Solutions of BATT1 and BATT2 disinfectants were prepared by dilution in sterile water at 10, 15, 20 and 25%. The bactericidal effects of BATT1 and BATT2 on multidrug-resistant nosocomial pathogens were assessed using a suspension-inhibition test as recommended by the European Committee for Standardization (CEN) with or without bovine serum albumin (BSA) 0.3 g/100 ml 13. The bactericidal effects of disinfectants were assessed using a slightly modified suspension-inhibition test as recommended by the CEN. The bactericidal effects of BATT1 and BATT2 on reference strains were tested from manufacturer (data not showed).

Environmental strains (Acinetobacter baumannii, Pseudomonas aeruginosa, methicillin-resistant Staphylococcus epidermidis, methicillin-resistant Staphylococcus aureus, high-level aminoglycoside-resistant Enterococcus faecalis, extended-spectrum beta-lactamase (ESβL)-producing Klebsiella pneumoniae, Candida albicans, Aspergillus fumigatus and Legionella pneumophila ser.1) were isolated from 2002 to 2005 in the intensive care units (ICUs) of the University Hospital "Federico II" in Naples, Italy, during sanitation checks following nosocomial outbreaks. These microorganisms are recognized as the major hospital-acquired pathogens that cause severe infections such as pneumonia, septicaemia, urinary tract infections and surgical site infections. Throughout this study, the isolates per species were of the same clonal type, showing identical macrorestriction PFGE pattern (data not showed).

Swabs moistened with Brain-Heart Infusion broth were used to sample horizontal surfaces and points of frequent hand contact, as well as monitoring equipment, drug trolleys, respirators and sinks. Contact plates (Rodac; International PBI) with selective agar were used to sample other surfaces (floor, walls and beds), and air was sampled through the Surface Air System (International PBI). Culture specimens were enriched overnight at 37°C in Brain-Heart Infusion broth and then subcultivated on agar plates.

Isolates were identified by a commercial microidentification system (API 20E; bioMèrieux, Marcy-L'Etoile, France).

Environmental Legionella isolates were obtained from multiple sites in patients' rooms. For each outlet, 2 l of hot water was collected in a sterile bottle that contained 1 ml of a 10 mg/ml solution of sodium thiosulphate. The water temperature and residual free chlorine were determined immediately after collection. Samples were concentrated by filtration through cellulose acetate membrane filters (0.22 μm pore size) and resuspended into 10 ml of the filtrate. Aliquots of the suspension were plated on to Buffered Charcoal Yeast Extract (BCYE) agar with Legionella Glycine, Vancomycin, Polymyxin B, Cycloheximide, and Wadowsky and Yee selective supplements (Oxoid, Basingstoke, UK). Plates were incubated in 2.5% CO₂ for 5 days at 37°C and examined daily for evidence of growth. Gram-negative typical colonies that required L-cysteine for growth were harvested, centrifuged at 3000 rpm for 20 min and resuspended in sterile distilled water prior to serological identification.

Resistance patterns of these microorganisms have been determined as following.

Susceptibility tests of K. pneumoniae, A. baumannii and P. aeruginosa were performed using the BD Phoenix (Phoenix Technologies Ltd, San Jose, CA, USA) system (MIC/panel susceptibility card) according to the manufacturer's instructions (E. coli 25922 ATCC QC, P. aeruginosa ATCC QC 27853, E. coli ATCC QC 35218)

ESβL activity in K. pneumoniae was evaluated using the double-disc synergy test between cephalosporins or monobactam and clavulanate disks on Muller Hinton agar plates. ESβL activity was confirmed by Etest cefotaxime/cefotaxime+clavulanic and ceftazidime/ceftazidime + clavulanic acid strip as recommended by the manufacturer (K. pneumoniae ATCC QC 700603).

High-level aminoglycoside resistance in E. faecalis was determined in 96-well microtitre plates using Gentamicin and streptomycin. For QC of HLAR screen tests was used E. faecalis ATCC 29212.

Methicillin and vancomycin were used to test for resistance in Staphylococcus spp. by antibiotic disc diffusion method (S. aureus MRSA ATCC QC 43300).

Susceptibility testing of yeast and mould was detected by E-test using C. albicans ATCC QC 90028 and A. fumigatus ATCC 204305.

Finally, MICs for L. pneumophila were determined using broth microdilution in liquid BCYE medium with selective supplement (L. pneumophila ser.1 ATCC QC 33152).

