OIG Sets Sepsis in its Sights: Part 1

Today, I am going to begin a two-part look at sepsis, starting with the U.S. Department of Health and Human Services (HHS) Office of Inspector General (OIG) focus on Medicare Inpatient Hospital Billing for Sepsis, brought to our attention by Dr. Ronald Hirsch; next week, I am going to write about updates to pediatric sepsis.

The introduction to the OIG’s plan to analyze Medicare claims for sepsis says some very impactful things. It asserts that “sepsis is the body’s extreme response to an infection,” and that “it is a life-threatening, emergency medical issue that often progresses quickly and responds best to early intervention.” It acknowledges that “the definition of and guidance for sepsis have changed over the years,” in an attempt to capture sepsis better. It identifies the issue that the Centers for Medicare & Medicaid Services (CMS) and the Centers for Disease Control and Prevention (CDC) “currently recognize an older, broader definition,” and the OIG expresses a concern that hospitals may take advantage of the broader definition because they are financially incentivized to land patients in the relatively higher-weighted sepsis Medicare-Severity Diagnosis Related Groups (MS-DRG).

Their study will analyze patterns in inpatient hospital billing for 2023 and assess the variability of sepsis billing among hospitals. They plan to compare costs using the broader definition (i.e., Sepsis-2, according to SIRS, or systemic inflammatory response syndrome, criteria) versus the narrower definitions of sepsis, that is, Sepsis-3.

On Feb. 23, 2016, The Third International Consensus Definition for Sepsis and Septic Shock (Sepsis-3) was published in the Journal of the American Medical Association (JAMA), written by Mervyn Singer, Clifford Deutschman, et al. Surviving Sepsis Campaign followed up in March 2017, with their acceptance of the definition as “life-threatening organ dysfunction caused by a dysregulated host response to infection.” Whereas Sepsis-3 defined the condition, Surviving Sepsis Campaign operationalized how to treat sepsis, issuing best-practice statements and utilizing the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) system. Prior to 2016, Surviving Sepsis Campaign (SSC) was where we had derived Sepsis-2.

This was over seven years ago. Why are we still adjudicating this? If the sentinel organizations use the same definition, why don’t all hospitals everywhere use it? Why hasn’t SIRS been put to bed?

There are multiple reasons, and the reason the OIG is leveraging is the money. The relative weight (RW) for sepsis without mechanical ventilation > 96 hours with a major comorbid condition (MS-DRG 871) in 2024 is 1.9826, and its “without MCC” counterpart (MS-DRG 872) has a RW of 1.0299. For comparison, Respiratory Infections and Inflammations with MCC (MS-DRG 177) has a RW of 1.6964, with CC (MS-DRG 178) is 0.9867, and MS-DRG 179 without CC or MCC has a RW of 0.7633. The MS-DRG for urinary tract infections w/wo MCC (MS-DRGs 689 and 690) have RWs, respectively, of 1.1744 and 0.8069. Hence, the most profitable DRG for a patient who is admitted with an infection is in the sepsis set.

What other reasons enter into this persistent utilization of SIRS? Clinician ignorance or clinicians clinging to “the way it has always been done” are obvious factors. The fact that the core measures bundle and New York has its own criteria that don’t align with either Sepsis-2 or Sepsis-3 are others. There are also practitioners who, out of an abundance of caution, would rather err on the side of picking up “early sepsis” than missing the boat and having a patient die, so they would rather liberalize the criteria to catch cases that turn out not to be sepsis. I am supportive of making a tentative diagnosis early, but the key to compliance is to remove the diagnosis once it has been ruled out.

We all designed our sepsis alerts to use the general variable SIRS criteria because it was easy, convenient, and ubiquitous, but the diagnosis of sepsis always included other clinical indicators. For instance, altered mental status, hyperbilirubinemia, thrombocytopenia, and coagulopathy were included in the Surviving Sepsis Campaign 2012 table of diagnostic criteria for sepsis. But everyone gets vital signs taken and a white blood cell (WBC) count drawn if they have a potential infection, so those were attractive as a screening diagnostic tool.

SIRS is a great marker for clinically significant disease; however, it is very non-specific. Conditions not infectious in etiology may demonstrate tachycardia, tachypnea, fever, or elevated WBCs. Patients with infections may demonstrate those symptoms without having progressed to sepsis. It may represent an appropriate response to the infection.

I once was rounding with a provider who documented sepsis as her clinical impression on a patient who was on the fourth day of their admission and was sitting in bed smiling and eating a sandwich. I asked the provider if the patient met the criteria of sepsis – were they “sick” with a capital S? The provider replied that this wasn’t part of the definition. I disagreed, saying it was so integral to the definition of sepsis that the experts didn’t think they needed to explicitly say it. It is my opinion that the indication of being “sick” with a capital S is organ dysfunction.

I had a reader ask me once why I don’t want providers to diagnose “sepsis without organ dysfunction. Don’t you think that it is better to catch it early than to miss it?” My response was that patients who have infections should be treated aggressively and appropriately, whether or not they have sepsis. If you nip the infection in the bud and avert the development of sepsis, good for you!

Dr. Hirsch used a great analogy that I would like to reuse. He said lots of patients have chest pain without enzyme markers for heart attack. They will be monitored and might be catheterized and stented. We don’t make the diagnosis of impending myocardial infarction (MI) and get paid in an MI DRG.

When I review records in the context of clinical validation denials, invariably, most of the cases I find righteously denied are billed as sepsis. Sepsis without organ dysfunction is…pneumonia or urinary tract infection or cellulitis. It doesn’t belong in the sepsis DRG, and I am going to predict that the OIG is going to agree with me.

If your hospital still uses SIRS criteria, it’s time for them to transition to Sepsis-3. It’s time for the state of New York to transition to Sepsis-3. If your institution uses Sepsis-3, but your providers document poorly or inconsistently, they should be educated and monitored.

Clinical documentation integrity specialists (CDISs) should put a program in place to perform clinical validation of the diagnosis of sepsis. Clinical validation denials are predictable, and a pain, but nothing compared to an unfavorable OIG determination.