CCS-P Medical Coding common mistakes
CCS-P Medical Coding · Understand the traps and how to avoid them.
Assigning a diagnosis code from the EHR problem list or past medical history instead of from the conditions documented as addressed at the current encounter.
The problem list is auto-populated, often stale, and contains 'history of' and 'rule out' entries; candidates scan the top of the note where the list appears and never reach the assessment and plan.
What to do instead
Read the Assessment and Plan first and code only conditions that were evaluated, treated, or monitored at this visit. Treat 'history of', 'ruled out', and 'suspected' entries as non-codable for the encounter, and confirm each chosen code appears in the current note's assessment.
Sequencing an etiology/manifestation pair in the wrong order, or omitting the required secondary code when the tabular list says 'code first' or 'use additional code'.
The Alphabetic Index lookup lands the candidate on the manifestation term, which looks like the more specific and therefore 'better' code, and the instructional note is in small type at the top of the category.
What to do instead
Look up the underlying condition as the main term in the Alphabetic Index, then check the Tabular List category header for 'code first' and 'use additional code' notes. Underlying condition is sequenced first unless the note states otherwise, and the secondary code is always added.
Stopping at an unspecified code when the documentation supports greater specificity, or leaving off a required 7th character such as laterality or encounter type.
Candidates grab the first code that matches the written diagnosis word and submit before verifying the full code length; unspecified codes feel 'safe' because they are rarely wrong in the literal sense.
What to do instead
Before finalizing, re-read the documentation for type, site, laterality, severity, and encounter (initial/subsequent/sequela) and re-check the code to its full character count. If a specific option exists and is documented, use it; if the documentation is genuinely silent, query rather than defaulting to unspecified.
Mixing E/M leveling methods within one encounter - counting history and exam elements from the 1995/1997 guidelines while also claiming medical decision making, or selecting a level by time without documenting total time.
Long-standing habit from older study materials and code sets that were replaced by the current E/M framework; time-based selection looks like an easy shortcut but requires an explicit time statement.
What to do instead
Choose one method per encounter - medical decision making or total time - and apply the current code-set definitions for that method only. If using time, the note must state the total time on the date of the encounter and that counseling/coordination dominated.
Appending modifier 25 to every E/M performed on the same day as a procedure, or appending modifier 59 to bypass an edit without a documented distinct service.
Modifiers are treated as a payment workaround rather than a documentation claim, and candidates do not check the NCCI edit pairs or read what the modifier actually attests to.
What to do instead
Apply modifier 25 only when the E/M is significant, separately identifiable, and above the usual pre- and post-procedure work, and the note supports it independently. Use the most specific distinct-procedure modifier (XE/XS/XP/XQ) supported by a different site, session, or encounter before falling back to 59.
Selecting an unlisted or miscellaneous procedure code (for example a CPT unlisted code or a HCPCS Level II code) when a specific Category I code exists, or submitting an unlisted code without the required special report.
The unlisted code is easy to find by keyword in the index and avoids the harder work of matching the procedure description to a specific code; candidates also confuse Level II supply/drug codes with procedure codes.
What to do instead
Search by the procedure's documented technique, approach, and anatomic site in the CPT index and read the full code descriptors before defaulting to unlisted. If an unlisted code is truly the only option, attach the operative report/special report and cross-reference the comparable procedure used for pricing.
Missing a documentation element that must be present for the code to stand - an attending authentication of a resident's note, a co-signature, a required interpretation, or the physician's own documentation of a service performed by ancillary staff.
Candidates code from the clinical content and assume that if the content exists the service is billable, without checking who documented and signed it under teaching-physician and incident-to rules.
What to do instead
Before assigning the code, verify the note is authenticated by the billing provider, that resident documentation carries the required attending attestation of the key components, and that incident-to services meet the supervision and plan-of-care requirements.
Sending a leading query that suggests the diagnosis or code the coder wants, or documenting a condition the physician never confirmed and coding it anyway.
There is pressure to reach a higher-specificity or higher-paying code, and a query that offers the answer feels faster than one built from clinical evidence.
What to do instead
Write queries that cite the clinical indicators (lab values, findings, treatment) and offer only clinically supported, non-leading options, including the option to reply 'unable to determine'. Code only the response that the provider documents and retains in the record.
Coding a service without checking the payer's coverage determination, or issuing an Advance Beneficiary Notice for an item that is statutorily excluded rather than for one that is covered but not reasonable and necessary in this case.
Coverage rules live outside the code books and are learned by payer experience; candidates focus on code accuracy and treat coverage as someone else's step.
What to do instead
Check the applicable NCD/LCD or payer policy for the diagnosis-service pairing before finalizing, and understand that an ABN applies to items expected to be denied as not medically necessary, not to services Medicare never covers.
Treating charge entry as data entry - missing charges for supplies, drugs, or add-on services, or failing to reconcile the encounter form against the operative/procedure note.
The coder codes the primary procedure and moves on, assuming the clinical staff captured everything; nothing in the workflow forces a line-by-line comparison.
What to do instead
Reconcile the coded record against the source documentation and charge master for every encounter: drugs and units, supplies, add-on codes, bilaterals, and multiple sites. Flag chronic misses (for example, a drug never on the charge sheet) to the revenue cycle owner for a system fix.
Never reviewing denial and edit data, so the same coding error recalculates month after month without correction.
Denial management is seen as a billing-department task, and coders rarely see the remittance outcomes of their own work.
What to do instead
Review denial and NCCI edit reports by reason code and by coder on a set schedule, feed recurring root causes into targeted education, and track the resulting accuracy rate rather than only total denials.