The common claim form errors that cause rejections are missing or invalid data: a member ID that does not match the card, an NPI that fails its check digit, a badly formatted service date or place of service code, and a diagnosis pointer that points at nothing. A payer cannot process such a claim, so it sends it back.
Claim rejections versus claim denials
A claim rejection means the payer refused to process the claim because data was missing or invalid. A claim denial means the payer processed the claim and decided not to pay. A rejected claim never entered adjudication, so it is fixed and resent. A denied claim was decided, so it is appealed or corrected. This post covers rejections.
Medicare spells out the difference in its own manual. Chapter 1 of the Medicare Claims Processing Manual, section 80.3.2, says a claim with a required data element missing or not accurately entered is "returned as unprocessable" by a Part B contractor or "returned to provider" by a Part A contractor. The same section says the contractor "shall not deny claims and afford appeal rights for incomplete or invalid information." In plain terms, a rejection carries no appeal, because there is no decision to appeal.
The practical difference shows up in the work queue. A rejection is usually cheaper to fix: correct the field and send the claim again. The cost is time, because the clock on the payer's filing limit keeps running while the claim sits in a rejection queue. A denial needs a different workflow, with records, letters and deadlines of its own. That work is covered in the post on denial management automation and appeal workflows.
Commercial payers use their own wording. Some call a front-end refusal a rejection, some call it a return, and some print it on a report from their clearinghouse. The test is the same everywhere: if the payer never assigned the claim a decision, it was rejected.
Which claim form errors cause the most rejections?
The claim form errors that cause rejections fall into four groups: patient and insurance data, provider identifiers, coding fields, and totals. Each group maps to a block of boxes on the CMS-1500 professional claim form. No public source ranks them by volume, so treat the list below as the places to look first rather than a league table.
Patient and insurance errors
Patient and insurance errors start in item 1a and items 2 and 3 of the CMS-1500. The NUCC 1500 Reference Instruction Manual tells billers to enter the insured's ID number as shown on the insured's ID card. For Medicare, the CMS-1500 instructions in chapter 26 of the Claims Processing Manual ask for the patient's name "as shown on the patient's Medicare card" and an 8-digit birth date. A dropped suffix on the member ID, a nickname instead of the legal name, or a birth date keyed in the wrong order will each stop a claim. Many of these mistakes begin at the front desk, which the post on patient registration errors that lead to denials walks through.
Provider identifier errors
Provider identifier errors sit in items 24J, 25, 32a and 33a. The CMS chapter 26 instructions make the billing provider NPI in item 33a a required field and ask for the rendering provider's NPI in the lower unshaded part of 24J. CMS describes the NPI on its National Provider Identifier Standard page as a 10-digit, intelligence-free number. The CMS paper on NPI check digit requirements says the tenth digit is computed with the Luhn formula, so a single mistyped digit produces a number that fails the check. Item 25 holds the federal tax ID, and the chapter 26 text warns that reimbursement of claims sent without it "will/may be delayed."
Coding field errors
Coding field errors live in item 21 and items 24D and 24E. The chapter 26 instructions allow up to 12 diagnosis codes in item 21 on the 02/12 form, labelled A through L, with an ICD indicator of 0 for ICD-10-CM. They also say not to put a period in the code. Item 24D holds the CPT or HCPCS code, and 24E holds the pointer back to item 21. A pointer that names line E when item 21 only has codes on lines A to C is a field error a scrubber should catch before the payer does. The rules for pointers, including how Medicare and other payers differ on how many letters a line may carry, are in the guide to diagnosis pointer errors that get a CMS-1500 rejected.
Totals and signature errors
Totals and signature errors come from items 24F, 28 and 31. Item 28 is the total of all charges in 24F, according to chapter 26, so a line added after the total was written leaves the claim out of balance. Item 31 needs a signature or "Signature on File" and a date. Medicare's instructions note the claim can still be processed when the signature is on file or computer generated.
Claim form date and modifier errors
Claim form date and modifier errors cause rejections when a service date, place of service code or modifier is missing, badly formatted or does not fit the line. These three fields sit in the item 24 service-line grid of the CMS-1500, where one wrong character on one line can stop the whole claim.
Dates of service go in item 24A. The CMS chapter 26 instructions ask for a 6-digit or 8-digit date for each service and say to return the claim as unprocessable "if a date of service extends more than 1 day, and a valid 'to' date is not present." The NUCC manual adds that a line covering a series of identical services on consecutive days must show a number of days that matches the units in 24G, so a three-day span with one unit, or a to date earlier than the from date, is a common trap. Chapter 26 also tells billers not to mix ICD-9-CM and ICD-10-CM codes on one claim, and to use the from date of a span to pick the code set.
Place of service codes go in item 24B. The CMS place of service code set defines two-digit codes such as 11 for an office, 02 for telehealth provided other than in the patient's home, 10 for telehealth in the patient's home, 19 and 22 for off-campus and on-campus outpatient hospital, and 21 for inpatient hospital. CMS says the codes "should be used on professional claims to specify the entity where service(s) were rendered." A telehealth visit billed with 11, or a hospital outpatient service billed with 11, is a mismatch between the code and the setting. Payer policies on these codes vary, so a scrubber needs the payer's own rules, not only the code list.
