Epidemiological impact of prolonged systematic use of topical SDD on bacterial colonization of the tracheobronchial tree and antibiotic resistance. A three year study.

OBJECTIVE: to evaluate the effect of the prolonged systematic use of topical SDD (tobramycin 80 mg, polymyxin E 100 mg, amphotericin B 500 mg) on ICU ecology as expressed by changes in tracheal colonization and bacterial resistances.
DESIGN: Prospective microbiological survey.
SETTING: Polyvalent ICU of a 2000 beds general hospital. PATIENTS: Data concerning bacterial strains isolated from the tracheo-bronchial aspirates of all the patients admitted to a polyvalent ICU over 3 consecutive periods of 12 months ('88, '89, '90) were prospectively entered in a database and subsequently analyzed. During a 3-year period 502 patients required artificial ventilation for more than 72 h and 332 of them ('89 and '90) were treated with SDD. All samples collected within 72 h from ICU admission were excluded as well as duplicate samples from the same patients. INTERVENTION: All the patients admitted to the ICU in '89 and '90 and submitted to artificial ventilation for at least 24 h were routinely treated with topical SDD without i.v. antibiotic prophylaxis; in '88 SDD was not employed. MEASUREMENTS AND
RESULTS: Criteria for collecting sputum samples and microbiological procedures remained unchanged throughout the study-time. Positive sputum were significantly less in '89 (80.8% versus 92.3% p < 0.001) and this was due to a very sharp decrease in the isolation of Gram-negative strains from 43-28% (-64% p < 0.0001) involving both: Enterobacteriaceae (-45%) and Pseudomonaceae (-77%). In 1990; however, a new increase in Gram negative was observed, although the overall amount of Gram-negative was still 49% lower in '90 if compared to '88 (p < 0.0001). A dramatic increase in Pseudomonas isolation was the only factor responsible for the "rebound" observed. An increasing percentage of Pseudomonas developed a resistance towards tobramycin and only 45% of Pseudomonas strains turned out to be sensible to tobramycin in '90 against 79% in '88. A similar trend was registered for all aminoglycosides with the exception of amikacin. Gram-positive colonizations tended to increase (+63%) (p < 0.0001) and this was mainly due to Coagulase negative Staphylococci (+290% p < 0.0001) and S. pneumoniae, whereas S. aureus isolations decreased (-18%) but not significantly.
CONCLUSIONS: Our data suggest that the prolonged use of SDD is associated with dramatic changes in ICU ecology: the incidence of Gram negative colonization is significantly diminished by SDD whereas Gram positive tend to increase. Pseudomonas developed an increasing resistance towards tobramycin one of the components of the SDD formula we used.