Predictive Accuracy of Sepsis-3 Definitions for Mortality

0
0
0

Predictive Accuracy of Sepsis-3 Definitions for Mortality Among Adult Critically Ill Patients With Suspected Infection

Sepsis is a life-threatening condition where the body’s response to infection damages its own tissues and organs. It’s a leading cause of death in intensive care units (ICUs), but defining it accurately has long been a challenge. For decades, clinicians used Sepsis-1—a definition based on systemic inflammatory response syndrome (SIRS) criteria like fever, high heart rate, or abnormal white blood cell counts. But SIRS is non-specific—it can happen with non-infectious conditions like trauma or burns—leading to overdiagnosis of sepsis.

In 2016, the Sepsis-3 consensus updated the definition to focus on organ dysfunction: sepsis is an acute increase in the Sequential Organ Failure Assessment (SOFA) score (a measure of organ damage) of 2 or more points due to infection. Septic shock, a more severe form, requires vasopressors (blood pressure drugs) to maintain circulation and a high lactate level (a sign of tissue damage) even after fluid resuscitation.

But does Sepsis-3 actually do a better job of predicting who will die from sepsis? A 2019 study by researchers from Yangzhou University-affiliated hospitals in China set out to answer that question.

The Study: Who, How, and Why

The team conducted a multicenter prospective cohort study—tracking patients over time—at five ICUs in four teaching hospitals in Jiangsu Province. From November 2017 to October 2018, they enrolled 749 adult ICU patients with suspected infection (defined as a doctor suspecting infection and starting antibiotics). Patients under 18 or with trauma, epilepsy, cardiogenic pulmonary edema, stroke, or active bleeding were excluded.

The goal was to compare how well Sepsis-1 and Sepsis-3 stratified mortality—meaning, could each definition correctly group patients by their risk of dying within 30 days? The team used:

  • Logistic regression to analyze risk factors for death.
  • Area under the receiver operating characteristic curve (AUROC)—a measure of how well a tool predicts an outcome (higher scores = better prediction, with 1.0 being perfect).

Key Results: Sepsis-3 Identifies Higher-Risk Patients

The study yielded three critical findings:

  1. Fewer patients were classified as sepsis with Sepsis-3—but those who were had higher mortality.

    • Sepsis-1 labeled 85.9% of patients as sepsis (644/749) and 48.3% as septic shock (362/749).
    • Sepsis-3 labeled 64.5% as sepsis (483/749) and 39.9% as septic shock (299/749).
    • But Sepsis-3 sepsis patients had a 41.8% 30-day mortality rate—10% higher than Sepsis-1 sepsis patients (31.8%).

  2. Sepsis-3 was more accurate at predicting death.

    • The AUROC for Sepsis-3 was 0.746 (meaning it correctly predicted 74.6% of deaths), compared to 0.620 for Sepsis-1.
    • Sepsis-3 also had higher sensitivity (72.8%—caught more true cases of high-risk sepsis) and specificity (69.0%—fewer false positives) than Sepsis-1 (63.3% sensitivity, 57.8% specificity).

  3. The SOFA score outperformed other tools.

    • The SOFA score (which measures organ dysfunction) had an AUROC of 0.828—far better than the SIRS score (0.609) or quick SOFA (qSOFA, a simplified screening tool, 0.694). This means SOFA is the most reliable way to predict death in ICU patients with suspected infection.

What This Means for Clinicians and Patients

Sepsis-3’s strength lies in its specificity: it excludes patients with mild inflammation (like those with SIRS but no organ damage) and focuses on those with life-threatening organ dysfunction. For example:

  • 22.6% of patients labeled as sepsis by Sepsis-1 were excluded by Sepsis-3—and these patients had a much lower mortality rate (15.3% vs. 38.2% for those meeting both definitions).
  • Sepsis-3’s requirement for lactate >2 mmol/L in septic shock helped identify patients at the highest risk of death—even though fewer patients were classified as septic shock, their mortality rate was higher.

In short: Sepsis-3 helps clinicians prioritize care for patients who need it most, while avoiding overtreatment of those with less severe illness.

Limitations to Consider

No study is perfect. This one had three key limitations:

  1. Short follow-up: Patients were only tracked for 30 days—long-term survival (e.g., 6 months or 1 year) wasn’t measured.
  2. ICU-only: The results apply only to ICU patients—not those in regular wards or emergency rooms.
  3. Mortality endpoint: The study used 30-day mortality as the main outcome, but some studies use 28-day or ICU mortality, which could affect results.

Conclusion: Sepsis-3 Is a Step Forward for ICU Care

For adult ICU patients with suspected infection, Sepsis-3 is more accurate at stratifying mortality than the old Sepsis-1 definition. It better identifies patients at high risk of death, helping clinicians make faster, more targeted decisions about treatment (like fluids, vasopressors, or antibiotics).

The study also confirms that the SOFA score—though more complex than SIRS or qSOFA—is the most reliable tool for predicting outcomes in this population.

Qi-Hong Chen, Jun Shao, Wei-Li Liu, Hua-Ling Wang, Lei Liu, Xiao-Hua Gu, Rui-Qiang Zheng. Predictive Accuracy of Sepsis-3 Definitions for Mortality Among Adult Critically Ill Patients With Suspected Infection. Chinese Medical Journal. 2019;132(10):1147–1153. doi.org/10.1097/CM9.0000000000000166

Was this helpful?

0 / 0