Detecting Sepsis: Are Two Opinions Better Than One?

Area of Science:

• Hospital Medicine
• Infectious Disease
• Critical Care Medicine

Background:

• Sepsis diagnosis relies on objective criteria and subjective suspicion of infection.
• Discrepancies in provider suspicion can impact patient outcomes.
• Early and accurate sepsis detection is crucial for effective treatment.

Purpose of the Study:

• To evaluate agreement between Registered Nurse (RN) and Medical Doctor/Advanced Practice Professional (MD/APP) suspicion of infection.
• To correlate provider suspicion with patient outcomes, including progression to severe sepsis or shock.
• To identify patterns of provider disagreement in sepsis suspicion.

Main Methods:

• Prospective data collection from a general medicine ward.
• RN suspicion assessed every 12 hours.
• MD/APP suspicion defined by antibiotic orders or new blood/urine culture orders within 12 hours prior to nursing screen.
• Systemic Inflammatory Response Syndrome (SIRS) criteria used for patient stratification.

Main Results:

• 11,489 screens from 1386 patients analyzed; 32.6% met SIRS criteria.
• Infection suspected by both RN and MD/APP in 5.8%, RN only in 22.2%, MD/APP only in 7.2%, neither in 64.7%.
• Overall agreement rate for suspicion was 80.7% (κ = 0.11).
• Progression to severe sepsis/shock was highest (17.7%) when both providers suspected infection in SIRS-positive patients, reduced with single-provider suspicion (6.0%), and lowest (1.5%) when neither suspected infection (P < 0.001).

Conclusions:

• Provider disagreement in suspected infection is frequent.
• RNs tend to suspect infection more often than MDs/APPs.
• A collaborative approach to sepsis detection may enhance diagnostic accuracy and timeliness, potentially improving patient outcomes.