Early versus Delayed Enteral Nutrition in Septic Shock: A Target Trial Emulation Study

Why It Matters

Early initiation of enteral feeding can improve survival and respiratory outcomes in septic‑shock patients, especially those with modest hemodynamic instability, guiding clinicians on timing decisions. The study also demonstrates how target‑trial emulation can yield more reliable evidence from observational ICU data.

Key Takeaways

• Early EN (0‑48 h) cut 28‑day mortality by 5% points.
• Delayed EN increased mortality risk (HR 1.21) and reduced ventilator‑free days.
• Benefit of early EN strongest at lower norepinephrine doses.
• Target trial emulation reduces immortal time bias in ICU studies.

Pulse Analysis

Guidelines have long advocated early enteral nutrition for critically ill patients, yet evidence specific to septic shock has been fragmented. Traditional observational studies often suffer from immortal time bias—where patients must survive long enough to receive the intervention—skewing results. By employing a clone‑censor‑weight framework, the researchers recreated a randomized trial environment within the MIMIC‑IV dataset, aligning baseline characteristics and time‑varying clinical variables to isolate the true effect of feeding timing.

The analysis revealed that initiating EN within the first 48 hours lowered 28‑day mortality by five percentage points compared with feeding between 48 and 96 hours. Moreover, patients receiving early nutrition enjoyed nearly two additional ventilator‑free days, a clinically meaningful gain in ICU resource utilization. A striking interaction emerged: the mortality advantage of early EN waned as norepinephrine doses rose, suggesting that hemodynamic stability modulates the safety and efficacy of early feeding. This nuance challenges a one‑size‑fits‑all approach and underscores the need for individualized nutrition protocols based on vasopressor requirements.

For clinicians, the findings advocate a proactive yet tailored EN strategy—start feeding promptly when vasopressor support is modest, but consider delaying when high‑dose norepinephrine signals profound circulatory compromise. From a research perspective, the study showcases the power of target‑trial emulation to generate high‑quality evidence from existing electronic health records, potentially accelerating guideline updates without costly randomized trials. Future work should validate these results across diverse health systems and explore mechanistic links between gut perfusion, immune modulation, and outcomes in septic shock.