CRC Risk Adjustment Coding common mistakes
CRC Risk Adjustment Coding · Understand the traps and how to avoid them.
Assigning an HCC based on a diagnosis keyword in the note without confirming the condition was assessed, treated, or monitored during the encounter (e.g., coding 'history of CHF' or a problem list entry as active heart failure).
Candidates learn code-to-HCC mappings faster than they learn the documentation standard, and EHR problem lists and past medical history sections are dense with chronic diagnoses that look codable at a glance.
What to do instead
Before capturing any chronic condition, locate the provider's assessment/plan statement for that condition in the current encounter. If the only mention is in history, problem list, or a 'history of' phrase, do not assign the HCC; instead flag the record for query and practice writing a compliant clarification question.
Treating a 'history of' code (e.g., Z86.79 for history of circulatory disease, Z85.x for history of malignancy) as equivalent to the active condition and mapping it to an HCC.
The wording looks similar and candidates under time pressure skip the distinction between personal history codes, which describe resolved conditions, and active diagnosis codes.
What to do instead
Memorize the rule that Z85–Z87 history codes never carry HCC weight. Drill a short list of pairs (active neoplasm vs. history of neoplasm, active CHF vs. history of CHF) and confirm active status in the assessment before selecting a code.
Missing hierarchical condition category interactions - capturing a lower-specificity code when a more specific code in the same hierarchy would be trumped, or failing to recognize when a condition is grouped into a higher category and no longer counts separately.
Candidates often code each documented condition in isolation and never build a mental model of the HCC hierarchy, so they either over-count or under-count risk weight.
What to do instead
Study the hierarchy tables for the model year being tested and practice with paired examples (e.g., a condition that is subsumed by a higher-level HCC). For each case, list documented conditions, then map to HCCs and explicitly note which were trumped.
Using an unspecified ICD-10-CM code when the documentation supports a more specific one (e.g., unspecified diabetes without complication when the note documents diabetic CKD, or unspecified heart failure when the note states systolic vs. diastolic).
Candidates default to the shortest code they can find in the index and stop at the first match rather than following the tabular list's specificity requirements.
What to do instead
Always run the code through the tabular list to check for 'code also', 'use additional code', and specificity notes. Practice extracting the specificity elements (type, stage, laterality, acuity) from a note before opening the index.
Failing to apply combination-code and 'code also' rules for diabetes, CKD stage, and other conditions that require linked coding (e.g., coding diabetic CKD as separate diabetes and CKD codes when a combination code exists, or omitting the CKD stage code).
The index leads with the diabetes code and candidates do not check the tabular list instructions that mandate additional codes for manifestations.
What to do instead
Build a checklist for diabetes, CKD, and other high-frequency combination scenarios. For every diabetes encounter, verify whether a manifestation is documented and select the combination code plus any required additional code (such as the CKD stage).
Ignoring the requirement that a condition be documented and supported in the record for the date of service being audited, and instead coding from a prior year's note or a longitudinal problem list.
In real-world chart review, candidates see a rich problem list and assume the condition was addressed at the visit; exam scenarios test whether they can separate encounter-specific documentation from historical data.
What to do instead
Anchor every code to a dated encounter. Read the note for the specific DOS and confirm the condition was addressed that day. When auditing, reject any code whose only support is a prior encounter or the problem list.
Overlooking the MEAT (monitor, evaluate, assess, treat) standard and accepting a bare diagnosis statement with no evidence of evaluation or management as sufficient support.
Candidates focus on whether the diagnosis is present rather than whether it was actively managed, which is the actual documentation threshold for risk adjustment.
What to do instead
For each chronic condition, highlight the sentence showing monitor, evaluate, assess, or treat. If none exists, do not capture the HCC. Practice annotating notes with MEAT evidence before assigning codes.
Misapplying the 'first-listed' or principal diagnosis logic in scenarios where the reason for the encounter differs from the chronic conditions being managed, causing the wrong condition to drive the encounter.
Candidates conflate outpatient coding conventions with inpatient principal diagnosis rules and apply the wrong hierarchy to the scenario.
What to do instead
Clarify the setting in each question before coding. For outpatient encounters, code the reason for the visit first, then additional conditions addressed. Review the official guidelines for first-listed diagnosis and apply them to practice scenarios.
Failing to recognize when a documented condition is a chronic condition that recurs or is a manifestation requiring a two-code structure, and coding only one code.
Candidates memorize single codes for common conditions and do not learn the manifestation pairs and etiology/manifestation conventions tested in the exam.
What to do instead
Study the etiology/manifestation convention and build a list of common pairs (e.g., diabetic retinopathy, hypertensive CKD). For each, confirm both the underlying condition and the manifestation are coded per the guidelines.
Not documenting or querying ambiguous provider statements (e.g., 'probable', 'suspected', 'consistent with') and instead coding the condition as confirmed, or failing to query when the record is unclear.
Candidates want to capture the condition and treat uncertain language as confirmation, which is a compliance violation in risk adjustment.
What to do instead
Apply the rule that uncertain diagnoses are not coded as confirmed in outpatient settings; issue a compliant query and code only what is supported. Practice writing non-leading queries that offer clinically reasonable options.