Vacation coverage for radiology reads: a planning guide

Every radiology group solves vacation coverage eventually; the question is whether it solves it in February on a spreadsheet or in July on the phone. A single radiologist's time off removes a fifth of a five-person group's reading capacity, vacations stack in summer and around school holidays, and the recruiting market offers no quick rescue: the ACR's 2026 workforce update describes open positions persistently outrunning the radiologists available to fill them.

This guide covers vacation coverage for radiology reads as a planning problem: the capacity math, the three coverage models with real 2026 figures, and a 90-day calendar you can lift directly. If the per-report option is on your shortlist, a request before the calendar fills brings back a written rate card within one business day.

The vacation math most groups have never written down

Radiologists take substantial time off, by design and by contract. The trend data goes back decades: a study of US radiologist work patterns indexed on PubMed tracked mean vacation rising from 27 days in 1995 to 39 in 2007, and current private practice packages commonly run 8 to 12 weeks for partners. Generous time off is one of the specialty's main recruiting levers, which means the coverage problem is permanent and growing, and no group should plan as if it will shrink.

Aggregate it and the shape gets clear. Five radiologists at 10 weeks each is 50 radiologist-weeks of absence a year: essentially one full-time reader who never shows up, distributed unevenly across the calendar. The distribution is the hard part. Weeks with nobody out require no plan; the two weeks in July with two partners out at once are the entire problem, and they recur every year in roughly the same places: summer, spring break, the winter holidays.

Cross-cover, the default plan, has arithmetic limits. With one of five out, the remaining four must absorb 25 percent more volume each to keep turnaround flat. Some weeks they do. But the 2022 ACR/RBMA workforce survey documented a market where most practices are already short-staffed and hiring, which means the slack cross-cover assumes mostly does not exist. What actually happens is quieter: routine turnaround stretches, the backlog grows by a few days, and the group's most conscientious readers work through lunch for two weeks. Repeat that six times a summer and it compounds into the burnout dynamics that make the next contract negotiation harder.

How does a two-week absence become a six-week backlog?

Work the numbers for a concrete group: five radiologists, 400 studies a day between them, 80 per reader. One partner takes two weeks. The group must now place 400 studies a week onto four readers who were already at 80 a day, so each needs to hit 100, a 25 percent lift, for ten consecutive working days.

Some of that lift happens; call it half. The four readers push to 90 a day, and the remaining 40 studies a day accumulate. After two weeks the group is 400 studies behind, which is one full group-day of reading. The vacationer returns and the group is five again, but nobody has a spare day: clearing 400 studies at a plausible 5 percent overcapacity takes four more weeks, during which routine turnaround runs a day late and referrers notice. One two-week vacation, six weeks of degraded service. Stack a second overlapping vacation in the same month, the standard July situation, and the deficit roughly triples while capacity drops by another fifth.

The failure is structural: cross-cover asks the group to store work, and imaging work stores badly. Reports age poorly for referrers, follow-ups queue behind them, and the recovery consumes exactly the slack the next absence needs. External coverage works because it removes the storage step; studies are read the day they arrive, by someone whose capacity does not depend on your PTO calendar. For calibration on what that costs: AstraRad prices vacation overflow the way it prices everything, per signed report by study type, so the 400 studies in this example cost 400 reads at rate-card prices, and the weeks on either side of the vacation cost nothing at all.

What are the options for vacation coverage of radiology reads?

Three models, and a hybrid that most groups end up with.

The locum figures come from the 2026 radiology locum pay guide, which puts diagnostic rates at $320 to $520 per hour; agency fees and travel sit on top. The full structural comparison, including malpractice and supervision details, is in locum radiologist versus teleradiology.

The honest division of labor: if the vacationing radiologist performs procedures, covers fluoroscopy, or satisfies an onsite supervision requirement, some onsite replacement is non-negotiable, and a locum earns the rate. If the gap is diagnostic reading, paying $4,000 a day for a seat is paying for geography. The per-report model covers the worklist at a per-study price, routes each study to a subspecialist in that anatomy, and costs nothing in the weeks nobody is away.

That last property is what makes teleradiology fit vacation coverage specifically, and it depends entirely on the contract. A vendor with monthly minimums charges you for the quiet weeks between vacations, which converts a variable-cost model back into a fixed one; minimum-free contracting is the term to insist on. AstraRad's structure is built for exactly this intermittency: per-report billing, no minimums, no subscriptions, no platform fees, so an account that sends 400 studies in July and 12 in October pays for 412 signed reports. Turnaround holds regardless of your volume pattern: STAT under 1 hour with a 30-minute measured median, urgent under 4 hours, routine under 24, measured monthly and documented on the SLA page.

The 90-day vacation coverage calendar

Credentialing and licensure, not technology, set the lead time, so the calendar runs backward from the first vacation day. This version assumes a hospital setting; imaging centers can compress most steps.

Two steps deserve emphasis. The credentialing step is the one that blows schedules: 30 to 90 days is normal for hospital medical staff offices, and credentialing by proxy exists precisely to shorten it. And the pilot step is the one groups skip: twenty studies read in a quiet week tells you how the vendor writes, how fast the reports really land, and how the impressions read to your referrers, all before anyone is on a beach. AstraRad's onboarding is built to fit inside this calendar, with a first signed report within 10 business days of countersignature; the pipeline is described on how we work.

The efficient version of all this is to run the calendar once, then keep the arrangement standing. The first vacation costs you the setup; every subsequent one is a routing change. Groups that establish coverage in spring have it ready for summer stacking, the winter holidays, and the unplanned absences, illness, jury duty, a family emergency, that no PTO calendar predicts and that arrive with zero days of lead time instead of ninety.

A note on the absences no calendar predicts. Illness, family leave, and sudden resignations present the identical coverage problem with the lead time removed, and a group whose only options require 60 days of setup meets them with pure backlog. A standing per-report arrangement is the difference between "reroute the worklist this afternoon" and "start credentialing while the queue grows." Groups that have lived through an unplanned three-month absence tend to describe the standing arrangement as insurance they were glad to have priced at zero in the months they never used it; the fuller version of that scenario, a group losing a radiologist on 30 days notice, has its own playbook.

When vacation coverage is really a capacity problem

Some groups discover, running this exercise, that the vacation weeks are just the visible edge of a standing shortfall: turnaround drifts even with everyone present, and time off merely makes it undeniable. The tell is a backlog that never quite clears between vacations. In that case the plan above still applies, but the standing arrangement should be sized for overflow reading all year, with vacation weeks as the peaks, and the weekend coverage math folded into the same contract, since the same panel and rate card serve both.

The reverse discovery also happens: a group with real slack finds cross-cover genuinely cheapest for single-absence weeks and needs external coverage only for the stacked ones. That is a fine outcome. The point of the calendar is that the decision gets made per gap, on numbers, in February.

One warning from groups that have run hybrid models for years: assign an owner. Vacation coverage that belongs to everyone belongs to no one, and the failure mode is a partner booking three weeks in August that nobody cross-references against the calendar until July. The owner needs three standing artifacts: the 12-month PTO calendar with coverage model marked per gap, the vendor relationship kept warm with occasional volume so credentialing never lapses, and the post-gap review notes that make each year's plan better than the last. In most groups this is two hours a month, and it is the difference between vacation season as routine operations and vacation season as an annual emergency.

Either way, the procurement step is identical: send a vendor your PTO calendar's worst two weeks, your study mix, and your volumes, and ask for the price of covering exactly that. AstraRad returns a written per-report rate card within one business day of a request, with every study type priced separately and the full pricing model in print, so the vacation line in your budget becomes a multiplication instead of an estimate.