Gastrointestinal dysfunction in critically ill patients: A prospective observational study of frequency, severity,...

Gastrointestinal dysfunction (GD) is common in intensive care patients with a wide range of admission diagnoses. Whether GD increases the severity and worsens outcomes from critical illness remains contentious. The aim of this study was to determine the frequency, severity, and risk factors associated with the development of GD in intensive care patients and to correlate these with clinical outcomes.

Adult critically ill patients receiving enteral and/or parenteral nutrition with an expected ICU stay ≥72 h were prospectively studied between February 2019 to July 2020. Predefined GI signs and symptoms, ICU interventions, organ scoring, and clinical outcomes were documented from admission to ICU discharge or at 90 days. Data on GD using the Acute Gastrointestinal Injury (AGI) and Gastrointestinal Dysfunction Score (GIDS) scoring systems were collected, and associations between GD and clinical outcomes (ICU length of stay, mechanical ventilation duration, and mortality) were analysed using logistic regression and Poisson mixed-effects models with fixed and random effects, adjusting for age, illness severity, and other covariates.

Of 2247 ICU patients screened, a convenience sample of 100 patients were enrolled (75 general ICU, 25 cardiac ICU; 61 % male; median age 53 years [range 41-82]). All patients had at least one GI dysfunction sign/symptom. Gut dysfunction was present in all patients based on the AGI and 79 % by the GIDS. Severe GD occurred in 46 % (AGI) and 25 % (GIDS). A gut-related ICU admission diagnosis was present in 23 % of patients, which increased the odds of severe GD (AGI OR 9.8, 95 % CI 2.66-31.83, p < 0.001; GIDS OR 4.3, 95 % CI 1.4-13.0, p = 0.01). Elevated serum lactate was associated with GD severity (AGI OR 1.30, p = 0.04; GIDS OR 1.32, p = 0.02). Severe GD was associated with longer ICU stays (median 13 vs 10 days, p = 0.02) and mechanical ventilation duration (12 vs 10.5 days, p = 0.05). A 1-Litre fluid balance was associated with an increase in AGI grading of 5.1 % (p = 0.02) and GIDS of 7.9 % (p < 0.001). Aggressive enteral feeding increased AGI odds by 82 % (OR 1.82, p = 0.015). Multiple inotropes were associated with higher GIDS at 72 h (p = 0.023). ICU mortality was 21 %, with no differences by GD.

These findings support the importance of consistent GD assessment to guide clinical decision-making in critical care. There is an urgent need for a robust, standardised, and objective approach to GD assessment in ICU practice, one that accounts for severity, dynamic risk factors, and the potential to alter clinical outcomes through timely recognition and intervention.