1 Universidad nacional de Colombia, 2 Wyeth Lab, Bogotá, Colombia. Corresponding author: Aura Lucia Leal, E-mail: allealc@unal.edu.co Tel: (+57-1) 2692662, Address: Calle 29, no. 34 A–33, 2do. Piso, Colombia. Gram-negative and Gram-positive microorganism breakpoints and interpretation criteria suggested by the Food and Drug Ad- ministration (FDA) were used; those corresponding to enter- obacteriaceae were used for Acinetobacter baumannii. The data from the susceptibility test results were analyzed using Whonet (see 5.4, WHO, Geneva, switzerland). A rank sum test (signed-rank test, i.e. a non-parametric statistical hy- pothesis) (Wilcoxon) was used to compare the medians of the re- sults in terms of zone diameter millimeters obtained with each agar for each type of microorganism. stata (9.2, Texas, UsA) was used for all analysis. Minor and major error percentages ob- tained by the disc diffusion method were compared according to the type of agar used and the microdilution result. Very major errors were not analyzed, given that tigecycline isolates were not tested by the microdilution technique. isolate distribution was as follows: 34% Gram-positive cocci (228 Staphylococcus aureus and 52 enterococcus spp), 55% enterobacteriaceae (208 e. coli, 143 Klebsiella pneumoniae, 52 enterobacter cloacae and 32 Serratia marcescens) and 10% Acinetobacter baumannii (80). All isolates were tigecy- cline-sensitive according to the broth microdilution method. With the disc diffusion method using Becton Dickinson medium, 100% of Gram-positive cocci were susceptible and 1.6% were resistant; 27.1% of enterobacteriaceae were inter- mediate compared to Acinetobacter baumannii (58.8% inter- mediate and 1.2% resistant). With the Oxoid medium, 76.4% of Gram-positive cocci were susceptible and 7.9% were resistant; 52.5% of enterobacteriaceae were intermediate while 20% were resistant; and 70% of Acinetobacter baumannii were in- termediate. Figure 1 compares the zone diameter medians obtained for each commercial medium used for the enterobacteriaceae, Gram-positive and Acinetobacter baumannii groups. There were larger zone diameters with the Becton Dickinson medium than the Oxoid medium for all bacteria studied. table 1 presents minor and major percentage errors calcu- lated. no minor and/or major errors were found for Gram-pos- itive cocci with the Becton Dickinson medium while 25% and 23.2% major errors were observed for enterococcus spp and Staphylococcus aureus with Oxoid medium, respectively. Minor and major errors occurred with Gram-negative bacteria with both media and a major percentage error occurred with the Oxoid medium. Only major errors occurred with es- cherichia coli using the Becton Dickinson medium while 28.4% minor errors occurred with the Oxoid medium. Minor and major errors were greater with the Oxoid medium than the Becton Dickinson medium for Klebsiella pneumoniae, enterobacter cloacae and Serratia marcescens, a higher percentage of minor errors being presented by Serratia marcescens with the Oxoid medium (69.2%). Our results show differences in bacterial susceptibility to tige- cycline when using the disc diffusion and microdilution meth- ods. Errors produced by the disc diffusion method were major for enterobacteriaceae and Acinetobacter baumannii. We noted differences according to the manufacturer supplying the Mueller Hinton agar used, with a major percentage of errors oc- curring when the Oxoid agar was used. A higher error per- centage rate was observed when the Oxoid medium was compared to BD medium, mainly for Acinetobacter bauman- nii, Klebsiella pneumoniae, enterobacter cloacae and Serra- tia marcescens. studies have been reported where Acinetobacter baumannii susceptibility varied according to the manufacturer supplying the medium used for disc diffusion tests, probably due to the high manganese concentration in Oxoid agar; however, these data were not compared to the gold stan- dard, meaning that percentage agreement could not be estab- lished 3 . bRief COmmUNiCAtiON Differences in Determining In Vitro tigecycline Susceptibility According to the Agar medium Used A.L. LEAL 1 - G. BUiTRAGO 1 - M.V. OVALLE 1 J.A. CORTÉs 1 - C.A. ÁLVAREZ 1 - J. LARROTA 2 COLOMBiAn TiGECYCLinE sUsCEPTiBiLiTY sURVEiLLAnCE GROUP Tigecycline is the first of a new class of antibiotics having broad-spectrum activity against clinically-relevant microorgan- isms. it is used for treating patients suffering from infections of the skin and soft tissues and complicated intra-abdominal infec- tions and ventilator-associated pneumonia (VAP) 1,2 . The literature contains evidence that the disc diffusion method does not have good agreement with the broth microdi- lution method (BMD) due to several factors related to the an- timicrobial susceptibility technique which could affect the final result, such as the medium used, pH, date of preparing the cul- ture medium and, especially, the concentration of cations such as Ca++, Mg++ and Mn 1 . This study was designed to determine differences in tigecy- cline susceptibility patterns using Mueller Hinton from two com- mercial suppliers: Oxoid (Cambridge, UK) and Becton Dickinson (new Jersey, UsA). A total of 802 isolates were processed; they came from an in vitro study conducted by the Colombian Tigecycline sus- ceptibility surveillance Group in 2008 in which clinical isolates were collected from 13 third-level hospitals (5 in Bogotá and 8 from other cities throughout Colombia). All the isolates were sent to the Universidad nacional de Colombia’s Microbiology Reference Laboratory and identified using an automatic system (Microscan, Dade Behring, sacra- mento, UsA). The antimicrobial tigecycline susceptibility tests were carried out by the disc diffusion method, following Clinical and Laboratory standards institute (CLsi) recommendations 2 , using 15mg tigecycline discs (Becton Dickinson, UsA) and Mueller Hinton agar culture medium from two commercial sup- pliers (Oxoid and Becton Dickinson, UsA). These tests were carried out by an experienced microbiologist; a blind study was made regarding the type of agar used. All isolates were processed by the broth microdilution method for determining minimum inhibitory concentration (MiC) using prepared panels (Treck Diagnostic system, Cleve- land, UsA) which included tigecycline at increasing concentra- tions (0.08-16 mg/mL). These were inoculated according to the manufacturer’s recommendations; CLsi recommendations were followed for preparing the inoculum and incubation. escherichia coli ATCC 25922, pseudomonas aeruginosa ATCC27853, Staphylococcus aureus ATCC 25923, Staphy- lococcus aureus 29213 and enterococcus faecalis ATCC 29212 strains were used for susceptibility test quality control for both disc diffusion and broth microdilution methods. The Journal of Chemotherapy Vol. 22 - n. 2 (137-138) - 2010 © E.s.i.F.T. srl - Firenze issn 1120-009x