Study of bacterial, viral and fungal agents affecting pediatric patients with lower respiratory tract infection admitted to PICU in Alexandria

ABSTRACT

This prospective study included all patients with acute lower respiratory tract infections [LRTI] admitted to the Pediatric Intensive Care Unit [PICU], in Alexandria, from December 1999 till the end of February 2001. PICU nosocomial infections were excluded. The pediatric risk of mortality score [PRISM] and number of organ failure, on admission, were recorded. Nasopharyngeal swabs and endotracheal aspirates were obtained on admission for viral, bacterial, chlamydial and fungal detection. Disc diffusion antimicrobial susceptibility tests using 15 different antibiotic discs for isolated bacteria, were performed. Results showed that LRTI admissions represented 11.8% [55/465] with Pneumonias representing 89% of them. The mortality rate was 43.6%. Pathogen identification was successful in 91% of cases. Nine different bacteria, with Pseudomonas aeruginosa the commonest [29.1%] and Acinetobacter spp. the least common [1.8%], were isolated. Five different viruses, with respiratory syncitial virus [RSV] the commonest [12.7%], were identified. Chlamydia pneumoniae and Candida albicans were detected in 40% and 9% of cases respectively. Multiple / mixed infection represented 58% of cases. Antibiogram showed that, the most resistant isolates to all 15 antibiotics used were pseudomonas aeruginosa [0/15] followed by Klebsiella and Acinetobacter [1/15] while, the most sensitive were H. influenzae and M. catarrhalis [9/15], followed by pneumococci [8/15]. It showed also that the most effective antibiotic, in vitro, was ciprofloxacin affecting 6 out of 9 isolated bacteria [6/9] followed by ofloxacin [5/9]; the least effective were ampicillin and amoxicillin clavulinate [0/9]. The PRISM score and number of patients with multiple organ dysfunction syndrome [MODS], on admission, were significantly higher among non-survivors. A multiple logistic regression analysis revealed that PRISM score, on admission, was the only significant parameter in mortality prediction with an overall accuracy of 89% and it did not correlate neither with age [r = - 0.0872] nor with length of hospital stay [LOS] [r = -0.0622]. LRTI admissions to PICU showed high PRISM score, MODS, high percentage of multiple / mixed pathogens, remarkable in vitro resistance of bacteria to antibiotics and high mortality rate. Pre-ICU pediatric medical services should be improved