Top

Underpayment

Underpayment is the difference between what a payer actually reimburses and what the contract or fee schedule says should have been paid, when the payer pays less than expected. Underpayment is usually tracked as both a dollar amount and a percentage of expected reimbursement.

Kathryn Thompson
Reviewed by Kathryn Thompson · Updated September 2026
Formula
Underpayment rate = (Total expected reimbursement − Total actual reimbursement) for paid claims with a shortfall, divided by Total expected reimbursement for those same claims.

In plain terms: Underpayment rate looks at all the claims that were paid short, adds up how much should have been paid versus how much was actually paid, and expresses that gap as a percentage of what should have been paid.

What it means

What underpayment is

Underpayment is the gap between expected reimbursement and actual reimbursement for a claim line, claim, or group of claims, when the payer paid less than the contracted amount. Expected reimbursement typically comes from the payer contract, fee schedule, per-diem rate, or case rate.

Underpayment is different from a denial. In an underpayment, the payer processed and paid the claim, but the payment is short. The underpaid portion may or may not be collectible, depending on whether the payer followed the contract and whether you can successfully appeal.

Most teams measure underpayment at two levels:

  • Claim or line level: where a specific service, per-diem, or bundle is short paid.
  • Aggregate level: total underpaid dollars and underpayment rate by payer, contract, location, program, or clinician.

Why underpayment matters operationally

Underpayment hides in "paid" claims that look fine if you only watch denial rate and days in A/R. Chronic underpayment pulls down net collection rate and can cost six to seven figures annually in a midsize behavioral-health organization.

Underpayment patterns often point to deeper issues: incorrect contract terms loaded in your system, wrong expected allowed amounts, or payers quietly repricing specific codes or levels of care. They also show where staff are writing off short pays as contractual without checking the contract first.

When you know your underpayment by payer and level of care, you can prioritize: which payer to escalate with contracting, which service lines need clinical or coding cleanup, and where appeals are worth the time. Without that visibility, you are flying blind and may accept chronic short pays as "normal."

How underpayment is used and read

Operationally, underpayment is used in three main ways:

  • Revenue integrity: compare paid amounts to contract terms and spot systematic payer issues or internal configuration errors.
  • Payer performance: track which payers consistently short pay specific codes, per-diems, or bundles and use that data in contract meetings.
  • Worklists and workflows: drive focused underpayment appeals for high-dollar or high-volume variances.

You can measure underpayment at different points in the claim life cycle. Some teams calculate based on the latest payment and adjustment data on the 835 remittance. Others lock the expected amount at first adjudication and treat later recoupments or corrections separately. The rule is to be consistent, and to document the method so finance and revenue cycle are speaking the same language.

Other ways to measure it

Beyond the main formula, a few variations are worth knowing. Each answers a slightly different question.

Claim-level underpayment amount
Underpayment amount for a claim = Expected allowed amount for that claim − Actual payer payment for that claim (if positive).
Use this on individual claims or lines to decide whether an appeal is worth the effort.
Payer-level underpayment rate
Payer underpayment rate = (Sum of underpayment amounts for a payer in a period) ÷ (Sum of expected allowed amounts for that payer in the same period).
Use this to compare how different payers perform and support contracting discussions.
Service-line underpayment rate
Service-line underpayment rate = (Total underpayment for a given service type or level of care) ÷ (Total expected allowed for that same service type or level of care).
Use this to flag specific programs or codes like IOP, PHP, or residential that are being consistently short paid.

Worked example

Imagine a residential program bills a 10-day stay to a commercial payer. The payer contract states a per-diem reimbursement rate of 1,000 dollars, so the expected allowed amount is 10 days × 1,000 dollars = 10,000 dollars.

The claim is processed and the payer pays 8,500 dollars. Your underpayment amount for this claim is 10,000 dollars expected − 8,500 dollars paid = 1,500 dollars. The claim-level underpayment rate is 1,500 ÷ 10,000 = 15 percent.

If you review 20 similar residential stays for the same payer over the month and see a total expected allowed of 200,000 dollars and total payment of 184,000 dollars, your aggregate underpayment is 16,000 dollars. The payer-level underpayment rate for that program would be 16,000 ÷ 200,000 = 8 percent, which is a concrete talking point with both your internal leadership and the payer.

