Case 1: A 72 year old patient with severe COPD and history of CHF has uncomplicated bowel resection for colon cancer. He requires three days of mechanical ventilation due to decompensated heart failure and COPD with respiratory failure. 

Case 2: A 55 year old patient with CAD undergoes CABG. He is on a ventilator following surgery and extubated within 24 hours with no apparent respiratory complications. “Ventilator dependent respiratory failure” is documented by intensivist following surgery in the ICU.

Should respiratory failure be coded as a secondary diagnosis and what codes should be assigned?

Understanding when to correctly and compliantly assign codes 518.5x is very challenging and can have significant consequences.  Because the 518.5x codes are designated as MCCs, they are associated with a substantial increase in DRG reimbursement.  On the other hand, these conditions are also classified as major complications of care in publically-reported quality data.  Assignment of these codes may adversely affect quality scores and patient safety indicators for both surgeons and the hospital.

Case 1: Acute respiratory failure following surgery attributed to other underlying conditions

Code 518.5 was subdivided into three codes in 2011:

518.51 - Acute Respiratory Failure or Respiratory Failure (NOS) following trauma and surgery

518.52 - Other Pulmonary Insufficiency, not elsewhere classified, following trauma and surgery

518.53 - Acute and Chronic Respiratory Failure following trauma and surgery

The above new codes separated "respiratory failure" following trauma and surgery from "ARDS" and "pulmonary insufficiency" following trauma and surgery.  Index entries were revised from "due to" trauma and surgery to "following" trauma and/or surgery.

The Tabular listing for code 518.51, Acute respiratory failure or respiratory failure NOS following trauma and surgery, and code 518.53, Acute and chronic respiratory failure following trauma and surgery, has an exclusion
note:

Excludes: Respiratory failure in other conditions (518.81, 518.83-518.84)

This indicates that codes 518.51 or 518.53 should not be assigned when post-op respiratory failure is attributed by the provider to a pre-existing or other underlying condition (e.g., severe COPD, heart failure, aspiration pneumonia, pneumothorax).  This corresponds with the Coding Clinic, 1988, Third Quarter, reference for use of 518.5: "another description of ARDS is respiratory failure due to shock and trauma occurring in the presence of previously normal lungs."

However, acute respiratory failure (518.81) and acute and chronic respiratory failure (518.84) also have an exclusion note: Excludes: "Acute respiratory failure following trauma and surgery", creating a circular reference. This same circular reference continues in ICD-10.

Absent any further clarification, it appears that patients who have other pre-existing or underlying conditions documented as the cause of respiratory failure following surgery, should be assigned code 518.81 or 518.84 together with the associated condition rather than code 518.51 or 518.53.

This would result is assignment of codes 518.81 and 491.21 as secondary diagnoses in Case 1.

Case 2: Postoperative respiratory failure not supported by the medical record

Some physicians have been inclined to document or diagnose a pulmonary problem (e.g., "respiratory failure" or "pulmonary insufficiency") when patients require a ventilator support following surgery, often as an expected or routine practice inherent to the procedure, with no apparent confirmatory evidence of an acute pulmonary
problem.

Since Medicare does not permit hospitals to submit claims with codes for conditions that cannot be clinically validated, how should the CDS or coder handle such situations?

To validate the diagnosis, the patient must have acute pulmonary dysfunction requiring non-routine aggressive measures.  A patient who requires a short period of routine ventilatory support during surgical recovery does not have acute respiratory failure, and a code for it should not be assigned on the claim.

In discussing post-op mechanical ventilation, Coding Clinic 2006, Quarter 2, page 8 states:

"A code should not be assigned for the mechanical ventilation when it is considered a normal part of surgery."

Following this same logic, a code should not be assigned for respiratory failure following surgery when mechanical ventilation is considered a normal part of surgery (routine, expected), unless there is verifiable clinical evidence of an acute pulmonary or other complicating condition actually causing respiratory failure.

Therefore in Case 2, even though the provider documented "Ventilator dependent respiratory failure", a code for respiratory failure (post-op or otherwise) would not be assigned since the diagnosis is not substantiated and the patient was "routinely" extubated within 24 hours having had no apparent respiratory complications.

Hospitals should have a specific written policy addressing situations where a diagnosis documented by a provider cannot be clinically substantiated by verifiable evidence within the medical record. It is no longer acceptable to simply assign codes for documented conditions that cannot by clinically validated. To do so, risks regulatory scrutiny with possible sanctions or penalties on the hospital.