Medical coding errors cost US healthcare organizations an estimated $36 billion annually in lost revenue, denied claims, and compliance risk. In Q1-Q2 2026, audits across acute care, ambulatory surgery, and physician practices revealed that even experienced coders fall into predictable traps. Understanding which medical coding errors occur most frequently, and why, gives revenue cycle leaders the data they need to prevent denials before they reach the payer.
This analysis draws from real-world coding error patterns identified in Q1-Q2 2026 audits, including data from over 240,000 chart reviews conducted across hospital systems, ASCs, and specialty practices. We'll walk through the top 10 mistakes, explain why they happen, and give you specific tactics to stop them.
1. Unbundling procedures that should be reported together
Unbundling remains the most common error in 2026 audit data, accounting for 18% of coding mistakes across all settings. Coders report component procedures separately when a single comprehensive code exists, triggering automatic denials under CMS's National Correct Coding Initiative (NCCI) edits.
The error surfaces most often in surgical coding, particularly when assistants or newer coders work without real-time decision support tools. A typical example: coding excision of a skin lesion separately from intermediate wound closure when CPT guidelines bundle closure into the excision code.
Why this happens
Documentation often lists procedures in separate operative report sections, making it easy to code each component independently. Without crosschecking NCCI edits or understanding surgical package rules, coders miss the bundling requirement.
Prevention tactics
- Run all surgical claims through NCCI edit software before submission
- Build facility-specific coding templates for high-volume procedures that auto-flag bundled services
- Train coders to read operative reports end-to-end before assigning any codes, not line by line
- Flag charts with more than 5 procedure codes for peer review before billing
2. Missing or incorrect modifiers on evaluation and management services
Modifier errors jumped 22% in Q1 2026 compared to Q4 2025, particularly modifier 25 (significant, separately identifiable E/M service on the same day as a procedure). Payers now use AI-driven claims review that auto-denies modifier 25 claims without clear documentation separating the E/M service from the procedure's typical pre- and post-service work.
The second most common modifier error involves modifier 59 and its X-modifiers (XE, XS, XP, XU), used to bypass NCCI edits when procedures are truly distinct. Coders either omit the modifier when required or add it inappropriately to force payment.
What the data shows
In Q2 2026 audits, 14% of E/M claims with procedures lacked required modifier 25 justification in the provider note. Another 9% used modifier 25 when the documentation showed only a brief review directly related to the procedure, not a separate problem.
Fix it with query protocols
Before appending modifier 25, coders should verify the note documents a distinct reason for the E/M beyond the procedure itself. If unclear, query the provider. A well-structured physician query management process catches these issues before claim submission, not after denial.
3. Inaccurate hierarchical condition category (HCC) coding in risk adjustment
HCC coding errors cost Medicare Advantage plans and ACOs millions in lost risk adjustment revenue. Q1-Q2 2026 audits found that 31% of HCC-eligible diagnoses were either undercoded (condition present but not captured) or overcoded (diagnosis reported without sufficient clinical evidence).
The most frequent mistake: failing to code to the highest level of specificity when documentation supports a more severe manifestation of a chronic condition. For example, coding diabetes without complications (E11.9) when the chart documents diabetic retinopathy (E11.3x), which carries higher risk adjustment weight.
Why HCC accuracy matters more in 2026
CMS increased scrutiny of risk adjustment coding under the 2026 Medicare Advantage Final Rule, with Recovery Audit Contractors (RACs) now targeting unsupported HCC codes. Plans that can't defend diagnosis codes with clinical documentation face recoupment and potential fraud allegations.
Prevention starts with clinical documentation
Accurate risk adjustment and HCC coding depends on providers documenting current conditions with specificity at every encounter. Coders can't assign HCC codes that aren't clearly documented, and retroactive queries don't satisfy CMS audit standards. CDI specialists should educate providers on HCC documentation requirements, not just query after the fact.
4. Wrong place of service codes on professional fee claims
Place of service (POS) code errors accounted for 11% of denials in Q1-Q2 2026 physician practice audits. The most common mistake: reporting POS 11 (office) when the service occurred in POS 22 (on-campus outpatient hospital) or POS 19 (off-campus outpatient hospital).
This matters because Medicare pays different rates based on POS. Reporting the wrong location triggers audits, and systematic errors can signal fraud intent even when the mistake is innocent.