Common mistakes

  • Treating every short pay as a contractual adjustment without checking the fee schedule or per-diem rate, so true underpayment never shows up and you lose appealable dollars.
  • Calculating underpayment against charges instead of the contracted allowed amount, which inflates the apparent problem and makes payer conversations less credible.
  • Mixing bundled services and standalone codes in one underpayment rate (for example, OTP bundled codes and individual therapy codes), which hides that only one category is actually being short paid.
  • Ignoring small-dollar underpayments on high-volume codes like 90834 or H0038, where a few dollars short per session can add up to significant monthly revenue loss.
  • Failing to update expected allowed amounts after a new contract or rate increase, which makes underpayment reports look worse or better than reality and sends your team after the wrong claims.

Why it matters in behavioral health

Underpayment is especially common in behavioral health when payers do not honor the agreed per-diem or case rate for IOP, PHP, or residential treatment, or when they apply medical-surgical fee schedules to behavioral-health codes. For example, a payer might pay IOP group sessions as if they were standard outpatient group therapy, ignoring the contracted IOP per-diem or case rate.

Carve-outs create extra risk. When behavioral-health benefits are administered by a separate vendor, claims often price incorrectly if the medical carrier and the carve-out vendor are not aligned on rates and levels of care. Underpayments show up as odd allowed amounts on 835s, while the claim still appears "paid."

Long per-diem episodes, like residential or withdrawal management, are also exposed to underpayment through concurrent auth and continued-stay review. Even when early days are paid correctly, later days might price at a lower tier than the auth allows, or a portion of days might be cut but still appear partially paid. Without line-level underpayment tracking by day-of-stay, your program may never notice that days 8-14 are consistently short.

State Medicaid and Medicaid MCOs introduce another layer. Many use fixed fee schedules or state-set per-diems, and underpayment can stem from outdated state rates in your system, incorrect revenue codes, or misapplied modifiers like telehealth modifiers 95 or 93. Systematic short pays can distort the margins on high-Medicaid programs and can easily go unnoticed if your team assumes "Medicaid always pays what it pays."

How AI can help with Underpayment

AI can help with underpayment by reading every 835 remittance, comparing paid amounts to expected allowed amounts from contracts or fee schedules, and flagging gaps in real time. An agent can normalize CARC and RARC codes, recognize patterns like a specific payer underpaying 90837 by a few dollars, and generate prioritized worklists of high-dollar or systematic short pays for human review.

Supabill's claims-scrubbing and denials agents can hold payer rules, contracted rates, and expected allowed amounts, then watch for underpayments as remittances arrive. The agents can classify variances, suggest whether an appeal is likely to succeed, and prepare draft disputes or reconsiderations. A human still needs to own contract interpretation, higher stakes payer negotiations, and the decision to escalate patterns to leadership or legal, since AI cannot replace judgment on contract nuance or long-term payer strategy.

FAQ

Is underpayment the same as a denial?

No. A denial means the payer refused to pay some or all of the claim, while an underpayment means the payer paid, but less than the contracted or expected amount. Many underpayments sit hidden in paid claims unless you compare paid amounts to the expected allowed.

How do I figure out the expected allowed amount?

Use the payer contract, fee schedule, or per-diem/case-rate agreement to determine what should have been paid for the billed codes, units, or days. Your practice management or billing system should store these rates so expected allowed amounts calculate automatically.

Should I appeal every underpayment?

No. Focus on high-dollar variances, systematic short pays, and underpayments that clearly violate contract terms. Many teams set a dollar or percentage threshold and concentrate on payer or service-line patterns instead of chasing every small variance.

Can a patient be billed for an underpayment?

Usually not, if the shortfall is due to the payer not following the contract or your internal configuration errors. Balance billing is limited by contracts and regulations, so underpayment is typically a payer issue, not a patient responsibility. Review your payer contracts and any state balance-billing rules before billing patients.

How often should underpayment reports be reviewed?

At minimum, review underpayment trends monthly by payer and level of care, and more frequently if you recently changed contracts or launched a new program. For high-risk payers or new behavioral-health lines like residential, weekly spot checks can prevent months of missed revenue.

Sources

AI agents that run your billing.

The first agentic RCM that actually works.

Book a live demo