Classifying diarrhea in critically ill patients through various criteria: a cohort study.

Background: The absence of consensus criteria for diarrhea in critically ill patients poses challenges, including an uncertain prevalence and inconsistent findings regarding the impact of diarrhea on mortality. This study aimed to examine the prevalence of diarrhea, the agreement among different diagnostic criteria, and their association with mortality.

Methods: A single-center cohort study was conducted among consecutive adult patients admitted to the intensive care unit (ICU) for at least three days between January 2017 and December 2018. The six diarrhea criteria evaluated were based on frequency, quantity, and consistency. These included the European Society of Intensive Care Medicine (ESICM) criteria (≥ 3 times and > 200 g/day loose or liquid stool), the World Health Organization (WHO) criteria (≥ 3 times loose or liquid stool), the Bristol Stool Chart Scale (BSCS) score of 6 or 7, and other quantity- or frequency-based definitions. Outcomes included: (1) prevalence of diarrhea according to each definition, (2) agreement between criteria, and (3) association between diarrhea and in-hospital mortality. Associations were assessed using multivariable Cox proportional hazards models, yielding hazard ratios (HRs) and 95% confidence intervals (CIs).

Results: Among 700 participants, 61% were men; the median age was 71 years. The prevalence of diarrhea ranged from 9 to 39%, depending on the criteria used. The WHO and ESICM criteria showed similar prevalences (18.7% and 15.1%, respectively) and high agreement (Kappa 0.87). However, both had weak agreement with the BSCS criteria (prevalence 39.3%; Kappa 0.52 and 0.43, respectively). In univariable analyses, the presence of diarrhea was associated with in-hospital mortality, regardless of the criteria used. In multivariable analyses, only the > 400 g/day loose or liquid stool, > 200 g/day loose or liquid stool, and BSCS criteria maintained this association; the adjusted HRs (95% CI) were 1.93 (1.29‒2.90), 1.78 (1.19‒2.64), and 1.73 (1.15‒2.60), respectively.

Conclusions: Diarrhea prevalence varied from 9-39% across definitions. WHO and ESICM, both frequency‑based, differed from BSCS and weight‑based criteria. Given the difficulty of accurate frequency counting in ICU patients, consistency‑ or weight‑based definitions may offer a more practical alternative for both clinical practice and research.