[Bacteriologic and clinical analysis of nosocomial infections in patients from the intensive care unit]

The aim of this study is to evaluate what pathogens are mainly responsible for infection among all hospitalised at our ITU patients, to define the influence of antibiotic use on the aetiology of nosocomial infection. The research was conducted on a six-bedded surgical ITU in a 700-bed teaching hospital from January 1995 till June 1996. In August 1995 we changed infection control protocols (more stress on: handwashing with antiseptic soaps and routine microbiological culture for early prediction of infection) and antibiotic guidelines on our ITU (third generation cephalosporins, fluoroquinolones and Vancomycin were used only as the last option and never in prevention). 1276 samples for microbiological culture were obtained in routine manner. From 60% positive cultures 1216 strains were isolated (Tab. 1) and resistance to antibiotics were defined (Tab. 3). Monthly antibiotic consumption was expressed in defined daily dose (DDD) for 1000 hospitalisation-days. DDD = (X/Y)/Z x x 1000, were: X-cumulative antibiotic consumption during analysed period (g), Y-standard daily dose (g/24 h), Z-number of hospitalisation-days during analysed period (Tab. 2). Values were expressed as the mean +/- standard error (SE). Relationships between variables were analysed using linear correlation. All data were categorised for the frequency table. Statistically significant differences were considered to exist when calculated p values were less than 0.05. There were no statistically significant differences in the number of treated patients, length of stay and mortality rate on our ITU in 18 months. 58% of isolated strains were Gram-negative bacteria especially Pseudomonas aeruginosa (22%) and Acinetobacter spp. (16%) and Proteus spp. (9%). The commonly isolated Gram-positive bacteria were Enterococcus faecalis (14%), Staphylococcus aureus (12%)--of which 90% were MRSA. In 8% of cases we have isolated Candida spp. Monthly antibiotic consumption was displayed in table 2. Trend analysis confirmed reduction in Ofloxacin, fluoroquinolones and Colistin consumption over 18 months period (Fig. 2). We observed statistically significant decrease in amount of isolated Proteus spp. strains from 70 in I'95 to 31 in II'95 (p < 0.05) and 10 in I'96 (p < 0.005). This observation was confirmed in trend analysis (Fig. 1). We have observed in analysed period improvement in activities of third generation cephalosporins and fluoroquinolones (Tab. 3). We have analysed the influence of antibiotic use on the aetiology of nosocomial infection. Table 4 shows statistically significant correlation between Acinetobacter spp., MRSA, MRSE isolates and antibiotics consumption. Crosstabulated variables analyses confirm MRSE outbreaks in periods when excessive amount of Amikacin (p < 0.05 for chi 2 test, D Somer rate 0.59, V Cramer rate 0.67), aminoglicosides (p < 0.05 for chi 2 test, D Somer rate 0.57, V Cramer rate 0.59), imipenem (p < 0.05 for chi 2 test, D Somer rate 0.44, V Cramer rate 0.60) and total antibiotics consumption were high (p < 0.005 for chi 2 test, D Somer rate 0.67, V Cramer rate 0.81) (Fig. 3). This study illustrates the influence of antimicrobial therapy on the species and the resistance of strains isolated in nosocomial infection. Restrictive antibiotics policy does not affect ITU outcome. Better strategies for antibiotic administration in the ITU setting may improve their efficacy and control the spread of nosocomial infection caused by multi-resistant organisms. Therefore, restrictive antibiotic policy would be mandatory in each hospital and department.