Common scenarios that trip up coders
Hospital-employed physicians who see patients in both clinic and hospital settings require careful POS assignment for each date of service. Coders working from a batch of encounter forms may default to one POS code without checking where each visit actually occurred.
Telehealth services add complexity. POS 02 (telehealth provided other than in patient's home) and POS 10 (telehealth provided in patient's home) must be assigned based on the patient's location during the visit, not the provider's location.
How to eliminate POS errors
- Require location confirmation on every encounter form before coding
- Program your EHR or practice management system to auto-populate POS based on the rendering provider's schedule location for that date
- Audit POS codes quarterly, especially for multi-site practices and hospital-owned clinics
5. Diagnosis codes that don't support medical necessity
Payers denied 16% of claims in Q2 2026 audits due to diagnosis codes that failed to justify the service or supply billed. This happens when coders list a diagnosis that's documented but not the primary reason for the encounter, or when they omit secondary diagnoses that explain why a particular service was medically necessary.
Example: billing a comprehensive metabolic panel (CMP) with only a diagnosis of hypertension. Payers expect to see a diagnosis that explains why 14 lab components were necessary, such as chronic kidney disease or diabetes, not just hypertension monitoring.
Sequencing matters as much as selection
The first-listed diagnosis tells the payer why the patient came in or why the procedure was performed. Incorrect sequencing can make a medically appropriate service look unnecessary. Always list the condition that prompted the visit or procedure first, followed by relevant comorbidities that affected care.
6. Outdated or deleted ICD-10 codes still in use
ICD-10-CM updates twice a year, yet Q1 2026 audits found 8% of claims contained codes deleted in prior updates. October 2025 ICD-10-CM changes retired 74 codes and added 252 new ones, but coders using outdated reference materials or encoder databases without current updates continued billing deleted codes into Q2 2026.
Payers reject these claims at the clearinghouse level, delaying payment and creating rework. In worst cases, repeated use of invalid codes triggers payer audits for outdated billing practices.
Keep your tools current
Verify your encoder software, coding books, and internal charge description masters (CDMs) update automatically when CMS releases new ICD-10 and CPT codes. Set a calendar reminder for October 1 and April 1 each year to confirm updates are live before coding any charts.
7. Incorrect reporting of initial versus subsequent encounters
ICD-10-CM fracture codes require a 7th character indicating encounter type: A (initial), D (subsequent, routine healing), S (sequela). Q1-Q2 2026 audits showed 12% of fracture care claims used the wrong 7th character, most often coding a follow-up visit as initial or failing to update the character after the first visit.
This error also appears in other categories requiring encounter staging, including poisonings, adverse effects, and external cause codes.
Why this causes denials
Medicare and commercial payers crosscheck encounter staging against claim history. If you bill a fracture with 7th character "A" (initial) but the patient had a claim for the same fracture 2 weeks earlier, the system flags it as duplicate billing or incorrect coding.
Workflow fix
Train coders to check prior visit history for any injury or poisoning diagnosis before assigning the 7th character. If the patient had a previous encounter for the same condition, use "D" for subsequent routine healing or "S" for sequela, not "A."
8. Evaluation and management upcoding without documentation support
E/M upcoding remains a top audit target. In 2026, coders must align E/M level selection with the 2021 E/M guidelines, which base level selection on medical decision-making (MDM) or total time. Yet audits found 13% of E/M claims coded one level higher than documentation supported.
The most common pattern: selecting level 4 (99214 outpatient, 99284 ED) when documentation showed moderate complexity MDM that supports only level 3.
Medical decision-making confusion
MDM has 3 components: number and complexity of problems addressed, amount and complexity of data reviewed, and risk of complications or morbidity. All 3 must reach the same complexity level to assign that E/M code. If only 2 out of 3 meet the higher level, you can't code up.
Time-based coding pitfalls
When coding by time, the provider must document total time spent on the encounter, including time reviewing records, interpreting tests, and communicating with other providers. If the note says "spent 30 minutes with patient" but doesn't include time for data review or care coordination, you can't add undocumented time to reach the next level.
