You are on page 1of 5

Antibiotic resistance in the intensive care unit

AJ Varley BSc MRCS FRCA H Williams MB BCh FRCPath S Fletcher FRCA FRCPE

Downloaded from http://ceaccp.oxfordjournals.org/ at Primary & Community Health Services Library on September 29, 2012

Key points Gene transfer is the predominant mechanism of acquisition of resistance. The use of broad-spectrum agents is a signicant factor in the development of resistance in bacteria and fungi. The risk of acquiring infection with antibiotic resistant organisms is positively correlated with increasing age, illness severity, debility, and length of ICU stay. No single measure can be effective in the prevention of infection due to antibiotic resistant organisms or in the reduction of resistance. Knowledge of the mechanism of transmission of resistance enables targeted efforts to control outbreaks.
AJ Varley BSc MRCS FRCA Senior Registrar, Intensive Care Unit Alfred Hospital Melbourne, VIC Australia H Williams MB BCh FRCPath Consultant Microbiologist Norfolk and Norwich University Hospital Colney, Norwich, UK S Fletcher FRCA FRCPE Consultant in Anaesthesia and Intensive Care Norfolk and Norwich University Hospital Colney, Norwich, UK E-mail: simon.etcher@nnuh.nhs.uk (for correspondence)

The use of antibiotics is unique in medical practice in that the treatment given to an individual may have consequence for the wider population. Pathogens may be intrinsically resistant to antibiotics, but the problem of induced or evolving resistance should not to be underestimated. Increasingly, it is recognized that the use of broad-spectrum agents, even when appropriate, is a signicant factor in the development of resistance in bacteria and fungi. This has major implications for healthcare in general, and particularly in the ICU where resistant organisms can present major challenges, as patients tend to be debilitated and particularly susceptible to nosocomial infection. Such infections often lead to prolonged ICU and hospital stay, and consequent increased healthcare costs. There is a degree of inevitability to the development of antibiotic resistance. However, strategies have been proposed and evaluated that attempt to limit the development of antimicrobial resistance, maintain the usefulness of existing antimicrobials, and inuence the development of novel agents. Much of what follows relates to bacterial resistance and pathogens selected by antibiotic use.

spectrum agents and the risks associated with a poorer outcome from delayed or inappropriate treatment.2

Molecular genetics of antibiotic resistance


Analysis of bacteria collected before widespread introduction of antibiotics reveals, excluding intrinsic, almost complete sensitivity. Organisms with intrinsic resistance are often of low virulence but do become a problem in vulnerable patients managed in selection pressure environments (Pseudomonas sp., Acinetobacter). Acquisition is based on the mechanisms of genetic mutation and inter-cell transfer. Mutation is often disadvantageous to the bacteria but will, infrequently, affect antibiotic resistance. However, the transfer of resistance between bacteria is of greater importance, the mechanisms of which are not mutually exclusive: 1. Naked DNA (transformation) Naked DNA is released from killed bacteria and as such is common in the ICU patient on antibiotics. DNA in this form is unprotected and is quickly degraded. Bacteria have a varying ability to take up this DNA. Transformation is the process where this DNA is incorporated into the genome of another bacterium. 2. Bacteriophages (transduction) These are viruses that infect bacteria. A protein coat protects the DNA within and the virus relies on the bacterias cellular machinery to propagate it. The DNA within the virus may be exchanged or transferred to the host, and through this mechanism can transfer genetic information encoding resistance. This is known as transduction. Bacteria vary in their

Bugs, drugs, and the patient


The success or failure of antimicrobial treatment depends on many things, including the vulnerability of the host, the virulence of the organism, and the use of the appropriate antimicrobial (sensitivity, tissue penetration), and other clinical interventions (removal of foreign bodies, devitalized tissue, drainage of abscesses).1 There is a tension between ensuring prompt treatment of infection, which unless the cause is obvious requires the use of broad-

114

doi:10.1093/bjaceaccp/mkp017 Advance Access publication 26 June, 2009 Continuing Education in Anaesthesia, Critical Care & Pain | Volume 9 Number 4 2009 & The Author [2009]. Published by Oxford University Press on behalf of The Board of Directors of the British Journal of Anaesthesia. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournal.org

Antibiotic resistance in the intensive care unit

susceptibility to infection with bacteriophages; Corynebacterium diptheriae and Vibrio cholerae are examples of bacteria that commonly receive genetic information through this route. 3. Plasmids (conjugation) These are self-replicating circles of DNA that exist within the bacteria but are separate from the chromosome. They lack the protective coat of the bacteriophage and are unable to move independently of the bacteria. Despite these limitations, they are the most important routes of transmission of genetic information within the intensive care unit. 4. Transposons These are small segments of DNA that can encode resistance genes. They also encode for their mobility, thus allowing them to move from plasmid to plasmid, or within the main genotype. As such they can transmit resistance between cells.

