complicated—despite the fact that the patient may have
documented complications such as chronic kidney disease (CKD) or neuropathy. All non-applicable/redundant
auto-suggested codes must be edited or deleted.
CAC is unable to identify documentation inconsistencies
in a patient’s record and may erroneously auto-suggest
codes for items that further require clarification for the
purposes of clinical truth and revenue integrity. For example, CAC may auto-suggest a code for a historically
acute condition that is now resolved and no longer pertinent to the current encounter.
Attention to detail will prevent application of incorrect
codes due to mapping inconsistencies. For example, a
CAC product was known to apply the code for novel influenza A to influenza A. Close review of codes assigned
while using the patient record as a whole is suggested.
Pay close attention to the validity of CAC-generated codes.
CAC may suggest codes that are based on words within
X-ray and lab reports which cannot be coded without provider corroboration on inpatient cases.
Ensure auto-suggested codes that are based on cloned
documentation are relevant and reportable. A cloned
progress note may document “pneumonia,” but further
examination determined that this condition occurred on
a previous admission, is now resolved, and should not be
reported as an acute condition for this encounter.
Remember that coded data has a long and influential life
span, and its importance goes beyond today’s reimbursement. Not only is correct coding essential to a healthy revenue cycle, it is key to healthcare initiatives such as quality
outcomes, risk adjustment, predictive analytics, population health, medical research, institutional longevity, and
provider/hospital ratings. Codes follow patients for a long
time, so correct use of CAC is essential.
Refer back to CAC training materials periodically to ensure
understanding of the nuances and mechanics of the product. Consult with a trusted colleague, manager, or CAC vendor when in doubt about how to use the product optimally.
For example, coding professionals can confuse the methods
of accepting and declining codes. By referring back to the
CAC training materials this problem can be resolved quickly.
CAC Doesn’t Replace, But Assists
People should always keep in mind that CAC is there to assist—not replace—the coding professional and that CAC and
any coding is only as good as the documentation on which it is
based. Coders who work with CAC are afforded the opportunity
to hone their critical thinking skills by the process of validating
(auditing) CAC-suggested coding data.
As with traditional coding, the full patient record must be read
in order to contextualize the CAC-identified verbiage. It is in-
cumbent upon the coding professional to never blindly accept
CAC’s suggestions, but—rather—validate them. CAC is an ever-
evolving tool that is also fallible. The degree of effective human
interaction with CAC varies directly with the quality of the fi-
nal coded product. Ideally, all CAC-generated codes should
be validated for accuracy. Some HIM professionals report that
the thought processes required to work with CAC have inspired
them to career-bridge into auditing. ¢
1. Levinson, Paul. Digital McLuhan: A Guide to the Information Millennium. New York, NY: Routledge, 1999.
Grinder, Deborah. “Perils and Pitfall of Computer-Assisted
Coding in Our ICD- 10 World.” ICD10monitor. November
28, 2016. www.icd10monitor.com/perils-and-pitfall-of-
Daniel Land ( firstname.lastname@example.org) is director of revenue integrity and compliance review services at MedPartners.