BackgroundWhile gastrointestinal problems are common in ICU patients with multiple organ failure, gastrointestinal failure has not been given the consideration other organ systems receive. The aim of this study was to evaluate the incidence of gastrointestinal failure (GIF), to identify its risk factors, and to determine its association with ICU mortality.MethodsA retrospective analysis of adult patients (n = 2588) admitted to three different ICUs (two ICUs at the university hospital Charité-Universitätsmedizin Berlin, Germany and one at Tartu University Clinics, Estonia) during the year 2002 was performed.Data recorded in a computerized database were used in Berlin. In Tartu, the data documented in the patients' charts was retrospectively transferred into a similar database. GIF was defined as documented gastrointestinal problems (food intolerance, gastrointestinal haemorrhage, and/or ileus) in the patient data at any period of their ICU stay. ICU mortality, length of stay, and duration of mechanical ventilation were assessed as outcome parameters.ResultsGIF was identified in 252 patients (9.7% of all patients). Only 20% of GIF patients were identifiable at admission. GIF was related to significantly higher mortality (43.7% vs. 5.3% in patients without GIF), as well as prolonged length of ICU stay (10 vs. 2 days) and mechanical ventilation (8 vs. 1 day), p < 0.001, respectively. Patients' profile (emergency surgical or medical), APACHE II and SOFA scores and the use of catecholamines at admission were identified as independent risk factors for the development of GIF. Development of GIF during ICU stay was an independent predictor for death.ConclusionGastrointestinal failure represents a relevant clinical problem accompanied by an increased mortality, longer ICU stay and mechanical ventilation.

The present results strongly suggest that data collected postoperatively using established scoring systems as well as documented events of high clinical impact for risk assessment and quality control are reliable predictors of prolonged ventilation.

We studied how lowering a syringe pump and changing the outflow pressure could affect syringe pump output. We experimentally reduced the height of three different syringe pump systems by 80 cm (adult setting) or 130 cm (neonatal setting), as can happen clinically, using five flow rates. We measured the time of backward flow, no flow and the total time without flow. An exponential negative correlation was present between infusion rate and time without flow (r2=0.809 to 0.972, P<0.01). Minimum flow rates of 4.4 and 2.6 ml h(-1) respectively were calculated to give 60 and 120 s without infusion. The compliance of the different syringe pumps and their infusion systems was linearly correlated with the effective time without infusion (r2=0.863, P<0.05). We conclude that the height of the syringe pumps should not be changed during transportation. If vertical movement of the syringe pump is necessary, the drugs should be diluted so that the flow rate is at least 5 ml h(-1).