Common and clinically important. .300 variants with selective to extended spectrum Inhibits uptake and transport of antibiotic

Glycopeptide-resistant Enterococci (VRE) Gram 2ve bacteria Imipenem resistant pseudomonas Tetracyclines Tertacyclines in enterobacter, linked with the above

Sulphonamide and trimethoprin resistance MRSA

Aminoglycosides, macrolides

Downloaded from http://ceaccp.oxfordjournals.org/ at Primary & Community Health Services Library on September 29, 2012

Notes

b-Lactams

Mechanisms
Different mechanisms of antibiotic resistance have been described in bacteria. These include enzymatic inhibition, decreased permeability, antibiotic efux, target modication (ribosomes, cell walls, enzymes), and development of alternative targets or pathways. They may be present singly, or in any combination, within a particular bacterial species. These are summarized in Table 1.

Table 1 Mechanisms by which bacteria may acquire resistance

Common resistance problems


Methicillin-resistant Staphylococcus aureus
This organism has a high political prole, with trusts in England recently set targets for bacteraemia rates.4 Patients developing

Cell wall permeability; decreased permeability of bacterial membranes;

Active expulsion; Promotion of antibiotic efux Metabolic

Inactivation; enzymatic inhibition

The risk of acquiring infection with antibiotic resistant organisms is positively correlated with increasing age, illness severity and debility, prolonged ICU stay, prior antibiotic usage, and exposure to indwelling prosthetic devicesprincipally central venous catheters and tracheal tubes.3 Unsurprisingly, the most common infections acquired on the ICU are ventilator-associated pneumonia (Gram ve and Gram 2ve), and line-associated bacteraemia ( principally Gram ve). Inappropriate antibiotic selection is strongly associated with increased mortality; therefore, infection with resistant organisms can be associated with poor outcome if initial antibiotic selection does not provide coverage. There is an increased mortality rate in patients with ventilator-associated pneumonia caused by multi-drug-resistant organisms. Most ICUs will have antimicrobial policies that guide the selection of antibiotics. These are based on risk factors for the patient (including previous therapy, underlying condition, and relevant microbiology) and local epidemiology of resistant organisms. Regular input from medical microbiologists or infectious disease physicians is essential.

1. b-lactamase

The problem of resistance


Examples

2. Acetylation /phosphylation 1. Penicillin-binding protein 2. Ribosomal proteins

Structural changes to bacterial components

Enzyme production

Target site modication

Method

Ribosomal target; Cell wall precursor; Target enzyme

Modication of naturally occurring efux pumps Development of alternative pathway or target

Structural changes to cell wall or protein channels (Porins) or transporters

Mode

3. Cell wall peptidoglycan Resistant to degradation 1. Cell wall Reduced permeability 2. Porins Inux blocked 3. Active transporters Reduced uptake Removes antibiotic molecules maintaining sub-therapeutic concentrations 1. Dihydrofolate reductase Resistant enzyme 2. Peptidoglycan Alternative pathway production

Hydolyses structural ring in b-lactam antibiotics Structural modication

Reduced afnity

Reduced afnity

Action

Continuing Education in Anaesthesia, Critical Care & Pain j Volume 9 Number 4 2009

115

Antibiotic resistance in the intensive care unit

methicillin-resistant Staphylococcus aureus (MRSA) infection commonly have more underlying diseases and a greater number of previous infections and antibiotic treatments. Several authors have investigated whether MRSA infection has an excess mortality. Overall mortality rates in patients with MRSA infection are higher than in those with methicillin-sensitive infection; however, once corrections are made for co-morbidities, there is no signicant excess mortality and patients tend to die with rather than of their MRSA infection.5 Despite this, it is clear that MRSA infection is associated with greater healthcare costsbe that from longer hospital stays, greater numbers of interventions and more expensive antibiotic therapies. Recent comprehensive guidance on the management of MRSA infection has been issued.6

Other Gram 2ve organisms


Acinetobacter species (commonly Baumannii) are ubiquitous in the environment and, in general, of low virulence. Their targets are similar to those already described with preponderance to nosocomial pneumonias. They seem particularly skilled at acquiring multi-drug resistance and, worryingly, have been linked to infection outbreak in healthy populations, suggesting the acquisition of virulence. Carbepenem resistance, while about 5% in the UK, is .50% in South America and the Middle East. Stenophomonas maltophilia, a bacterium originally classied as a pseudomonas, is also found in many environments and has very low virulence. Immuno-compromised patients are at increasing risk with the predictable spectrum of infection. Treatment is difcult because of multiple drug resistance and strict hygiene/isolation is essential to prevent dissemination. Of the many strains of Pseudomonas, Pseudomonas aeruginosa is most frequently associated with human infection. Again a ubiquitous Gram 2ve rod with low natural virulence, infection is truly opportunistic with patients in ICU a common target. In many parts of the world the number one cause of nosocomial pneumonia, this organism has evolved multidrug resistance and may be very difcult to treat. There are many other enterobactreacia responsible for nosocomial infections, all largely opportunistic and with varying degrees of multiple resistance. Table 2 describes resistant organisms.

