Medical bill review software checks provider bills before payment: it confirms codes are valid and allowed together, prices each line against a fee schedule or contract, and flags duplicates. Compare options on three things: which checks it runs, how it reads paper and PDF bills, and whether it shows the reason behind every flag or reduction.
What does medical bill review software check?
Medical bill review software checks a bill in three broad groups: coding, pricing and duplicates. Coding checks ask whether the codes are valid and belong together. Pricing checks compare each charge with what a fee schedule or contract allows. Duplicate checks look for the same service billed twice. Most tools also check basic completeness first.
Coding checks often mirror the edits Medicare publishes. The CMS National Correct Coding Initiative page describes procedure-to-procedure edits, which stop pairs of codes that should not be billed together, and medically unlikely edits, which catch units of service reported in implausible amounts. A bill review tool can apply the same logic, either by loading the CMS edit files or by running its own rule set.
Pricing checks compare the billed charge with an allowed amount. For a group health plan, the allowed amount usually comes from a contract with the provider. For workers compensation and some government programs, it comes from a published fee schedule. The difference between billed and allowed becomes a reduction on the line.
Duplicate checks catch the same service on the same date for the same patient. They need care, because legitimate repeats exist: two units of the same service, the same procedure on two different body sites, or a daily charge across a hospital stay. A good tool compares date, code, modifier, units and provider before calling two lines duplicates.
Completeness checks come first. A bill missing a date of service, a diagnosis or a provider identifier cannot be priced. Professional bills use the CMS-1500 layout, and the guide to CMS-1500 form fields explained lists what each box holds.
When a check reduces or denies a line, the result is often expressed as a reason code. The X12 Claim Adjustment Reason Codes list includes codes that match these checks closely. Code 18 marks an exact duplicate claim or service. Code 45 means the charge exceeds the fee schedule, maximum allowable or contracted fee arrangement. Code 236 covers a procedure or modifier combination that is not compatible with another one on the same day under the National Correct Coding Initiative or workers compensation state rules.
How bill review software gets data from paper and PDF bills
Bill review software gets data from paper and PDF bills through an intake step that turns each page into fields before any rule runs. Electronic bills arrive already structured. Paper and PDF bills must be scanned, read and keyed, by people or by document extraction software. Errors made at this step pass straight into the rules.
The input matters because rules trust their data. A pricing rule given a units value of 10 instead of 1 will price ten services. A duplicate check given a misread date will miss a real duplicate or invent a false one. A coding edit given a code with one wrong digit may deny a valid service.
Two bill formats dominate. Professional bills follow the CMS-1500, which puts six service lines in item 24 according to the CMS claims processing manual chapter on the CMS-1500. Facility bills follow the UB-04, where the CMS claims processing manual chapter on the UB-04 pairs a revenue code in FL 42 with each charge in FL 47. The post on CMS-1500 vs UB-04 explains how to tell them apart and who files each. A tool that reads both needs a separate field map for each.
Itemized hospital bills are the hardest input. They can run to many pages, and each page break risks a dropped or doubled line. The guide to itemized medical bill extraction line by line covers how to capture those tables without losing rows.
Ask any vendor how it handles intake. Does the tool read paper itself, rely on a separate scanning service, or require the payer to key bills first? Does each extracted value carry a confidence score? Can a reviewer see the original page next to the value? Does the extracted line total get checked against the printed total before rules run?
Schema Care works at this intake step rather than at the rules step. It reads claim fields and billing statements into data, scores each value, and sends the doubtful ones to a reviewer who can see them highlighted on the source page, as the Schema Care review workflow shows.
Workers compensation bill review software and fee schedules
Workers compensation bill review software differs from group health tools mainly in its pricing rules. Workers compensation medical bills are often priced against a fee schedule set by a state or federal program, not a private contract. A tool must load the right schedule for the right jurisdiction and date of service, and keep it current as the schedule changes.
Federal programs publish their own schedule. The U.S. Department of Labor page for OWCP medical providers says the OWCP Medical Fee Schedule applies to claims under the Federal Employees' Compensation Act, the Energy Employees Occupational Illness Compensation Program Act and the Longshore and Harbor Workers' Compensation Act, with a modified version for the Federal Black Lung Benefits Act.
States publish theirs separately, and the structure varies. California is one example. The California Division of Workers' Compensation Official Medical Fee Schedule page describes the OMFS as issued under Labor Code section 5307.1, with separate sections for physician and non-physician practitioner services, inpatient hospital services, hospital outpatient departments and ambulatory surgical centers, pharmaceuticals, durable medical equipment, pathology and clinical laboratory, and ambulance services. The page says the schedule is adjusted by orders that conform it to relevant Medicare and Medi-Cal changes.
