High prevalence of bla(OXA)-23 in oligoclonal carbapenem-resistant Acinetobacter baumannii from Siriraj Hospital, Mahidol University, Bangkok, Thailand.

Acinetobacter baumannii has emerged in health care settings as a pandrug-resistant pathogen. Carbapenems are ineffective for treatment of this pathogen. Here we explored the molecular epidemiology and mechanism of carbapenem resistance in clinical isolates of carbapenem-resistant A. baumannii (CRAB). Antibiotic susceptibility by disk diffusion test was performed using imipenem and meropenem disk on 200 different clinical CRAB isolates. All isolates were resistant and gave inhibition zones of both antibiotic disks < or = 13 mm. Polymerase chain reaction (PCR) was carried out on 37 randomly selected isolates to amplify the common carbapenem hydrolyzing beta-lactamase genes (bla(OXA23)-like, bla(OXA-24/40)-like, bla(OXA-58), bla(IMP), and bla(VLM)). Clones were resolved by PCR-randomly amplified polymorphic DNA (PCR-RAPD) and plasmid profiling. PCR amplification and DNA sequencing revealed the existence of bla(OXA-23) downstream of the insertion element, ISAba1, in all 37 isolates tested. This segment was present in the carbapenem-resistant genomic resistant island AbaR4. These isolates were resolved into three RAPD types (Type I, 20 isolates; Type II, 16 isolates; and type III, 1 isolate) and 10 plasmid profiles. The CRAB isolates investigated here were oligoclonal and carbapenem resistance was conferred by the presence of bla(OXA-23). The presence of this beta-lactamase gene in many clonal isolates indicated its wide spread.

Antibacterial activity of carbapenem-based combinations againts multidrug-resistant Acinetobacter baumannii.

BACKGROUND: Multidrug-resistant (MDR) Acinetobacter baumannii are increasingly encountered and frequently susceptible only to colistin with their MIC values close to resistance breakpoint. Antibacterial activity of two carbapenem-based combinations were explored in order to overcome the bacterial resistance. MATERIAL AND METHOD: Thirty clinical isolates of MDRA. baumannii were employed to assess in vitro antibacterial activity of two carbapenem-based regimens. Imipenem combined with colistin and meropenem combined with colistin and sulbactam were the first and second regimens, respectively. All isolates were resistant to imipenem (MIC range: 8-128 microg/ml) and meropenem (MIC range: 64-256 microg/ml) but still susceptible to colistin (MIC range: 0.5-2 microg/ml). The MIC range of sulbactam was 4-64 microg/ml. None of the isolates produced metallo-beta-lactamase. RESULTS: Synergistic antibacterial effect of imipenem combined with colistin was observed against 100 percent of A. baumannii isolates by the checkerboard microdilution panel method. In a subsequent time kill study, the most active concentration of this regimen was the combination of imipenem at the fixed concentration of 32 microg/ml and colistin at the 1/4 of the MIC values of each isolate that exerted significantly higher bactericidal activity than imipenem at 32 microg/ml alone and colistin alone at the 1/4 of the MIC values. The scanning electron micrographs demonstrated major cell morphological change and cell wall destruction after 2-hour exposure to this combination. The triple combinations of meropenem, sulbactam and colistin showed synergy against 96.7 percent of MDR A. baumannii while double combinations of either meropenem and sulbactam, meropenem and colistin, and sulbactam and colistin showed synergy effects of 70%, 73.3% and 53.3%, respectively The time kill study using ten isolates also showed better killing effect by the triple combination than any of the double combinations. CONCLUSION: Antibacterial activity against MDR A. baumannii of imipenem plus colistin was superior over any single of the two agents. The addition of sulbactam to meropenem and colistin may further improve their antibacterial activity. The double or triple carbapenem-based combinations offer promising alternatives in the treatment of infections due to MDR A. baumannii.

Discriminatory powers of molecular typing techniques for methicillin-resistant Staphylococcus aureus in a university hospital, Thailand.