Downloaded from http://ceaccp.oxfordjournals.org/ at Primary & Community Health Services Library on September 29, 2012

Extended-spectrum b-lactamase-producing organisms


Extended-spectrum b-lactamase-producing organisms (ESBL) is a plasmid-mediated form of resistance in Gram 2ve bacteria to the b-lactam antibiotics. Until relatively recently, the clinical problem was small, stable (at least in the UK), and largely restricted to Klebseillae isolates in high-risk hospital patients. Since 2004/5, a new, highly resistant ESBL (CTX-M) has emerged in Escherichia coli and Enterobacter, responsible for signicant outbreaks of resistant, community-based infection, often of the urinary tract. Mortality in these patients has been high. ESBLs (a term for any infection with bacteria possessing ESBL) have been reported in almost every European country. The global prevalence is approximately 25%, with higher gures for Mediterranean countries and as little as 3% for the USA. ESBL-carrying bacteria can also have other resistance mechanisms (e.g. membrane protein deciencies) and resistance to other classes of antibiotic (e.g. the aminoglycosides, and quinolones trimethoprim/sulfamethoxazole). Carbapenems are currently the most active class against ESBL bacteria, because of their stability against the hydrolytic effects of the b-lactamase.

Management strategies
From this brief consideration of possible mechanisms, it is apparent that no single measure can be effective in the prevention of infection due to antibiotic resistant organisms or reduction of resistance. Knowledge of the mechanism of transmission of resistance enables targeted efforts to control outbreaks. For example, if resistance is being imported, greater isolation and screening measures should be employed. Alternatively if resistance is being disseminated from within the unit, better adherence to infection control measures should be enforced. A wide range of non-pharmacological methods are employed to try and limit the spread of resistant infections. These are to be covered elsewhere, and will not be discussed further in this review.

Vancomycin-resistant Enterococci
The enterococci, Enterococcus faecalis and Enterococcus faecium, are normal bowel commensals in humans and animals. They are generally of low virulence causing serious infection in debilitated, immunosuppressed patients, with little risk to healthy individuals. Common sites of infection are the urinary tract, peritoneal cavity, and surgical wounds. Rarely, they present as a bacteraemia or endocarditis. Vancomycin-resistant enterococci (VRE) were unknown before 1986 (identied in France), but have since populated the world, comprising 47% of isolates in one recent study. A direct effect on mortality is debatable once confounding variables are considered. Recent studies correcting for these have concluded that Vancomycin resistance per se does not carry an excess mortality; these isolates are no more virulent, just more difcult to eliminate. However, infection is associated with longer ICU and total hospital stay.

Selective decontamination of the digestive tract


Selective decontamination of the digestive tract (SDD) is not commonly practiced within the UK health care environment despite evidence that the incidence of ventilator-associated pneumonia may be reduced. Meta-analysis have demonstrated a reduction in the odds ratio for lower airway infections to 0.35 (0.290.41) and a 6% overall mortality reduction to 24% without an increase in superinfections because of resistant bacteria.7

116

Continuing Education in Anaesthesia, Critical Care & Pain j Volume 9 Number 4 2009

Antibiotic resistance in the intensive care unit

Strict isolation etc. linezolid, tigecycline Antibiotic restriction Hygiene and isolation. carbapenems, otherwise depending on sensitivity (e.g. gentamicin, quinolones) Strict hygiene and isolation Combination therapy; aminoglycoside and tazosin or quinolone or carbapenem. High incidence clinical failure

Isolation, limiting staff access, aprons, alcohol gel hand washing glycopeptides such as vancomycin teicoplainin

Antibiotic restriction (third generation cephalosporins, glycopeptides (and vetinarary)).

A recent comprehensive recommendations:

review8

proposed

four

key

SDD should be considered when it is anticipated that mechanical ventilation will be required for more than 48 h. SDD regimens should include topical and parenteral agents with activity against Gram 2ve bacilli; the choice of treatment should depend on local pathogen antimicrobial susceptibility proles. The use of SDD should not be withheld because of concern about the development of antibiotic resistance. SDD should be supported by good infection control and planned prospective susceptibility surveillance so that problems can be identied early and addressed.

