busy to ignore and too quiet to justify a fully staffed shift. Run the four-row model above with your own Saturday and Sunday counts and the fixed-cost row looks worse than it did overnight, because the same salary is now spread across even fewer studies. The staffing patterns and the call-rota alternatives are covered on the weekend and holiday radiology coverage page, and the burnout question behind all of it has a guide of its own; this page stays on the money.

AstraRad applies no weekend fee, no holiday fee, and no after-hours multiplier tied to the clock. The turnaround tier you select for each study is the only thing that moves the price, and the published turnaround and quality figures are measured the same way on a Sunday in December as on a Tuesday in March. The four printed rules of the model are on the per-report pricing page.

Price these 11 inputs before comparing quotes

Eleven inputs, filled in with your facility's own numbers, turn any two overnight quotes into a comparison you can defend. Most vendor cost models go wrong at inputs 5, 8, and 11.

1. Overnight study count per night, by modality, averaged over the last 12 months.
2. The same count on the 90th percentile night.
3. Share of overnight studies routed to a subspecialist: neuro, MSK, body, cardiac, breast, pediatric, chest, ER, oncologic, nuclear.
4. Current time from the arrival of the last image of the study to a final signed report, at median and at P90.
5. Fully loaded cost of one nocturnist FTE: base, night differential, benefits, taxes, malpractice, amortized recruiting.
6. The coverage factor: FTE required for 365 nights at your shift pattern.
7. Cost of the gap solution on nights the in-house reader is off.
8. Every recurring fee in each quote: minimums, platform, interface, per-facility.
9. Every one-time fee: integration, onboarding, per-modality connection.
10. Every conditional fee: after-hours multiplier, weekend or holiday surcharge, prior comparison, critical-results call, corrected report.
11. The contractual service level and its remedy: measured from which event, at median or P90, and what happens when it's missed.

Input 11 turns a price into a comparison. A report that arrives in four hours and a report that arrives inside the hour carry different clinical value, and pricing them as interchangeable is how cheap quotes win bad contracts. AstraRad's service level commitments publish the tiers in writing, measured from last-image arrival to radiologist signature, with compliance reported against each one.

AstraRad staffs the night with subspecialists on shift

Your 3 a.m. study is read by a board-certified, fellowship-trained subspecialist who is on a scheduled shift, mid-workday and wide awake, physically in the United States and licensed in the state where your patients are located. The study routes by subspecialty, so a pediatric abdominal CT reaches a pediatric radiologist at 4 a.m. the same way it would at 4 p.m. Every report comes back final and signed; nothing waits for a morning overread. One study in 20 goes to an independent double-read, and major discrepancies run under 0.3% of signed reports, each reviewed at the monthly discrepancy meeting and closed with the reader. The panel carries spare capacity of 25,000 studies a month, so a surge night at your facility is an ordinary night on the worklist. Every figure in this paragraph is defined, dated, and measured on the SLA and quality methodology page.

AstraRad's overnight radiology coverage is described in full on the nighthawk radiology page. To make the comparison concrete, send your last three months of overnight volume with a rate card request and drop the written per-report prices into the worksheet above. The right cost for the night shift has the shape of the night itself: nothing on the quiet ones, exactly what you used on the loud ones.