Discriminatory powers of various molecular techniques were evaluated for typing of methicillin-resistant Staphylococcus aureus (MRSA) isolated in Siriraj Hospital, Bangkok, Thailand. Thirty MRSA isolates were randomly selected in this study. They were characterized by pulsed-field gel electrophoresis, Clal-mecA and Clal-Tn554 polymorphisms, ribotyping, and PCR-based methods including SCCmec typing, spa and coa gene polymorphism, and repeat units in hypervariable region downstream of mecA. Individual molecular typing technique distinguished those MRSA isolates into 2 to 5 types. Eleven genetic backgrounds of MRSA isolates were elucidated by combination of typing methods with trimethoprim/sulfamethoxazole (TMP/SXT) susceptibility. Combination of all typing methods including TMP/SXT susceptibility yielded a discriminatory index of 0.94. Combination of PCR-based methods and TMP/SXT susceptibility, with the discriminatory index of 0.89, is a practical typing approach suitable for rapid epidemiological investigation of MRSA isolates in a hospital setting.

Intact pks15/1 in non-W-Beijing Mycobacterium tuberculosis isolates.

To determine whether intact pks15/1 is unique to the W-Beijing family, we investigated 147 Mycobacterium tuberculosis strains with different IS6110 genotypes. Intact pks15/1 was found in 87.8% of cerebrospinal fluid and 84.9% of sputum isolates. It was found not only in W-Beijing strains (approximate, equals 97%) but also in other genotypes (38.5%-100%).

Prevalence and impacts of nosocomial infection in Thailand 2001.

OBJECTIVES: To study the prevalence and impacts of nosocomial infection (N.I.) in Thailand. MATERIAL AND METHOD: A point prevalence study on N.I. was carried out in 42 hospitals across Thailand in March 2001. The impacts of N.I. were done in the same hospitals by matched control groups in a period prevalence study March 12-25, 2001. RESULTS: The point prevalence rate of N.I. in 42 hospitals involving 18,456 patients across Thailand in March 2001 was 6.4%. The prevalence was higher in male than female patients (7.8% vs 5.0%). The prevalence rates of over 10% were found in 4 hospitals. The infection rate was highest in surgical followed in rank by medical, pediatric and orthopedic departments (9.1%, 7.6%, 6.1% and 5.8%) respectively. The commonest site of the infection was the lower respiratory tract, followed by urinary tract, surgical site and skin and soft tissue (34.1%, 21.5%, 15.0% and 10.5%). Gram-negative bacteria were isolated in 75.3% and gram-postive 18.4%. Penicillins, cephalosporins, aminoglycosides were the most used antimicrobials (31.2%, 25.2%, 12.3%). A period prevalence study on 53,882 patients during a 2 week period in March 2001 showed an infection rate of 2.5%. By matched control group study, an episode of N.I. was associated with 10.1 to 12.5 extra hospital days. The cost of antimicrobials for treatment of an episode of N.I. was 5919.50 baht (148 U.S. dollars). Thirteen point eight per cent of patients with N.I. died, 6.7% directly due to N.I. CONCLUSION: Nosocomial infection is common in hospitalized patients in Thailand and is associated high mortality rate and economic burden.

Endemic Serratia marcescens in a neonatal intensive care unit.

OBJECTIVE: To study the endemicity of Serratia marcescens in a neonatal intensive care unit (N.I.C.U). MATERIAL AND METHOD: During the first 4 months of 2001, neonates in the N.I.C.U. in a teaching hospital were screened for S. marcescens by serial throat swabs and collections of other appropriate clinical specimens. Environmental cultures were also done in the same period. Isolated S. marcescens were tested for antimicrobial susceptibility and for genotyping by pulsed field gel electrophoresis. RESULTS: During the period, 104 neonates were studied. S. marcescens were isolated in 34.6% of the cases. Environmental cultures were positive for S. marcescens in 1.4%. There were 10 patterns of antibiogram of the 190 strains isolated. All strains belonged to pulsotype A. CONCLUSION: The study confirmed that S. marcescens was endemic in the N.I.C.U. and belonged to one genotype.

Antimicrobial susceptibility of community and hospital acquired bacteria.

