The high prevalence of nosocomial infections in critically ill ICU patients is associated with high antibiotic consumption. Besides its economic impact, there is the constant threat of selection and induction of antibiotic resistance. Surveillance studies recording the incidence of infections, antibiotic use, and antimicrobial susceptibilities of pathogens supply vital information regarding infection control and prevention of antibiotic resistance. In order to analyse antibiotic consumption we recorded antibiotic use in a general ICU during one year by categorizing the indications for antibiotic use into three groups; (i) prophylaxis; (ii) therapy for a bacteriologically proven infection (BPI); (iii) therapy for a non-bacteriologically proven infection (non-BPI). Bronchoscopic techniques were used to diagnose pneumonia. In practice, BPI must be treated, but a proportion of antibiotics prescribed for non-BPI may be unnecessary. The subdivision in BPI and non-BPI may help to identify these cases. In all, 515 patients were admitted to ICU and 36% of these had at least one infection. Of all infections, 53% were ICU-acquired and 99% of these occurred in intubated patients. Antibiotics were prescribed in 61% of admissions. Of all antibiotics prescribed for therapy, 49% were for respiratory tract infections, 19% for abdominal infections and 13% for sepsis eci. Categorized by indication, 59% of all antibiotic prescriptions were for BPI, 28% for non-BPI and 13% for prophylaxis. A theoretical reduction of 25% in the number of non-BPI prescriptions would result only in a 7% decrease of total antibiotic use. We conclude that almost all antibiotics prescribed were for intubated patients and for BPI. Respiratory infections were the single most common infection and accounted for 49% of all antibiotics used. Therefore, in our setting, prevention of respiratory tract infections is probably the most effective mode to reduce antibiotic use.

Risk factors for the development of ventilator-associated pneumonia (VAP) and colonization of the respiratory tract and stomach with enteric gram-negative bacteria (EGB) and Pseudomonadaceae were determined in 141 ventilated patients using univariate analysis and the Cox proportional hazards model. VAP was caused by EGB in 14 patients (10%), and by Pseudomonadaceae in 19 patients (13%). The duration of ventilation was a significant risk factor for VAP caused by EGB and Pseudomonadaceae, and for acquired colonization in oropharynx, stomach, and trachea with these species. Of 20 other variables, oropharyngeal colonization with EGB on admission (hazard ratio [HR] = 4.5) and an infection on admission (HR = 2.7) were selected as risk factors for VAP caused by EGB. Acquired colonization with Pseudomonadaceae in oropharynx (HR = 5.0) was the most important risk factor for VAP caused by these species. Gastric colonization with EGB or Pseudomonadaceae were no risk factors for VAP. For acquired oropharyngeal colonization with EGB only the duration of ventilation was a risk factor, whereas preceding colonization of the trachea with Pseudomonodaceae and duration of ventilation were risk factors for acquired oropharyngeal colonization with these species. In the Cox model, only the duration of ventilation was significantly related to acquired gastric colonization with EGB. Preceding colonization of the orophayrnx and of the trachea with Pseudomonadaceae were risk factors for acquired colonization with these species in the stomach. Twelve patients with VAP (46%) and 38 without VAP (33%) died (p = 0.21). In conclusion, duration of ventilation and colonization of the upper respiratory tract are the most important risk factors for VAP caused by EGB or Pseudomonadaceae.

No abstract

After stratification on initial intragastric pH into two groups, patients from both groups were randomly assigned to receive either antacids (a suspension of aluminum hydroxide and magnesium hydroxide), 30 mL every 4 h, or sucralfate, 1 g every 4 h. Continuous intragastric pH monitoring was performed in 112 patients (58 antacids, 54 sucralfate). Using predetermined criteria, colonization of stomach, oropharynx, and trachea, and the incidence of VAP were assessed. Altogether, 141 patients were included (74 receiving antacids, 67 sucralfate) and continuous intragastric pH monitoring was performed in 112 patients, with a mean of 75 h per patient. The median pH and the percentage of time with a pH < 4.0 were calculated from each measurement. No significant differences in median pH values (4.7 +/- 2.2 and 4.5 +/- 2.0 for antacids and sucralfate, respectively) were observed. Median pH values were higher in patients with gastric bacterial colonization than in noncolonized patients (5.5 +/- 2.1 and 3.3 +/- 2.0, p < 0.01), but colonization of oropharynx and trachea was not related to intragastric acidity. Thirty-one patients (22%) developed VAP, with a similar incidence in both treatment groups. In addition, antibiotic use, duration of hospitalization, and mortality rates were similar in both groups. Enteral feeding did not change intragastric acidity significantly but increased gastric colonization with Enterobacteriaceae, without influencing oropharyngeal and tracheal colonization. Antacids and sucralfate had a similar effect on intragastric acidity, colonization rates, and incidence of VAP. Intragastric acidity influenced gastric colonization but not colonization of the upper respiratory tract or the incidence of VAP. Therefore, it is unlikely that the gastropulmonary route is important for the development of VAP.

Continuous enteral feeding (CEF) has been associated with decreased gastric acidity, thereby stimulating gastric colonization and ventilator-associated pneumonia (VAP). Intermittent enteral feeding (IEF) could induce a temporary increase in gastric acidity and decrease the risk of VAP. We studied the influence of IEF (18 h/d) and CEF (24 h/d) on gastric and oropharyngeal colonization. Sixty patients were randomized to receive either IEF or CEF, and continuous intragastric pH monitoring was performed in 50 patients. Median intragastric pH levels were similar before enteral feeding was instituted (pH 2.5 for CEF and pH 2.4 for IEF), and median pH values increased slightly after institution of nutrition (NS). In patients receiving IEF, median pH decreased from 3.5 to 2.2 (p = 0.0002) when enteral feeding was discontinued. However, despite this, 80% of the patients in both study groups were colonized in the stomach after 7 d in study. In addition, colonization rates of the oropharynx and trachea, the incidence of VAP, and mortality were similar in both study groups. IEF was less well tolerated than CEF. We conclude that almost all patients receiving enteral feeding are colonized in the stomach with gram-negative bacteria. IEF resulted in a slight decrease in intragastric pH without influencing rates of colonization and infection of the respiratory tract.