Clinical coverage creates booked expenses such as payroll, overtime, agency invoices, locums, and stipends. Vacancies can also reduce capacity or revenue, but those effects do not belong in the same accounting stack. A useful coverage view reconciles actual spend first, models vacancy impact separately, and links both to the service line and shift.
Key takeaways
- Labor is the base. The healthcare consultancy Kaufman Hall puts it at 84.4 percent of physician-enterprise expenses and reports a median subsidy of $315,358 per employed physician FTE (Physician Flash Report, Q4 2025). Those are physician-enterprise benchmarks, not measures of hospital nursing or contracted coverage. Separately, the median hospital operating margin closed 2025 at 1.3 percent including system allocations (National Hospital Flash Report, December 2025).
- Most coverage-cost figures are published by the companies that sell coverage, so read locum rates and vacancy costs as sales material, not neutral data.
- Important drivers sit outside scheduling: national labor supply, reimbursement, payer mix, and hospital capacity. Scheduling still affects how available labor is deployed and how often a gap requires premium coverage.
- Montefiore Einstein cut contract labor 68 percent between 2023 and 2025 and nursing premium pay 23.5 percent. The initiative combined unit-level reporting, financial training, contract controls, faster onboarding, recruitment and retention, and demand-aligned staffing (HFMA, 2026).
A note on sources, because this topic needs one
Most coverage-cost content is produced by the businesses that profit from coverage being expensive. Locum market sizes, day rates, and vacancy-cost figures are largely published by staffing agencies and their research arms. That is a structural conflict, and it matters more here than in almost any other topic.
The figures below appear with their provenance attached. Peer-reviewed and government sources are labeled as such; staffing-industry and hospital-lobby figures are labeled too, so you can weigh them differently. A CFO already lives surrounded by vendors quoting their own numbers, so this is the least we can do.
How should coverage cost be measured?
Start with booked expense. Base compensation, benefits, ordinary differentials, and the expected cost of the coverage model belong in the baseline. Kaufman Hall’s 84.4 percent figure describes labor across physician enterprises, including physicians, APPs, and staff; it should not be applied as a universal share for nursing units, locum coverage, or an individual specialty.
Then calculate incremental coverage expense above that baseline: overtime premiums, extra-shift bonuses, agency rate premiums, locums, and call payments not already included in regular compensation. Under the Fair Labor Standards Act, hourly registered nurses generally receive overtime after 40 hours, while salaried RNs who meet the learned-professional tests may be exempt; state law and collective agreements can add requirements (US Department of Labor). Most physicians qualify as exempt professionals, so their incremental cost usually appears as call pay, extra-shift compensation, or locum coverage rather than statutory overtime.
Agency and travel labor can carry a large premium. Kaufman Hall tracked the median contract nurse wage rising from $64 an hour in 2019 to $132 at the 2022 peak while employed nurses went from $35 to $39. NSI’s 2026 report puts the average travel RN fee at $189,758 a year against $123,676 in fully loaded pay for an employed one. NSI sells nurse-recruitment services, so the direction is more useful than the decimal.
Vacancy impact needs a separate model. An open position does save salary and benefits while it remains vacant. Compare that saving with the incremental cost of backfill and, only when the vacancy actually constrains activity, the net contribution lost from closed beds, canceled procedures, reduced visits, or delayed throughput. Do not count both an agency invoice and revenue loss for capacity the agency worker preserved.
NSI estimates the cost of turning over one staff RN at $60,090 and each percentage point of RN turnover at about $295,000 a year for its average hospital. Those are survey-derived nursing estimates from a recruitment company. They are not physician benchmarks and should not replace an organization’s event-level calculation.
Why is the shift-level connection hard to see?
The detail available depends on how each organization has built its general ledger, payroll, timekeeping, position-control, and contract-labor systems. Base pay may sit in salaries, overtime in premium-pay codes, agency and locum coverage in purchased services, and stipends in professional fees. Finance can usually see the expense by cost center; the harder step is connecting an incremental dollar to the service line, date, shift, vacancy, and approval that produced it.
The Healthcare Financial Management Association named the mechanism in its account of one health system’s turnaround: leaders were “navigating contract labor decisions without complete visibility,” and frontline managers lacked “real-time visibility into overtime, late punch-outs, contract hours and associated costs.” The same piece notes that many unit leaders “are experts in patient care, not budgets.”
Traditional top-down cost accounting often allocates labor across departments or encounters without preserving the operational event that caused incremental spend. More granular methods, including time-driven activity-based costing, can retain that connection, but they are not universal. When scheduling and finance systems remain separate, operations sees the gap live while finance may receive an aggregate variance later. Joining those records is the practical problem.
What scheduling cannot fix
Skip this part and a finance leader stops reading, because they already know it.
The workforce shortage is real. The AAMC projects a shortfall of up to 86,000 physicians by 2036, and no schedule creates clinicians who do not exist. Reimbursement also sits outside scheduling: the American Hospital Association reports Medicare paying hospitals 83 cents on the dollar in 2023, with more than $100 billion in Medicare underpayments that year and $130 billion counting Medicaid. These are hospital-lobby estimates calculated on a cost basis, not reimbursement ratios for every service line.
Payer mix matters too. A peer-reviewed Health Affairs study found that a higher public-payer share was associated with both the probability and size of physician stipends (Duffy, Green and Trish, 2025). Boarding and other capacity failures cannot be solved by a clinician roster alone, although staffing can worsen or relieve parts of the throughput problem. In teaching hospitals, fixed duty-hour limits also cap how far resident coverage can be reshuffled, as the ACGME guide explains.
Scheduling can mitigate the operational effect of these constraints. It cannot remove their underlying cause.