OBJECTIVES: To study the antibiotic susceptibility of common community- and hospital-acquired bacteria in Thailand. MATERIAL AND METHOD: Eight common bacterial pathogens were studied in 24 hospitals across Thailand in 2002-2003. Isolates of clinically proven infections were tested for their susceptibility by agar-based disc diffusion method. RESULTS: A total of 9,091 isolates of target bacteria were studied. Community and hospital acquired bacteria accounted for 54.9% and 45.1% respectively. Community acquired Escherichia coli, Klebsiella pneumoniae, Acinetobacter spp., Enterobacter spp., Staphylococcus aureus were more susceptible to antimicrobials compared to hospital acquired strains. The difference in susceptibility of community-acquired vs hospital acquired Pseudomonas aeruginosa, Coagulase-negative staphylococci and Enterococcus spp. was less impressive indicating the spread of hospital strains into the community. Bacteria isolated from the blood stream were more susceptible to antimicrobials compared to those from the lower respiratory tract, urinary tract and surgical sites. Acinetobacter spp. and Enterococcus spp. were less susceptible to antimicrobials compared to others. CONCLUSION: Decreased susceptibility to antimicrobials was found in all bacteria tested. The susceptibility to commonly used antimicrobials of community-acquired bacteria decreased to a critical level indicating the widespread resistant bacteria to the community.

Catheter-associated urinary tract infection.

OBJECTIVES: To study the incidence, risk factors, clinical outcomes and antibiotic costs of catheter-associated urinary tract infections (CAUTI) in patients with indwelling urinary catheter for one week and longer MATERIAL AND METHOD: Patients in neurology and neurosurgery wards in a teaching hospital were studied. Patients with UTI before catheterization and in whom the catheter was removed before one week were excluded. Urine cultures were done immediately after catheterization and 3 times a week there after. Patients were followed for symptoms of UTI for 1 week after catheter removal, for 4 weeks without evidence of UTI or until discharge. RESULTS: One hundred and one patients met the inclusion criteria. The incidence of CA UTI was 73.3%. High incidence of CAUTI was found in the first 2 weeks after catheterization. About one-half of the patients with CAUTI had a single episode and were symptomatic. None of the 132 episodes of CAUTI were associated with secondary bacteremia. Risk factors for CAUTI identified were prolonged catheterization and change of the catheter Nosocomial pathogens were found in urine and yeast was the commonest. Eleven patients (14.9%) with CAUTI died and only in 2, CAUTI was considered a contributory factor for mortality. The cost of antimicrobials for treating one episode of CAUTI was 8,180 baht and this rose to 49,983 baht for CAUTI associated with concurrent infections at other sites. CONCLUSION: Catheter-associated urinary tract infection was common. Uropathogens were nosocomial micro-organisms with high incidence of resistance to antimicrobials. Impacts on morbidity, mortality and costs were substantial. Better management of urinary catheter is to be explored and implemented.

Bacterial contamination of vegetables served in hospitals.

OBJECTIVES: To study bacterial contamination of fresh vegetables before cleaning and before serving to patients in 14 hospitals. MATERIAL AND METHOD: Aerobic plate count was performed and emphasized on total viable aerobic bacteria, fecal coliform, fecal Escherichia coli and enteric pathogens in fresh vegetables including romaine lettuce, onion, parsley, celery and tomato before cleaning and before serving. Hospital nutrition officers who were involved in food purchasing and processing were interviewed. RESULTS: One hundred and six of 403 of fresh vegetable samples (26.3%) before cleaning were contaminated with > 10(7) colony forming unit per gram (CFU/gram) of viable aerobic bacteria, 106 of 178 samples (59.6%) contained MPN/fecal coliform >1,100 /gram, 78 samples (43.8%) contained MPN fecal E. coli >10/gram. Enteric bacteria were isolated from 7.2% of the total 304 samples including non typhoid Salmonella (1 sample), Vibrio cholerae non O1/O139 (7 samples) and Aeromonas species (14 samples). Forty of 396 ready to serve vegetable samples (10.1%) contained > 10(7) CFU/gram of viable aerobic bacteria. Seventy five of 183 (40.9%) samples contained >1,100 MPN fecal coliform/gram and 43 (23.5%) contained >10 MPN fecal E. coli/gram. Enteric bacteria were also detected in 7.6% of the samples including V. cholerae non O1/O139 (6 samples) and Aeromonas species (17 samples). There were three different ways in obtaining fresh vegetables to the hospitals: by auction (50%), wholesalers (21.4%) and retailers (14.2%). There were also different standards of transportation, packaging, delivery and food processing, particularly cleaning methods. CONCLUSION: Ready-to-eat fresh vegetables were contaminated in high percentages with microorganisms in the number that exceeded the standard. Better management is required to safeguard patients.

