Narrative-first, evidence-driven

Back to case studies [pdf]

Receipts: tracked in Methods and Sources by type: Official data | Primary documents | Independent analysis

Pattern summary

American hospitals spend more per person on administration than hospitals in any other country. That is not a coincidence. The system runs through thousands of separate payer contracts, heavy coding and billing complexity, and market structures that reward opacity and delay.

Two things can be true at once:

• hospitals can be under real financial strain
• complexity can still be the reason the bill keeps climbing

Those facts do not cancel each other out. They reinforce each other.

When labor, supply, and debt-service costs rise inside a system that already wastes time and money on billing games, prior-auth churn, and opaque pricing, patients and clinicians end up carrying the full weight of both failures. That is why “more spending” does not automatically mean “more care.”

What this looks like in human terms

A patient gets hurt on a Saturday, goes to the ER because everything else is closed, asks what it will cost, and gets no clear answer. Weeks later the bill arrives. The deductible eats the negotiated rate. A follow-up MRI is delayed by prior authorization. The doctor spends unpaid time fighting with the insurer. The patient gets a collections notice from a company they have never heard of.

None of that requires a cartoon villain. It only requires a system where no one has to give a clear price up front, no one has to move fast enough for the answer to matter, and the collections pipeline starts before the patient can make sense of the bill.

That is the complexity tax: money and time get consumed before the patient gets much more actual care.

Mode 1: The billing maze

What it is: The United States runs one health system through thousands of separate billing relationships. Hospitals negotiate separately with each insurer, code care through a huge administrative layer, and employ entire teams to move claims through the system.

Examples:

• U.S. hospitals spent $933 per person on administration in 2017; Canada spent $196. [confirmed]
• 25.3% of U.S. hospital expenditures went to administration in the 2014 cross-country comparison, versus 15.5% in England. [confirmed]
• Trilliant Health’s 2023 Medicare cost-report analysis found roughly $687 billion in hospital administrative cost versus $346 billion in direct patient care. [plausible]
• McKinsey estimated hospitals spend at least $40 billion annually on billing and collections alone. [plausible]

Missing guardrail: There is no federal requirement to simplify hospital billing transactions, standardize contract structures, or benchmark administrative spending in a way patients and purchasers can actually use.

Mode 2: Opacity by default

What it is: Patients usually do not know what hospital care will cost before they receive it. Hospitals post files because they have to, but that is not the same thing as giving a normal person a usable price.

Examples:

• Cash prices average about 60% above negotiated prices, and chargemaster prices average about 164% above negotiated prices. [confirmed]
• In 2024, HHS OIG found 46% of hospitals still not fully compliant with the hospital price transparency rule. [confirmed]
• GAO found in 2025 that CMS still lacks enough information to know whether posted pricing data is complete and accurate. [confirmed]
• The No Surprises Act fixed some emergency and in-network balance-billing problems, but ground ambulance billing is still outside the federal protection. [confirmed]

Missing guardrail: The transparency rule exists, but the penalty structure is still too weak and the data quality problem is still too easy to hide inside.

Mode 3: Market power pricing

What it is: When one hospital system controls most of the beds in a region, insurers do not have much leverage. That lets dominant systems charge prices that would be harder to sustain in a competitive market.

Examples:

• By 2016, 90% of metropolitan areas were highly concentrated for hospital services. [confirmed]
• 1,164 hospital mergers occurred between 2002 and 2020; FTC enforcement touched only about 1% of them. [confirmed]
• The average hospital merger was raising prices by about 5.2% by 2023, with higher effects in some concentrated markets. [confirmed]
• Private health plans paid hospitals an average of 254% of Medicare rates for the same services in 2022. [confirmed]

Missing guardrail: Merger review still does not match the real geography of hospital market power, and dominant systems do not have to prove competitive benefit before consolidating.

Mode 4: Prior auth and denial churn

What it is: Prior authorization can serve a legitimate role, but the volume of requests, low appeal rates, and high overturn rates suggest the process has drifted well beyond clinical gatekeeping.