What scheduling does drive
Inside those limits, three things move with how the schedule is built and maintained: the match between staffing and demand within the day, the fill discipline that decides whether a gap gets covered internally or at premium, and the ratio of premium hours to total hours. The emergency department is where the first of those is most visible, and the ED scheduling guide works through it hour by hour.
The strongest peer-reviewed evidence is a simulation drawn from 81 general medical and surgical units across three English hospitals. Plans with a low baseline roster and heavy reliance on flexible or temporary staff left more shifts understaffed, while higher baseline staffing improved outcomes and appeared cost-effective (Griffiths et al., International Journal of Nursing Studies, 2021). The finding is about inpatient nursing under the modeled assumptions, not all clinical coverage. It supports matching the baseline roster to variable demand and treating temporary labor as a constrained resource.
Montefiore Einstein provides a documented before-and-after example that HFMA published in 2026: contract labor down 68 percent between 2023 and 2025, nursing premium pay down 23.5 percent across its Bronx hospitals, holiday contract overtime down 88.3 percent, and RN vacancies down 42 percent. The program included simplified reporting, financial training, vendor-contract changes, approval controls, faster onboarding, recruitment and retention, interdisciplinary review, and demand-aligned staffing. The report does not isolate how much each intervention contributed.
The extreme case
Anesthesia is where the invisible cost became impossible to ignore, which is why it gets its own treatment. A peer-reviewed analysis of two decades of California disclosures found the share of hospitals paying anesthesiology stipends rising from 35.5 percent in 2002 to 57.4 percent in 2021, with mean payments reaching $2.89 million among hospitals paying them (Duffy, Green and Trish, Health Affairs, 2025). The drivers are the same ones described above, in unusually pure form.
What should finance and operations measure together?
Use a small set of measures with explicit denominators and peer groups.
| Metric | What it tells you |
|---|---|
| Incremental premium spend over baseline, by unit and shift | Shows where overtime, bonuses, or contract premiums exceed the planned cost of coverage |
| Agency and locum hours and rate premium | Separates reliance on contract coverage from the price paid for it |
| Coverage cost per relevant productive unit | Uses the denominator the service produces, such as patient-days, visits, wRVUs, ASA units, or staffed-location hours |
| Vacancy impact waterfall | Keeps saved salary, backfill expense, and any lost net contribution separate |
| Physician-enterprise subsidy by specialty peer group | Avoids comparing one service with the aggregate $315,358 benchmark across dissimilar specialties |
No single ratio is a universal leading indicator of margin. Premium hours can flag staffing stress, but open shifts, census, acuity, fill rates, agency use, productivity, and rate variance explain why it moved.
Frequently asked questions
What counts as premium labor? For this analysis, count the incremental amount paid above the regular planned cost of coverage: overtime premium, extra-shift bonuses, call-back, agency or locum rate premium, and other avoidable additions. A routine night or weekend differential built into the normal staffing model belongs in baseline labor unless the analysis specifically asks a different question.
How much of my coverage cost can scheduling actually move? The part created by how baseline coverage matches demand, how open shifts are filled, and how assignments change after publication. Scheduling can mitigate some capacity pressure, but it does not set reimbursement, payer mix, market wages, or national labor supply.
Should we aim for zero agency use? No. Agency labor can be rational for demand spikes, leaves, hard-to-recruit roles, and temporary transitions. Evaluate its incremental rate against overtime, internal float capacity, vacancy duration, quality, and the cost of leaving demand uncovered. The NSI turnover estimate is context, not a fixed charge generated by every contract reduction.
Where to start
Build the booked-expense view before optimizing anything. Attribute baseline payroll, incremental premium, agency, locum, and stipend spend to the service line and, where possible, the shift. Reconcile it to the general ledger so operations and finance trust the same numbers.
Model vacancy impact in a separate waterfall: salary and benefits not paid, backfill expense incurred, and net contribution lost only where capacity actually went unused. Then give frontline leaders the measures they can influence. Montefiore’s results came from a broader workforce program, with simple overtime and agency reports as one component.
Separate structural from schedulable drivers in every budget conversation. If premium and agency use fall while cost rises, test market compensation, volume, productivity, contract rates, service mix, and required coverage before labeling the residual structural.
YouShift builds clinical schedules from an organization’s own rules and shows booked coverage cost by site, clinician type, and shift type, including overtime and premium, before the month closes. Operations and finance can review the same reconciled view while keeping vacancy and lost-capacity assumptions in a separate model. More than a thousand physicians schedule on it across the United States and Europe.
See what your coverage costs before the month closes. Tell us your shift structure and we will show you the view.
Sources
-
Kaufman Hall, Physician Flash Report: Q4 2025 Metrics
-
Kaufman Hall, National Hospital Flash Report: December 2025 Metrics
-
Kaufman Hall, “Reliance on Contract Labor During Pandemic Means Higher Hospital Expenses,” 2022
-
HFMA, “How a Health System Can Reduce Premium Labor While Building a Sustainable Workforce,” 2026
-
NSI Nursing Solutions, 2026 National Health Care Retention & RN Staffing Report
-
AAMC, “New AAMC Report Shows Continuing Projected Physician Shortage”
-
American Hospital Association, “Hospitals and Health Systems Squeezed by Persistent Economic Challenges” (Cost of Caring), 2025
-
Duffy, Green and Trish, “Stipends From Hospitals To Emergency Medicine And Anesthesiology Clinicians Increased In California, 2002-21,” Health Affairs, 2025
-
Griffiths et al., “Beyond ratios — flexible and resilient nurse staffing options,” International Journal of Nursing Studies, 2021
-
US Department of Labor, “Nurses and the Part 541 Exemptions Under the FLSA.”