Microbial contamination of antiseptics and disinfectants.

OBJECTIVE: To study the bacterial contamination of antiseptics and disinfections in-use and the risk factors for contamination. MATERIAL AND METHOD: Bacterial contamination of antiseptics and disinfectants was done by culturing in-use solutions. Eight commonly used solutions were studied: alcohol 70%, chlorhexidine 4%, and 0.5%, povidone iodine 7.5% and 10%, tincture iodine 1-2%, lysol 2% and sodium hypochlorite 0.5%. RESULTS: The following risk factors for contamination were found : preparation by unskilled personnel, improper containers and prolonged use. Contamination with bacteria were found in 1.8% of 16,142 samples tested Highest rate of contamination was found in Lysol 2%. There was no contamination of povidone iodine 10% and tincture iodine 1-2%. Bacterial contamination of antiseptics and disinfectants was highest in provincial hospitals and was not found in university hospitals. The rates of contamination correlated with the duration of use. Most bacteria isolated were those found in the environment. CONCLUSION: The contamination of in-use antiseptics and disinfectants was as high as 1.8%. Risk factors for contamination were improper preparation and prolonged use.

A study on the application of pasteurization of medical equipment.

OBJECTIVES: To study the need of pasteurization of medical equipment and the possibility of production of pasteurizer in Thailand. MATERIAL AND METHOD: The need of pasteurization of medical equipment was studied by a set of questionnaires to heads of the central sterile supply department (CSSD) and head ward nurses in 29 hospitals across Thailand. Efficacy of pasteurization was demonstrated by disinfection with an imported pasteurizer. A pasteurizer was later produced by the researchers and had it tested for efficacy in disinfection. RESULTS: There were 26 items of medical equipment that could be disinfected by pasteurization. The number of the equipment was 6.2 pieces per bed per week. Disinfection of the equipment was done in C.S.S.D. as well as in patient's wards. The imported pasteurizer was efficacious in disinfection. The pasteurizer made by researchers was convenient for use, not expensive to manufacture and the operating cost for disinfection was 2 to 6 folds less than that done by ethylene oxide gas. CONCLUSION: Pasteurization is effective in disinfection and is applicable to certain heat labile medical equipments. A pasteurizer is not difficult to produce, cheap and the operating cost is low. Pasteurization should be more widely applied in Thailand

Serological study of Mycoplasma pneumoniae infections.

Mycoplasma pneumoniae antibody was determined in 811 sera of different patients admitted to Siriraj Hospital with respiratory tract infection from July 1, 2000 to August 31, 2003 by agglutination with gelatin particle agglutination test kit (SERODIA-MYCO II, Fujirebio Inc. Japan) in microtiter plates. Three hundred and three sera were positive (37.36%). The five most positive titer were found in patients 5-9 yr (40.26%), followed by patients 1-4 yr (24.75%), 10-14 yr (19.80%), 30-39 yr (5.28%) and 20-29 yr (3.96%). The positive titers ranged from 40 to > 20,480. Female:male ratio in positive patients was approximately the same (1.19:1). High titers (> or = 320) were found in 146 out of 303 patients (48.18%). The infection was mostly found in children aged 5-9 yr. Detection of antibody to M. pneumoniae infection showed that 37.36% of patients who were suspected of having atypical bacterial pneumonia were positive.

Bacterial isolation with on-site inoculation of ascites fluid into hemoculture bottle in spontaneous bacterial peritonitis.

