Multidrug-resistant (MDR) strains have become progressively more common causes of nosocomial infections since 1980s1. Many Asian countries, including Republic of Korea, have also shown increasing rates of antibiotic resistance2. Especially, intensive care units (ICU) have become the main foci for developing multidrug-resistance such as methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococci, and multi-drug resistant gram negative bacilli3.

Among these pathogens, gram-negative bacteria resistant to carbapenem are great concern. Carbapenems are the drugs of choice for the treatment of infection by many gram-negative bacteria4. However, the increasing use of carbapenems has contributed to the development of carbapenem-resistance among these pathogens. The most frequently encountered microorganisms resistant to carbapenems are Pseudomonas aeruginosa, Acinetobacter baumannii, and Enterobacteriaceae.

P. aeruginosa causes severe invasive diseases in critically ill and immunocompromised patients. Pseudomonas species isolated from patients in ICUs show carbapenem-resistance rates of 28~37%5,6. A. baumannii is also a major cause of nosocomial infections. Acinetobacter initially exhibited almost uniform susceptibility to imipenem in most institutions. However, carbapenem-resistant strains are emerging worldwide and resistance rates have reached 50% to 60% in some institutions7,8. Among the Enterobacteriaceae, extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae which are resistant to carbapenems are great problems in ICUs because carbapenems are usually considered the last choice for the treatment of these organisms9. The major driving force behind carbapenem-resistance has been the heavy use of third generation cephalosporins, aztreonam, and imipenem.

Although there have been many articles regarding MDR pathogens in the ICU, few have evaluated the annual trend of carbapenem-resistance since the opening of new ICU. This study aimed to determine the annual trend of carbapenem-resistance patterns of major pathogens and their respective contribution at a newly opened ICU of a university-affiliated hospital in Seoul, Republic of Korea.

Carbapenems are one of beta-lactam drugs which contain bicyclic beta-lactam ring in the center. Although they share a common beta-lactam ring with penicillin, they are generally resistant to cleavage by most plasmid and chromosomal beta-lactamase11. Therefore, they retain activity against the chromosomal cephalosporinases and extended-spectrum beta-lactamases found in many gram-negative pathogens. However, the increasing use of carbapenems has contributed to the development of carbapenem-resistance in many pathogens12. In the present study, we evaluated annual trend of carbapenem-resistance at newly opened ICUs of a university-affiliated hospital for 4 years since 2005. We found that carbapenem-resistance increased constantly during the first 3 years and then decreased slightly in the fourth year in the ICU (Figure 3). As expected, annual consumption of imipenem and meropenem has increased constantly along with other important antibiotics (beta-lactam/beta-lactase inhibitors, cephalosporins, fluoroquinolones, and aminoglycosides) in our hospital from 2006 to 2008 (Table 2). Therefore, it is not surprising that even though standardized infection control measures have been applied at our ICUs, they could not prevent the proliferation of carbapenem-resistant pathogens. It is a great concern because optimal treatment of infection due to carbapenem-resistant organisms is uncertain and antibiotics options are very limited. It is well reflected in our data that only part of the initial antibiotics were adequate to cover pathogens with carbapenem-resistance (n=37, 9.3%) (Table 1).

Risk factors for developing multi-drug resistance are old age, male gender, cardiovascular diseases, mechanical ventilation, duration of the ICU stay, surgery, trauma, and the use of broad-spectrum antibiotics13. It is consistent with our finding that most of patients with carbapenem-resistance (88.2%) have been treated by other antibiotics before ICU admission, had long duration of ICU stay, high APACHE III score, and high mortality rate of 37.5% (Table 1). It is noteworthy that over forty percent of patients had diabetes as an underlying disease.

P. aeruginosa and A. baumannii were most common pathogens with carbapenem-resistance, as expected (Figure 1). It is consistent with the fact that Pseudomonas is the most widespread MDR gram-negative pathogen causing pneumonia in hospitalized patients14. P. aeruginosa and Acinetobacter have the potential of causing outbreaks in the hospital setting and most outbreaks occurred in adult ICUs15,16. Unfortunately it is evident that our ICUs are not the exception. It is somehow relieving that many isolates of P. aeruginosa (80.6%) and A. baumannii (72.6%) remain sensitive to colistin. However considering the high degree of nephrotoxicity, colistin as an alternative for the treatment of carbapenem-resistant pathogens is not a great relief.

Pseudomonas is the most common pathogen of hospital-acquired respiratory tract infection and the most common manifestation of Acinetobacter outbreak infection have been nosocomial pneumonia (primarily ventilator-associated)17. It is consistent with our findings that respiratory specimen were the most common source of isolation for carbapenem-resistant pathogens (Figure 2).

Carbapenem-resistant rate increased constantly during the first 3 years since the opening of ICUs (Figure 3). Our ICUs adopted standardized, meticulous infection control measures including, use of gown, glove, facial masks for the care of patients, repeated education for prevention of infection in the ICU, monitoring of hand washing among medical staffs, cohort isolation of patients infected by MDR pathogens, closed suction system of endotracheal tube, and oral hygiene using chlorhexidine three times a day. With all these measures, proliferation of carbapenem-resistant pathogens in the ICU seems to be an unavoidable phenomenon.

A. baumannii infection was not significant in the first two years. However, after explosive isolation on 3rd quartile of 2007, there was sustained isolation of A. baumannii until the 4th quartile of 2008 (Figure 4). Once the epidemic of A. baumanni happens, it is very difficult to eliminate it from the ICU. Koeleman et al.18 even suggested the complete 10-day closure of ICU to contain an A. baumannii. Although we did not adopt this policy during the period, it could be an option when there is no other way to control explosive isolation of A. baumannii in the ICU.

Of note, peak isolation of P. aeruginosa or A. baumannii in the MICU was followed by same epidemics at SICU (Figure 5). In our hospital, MICU with 21 beds and SICU with 16 beds are run side-by-side on the same floor. Although each ICU has its own operating staffs, sometimes doctors can take care of both sides of patients. Still, the direction of spread was always from MICU to SICU but not vice versa. It can be explained that generally patients in the MICU have more risk factors for MDR pathogens than those in the SICU.

This study has several limitations. First, we did not evaluate the risk factors for carbapenem-resistant organism systematically. Second, the isolated pathogen could be colonization. For example, A. baumannii is a ubiquitous commensal and it is widely distributed in ICU environments. Differentiation between colonization and true infection may be difficult. However, clinical samples were collected when the patients developed signs of infection. Therefore, the possibility of colonization might be low. In spite of some limitations, the present study is a rare report that presented the annual trend of carbapenem-resistance in newly-opened ICUs for a long period of time. In conclusion, carbapenem-resistant pathogens are growing concern in the ICUs and epidemic of carbapenem-resistance in the MICU can be followed subsequent epidemic in the SICU by the same pathogens when they are run in close proximity.