Updated on: November 28, 2016

Sepsis Yesterday, Now, and Tomorrow

Original story posted on: May 23, 2016
EDITOR’S NOTE: The late Robert S. Gold, MD tackled the issue of sepsis even before the advent of the 1992 SIRS criteria. He made certain that healthcare professionals should be aware that capturing the true clinical picture should be first and foremost on the agenda. This is the first in a series of articles on the subject of sepsis. Dedicated to Dr. Gold’s memory, this series has been written by Cesar M. Limjoco, MD, vice president of clinical services for DCBA, Inc., a consulting firm co-founded by Dr. Gold.

Decades ago, many providers were averse to giving a sepsis diagnosis unless blood cultures came back positive with the offending organism. This was also compounded by the fact that many infectious disease practitioners would never use the term “sepsis.” Instead, “bacteremia” was the preferred term. Medicare diagnosis-related group (DRG) denials were imparted to sepsis DRGs when there were negative blood cultures. It made for a really confusing quagmire that led to no-win situations for acute-care hospitals. 

Blood cultures are neither perfect nor absolute indicators of sepsis. Only about 40-60 percent of patients who have sepsis are found to have positive blood cultures, even with the presence of organ failure. Drug-resistant organisms, prior antibiotic therapy, and insufficient/incomplete workup may account for low rates of conclusive results. It’s similar to casting a fishing net in the ocean. There are a lot of factors that will determine a good catch rate. The nets may still come up empty despite the presence of fish in the area. 

Conversely, positive blood cultures may be part and parcel of an infection in a non-septic patient. For example, patients with bacterial growth on a prosthetic heart valve or patients with longstanding central venous catheters in place may have intermittent presence of bacteria retrievable in the bloodstream but be perfectly stable and asymptomatic. Or the blood cultures may have false-positive results, having grown contaminants or colonizers.

Sepsis has always been defined as an abnormal (also described as “toxic”) response to infection that leads to organ failure and death. The ominous predilection was a concern that demanded early determination and intervention. Systemic inflammatory response syndrome (SIRS) was the mechanism behind this deadly course of events. It involved activation of coagulation, prostaglandins, leukotrienes, and complement cascades leading to disseminated intravascular coagulation, endothelial cell damage, and, subsequently, organ failure. 

In 1992, the SIRS screening criteria were unveiled. They originated from cases studies in the critical care environment. These studies were formulated to gain the most impact in determining and obviating sepsis. There were four commonalities that emerged: fever (or hypothermia), leukocytosis (or leukopenia), tachycardia, and tachypnea. In critically ill patients, when two out of the four criteria were met in the presence of an infection, the likelihood that patients would slide down the slippery slope into multi-organ failure and death increased tremendously. It was a great screening tool to help impede infection mortality in critically ill patients being treated in critical care units. 

A great screening tool, it still is. But it is not by any means grounds for adherence to dogma! Two of the four SIRS criteria (e.g., fever and leukocytosis) can also be found in patients with uncomplicated infections. They do not necessarily define an abnormal or high-risk toxic response to infection. In the ICU setting, satisfaction of two of these four criteria has a high sepsis capture rate. But if you apply the same criteria to the general patient population, there will be a lot of non-septic patients that will be caught in the net. It is up to the astute provider to determine if a patient is truly septic (and it may take a day or so before one can truly be sure!) 

This screening tool has, in the past two decades, led to indiscriminate misdiagnosis of sepsis. Many patients going through the hospital doors that satisfied the SIRS criteria were being diagnosed with sepsis. People did not bother to delineate those patients whose criteria can be explained by other factors/conditions. Folks need to understand that abnormal values can mean a number of things. The inauspicious finding of lactic acidosis is not only seen in sepsis, but also in other cases of kidney or liver dysfunctions, poisoning, and severe asthma/COPD/asphyxia. One needs to have a keen clinical mind to decipher the root cause of any abnormal finding. 

In practice, even after sepsis was determined to be unlikely, the documentation of it being ruled out in favor of some other explanation was lacking. “Sepsis” continued to be copied and pasted forward, all the way to the discharge summary. Clinical documentation improvement (CDI) programs generated queries based on the SIRS criteria, and providers across the country may have been inadvertently led to the diagnosis of sepsis in error. The incidence of sepsis acute-care admissions more than quadrupled across the country for more than two decades. What a disaster!

The introduction of the 2001 international sepsis/SIRS definition marked an effort to try to fix the problem, but the damage was done. The recent sepsis definition version 3.0 further tightened the parameters to capture true sepsis, and now everyone is in an uproar. 

If only folks stuck to the clinical truth as the main goal of diagnosis capture.
Disclaimer: Every reasonable effort was made to ensure the accuracy of this information at the time it was published. However, due to the nature of industry changes over time we cannot guarantee its validity after the year it was published.
Cesar M Limjoco, MD

Dr. Limjoco has more than 25 years of experience as a consultant with significant expertise in the capture of severity of illness in clinical documentation.  He serves as vice president of clinical services of DCBA, Inc. a position he has held since 2005. Dr. Limjoco is a member of the ICD10monitor editorial board and makes frequent appearances on Talk Ten Tuesdays.

Related Stories

  • Supporting the Next Wave of CDI Professionals
    We often overlook the human component of metrics within our profession. I recently received some feedback on an article I wrote about the metrics used to measure clinical documentation improvement (CDI) performance. It reminded me that we often overlook the…
  • Proposed Biannual ICD-10 Implementation Dates: Weighing the Pros & Cons 
    Understanding the impact on business processes in the event that biannual ICD-10-CM and PCS updates are implemented is essential – as is your feedback.  With a comprehensive agenda covered at the virtual ICD-10 Coordination and Maintenance Committee meeting earlier this…
  • Some Risk-Adjustment HCC Basics for HIM Coding and CDI Professionals
    Compliance is a big part of the risk adjustment HCC. Everyone is welcoming the New Year, and I am among them. This is a good time to take a look at some basics that make up the Medicare Advantage (MA)…