OBJECTIVE: To compare the on-site bacterial inoculation of ascites fluid into hemoculture bottle with routine method in the patients who were preliminary diagnosed of SBP. MATERIAL AND METHOD: A retrospective analysis of case records during January-December 2001. RESULTS: A total of 673 specimens from 325 patients were retrieved from the data records at the Department of Microbiology, Siriraj Hospital in 2001. The neutrocytic ascites were found in 163 specimens (94 patients). The routine method and on-site inoculation into the hemoculture bottle were employed in 107 and 56 specimens respectively. Culture-positive neutrocytic ascites was found in the routine method 16 (14.9%) specimens and in the on-site inoculation 26 (46.4%) specimens (p < 0.0001). Among these samples, the two methods were simultaneously performed in 42 specimens of which 18 paired specimens were eligible for analysis. Positive culture was found in 2 samples in which the routine method and in an additional 5 samples in which on-site inoculation into hemoculture bottle method. Using Kappa analysis (e score = 0.328, 95% CI = -0.172 to 0.829) that can be interpreted the on-site inoculation method had a higher yield than the routine technique. Moreover, 21 cases also had their blood and ascites samples simultaneously collected and cultured. 4 of ten (19%) and 5 of eleven (23.8%) cases were found in the routine and on-site and direct inoculation groups respectively. These finding suggested that the severity of infection in among two groups were similar and unlikely to be the cause of the difference of the positive isolation rate in both groups. CONCLUSION: The on-site and direct inoculation of ascites into hemoculture bottle method had a significantly higher isolation rate than routine method (i.e. 46.4% versus 14.9% p < 0.0001) either with separated or paired samples of ascites.

An open label, randomized comparative study of levofloxacin and amoxicillin/clavulanic acid in the treatment of purulent sinusitis in adult Thai patients.

The objective of the study was to compare the clinical efficacy and bacteriological response of levofloxacin and amoxicillin/clavulanic acid (co-amoxiclav) in the treatment of purulent maxillary sinusitis. Sixty patients randomly received either levofloxacin 300 mg orally once daily (LEV group) or co-amoxiclav 625 mg three times a day (COA group) for 14 days. Thirty four patients were in the LEV group and 26 patients were in the COA group. The mean total symptom score was significantly decreased after treatment and was comparable between both groups. Radiological improvement was 61.8% in the LEV group (41.2% resolution, 20.6% improvement) and 61.5% in the COA group (26.9% resolution, 34.6% improvement). Pretreatment maxillary antral aspiration cultures were positive in 28 patients (82.4%) in the LEV group and 20 patients (76.9%) in the COA group. Bacteriological eradication was 78.5% in the LEV group and 70.0% in the COA group, which was not significantly different. In the LEV group, the eradication rate for major pathogens of acute sinusitis was 100% for H. influenzae (both betalactamase +ve and -ve), 100% for S. pneumoniae and S. aureus, 100% for Neisseria species, and 66.7% for P. aeruginosa. The eradication rate in the COA group was 75% for H. influenzae (both betalactamase +ve and -ve), 100% for S. pnumoniae and S. aureus, 50% for Neisseria species, and 0% for P. aeruginosa. There were no significant changes in vital sign measurements or hemato-biochemical parameters at the end of treatment as compared to baseline values, in both groups. Adverse events were found in 8.8% of patient in the LEV group and in 7.7% of patients in the COA group. Adverse events included nausea, abdominal pain, and diarrhea. All the adverse events in both groups were mild and resolved spontaneously. This study demonstrated that levofloxacin 300 mg orally once daily was as effective and safe as amoxicillin/clavulanic acid 625 mg three times a day in the treatment of maxillary sinusitis, either acute or acute exacerbation. Both drugs showed bacteriological efficacy that was not significantly different. The once daily dosage regimen is more applicable, convenience and has better compliance.

Prevalence and antimicrobial susceptibility of Pseudomonas aeruginosa mucoid and non-mucoid type.

Pseudomonas aeruginosa is a leading cause of nosocomial infections. One thousand two hundred and twenty strains of mucoid and non-mucoid types of P. aeruginosa isolated from different patients were examined at Siriraj Hospital from January 2001-October 2003. The prevalences of P. aeruginosa mucoid type and non-mucoid type were 3.6% and 96.4%, respectively. Susceptibility testing was performed by Kirby-Bauer disk diffusion method as recommended by NCCLS. The isolates with mucoid phenotypes were more susceptible than the non-mucoid isolates. The antimicrobial susceptibility pattern of both types should provide guidelines for the selection of appropriate drugs for treatment.

Detection of Clostridium difficile toxin A and B genes from stool samples of Thai diarrheal patients by polymerase chain reaction technique.

