Nosocomial bloodstream infection in pediatric patients: Siriraj Hospital, Bangkok; 1996-1999.

A retrospective study on nosocomial bloodstream infection (NBSI) in pediatric patients hospitalized at Siriraj Hospital from January 1996 to December 1999 was performed. Of the 18,087 blood specimens sent for culture, 533 (3%) were positive for organisms after 72 hours of hospitalization and were defined as NBSI. The rate of NBSI detected in blood culture specimens was highest among neonates (5.2%). Gram-positive cocci and gram-negative rods caused NBSI in an equal proportion (46% and 44% respectively) and Candida caused 10 per cent of NBSI. Coagulase-negative staphylococci was the most common pathogen followed by K. pneumoniae and Enterobacter. Antibiogram showed that 15 of the 35 (43%) S. aureus identified were methicillin-resistant. Only 35-38 per cent of Enterobacteriaceae were sensitive to cefotaxime or ceftazidime. Cefoxitin was still effective against 95 per cent of K. pneumoniae. Compared with other third generation cephalosporins, combination of cefoperazone and betalactamase-inhibitor (sulbactam) possessed an increased in vitro efficacy against K. pneumoniae, Enterobacter, E. coli, Acinetobacter and non-fermentative gram-negative rods. Resistant rate of amikacin among all gram negative rods was 25-69 per cent. Ciprofloxacin sensitivity varied from 62-100 per cent among all gram-negative rods. Imipenem was excellent against all gram-negative rods with the sensitivity of 80-100 per cent. Epidemiological data of this study is important for the decision of the appropriate empirical antimicrobial treatment in our hospital.

Evaluation of susceptibility status of invasive pneumococcal isolates to various antibiotics and risk factors associated with invasive penicillin-nonsusceptible pneumococcal infection: Bangkok 1997-1998.

The antibiotic susceptibility pattern of Streptococcus pneumoniae isolated from specimens of invasive infections was examined at Siriraj Hospital, a tertiary care center in Bangkok, during December 1996 April 1998. The percentage of S. pneumoniae isolates intermediate and resistant to various antibiotics were: penicillin, 25% and 21%; amoxicillin-clavulanate, 24% and 0%; cefuroxime, 6% and 36%; cefotaxime, 6% and 1.4%; ceftibuten, 5% and 42%; imipenem 22% and 0%; co-trimoxazole, 6% and 41%; chloramphenicol, 2% and 26%; erythromycin, 12% and 16%; azithromycin, 0% and 30%; and roxithromycin 0% and 33%. Most of the penicillin-nonsusceptible S. pneumoniae (PNSP) were also nonsusceptible to other antibiotics except cefotaxime, and imipenem. The isolates from respiratory specimens have a higher rate of resistance to all antimicrobial agents with a significant rise in MIC50 of beta-lactam antibiotics. There was no difference in the outcome of infections caused by penicillin-susceptible and -nonsuscetible S. pneumoniae. The only identifiable risk factor associated with PNSP infection was prior use of antibiotic within 3 weeks.

Comparison of BACTEC automated blood culture system and conventional system in hospitalized pediatric patients.

A prospective study in pediatric patients compared the BACTEC system, an automated blood culture system using Bactec Ped Plus/F broth media, with the conventional system using Brain Heart Infusion broth media. Each single blood sample drawn for culture from hospitalized children was evaluated by both systems simultaneously. Of 244 pairs of blood samples, 44 (18%) were positive for microorganisms. Of these isolates, 24 (55%) were detected by both systems, 10 (23%) were detected by the BACTEC system only, and 10 (23%) were detected by the conventional system only. The mean turn around time of the BACTEC system (0.56 +/- 0.34, range 0.08-1.55 days) was significantly shorter than that of the conventional system (3.36 +/- 2.72, range 1-7 days, p < 0.001). Seven isolates strongly suspected to be due to contaminants grew out after 5 days of incubation and were detected by the conventional system only. In this study the BACTEC system and the conventional system were equally effective in detecting microorganisms in the patients' blood samples even if antibiotic therapy had been previously administered. However, the results from the BACTEC system were available much sooner and less likely to be contaminants.

Bacterial pathogens (non-Mycobacterium) from sputum culture and antimicrobial susceptibility.

Sputum culture of patients at Siriraj Hospital, Bangkok was 49.84% positive for bacterial pathogens in 1994 and 40.95% in 1995. The average incidence of gram-negative rods was 3.11 fold more than the combination of gram-positive cocci and gram-negative cocci. The most common gram-negative rod was Pseudomonas aeruginosa, followed by either Klebsiella pneumoniae or Acinetobacter anitratus depending on year. The most common coccus was Staphylococcus aureus. From both years, the number of Haemophilus influenzae, Streptococcus pneumoniae, Burkholderia pseudomallei and Nocardia spp isolated were 122, 93, 13 and 11 strains respectively. For antimicrobial susceptibility, P. aeruginosa was sensitive to ceftazidime, imipenem, gentamicin, amikacin, netilmicin, ciprofloxacin (range 56-89%). S. aureus (MSSA) was sensitive to common used drugs. S. aureus (MRSA) was sensitive to co-trimoxazole, fosfomycin, vancomycin (range 57-100%) and resistant to most drugs.

Helicobacter pylori: a comparison of CLO test and Giemsa's stain with culture in dyspeptic patients.

The prevalence of H.pylori in Thailand is high compared with Western countries and is the same as in China. We suggest either rapid urease test (CLO test) or Giemsa stain to be a rapid, reliable and convenient detection method for H.pylori and is also suitable for use in follow-up studies by gastroenterologists.

Advantage of Sandiford's counterstain in detection of Gram negative bacteria in clinical specimens.

Confirmation of bacterial in clinical specimens using Gram's stain with the Sandiford's modification was found to be more useful, especially when Gram-negative bacteria with other forms coexisted, than ordinary Gram's stain. It was useful for staining histopathological specimens also. We believe that this method deserves wider recognition and should be used as a standard procedure in laboratories.