The impact and characterisation of clinical diarrhoea in Intensive Care Units – A retrospective study

Abstract

Objective: To evaluate the impact and clinically characterise patients who developed diarrhoea in an Intensive Care Unit (ICU).
Design: A 1-year (2004) retrospective, case-control study.
Methods: patients with an ICU stay longer than 24 hours and
with enteric tube feeding, were included. Diarrhoea was defined
as 3 or more loose watery stools during a 24-hour period. Patients
with colectomy or ileostomy were excluded. From a total of 461
patients admitted, 140 met the inclusion criteria; of these, 52
developed diarrhoea. For clinical characterisation the following
parameters were evaluated: age, gender, ICU stay, severity scores
(APACHE II and SAPS II), therapy with antibiotics, serum urea and
creatinine, haemoglobin (Hg), serum proteins, co-morbilities and
ICU mortality. Patients with and without diarrhoea were compared
regarding the above parameters. Comparison was performed with
student t- test; correlation between numerical variables was performed with linear regression analysis; logistic regression analysis
was performed to identify independent predictive parameters for
mortality in patients with diarrhoea.
Results: Patients with diarrhoea had lower Hg (p=0,007), higher
serum urea (p=0,014) and higher mortality (p<0,001). Deceased
patients with diarrhoea presented a lower Hg (p=0,001) and a
higher serum urea (p=0,0014) and creatinine (p=0,03). Hg was
independently related to mortality in patients with diarrhoea (p=0,002).
Conclusions: Patients who developed diarrhoea in an ICU presented higher mortality, lower Hg, higher serum urea and more
frequent cephalosporin use (p=0,02). Deceased patients with
diarrhoea presented lower Hg levels and higher serum urea and
creatinine. Hg was independently predictive of death in patients
with diarrhoea.