Examples:

• In 2024, physicians completed an average of 39 prior-auth requests per physician per week and spent about 13 hours on the process. [confirmed]
• Medicare Advantage plans received nearly 53 million prior-auth requests in 2024; 7.7% were denied in whole or in part. [confirmed]
• 80.7% of appealed Medicare Advantage denials were overturned, but only 11.5% of denied requests were appealed. [confirmed]
• 94% of physicians report prior authorization delays necessary care, and large majorities report that patients sometimes abandon treatment because of it. [confirmed]

Missing guardrail: There is still no binding federal floor for commercial-plan response timelines, denial disclosure by service category, or meaningful limits on how much prior authorization volume a plan can impose.

Mode 5: Admin drag on clinicians

What it is: Clinicians spend a large share of their day on documentation, billing support, inbox work, and prior-auth tasks instead of direct patient care.

Examples:

• Physicians spent 27% of office time on direct patient care and 49.2% on EHR and desk work in the classic time-and-motion study. [confirmed]
• A 2019-2023 longitudinal study found EHR time per clinic day increased by 28.4 minutes, with order and inbox time rising sharply. [confirmed]
• Nearly half of physicians report burnout, and burnout remains widespread among nurse practitioners as well. [confirmed/plausible]
• Higher nurse workloads are associated with higher patient mortality risk. [confirmed]

Missing guardrail: There is no standard for maximum administrative load or any requirement that documentation and insurer-imposed tasks be justified against their clinical value.

Mode 6: Medical debt and collections

What it is: When pricing opacity, billing complexity, and denial churn produce a bill a patient cannot pay, the collections system takes over.

Examples:

• CFPB found 15 million Americans had more than $49 billion in medical debt on credit reports. [confirmed]
• The CFPB’s 2025 medical-debt credit-reporting rule was vacated later that year, leaving federal protection weaker than it briefly appeared. [confirmed]
• Since 2023, 15 states have enacted some form of medical-debt credit-reporting protection. [confirmed]
• Lown Institute found 77% of nonprofit hospitals in its sample spent less on charity care and community investment than the estimated value of their tax exemption. [confirmed, methodology disputed]

Missing guardrail: No federal minimum charity-care standard is tied to nonprofit tax exemption, and national medical-debt protections remain partial and unstable.

Fallout

• Employers and workers pay the pricing spread through premiums and wages.
• Patients face opaque bills, denied care, delayed care, and collections they often cannot navigate.
• Clinicians lose time to non-care work and burn out inside a system that asks them to absorb the friction.
• Communities lose trust because the system is hard to understand even when people are trying to do the right thing.

What good looks like

Hospital care should feel more like care and less like a tollbooth.

What that looks like:

• Usable prices: clear, accurate estimates before non-emergency care whenever possible
• Faster decisions: prior-auth and appeals timelines short enough to matter
• Less billing waste: simpler transactions and visible administrative overhead
• Fairer guardrails: real merger scrutiny, clearer charity-care standards, and debt protections that do not punish people for getting sick

This is not a promise that hospital care becomes cheap. It is the baseline for making it less punishing and less wasteful.

What would reduce the harm

The near-term job is to make the system less punishing before trying to redesign all of it.

Short term: 0-12 months

• Enforce price-transparency rules with revenue-scaled penalties and data-quality audits.
• Publish standardized hospital administrative-cost dashboards using existing Medicare cost-report definitions.
• Require denial-rate disclosure and faster public reporting for Medicare Advantage prior authorization.
• Expand state-level medical-debt protections and adopt employer policies that ignore medical debt in credit checks.

Medium term: 1-3 years

• Extend gold-card exemptions where approval rates are consistently high.
• Apply 72-hour urgent and seven-day routine prior-auth response standards to more plans.
• Require commercial-plan denial, appeal, and overturn-rate disclosure by service category.
• Tighten hospital-merger review around real local concentration rather than just formal transaction size.

Long term: 3-10 years

• Close the ground-ambulance gap in the No Surprises Act.
• Set a real federal minimum charity-care standard for nonprofit hospital tax status.
• Simplify hospital billing and contract structures enough to reduce the administrative layer itself, not just report on it.
• Build a delivery system where documentation, price information, and approvals are designed around patient care instead of payment friction.

Related methods

• official-data: administrative-cost comparisons, Medicare Advantage prior-auth volume, medical-debt prevalence
• primary-documents: HHS OIG transparency enforcement, CFPB debt-rule timeline, CMS and court actions
• independent-analysis: merger-price evidence, hospital price variation, clinician admin-burden research

Back to case studies