Address these common Q&As related to SIRS documentation

Association of Clinical Documentation Improvement Specialists, April 6, 2009

Editor’s note: This is the first article in a two-part series. Part one discusses how to combat denials due to short stays and insufficient SIRS documentation. Part two will discuss the documentation challenges coders and clinical documentation specialists face, as well as how coders should assign principal diagnoses.

Sepsis coding often involves some of the most complicated documentation and intricate coding guidelines for inpatient coders to decipher. To further intensify this already challenging area of coding, physicians often fail to document proper ICD-9-CM terminology (e.g. bacteremia versus septicemia or urosepsis versus sepsis). As a result, coders and clinical documentation improvement (CDI) specialists fail to recognize clinical indications in the medical record that should trigger physician queries.

During a February 5 HCPro audio conference titled “Sepsis Documentation and Coding: Clinical Indications, ICD-9 Guidelines, and Queries for Clarity,” callers had the opportunity to ask James S. Kennedy, MD, CCS, director at FTI Healthcare in Atlanta for guidance on a variety of coding challenges. Below is a summary of selected questions and answers.

Q: Some insurance companies have denied our sepsis claims due to a short length of stay. The physician documents sepsis, and we code it according to ICD-9 Coding Clinic guidelines. How do we approach our insurance companies about these denials?

A: Coders have an ethical responsibility to follow the ICD-9 coding guidelines. To the extent that a physician has explicitly documented sepsis—and clinical indicators are present—coders can code the condition. One point that’s really interesting is that the sepsis criteria do not include length of stay. Some patients have systemic inflammatory response syndrome (SIRS) without organ dysfunction and will probably respond fairly quickly to treatment, especially when they have expedited home health or other good social support. It is possible that these septic patients have a relatively low short stay.

In 2005, Medicare considered creating a separate DRG for severe sepsis, which is sepsis with organ dysfunction. However, it chose not to do this, claiming that coders had not reported ICD-9 code 995.92 with enough frequency to warrant this new DRG. Read Medicare’s stance on the issue on page 215 of the 2005 IPPS Final Rule.

I think Medicare has created some of the problem because it has included the following three conditions in the same DRG:

* SIRS without organ dysfunction
* SIRS due to infection without organ dysfunction
* SIRS due to infection with organ dysfunction

If a Recovery Audit Contractor (RAC) or medical director of an insurance company wants to deny reimbursement due to a short length of stay, refer the RAC or medical director to the ICD-9-CM coding guidelines as well as the 2005 IPPS Final Rule that groups sepsis with and without organ dysfunction rule into the same DRG.

The bottom line is that the physician must document in the medical record that the septic patient is systemically ill. This will substantiate that the patient is not well. I think the physician must also document at least three sepsis indicators to support his or her diagnosis.

However, when a physician documents sepsis but not include any reference to other clinical indicators, the only option coders have is to turn the documentation over to their compliance officer or quality officer. When the facility has a robust peer review process or a physician advisor to clarify this issue, negotiation with the treating physician regarding the definition of sepsis can take place. Refer to Coding Clinic, September-October, 1984, pages 8-11 for more information.

Q: I’m a pediatric hospitalist. We have many children with cellulitis, pneumonia, or appendicitis who have high white counts, fevers, and tachycardia. In cases for which there is with no positive blood culture and no organ failure, what should I report as the principal diagnosis? Should I report a local infection, sepsis, or SIRS?

A: This is a great question. I want to refer you to the International pediatric sepsis consensus conference article published in the January 2005 issue of Pediatric Critical Care Medicine, which discussed definitions for sepsis and organ dysfunction in pediatrics.

Physicians should refer to the indicators outlined in the article to support their impression for whether a patient looks sick. Coders don’t report the diagnosis of sepsis solely based on elevated white counts or any other single criteria. Coders assign diagnosis codes based on the physician’s explicit documentation.

This documentation must include diagnoses that support the physician’s judgment of whether a patient with cellulitis or appendicitis appeared toxic. Did they “look” septic? When a physician documents sepsis due appendicitis along with clinical indications (based on the pediatric sepsis definition), it is not wrong to code sepsis due to appendicitis.

When an outside entity holds a facility accountable for this diagnosis, the most important supporting element is whether the physician explicitly documented sepsis as well as whether the patient appeared sick or toxic. He or she should also document sufficient clinical indicators as defined in the literature to support the concept of sepsis due to a localized infection.

Q: A patient with a urinary tract infection appeared septic, and she also fit the criteria for sepsis. The physician documented the patient had SIRS but not sepsis. From a coding standpoint, SIRS is sepsis. So I’m confused about the physician’s thought process. Would I incur any risk if I reported ICD-9 code 038.9 as the principal diagnosis and then the SIRS as a secondary diagnosis?

A: SIRS is not synonymous with sepsis. SIRS is sepsis only when the physician documents it as being due to an underlying infection. Report ICD-9 code 995.90 for SIRS alone. Follows the usual coding conventions for sepsis when reporting SIRS due to an infection.

When a physician only writes SIRS in the medical record, coders should automatically query for its underlying cause. Coders should ask the physician the underlying cause of SIRS to determine whether there is SIRS due to infection (i.e., sepsis) or whether there is SIRS that is due to non-infection. The fact that the physician only wrote SIRS is not specific enough to assign the appropriate code, and it could make the hospital vulnerable to a denial.

Note that just because SIRS and a urinary tract infection coexist does not mean that they’re linked. Look in the ICD-9-CM alphabetic index under SIRS, which states the following:

SIRS (systemic inflammatory response syndrome) 995.90
due to (emphasis added)
infectious processes 995.91
with organ dysfunction 995.92
non-infectious process 995.93
with acute organ dysfunction 995.94

There is no an automatic linkage between SIRS and infections like we have with diabetes with osteomyelitis or gangrene. The physician must explicitly state what condition causes the SIRS.

Query the physician to obtain the link. Also ask whether it was present on admission (POA). When it’s not crystal clear that sepsis or SIRS was POA, coders are required to query. Failure to do so is a violation of the ICD-9 guidelines.

When the physician does not document that the patient was systemically ill, he or she is losing one of the major clinical indicators that would support him or her during a retrospective review.

Editor’s note: James S. Kennedy, MD, CCS, is a director of FTI Healthcare in Atlanta. E-mail him at james.kennedy@ftihealthcare.com.

Most Popular