As meticulous as a coder may be, he or she is bound to make a mistake at some point in his or her career. After all, nobody is perfect. Mistakes aren’t necessarily a reflection on one’s abilities or attention to detail. Coders know that physician documentation often makes the job much more difficult. Add stringent productivity standards to that, and you’ve got a potential recipe for disaster.
Although humans are bound to make errors, the goal is to learn from those errors so as not to repeat them. Where are coders likely to veer off the compliance track? Our experts share their thoughts and provide tips for compliance.
Sequencing.James S. Kennedy, MD, CCS, a coding and clinical documentation improvement (CDI) expert near Nashville, says sequencing errors—or “opportunities”—account for nearly one-third of the coding problems he encounters when reviewing charts.
Even the American Hospital Association’s Coding Clinic acknowledges the difficulty of sequencing, stating in its Third Quarter 2002 edition (p. 28) that “determination of the appropriate principal diagnosis is not always an easy task.”
Coders are frequently presented with the challenge of selecting a principal diagnosis when two or more conditions equally meet the definition. Although the ICD-9-CM Official Guidelines for Coding and Reporting state this is an “unusual circumstance,” coders know that this situation occurs quite frequently. Kennedy believes that, as a general rule, some coders sequence the lower-weighted condition as the principal diagnosis, fearing criticism or rejection by retrospective auditors.
“Although this is not an error, there are opportunities to compliantly sequence other higher-weighted diagnoses that coexisted at the time of admission,” he says.
For example, Coding Clinic, First Quarter 2012, pp. 7-8 states that when documented and when clinically supported, either decompensated atrial fibrillation or heart failure may serve as the principal diagnosis. However, some coders sequence heart failure first, resulting in a lower-weighted DRG. The same concept applies to chronic obstructive pulmonary disease exacerbation and pneumonia, as discussed in Coding Clinic, First Quarter 2010, pp. 12-13.
This is mostly due to the coder’s fear and insecurity, says Kennedy. “A lower-weighted DRG is less likely to be retrospectively challenged and changed.”
Actual sequencing errors still occur frequently. Consider the following diagnoses:
- Myocardial infarction (MI) and coronary artery disease. If a patient has an MI at hospital A and is transferred to hospital B for a coronary artery bypass graft or percutaneous transluminal coronary angioplasty, coders must assign the MI as the principal diagnosis, says Kennedy. See Coding Clinic, Third Quarter 2009, pp. 9-10.
- Hepatic encephalopathy with a chronic noninfectious liver disease (e.g., alcoholic cirrhosis). Encephalopathy is the principal diagnosis, says Kennedy. Refer to Coding Clinic, First Quarter 2002, p. 3 for additional information.
Complications of care. The ICD-9-CM guidelines state that “when the admission is for treatment of a complication resulting from surgery or other medical care, the complication code is sequenced as the principal diagnosis.”
Failure to query. AHIMA’s newest query practice brief published in February 2013, Guidelines for Achieving a Compliant Query Practice, requires coders to query when clinical indicators for a documented diagnosis are absent from the record. The practice brief states the following:
When a practitioner documents a diagnosis that does not appear to be supported by the clinical indicators in the health record, it is currently advised that a query be generated to address the conflict or that the conflict be addressed through the facility’s escalation policy.
Failure to query when clinical validity is uncertain is now a coding error of omission, says Kennedy. “There is a new emphasis on coder accountability for the validity of documented diagnoses,” he says. “The 2008 brief inferred that coders do not challenge physician documentation. Well, that is turned upside down now. This is an area of tremendous risk and opportunity.”
Sepsis is another area of confusion, says Kennedy. The ICD-9-CM guidelines require coders to query when it’s unclear whether sepsis is present on admission. More specifically, the guidelines state the following:
Sepsis or severe sepsis may be present on admission, but the diagnosis may not be confirmed until sometime after admission. If the documentation is not clear whether the sepsis or severe sepsis was present on admission, the provider should be queried.
For example, a patient has an underlying infection (e.g., pneumonia). Sepsis is present but not documented until day three of the hospital stay. “It’s not clear that the doctor intended to say that sepsis was present on admission,” says Kennedy. “So while it was clinically there, it wasn’t explicitly documented as such.”
Laura Legg, RHIT, CCS, HIM director at Healthcare Resource Group in Spokane Valley, Wash., and AHIMA approved ICD-10 Trainer, , agrees that querying is important. “In the old days, coders were supposed to code what the physician documented. Now, coders have the responsibility of clinical validation,” she says. “I always tell coders to query, query, query. If you don’t query, it’s going to be counted as an error against you.”
Coders need to pay particular attention to CCs or MCCs documented only once in the chart. “It may be in a progress note or anywhere, and the coder codes it. Then the discharge summary doesn’t mention it. Recovery Auditors will deny that CC or MCC,” she says.
For example, a physician documents renal insufficiency throughout the chart. The discharge summary mentions acute renal failure as well as renal insufficiency. “If the coder hasn’t queried the physician to clarify the acute renal failure, the Recovery Auditor will call it renal insufficiency. It has to be consistently documented through the record,” says Legg.
Dropping the fifth digit or using invalid codes. Though these mistakes are relatively easy to spot, they can still occur frequently, says Legg. “I think this will be more of an issue in ICD-10 when we have more characters in the codes,” she adds.
Coding “history of” conditions as acute. Coders sometimes code conditions listed in a patient’s history as though they are acute conditions that existed during that particular stay, says Legg. Deep vein thrombosis is one example. “There’s a code for [history of] a deep vein thrombosis, and there’s also a code for the acute condition,” she says. If there’s any confusion as to whether a diagnosis is past or present, coders should query the physician directly.” Reference Coding Clinic, First Quarter 2011, pp. 20-21 for additional guidance.
Once coders are aware of any mistakes they make, they can take steps to avoid those errors going forward.
However, coders are only aware of their mistakes when the facilities in which they work invest the time and resources to perform internal audits, says Kennedy.
Hospitals should perform regular audits (ideally on a quarterly basis) and provide immediate feedback to coders, says Legg. This feedback should include a rationale for why the mistake is classified as an error. This includes applicable and specific references to the ICD-9-CM guidelines or Coding Clinic.
Editor’s Note: This article was originally published in the September issue of Briefings on Coding Compliance Strategies.