Calculation methods

What these tools are for

This page documents the five DueMD tools: the amounts they use, the formulas they apply and the limits of each result. Calculator inputs are supplied by you; the Medicare lookup uses the installed reference snapshot. Follow the links below to the relevant method.

Distinguish observed values from assumptions. The three calculators use entered amounts; the Medicare lookup retrieves components from the installed snapshot. No tool determines contract applicability, benefits or final payment on its own.

1. Reimbursement variance · v1.0.0

The responsibility split is supplied by the user. This is not a coordination-of-benefits calculation. Allocations greater than the expected allowed amount are blocked as inconsistent before evaluating the formula. Required fields cannot silently become zero; optional other-payer and supplemental values default to zero only as labeled.

Amounts are represented as whole cents. A difference from −$0.01 to +$0.01 is labeled within this tool’s tolerance; the precise difference remains shown. This tolerance is not a contractual write-off instruction. Negative variances are preserved and described as payments above the entered expectation, not confirmed overpayments.

Control example: $5,720 − $250 − $0 − ($4,730 + $0) = $740. If an additional $600 is received, the same model gives $140.

Excluded: reversals, provider-level offsets, pending payments, complex coordination of benefits, contractual applicability and bank reconciliation. Do not use a billed charge as an expected allowance. Use the variance calculator.

2. Denial cost · v1.0.0

Rates are percentages of a consistently defined claim population. The resolution rate is assumed for the same group of claims; no live follow-up is performed. Fractional modeled counts are preserved through the calculation and shown to up to two decimals. They are estimates, not assertions of fractional physical claims.

The per-claim rework estimate is the only administrative-cost method in this version. There is no additional labor-cost multiplication. Average additional payment delay is shown in days, separately. No cost-of-capital or permanent-loss estimate is calculated.

Control example: 20,000 × 8% = 1,600 initial denials. At $230 allowance, the associated amount is $368,000. At $21 rework, administrative cost is $33,600. With 48% resolution, modeled unresolved claims are 832 and the associated allowance is $191,360. An entered 34-day delay remains a separate assumption. Use the denial cost calculator.

3. Contract impact · v1.0.0

Volume, reference amounts and second-procedure share are shared by both scenarios. This keeps the population fixed while comparing supported terms. A factor of 75% means 75% of the modeled second-procedure amount is paid; it does not mean a 75% reduction. Base multipliers may exceed 100%, up to this tool’s input limit of 1,000%.

Current and proposed fixed adjustments can be positive or negative. They are explicitly entered annual assumptions, not automatic carve-out calculations. A fixed adjustment that makes a modeled total negative is blocked. Percentage change is not defined when the current total is zero; the dollar change remains available.

Effects are decomposed in a stated order: change the base multiplier first while retaining the current second-procedure factor; then change the second-procedure factor; then change the fixed adjustment. The interaction between the first two changes is allocated to the second step. A different order can allocate effects differently; the total change is unchanged.

Control example: 10,000 cases, $100 primary reference, $80 secondary reference, 40% second-procedure share. Current base 100% and second factor 75% give $1,240,000. Proposed base 106% and second factor 50% give $1,229,600. Base effect +$74,400, second-factor effect −$84,800, fixed effect $0: total −$10,400.

Each primary and second component is rounded to cents once, then summed with the fixed adjustment. This is an aggregate scenario convention, not a reproduction of claim-line rounding. Excluded: real fee schedule lookup, modifiers, DRG/APC, detailed bundling, patient benefits, balance billing, effective-date adjudication, appeal rights, taxes, expenses, profit and forecasts of demand. Use the contract impact calculator.

Number entry and rounding

Currency inputs accept up to two decimal places. Examples: 1234.56, 1,234.56, 1234,56 and 1 234,56. A grouped comma such as 1,234 means 1,234, not 1.234. Mixed notation such as 1.234,56 is rejected with a message rather than guessed. Scientific notation and more than two decimal places are not accepted.

Percentages accept up to two decimal places. Counts and days are whole-number inputs. Money is stored as integer cents inside the computation, percentage inputs as integer basis points and intermediate products as integer arithmetic. Half-cent ties in aggregate components round away from zero. Excessively large totals are blocked rather than silently losing numerical precision.

Sources and precedence

Official reference links support terminology and context. The educational formulas and worked examples are DueMD’s own stated models; no cited agency has approved them. Actual agreements, applicable rules and case-specific evidence require separate review and may not fit the models. See the checked source registry.

No patient information needed

Do not enter patient names, birth dates, member IDs, claim identifiers, medical record numbers or other PHI. Calculations happen in your browser. Optional usage measurement records approved event categories only, not calculator amounts. Copying or printing happens only when you request it. See the privacy notice.

Versions and corrections

The three original calculators use version 1.0.0. Reference lookup and checklist methods have separate versions below. A change to formulas, rounding or supported scope requires a method-version update.

To report a calculation problem, send the tool name and a short example that reproduces it. Do not send patient records or confidential files.

M-MB-001 · Medicare reference benchmark · v2.0.0

Imported-data mode: selected QP/non-QP RVUs × the three relevant GPCI components, summed and multiplied by the conversion factor in that source record. Non-facility and facility use different PE RVUs. Fixed-point arithmetic preserves source decimals; the base is rounded half-up to cents once. The contract percentage is applied to the rounded base.

Control example from the imported source: code 99214, non-facility, non-QP, Manhattan (13202-01), Q4 2026: [(1.92 × 1.064) + (2.00 × 1.162) + (0.14 × 1.586)] × 33.4009 = $153.27 after rounding. At 125%, the separate comparison is $191.59. This code is checked against source RVUs/GPCIs, not against the six-code PFREV payment revision.

Unadjusted mode: all three GPCIs equal 1 by explicit selection. This is not a weighted national price or a retrieved PFALL amount.

Manual mode: the visitor's reference cents × percentage / 100, rounded half-up. Its input is not verified against Medicare. The illustrative $100 × 110% example remains a manual example.

Status A, supported components and settings are required. OPPS-cap dependencies, anesthesia, other statuses and modifiers are not priced. NA is never treated as zero. Service dates outside October–December 2026, unknown ZIP+4 and missing QP records receive no amount.

Checks and limits: the base formula matches all 1,308 non-facility PFREV revision records supplied with this build. These cover six codes only. See the full source scope and provenance. No claim-level payment or coverage determination is performed.

M-CTG-001 · Contract test generator · v1.0.0

Three deterministic templates produce at least five distinct checks. Boundary dates are calculated in UTC. Percentage and fixed-fee examples use illustrative values and half-up cent rounding. Modifier, scope and component conditions without a supplied rule remain REVIEW REQUIRED; every exported actual-result field starts NOT RUN.

The second-procedure template applies whole-service coefficients. It is not a Medicare therapy MPPR engine. The generator does not read contracts or execute tests against a payer.

Method versions and changes

Saved results include their method version. Compare that version before reusing a calculation after a formula or scope change.