Complex case: Predictive analytics for audits

Wednesday, September 20, 2017

A two-year-old child is brought in to the emergency department with a case of croup, experiencing inflammation in the lungs as well as coughing. The child is treated with Dexamethasone, a corticosteroid that reduces inflammation. The child’s coughing subsides and he is discharged.

The revenue integrity department begins reviewing cases using an application for predictive analytics. When the program runs, an alert pops up for the patient with croup. The application indicates that the billing code does not match the care provided. The account is reportedly missing an injection billing code for the act of administering the Dexamethasone either via an intravenous or intramuscular route. In this case, injecting the medication seems unnecessary. If the patient was so ill that he needed to be injected with this medication, how could he be well enough to be discharged so quickly?

The documentation is reviewed for accuracy. Upon inspection of the chart, notes indicate that the nurse administered Dexamethasone orally, using the injectable form of the medicine mixed with juice. This doesn’t match up; the billing indicated that the injectable form of the medicine had been administered.

Recognizing that the facility should have billed an oral version of this medicine, an audit is performed to correct the billing. Subsequently, the facility built a medication charge into the system for the oral form of Dexamethasone. The oral form of the medication doesn't have a billing code attached which would trigger an alert indicating the injection billing code was missing from the claim.

Editor’s Note: Adapted from an interview with Felicia Ziomek, MBA, BSN, CHFP, nurse auditor and CDM coordinator at Stanford Healthcare – ValleyCare in Dublin, California.