Modifiers go next to the procedure code in item 24D. The NUCC manual says the field "accommodates the entry of up to four 2-character modifiers." Chapter 26 adds that when modifier 99 is used to signal multiple modifiers, all applicable modifiers should be listed in item 19. Errors here are usually a modifier the code does not allow, a missing modifier that the payer requires for a setting, or a fifth modifier that will not fit.
A worked example shows how small these errors are. On a fictional claim for an example patient, Jane Rivera, member ID XJH400112233 (an example value), line 1 reads 24A from 03/04/2026 to 03/06/2026, 24B 11, 24D 97110 GP, 24G 1. The three-day span with one unit is the error. The fix is either three units or three separate lines, depending on the payer's rule for that service.
What does claim scrubbing before submission check?
Claim scrubbing before submission is an automated check of every claim against format rules, code lists and payer edits before it leaves the office. A scrubber tests each field for presence and format, checks fields against each other, and compares code pairs and units with published edits. It runs after charge entry and before the clearinghouse.
Payers run the same kind of checks on their side. Chapter 24 of the Medicare Claims Processing Manual lists the front-end capabilities Medicare contractors use, including code set editing, Medicare-specific editing and duplicate submission checking. A scrubber's job is to run those checks first, inside the office, so a claim that would fail them never leaves.
Code pair and unit checks usually come from published edits. CMS describes the National Correct Coding Initiative as having procedure-to-procedure edits, meant "to prevent improper payment when incorrect code combinations are reported," and a separate program for units of service. A scrubber can run these code edits in the same pass as its format checks, so one report covers both.
The checklist below covers the rejection causes in this post. The box numbers follow the CMS-1500 02/12 form.
| Error | Box | How to detect it | How to fix it |
|---|---|---|---|
| Member ID does not match the card | 1a | Compare with the card image and an eligibility response | Rekey from the card, including any prefix or suffix |
| Patient name differs from the card | 2 | Compare spelling and order with the card | Use the name exactly as printed on the card |
| Birth date in wrong order or format | 3 | Pattern check for 8 digits, MM DD CCYY | Rekey from registration records |
| Diagnosis code with a period or wrong code set | 21 | Pattern check and ICD indicator check | Remove the period, set the indicator to 0 for ICD-10-CM |
| Pointer to an empty diagnosis line | 24E | Every letter in 24E must have a code on that line of 21 | Point to a filled line or add the missing code |
| Date span without a valid to date | 24A | Check the to date is present and not before the from date | Add the to date or split into single-date lines |
| Units that do not fit the date span | 24A, 24G | Compare days in the span with the units | Correct units or split the line |
| Place of service does not fit the setting | 24B | Compare with the encounter location and visit type | Use the code that matches where the service happened |
| Too many or wrong modifiers | 24D | Count modifiers, check each against the code | Keep up to four, list the rest per payer rules |
| Rendering NPI fails its check digit | 24J | Luhn check on the 10 digits | Rekey from the provider record |
| Line charges do not sum to the total | 24F, 28 | Add 24F and compare with item 28 | Recompute the total after the last line change |
| Billing NPI missing | 33a | Presence check | Enter the billing provider or group NPI |
Format checks like these also apply to claim data read from scanned forms. The post on format checks for extracted claim codes and identifiers covers that use.
Claim form errors scrubbers miss
Claim form errors that scrubbers miss are values that are valid in format but wrong in fact. A real member ID for the wrong family member, a correct procedure code on the wrong service line, or a real NPI that belongs to another provider will all pass a format check. Catching these needs a person or a second source.
The member ID case is the most common of the three. A subscriber's ID with a different dependent suffix still has the right length and characters. Only an eligibility response or the patient's card shows it is the wrong person. When a family shares a policy, compare the patient's name and birth date with the eligibility record, not only the ID.
The wrong-line code is harder. Suppose a fictional claim has two lines: line 1 for an office visit and line 2 for an injection, with pointers A and B. If the codes in item 24D are swapped between the lines, every field still passes. The pointers now link the visit to the injection's diagnosis. The claim may be rejected later by a payer edit, or it may be paid wrongly. The check is a person reading the line against the encounter.
The NPI case looks like this. A practice with two providers whose NPIs differ only in their middle digits may get a valid but wrong NPI in 24J. The Luhn check passes because the number is real. The check that catches it is a lookup against the practice's own provider list for the date of service.
Scanned paper claims add their own misses. An 8 read as a 3 in a member ID, or a dropped digit in a charge, can produce a value that still passes a format check. Errors like these start at the scan or the keying step, and the post on paper claims processing traces where they creep in. Schema Care reads CMS-1500 fields from the page, gives each one a confidence score with a link to its place on the form, and holds low-scoring fields for a reviewer, as the human review steps on the homepage show.
Paper claims that will be sent electronically meet one more layer of checks at the clearinghouse and payer. What those checks reject, and how the rejections come back, is the subject of the guide to converting paper claims to 837 files.