That structure has practical consequences for software:
- Many schedules. A payer handling claims in several states needs a schedule for each, plus any federal program it administers.
- Frequent updates. Schedules tied to Medicare change when Medicare changes. The tool must apply the version in force on the date of service, not the date the bill arrived.
- Different rules per section. An inpatient bill, an outpatient bill and a pharmacy bill may each be priced under a different section with its own method.
- Reason codes with state rules. The X12 list of claim adjustment reason codes notes that code 18 for an exact duplicate is used with group code OA except where state workers compensation regulations require CO.
Ask a vendor which jurisdictions it covers, how fast it loads a new schedule after an update, and whether it keeps prior versions so older dates of service price correctly.
How should you compare medical bill review software?
You should compare medical bill review software on a short list of criteria scored the same way for every option: the checks it runs, how it reads bills, how current its rules and fee schedules are, how it explains its decisions, and how it protects data. A written scorecard keeps a polished demo from deciding the outcome.
Run every candidate on the same sample of your own bills, de-identified, with a mix of clean bills, known errors and hard layouts. Score each criterion from 0 to 3. Weight the criteria that matter most to your team.
A vendor-neutral scorecard
| Criterion | Why it matters | Question to ask | How to score (0 to 3) |
|---|---|---|---|
| Coding checks | Catches code pairs and units that should not be paid | Which edits do you run, and where do the rules come from? | 0 none named; 3 named sources, updated on a stated schedule |
| Pricing and fee schedules | Wrong schedule means wrong payment | Which schedules and jurisdictions do you load, and how fast after an update? | 0 vague; 3 listed jurisdictions with version history by date of service |
| Duplicate detection | False duplicates deny real care; missed ones overpay | Which fields do you compare before calling a line a duplicate? | 0 code only; 3 date, code, modifier, units and provider |
| Paper and PDF intake | Rules are only as good as their input | How do paper bills become data, and how is each value checked? | 0 manual keying not checked; 3 per-field confidence and page view for review |
| Line reconciliation | Proves no line was dropped or doubled | Do you check extracted lines against the printed total? | 0 no; 3 automatic check with flagged gaps |
| Explanation of each flag | Staff and providers need the reason | Can a reviewer see the rule and source behind each reduction? | 0 score only; 3 rule, source and reason code on every flag |
| Human review and audit trail | Disputes need a record | Who approved each change, and when, and can we export that record? | 0 no log; 3 full log of every approval and correction |
| Data protection | Bills contain protected health information | Will you sign a business associate agreement, and what controls are in place today? | 0 no BAA; 3 BAA plus documented controls you can verify |
Total the weighted scores, but read the zeros first. A tool that scores 0 on explanation or data protection is a poor choice regardless of its total.
The data protection row deserves its own review. The HIPAA and PHI category covers the questions to ask a vendor about business associate agreements, access logs and retention before any real bill is shared.
Medical bill review software that hides its reasoning
Medical bill review software that hides its reasoning is a costly buying mistake. A tool that cuts a line or flags a bill without showing which rule fired, what data it used and which fee schedule it applied leaves your team unable to defend the decision, correct an error or answer a provider's appeal.
The cost shows up later. A provider disputes a reduction. Your reviewer opens the bill and sees only "reduced per review" and a lower amount. Nobody can say whether the tool applied the wrong schedule, misread the units or ran an edit that does not fit this case. The dispute takes longer, and the same error repeats on the next bill.
Signs of a tool that hides its reasoning:
- Flags show a score or a color but no rule name.
- Reductions show a new amount but no fee schedule section or version.
- Duplicate flags do not show which earlier line was matched.
- Extracted values cannot be traced back to the spot on the page they came from.
- There is no log of who overrode a flag, or when.
The opposite design is easy to test. Pick one reduced line from the demo and ask the vendor to walk back from the amount to the rule, the schedule version, the input values and the page where each value was read. If they cannot do it for one line, they cannot do it for the thousands your team will process.
A fictional example makes the point. Suppose a bill for an example patient, claim number WC-55021, shows a line billed at $480.00 and reduced to $312.00. A transparent tool would show the code and units read from the bill, the fee schedule and version applied, the allowed amount it calculated, the reason code it assigned, and the person who approved it. Every one of those facts can be checked. A tool that shows only the two numbers asks you to trust it.