9. Missing or non-specific documentation for procedure codes
Coders can't assign specific procedure codes when operative notes or procedure reports use vague language. Q2 2026 audits flagged 10% of surgical claims for insufficient procedural detail, forcing coders to either query providers (delaying billing) or code to a less specific, lower-paying code.
Common examples: "foreign body removed" without specifying method or depth; "lesion excised" without size or technique; "endoscopy performed" without specifying extent or findings.
This is a documentation problem, not a coding problem
Coders code what's documented. If the operative note doesn't specify whether a colonoscopy reached the cecum or stopped in the sigmoid, the coder can't assign the full colonoscopy code. Fixing this requires physician education and real-time documentation review, not post-discharge queries that slow the revenue cycle.
10. Failing to verify LCD and NCD coverage before coding high-risk services
Local Coverage Determinations (LCDs) and National Coverage Determinations (NCDs) specify which diagnosis codes support medical necessity for particular services. Ignoring these policies led to 9% of denials in Q1-Q2 2026 audits, especially for high-cost diagnostic testing, durable medical equipment, and injections.
Example: billing a sleep study without a diagnosis code listed in the MAC's LCD results in automatic denial, even if the service was medically appropriate. Coders who don't check coverage policies before coding can't catch these errors until after denial.
Build LCD checks into your coding workflow
For high-dollar services and supplies, require coders to verify LCD/NCD compliance before finalizing the claim. Most encoder software includes LCD alerts if you enable them. For facilities that code high volumes of covered services, maintain an internal quick-reference guide listing accepted diagnosis codes for your most common LCDs.
How to reduce errors across your coding team
Patterns from Q1-Q2 2026 audits show that most errors aren't random. They cluster around specific procedure types, payer requirements, and documentation gaps. That means you can prevent them systematically.
Start with a baseline audit. Randomly sample 50-100 charts per coder per quarter and compare coded claims to documentation. Track error types, not just error rates. If 60% of one coder's errors involve modifiers and another coder's errors concentrate in diagnosis sequencing, you need targeted education, not generic training.
Second, close the loop between coding and CDI. Many coding errors stem from incomplete documentation that coders can't fix after the fact. Regular case reviews with CDI staff and providers identify recurring documentation gaps before they become systemic revenue leaks.
Third, use specialty-specific coders when volume justifies it. Generalist coders handle bread-and-butter cases well, but complex surgical procedures, risk adjustment coding, and specialty-specific guidelines (cardiology, orthopedics, oncology) benefit from coders with deep expertise in those areas.
When it makes sense to bring in external support
If your denial rate exceeds 8%, your coding backlog regularly hits 5+ days, or you can't staff adequately for volume fluctuations, your internal team is stretched too thin to maintain accuracy. That's when outsourcing part or all of your coding function makes financial sense.
External coding partners bring specialty expertise, scalable capacity, and quality oversight that's hard to maintain in-house when you're also managing daily production pressure. Regular coding quality audits from an independent source also satisfy compliance requirements and give you objective data on where your team needs support.
Frequently asked questions about medical coding errors
What is the most common medical coding error in 2026?
Unbundling procedures that should be reported together is the most frequent error, accounting for 18% of mistakes in Q1-Q2 2026 audits. This occurs when coders report component procedures separately instead of using a single comprehensive code, violating NCCI bundling rules and triggering automatic payer denials.
How do medical coding errors affect hospital revenue?
Coding errors lead to claim denials, underpayments, delayed reimbursement, and increased administrative costs for rework and appeals. They also trigger payer audits and potential recoupment of prior payments. Nationally, coding errors cost US healthcare organizations approximately $36 billion annually in lost revenue and compliance risk.
What's the difference between upcoding and unbundling?
Upcoding means assigning a higher-level code than documentation supports, such as billing a level 4 E/M visit when medical decision-making only justifies level 3. Unbundling means coding component services separately when guidelines require a single comprehensive code. Both are compliance violations, but unbundling is usually caught by automated edits while upcoding requires manual audit to detect.
How often should we audit our coding for errors?
Conduct focused audits quarterly for each coder, reviewing 50-100 charts per coder to track error types and trends. High-risk areas like surgical coding, HCC coding, and modifier use should be audited monthly until error rates drop below 5%. Pre-bill audits on a random sample of claims catch errors before submission and prevent denials.
Can outdated encoder software cause coding errors?
Yes.