The prevalence of Clostridium difficile isolated from stools of Thai adult patients with suspected antibiotic-associated diarrhea (AAD) was 18.64 per cent. The recovery rate of toxin genes (tcdA and tcdB) by polymerase chain reaction (PCR) from stool samples yielded almost the same compared to the recovery rate of the toxin detection by enzyme immunoassay (EIA), which were 44.9 per cent and 46.7 per cent, respectively. Correlation of toxin gene detection by PCR and toxin detection by EIA was 90.6 per cent. All but one stool sample, the tcdA gene was detected together with the tcdB gene. Both genes were always detected together from tox gene-positive strains. Although, there were some discrepancy results for certain samples, the direct PCR-based-detection of C. difficile tox genes in stool samples seems to be the appropriate method for the diagnosis of C. difficile diarrhea. The PCR assay should be a recommended technique to be used routinely in laboratories. Further optimization of the technique to increase the sensitivity of the PCR assays is still needed. However, a quantitative isolation of the organism from stools of suspected antibiotic-associated diarrhea (AAD) or antibiotic-associated colitis (AAC) patients may give some evidence for clinicians in hospitals who cannot perform PCR-based or EIA-based techniques, since 48.6 per cent of the isolates were demonstrated as toxigenic strains.

The outbreak of Serratia marcescens bacteremia in a pediatric ward, Siriraj Hospital 1997.

Between October 20 and November 11, 1997, Serratia marcescens bacteremia was identified in 8 patients in a pediatric ward at Siriraj Hospital. The organism was isolated from 17 blood and 3 bone marrow specimens. The only common associated factor in these patients was that they all had received an intravenous fluid infusion. In the attempt to investigate the source of S. marcescens implicated in the outbreak, 108 specimens of intravenous fluid, 3 intravenous fluid bottle caps, 4 specimens from intravenous fluid tubing sets, 21 specimens of antiseptics used on the ward, 28 specimens of rectal swabs from patients on the ward, 1 sample of blood culture media prepared by the hospital for routine use, and 62 environmental specimens including hand swabs of the medical personnel, refrigerator, air conditioning, milk samples, room air, water sink, wooden splint and adhesive tape used to immobilize the intravenous access. Of 227 specimens sent for culture, S. marcescens was isolated from only one specimen collected from the in-use intravenous fluid given to a patient with Serratia bacteremia. S. marcescens was not found in any other surveillance culture. The 8 patients were placed under quarantine in the same room with an exclusive nursing team. With the investigation and intervention including monitoring for meticulous hand washing of the ward staff, the outbreak was stopped within 7 days. Although the investigation failed to discover the environmental reservoir of S. marcescens in this outbreak, the data suggested that intravenous fluid was probably the route of transmission and the medical personnel played an important role in spreading the infection.

Antiseptics for preventing omphalitis.

BACKGROUND: Omphalitis may cause serious complications and contribute to neonatal morbidity and mortality. From January 1997 to August 1998, the incidence of omphalitis in the Department of Obstetrics and Gynecology, Faculty of Medicine Siriraj Hospital had been increased from 0.9 to 17.4 per 1,000 live births. A prospective randomized trial using antiseptic applied directly to the umbilical stump was conducted aiming to reduce an epidemic outbreak of omphalitis in the newborn nursery. OBJECTIVE: To determine which antiseptic is appropriate for preventing omphalitis in the newborn infants. PATIENTS AND METHOD: Newborn infants delivered in the Department of Obstetrics and Gynecology, Faculty of Medicine Siriraj Hospital were randomized into group A (Triple dye) or group B (70% Alcohol). The infant with omphalitis was assessed by a pediatrician or a neonatology fellow. At home, the same antiseptic will be continually applied to the umbilical stump daily until a few days after cord detachment. Relative risk was calculated and statistical significance was tested by Chi-square test. RESULTS: Four hundred and twenty-seven infants were enrolled. Birth weight, gestational age and gender of the infants in both groups were not different. There were no known maternal risk factors for omphalitis. Omphalitis was observed in 9/213 (4.2%) infants in group A and 23/214 (10.7%) infants in group B. The relative incidence rate between each group was statistically significant (p<0.01). Triple dye group was 60 per cent less likely to develop omphalitis compared to 70 per cent Alcohol group (RR 0.39, 95% CI: 0.19-0.83). The mean duration for cord detachment were 13.6 and 11.5 days in group A and group B, respectively. CONCLUSION: During an epidemic outbreak of omphalitis, Triple dye was the most appropriate and effective antiseptic to prevent omphalitis but could delay cord separation.