Throughout this study, results were interpreted according to the Clinical and Laboratory Standards Institute (CLSI) criteria for broth microdilution and disk diffusion methods 14.

Antimicrobial susceptibility testing showed common multidrug-resistant antibiotypes for all the isolated microorganisms (Table 1). In particular, the antimicrobial susceptibility patterns of the ESβL-producing K. pneumoniae showed resistance to penicillins, monobactams and third-generation cephalosporins. Also the K. pneumoniae strain was identified as ESβL-producer by Etest analysis (Table 2). The antimicrobial susceptibility analysis showed high-level aminoglycoside-resistance in E. faecalis and methicillin-resistance in Staphylococcus spp. that were either coagulase positive or negative (data not showed).

Microbial suspensions in Luria-Bertani broth were prepared from fresh cultures of the above microorganisms, and subcultivated in non-selective media (Brain-Heart Infusion agar for Gram-negative bacteria, Enterococcus and Candida spp., Tryptone Soy Agar for Staphylococcus spp., Columbia Blood Agar for Aspergillus fumigatus). L. pneumophila was cultivated on a selective medium, BCYE Agar with selective supplements.

Final concentrations of inocula of 10⁸ CFU/ml bacteria and 10⁶CFU/mL yeasts and moulds were measured using a spectrophotometer (UV/VIS Lambda 2, PerkinElmer, Monza, Italy) (OD₆₆₀ 0.08). A 100-μl suspension for each test was added to tubes that contained 900 μl of sterile physiological solution (control) or 900 μl of each disinfectant solution at different concentrations (10, 15, 20 and 25%), and left for 5, 10 and 15 min at 20 ± 2°C in a thermostatic bath. Each inoculum was prepared twice, with and without BSA 0.3 g/100 ml.

After incubation, 100 μl were removed from each inoculum and smeared on Bacto D/E Neutralizing agar (Becton Dickinson) after serial dilution (1:10,1:100,1:1000,1:10000) in phosphate buffered saline (PBS ph 7.4), in order to count the number of colonies surviving on each plate at different dilution. According to the standards, an efficacious biocide must reduce the initial count by 4 or 5 log units, and the efficacy is estimated by the ratio between the number of microorganisms in the starting solution (inocula of 10⁸ CFU/ml bacteria and 10⁶ CFU/mL yeasts and moulds) and number of colonies surviving on neutralization plates. Plates were incubated at 37°C for 24 h for bacterial strains and at 32°C for 72 h for C. albicans and A. fumigatus strains. Only the aliquots of the reaction mixture (100 μl) that contained L. pneumophila cells and disinfectants after incubation were added to 900 μl of neutraliser solution and left at 20 ± 2°C for 3 min. After serial dilution as above, the aliquots were smeared on BCYE agar with supplements. These plates were incubated at 37°C in 2.5% CO₂ for 5 days. The test was also carried out on glass surfaces initially contaminated with the microbial inocula, prepared as above, for which the same concentrations were used as for the suspension test, with and without BSA. After 5, 10 and 15 min of sanitation with both disinfectant solutions at different concentration (10, 15, 20 and 25%), the surfaces were rinsed with a sterile physiological solution, and 100 μl was removed from the rinsing solution and smeared on D/E Neutralizing agar after serial dilution. Plates were incubated under the same conditions as above. For L. pneumophila, it was only necessary to use the neutraliser solution and selective medium. The experiments were repeated three times on different days.

Bacterial, mould and yeast concentration was evaluated by colony counts on agar plates (Bacto D/E Neutralizing agar, BCYE agar with supplements) after incubation under the described conditions.

The results of the in vitro tests were interpreted in accordance with the CEN Standards 13. According to the standards, an efficacious biocide must reduce the initial count by 4 or 5 log units, and the efficacy is estimated by the ratio between the number of microorganisms in the starting solution (inocula of 10⁸ CFU/ml bacteria and 10⁶CFU/mL yeasts and moulds) and number of colonies surviving on neutralization plates.

Bactericidal activity of BATT1 and BATT2 was observed at a concentration of 25% after 10 min contact. In vitro tests with BATT1 and BATT2 demonstrated an overall drop in microbial counts, without BSA (Table 3). The same results were obtained in presence of BSA. To reduce P. aeruginosa count, it was necessary to use 50%, 75% concentrations and finally both undiluted disinfectants, with and without BSA (data not shown). The same results were obtained in the tests carried out on glass surfaces for all nosocomial strains. Also in this test the biocide effect on P. aeruginosa was obtained by using undiluted disinfectants with and without an organic substance.