[Surveillance of resistance in the intensive care units using a cumulative antibiogram]

Empiric treatment with broad-spectrum antibiotics exerts condensed pressure in intensive care units (ICUs) for the selection of multidrug-resistant bacteria. Multidrug-resistant gram-negative bacteria became the focus of interest owing to limited treatment options and negative effects on patient survival. Cumulative antibiograms can guide selection of correct empiric treatment, de-escalation treatment according to antibiogram results and development of policies in fight against antibiotic resistance. In this study, we aimed to determine the antibiotic resistance rates of gram-negative bacilli in the intensive care units of the public hospitals in the region where we are connected by using the cumulative antibiogram result and to show the change of resistance over the years and to determine whether there is any difference between the hospitals. Gram-negative bacilli were isolated from ICUs of state hospitals in the second Istanbul State Hospitals Association area during 2014-2016. Isolates were identified using Vitek MS (bioMérieux, France) system and tested for antibiotic susceptibility with Vitek 2 (bioMérieux, France) system according to the Clinical and Laboratory Standards Institute (CLSI) criteria followed during those years. Cumulative antibiogram reports for these strains were prepared according to the CLSI M39-A4 guide. Gram-negative bacilli were divided into three groups: non-fermentative, urinary enteric, and non-urinary enteric bacilli. Total number of strains isolated during three years from these groups were 2626, 1390 and 2011, respectively. Annual trends of susceptibility during the aforementioned years were evaluated. Hospitals were also classified into five groups and differences were evaluated in the susceptibility profiles of these hospitals. Among the non-fermentative bacilli, Acinetobacter baumannii complex was the most commonly isolated species and the most resistant bacteria against antibiotics. The susceptibility rates of A.baumannii complex against the beta-lactam group of antibiotics were < 10%. Colistin susceptibility rates of A.baumannii complex and Pseudomonas aeruginosa isolates were over 98%. Among the non-urinary enteric bacilli, K.pneumoniae was the most commonly isolated species displaying maximum antibiotic resistance. Susceptibility rates for colistin, which is the last resort for treating resistant gram-negative bacteria, ranged between 73% and 80%. Escherichia coli, which was the second most common isolated species among non-urinary bacilli, had susceptibility rates over 90% to carbapenems along with colistin and tigecycline. Although E.coli was the most commonly isolated species among urinary enteric bacilli, K.pneumoniae and Proteus mirabilis were the most resistant isolates. A statistically significant decrease in susceptibility rates against all antibiotics was observed in P.mirabilis isolates between the years 2015-2016. Carbapenem susceptibility rates decreased below 70%. E.coli, Serratia spp., and Stenotrophomonas maltophilia had similar susceptibility profiles among different hospitals, indicating homogenous distribution, whereas other species had different profiles, indicating a more heterogenic distribution, among hospitals. The reports of this study were generated according to a standard guide and they clearly revealed the seriousness of antibiotic resistance in our region which represents approximately one-fourth area of Istanbul. When all the results were considered, best empiric treatment option for enteric bacilli except K.pneumoniae was carbapenems. For K.pneumoniae infections there is no reliable choice other than colistin but a de-escalation treatment can be planned according to antibiogram results. Similarly colistin is the first choice in empiric treatment of infecitons due to non-enteric bacilli. However, the heterogeneity of the susceptibility profile observed in the hospitals, which are geographically close to each other, indicated the difference in the flora of the intensive care unit of hospitals. It would be appropriate to prepare cumulative antibiogram reports similar to those in the present study, to prevent complications, reduce costs and improve patient prognosis in the intensive care units of hospitals and these reports should become part of the infection control policies applied in hospitals.