Controlling antibiotic resistance in intensive care units

Colonisation with multidrug resistant bacteria, especially resistant Gram-negatives, occurs frequently and increasingly in intensive care units (ICUs), also in the Netherlands. Infections caused by (resistant) microorganisms increase morbidity and mortality in this vulnerable patient group. This necessitates more effective control measures. Current infection control strategies include hand hygiene programs, body washing with chlorhexidine, screening plus isolation of identified carriers and selective digestive (and oral) decontamination. Hand hygiene is generally low in ICUs. Most evidence on the effect of improved hand hygiene on infection rates stems from observational studies. However, it is likely futile to implement costly, labour-intensive interventions without optimising basic hygiene. Chlorhexidine body washing has been proven effective in reducing methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococci, but not resistant Gramnegatives. Reported effects of rapid screening plus isolation are conflicting, and mostly include just MRSA. Selective digestive (and oral) decontamination reduced the 28-day mortality in Dutch ICUs in a large trial, and is considered standard of care in most ICUs in the Netherlands. Hand hygiene, despite a lack of rigorously performed trials, should be improved in ICUs as part of normal hygienic measures. Our findings do not support the use of chlorhexidine body washing in Dutch ICUs, as MRSA prevalence is low. For patients at high risk for MRSA carriage, rapid screening can reduce unnecessary isolation days. The control of resistant Gram-negative bacteria will be a major challenge in the coming years, also in the Netherlands. We will need new methods to control the spread of these microorganisms, as current strategies have not proven effective. Introduction Colonisation and infection with multi-drug resistant organisms (MDRO), such as methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococci (VRE) and highly resistant Enterobacteriaceae (HRE) occur with increasing frequency in hospitals worldwide, especially in intensive care units (ICUs).1-3 In Dutch ICUs, resistance to third-generation cephalosporins has increased to 8% in E. coli and 9% in K. pneumoniae.4 There is a strong increase in resistance to ciprofloxacin in P. mirabilis (7.7% in 2012) and the prevalence of highly resistant E. coli was 11% in 2012. Healthcare-associated infections (HCAI) caused by MDRO are associated with delayed initiation of appropriate therapy, failure of therapy, prolonged length of hospital stay, and increased morbidity and mortality.5-8 The exact burden of HCAI is difficult to quantify because of the use of many different and partly subjective diagnostic criteria, hampering comparison of infection rates between countries and studies. In developed countries, HCAI affect up to 15% of hospitalised patients and 9 to 37% of patients in ICUs.9-11 Consequences of HCAI for patient outcome are difficult to determine, and recent estimates suggest that the excess mortality of one of the most important HCAI, ventilator-associated pneumonia (VAP), is lower than previously assumed. Based on a meta-analysis of individual patient data from 24 randomised prevention studies, the attributable mortality of VAP was estimated to be 13%.12 The same holds for the consequences of HCAI caused by MDRO. It is obvious that resistance has a detrimental effect on patient outcome, but recent studies reach considerably lower estimates of attributable mortality than older studies.13 It is estimated that such infections would lead to 3.3 associated deaths per 100,000 Europeans in 2015, and that the observed increase in bacteraemia caused by HRE will become an even more pressing healthcare problem in the near future.14-17 Hospital-acquired bacteraemia caused by MDRO Netherlands Journal of Critical Care NETH J CRIT CARE VOLUME 19 NO 1 FEBRUARY 2015 13 Controlling antibiotic resistance in ICUs adds to the total burden of hospital-acquired bacteraemia, rather than replacing them, thus increasing to the total burden of disease.2,18 Bacterial colonisation is the first step in the development of HCAI.19 Admission of colonised patients contributes to the total burden of MDRO in the ICU.20 Acquisition of MDRO colonisation during ICU admission may occur through newly developed resistance in previously susceptible bacteria (i.e., de novo resistance), selection of resistant bacteria induced by antibiotic use (i.e., endogenous selection), acquisition from contaminated surfaces in the ICU, or through patient-topatient transmission of bacteria (i.e., cross-transmission).21 Cross-transmission mostly occurs through temporarily contaminated hands of healthcare workers.9 Risk factors for MDRO colonisation are listed in table 1. The global increase in MDRO infections necessitates more effective control measures, especially in intensive care units.5 Table 1. Risk factors for MDRO carriage Important antimicrobial resistant bacteria Infections caused by MRSA started to increase, worldwide, in the late 1980s, and became endemic in many countries. In Dutch ICUs, MRSA infections have remained rare due to a nationwide infection control policy (‘search and destroy’), which includes, amongst other things, pre-emptive isolation of patients at high risk for MRSA carriage. This program, though not based on well-designed prospective studies, has been highly successful, but the relative importance of the individual components is unknown.22 Some European countries, such as the United Kingdom and France, have achieved impressive reductions in the incidence of HCAI caused by MRSA, but proportions of MRSA among nosocomial S. aureus bacteraemias remain above 25% in one-third of the countries participating in the European Antimicrobial Resistance Surveillance System (EARSS).1 VRE were described for the first time in Europe in 198823, but emerged as nosocomial pathogens in United States hospitals in the 1990s. VRE infections started to increase in European hospitals around ten years ago, and since 2011 VRE hospital outbreaks have occurred in Dutch hospitals.24 Results from several studies suggest considerable attributable mortality of enterococcal infections, including VRE.25 Incidences of infections caused by resistant Gram-negative bacteria, such as extended-spectrum beta-lactamase producing Enterobacteriaceae (ESBL) have increased markedly worldwide during the past decade.26,27 Moreover, carbapenemaseproducing Enterobacteriaceae (CRE; notably K. pneumoniae) have recently emerged in different parts of the world, and represent an even bigger healthcare threat for hospitalised patients.28-30 All Enterobacteriaceae resistant to at least thirdgeneration cephalosporins, including ESBL and CRE, are considered highly resistant Enterobacteriaceae (HRE). ESBLs and most genes conferring carbapenemase production are plasmid-mediated, which allows efficient transfer of resistance to other bacteria.26 Furthermore, the different genes can be associated with different resistance phenotypes, hampering rapid detection in laboratories. These aspects complicate infection control strategies, especially in healthcare facilities. In the Netherlands, 5 to 10% of all Enterobacteriaceae causing bacteraemia harbour ESBL genes. Many of these isolates are also resistant to ciprofloxacin and aminoglycosides, and these figures are expected to increase in the coming years. The increased occurrence of infections caused by ESBL-producing bacteria coincides with an increased use of carbapenems.4 Current strategies to combat antibiotic resistance in the ICU Current strategies to control the spread of antibiotic resistance include hand hygiene programs, the use of chlorhexidine body washing, nasal decontamination with mupirocin (for MRSA only) and screening of patients on admission with contact precautions of identified carriers. All these interventions aim to interrupt cross-transmission of bacteria. Often, these interventions are combined in a variety of ‘infection control