Downloaded from http://ceaccp.oxfordjournals.org/ at Primary & Community Health Services Library on September 29, 2012

Antibiotic cycling
This is the planned withdrawal of commonly used antibiotics for a scheduled period (e.g. 6 months), to reduce selection pressure and the emergence of resistance. Different suites of antibiotics are then used which are themselves changed after 6 months. Raymond and colleagues9 prospectively studied the effect of a quarterly empiric antibiotic rotational schedule in a surgical and trauma ICU. They devised a simple rotating antibiotic class schedule for the empirical treatment of pneumonia and peritonitis or sepsis of unknown origin. Compliance was high, and they were able to demonstrate a statistically signicant decline in the incidence of resistant infections. This included gram positive (S. aureus, S. epidermidis, and S. enterococus) and gram negatives (Pseudomonas and Acinetobacter). A reduction in infectious (2.9 deaths per 100 admissions vs 9.6, P , 0.0001) and overall mortality (15.5% vs 38.1%, P , 0.0001) was seen. Rates of vancomycin resistance were unchanged, however, and there was no signicant decrease in duration of hospital stay.

Treatment strategies

Colonization is not uncommon. Bacteraemia more often seen in MFE and dialysis units

Up to one-third of nosocomial infections in USA. Sporadic in Europe 15% invasive E. coli Western Europe 2 incidence of VRE

Table 2 Common resistant organisms: relevant factors and possible management strategies

ICU incidence

Common nosocomial pathogens

Surgical wounds ICU, renal dialysis, immuno-comprised, generally debilitated

Immunocompromised, debilitated, ICU, burns etc

Restrictive antibiotic strategies


External oversight of antibiotic prescribing has been advocated by some authors. They argue that ICUs that are manned by relatively junior physicians that rely on a relatively small pool of antibiotics. Additionally, they are more likely to treat every bacterial isolate received on patients, when many such results may be evidence of colonization rather that true infection. Mechanisms proposed include the use of infection scoring for the diagnosis and treatment duration of ventilator-associated pneumonia, and the use of bronchoscopic lavage to isolate the causative organisms of pneumonia. If gram stain is negative or signicant bacterial culture is not seen with samples obtained at BAL, antibiotics are stopped. This strategy led to a signicant reduction in mortality (16.2% mortality vs 25.8% control, P 0.022) in a study by Fagon and colleagues.10 An alternative strategy involves external supervision of antibiotic therapy by an Antibiotic Stewardship Team, consisting of medical microbiologists/infectious disease physicians and pharmacists.
Continuing Education in Anaesthesia, Critical Care & Pain j Volume 9 Number 4 2009

Patient factors

Prolonged/previous hospitilization Presence of intravascular catheters.

Resistant organism

ESBLs Not Pseudomonas

As above Elderly

MRSA

VRE

Pseudomanas species

117

Antibiotic resistance in the intensive care unit

De-escalation
Many studies have shown in severe infections, particularly nosocomial pneumonia, that inadequate initial antibiotic therapy is associated with increased mortality.11 As other risk factors for mortality may not be amenable to modication by the treating physician (these being length of inpatient stay and presence of co-morbidities), adequate initial antimicrobial treatment is paramount. Furthermore, research has shown that subsequently altering prescriptions to provide adequate cover does not signicantly alter outcomes. Consequently, broad-spectrum agents are often used as initial therapy. The disadvantage of this is the increased risk of developing resistant infections when broad-spectrum use is widespread. Hoffken11 reviewed the literature regarding de-escalating strategies in the treatment of pneumonias in ICU, and demonstrated that targeting treatment and limiting duration may reduce cost without affecting outcome.

Microora surveillance
Microora surveillance has shown that the range and frequency of micro-organisms isolated varies between institutions and even within an ICU over time. The collection of regular microbiological data allows a picture of common pathogens and their antibiotic sensitivities to be collected, with clear therapeutic advantage.

Strict infection control measures including isolation, hand washing, and minimizing the number of nursing and medical staff involved in each patients care. Regular surveillance of the local microora and resistance patterns occurring within the ICU should be performed, leading to regular review of empirical antibiotic guidelines. Before administration of antibiotics, sufcient cultures should be sent to facilitate microbiological identication. Bronchoscopy should be considered if pneumonia is suspected to enable better sputum samples to be sent and improve microbiological isolation. Prescriptions should be reviewed after 2 3 days. Antibiotic therapy may then be altered according to sensitivity information, or stopped altogether if no organisms have been isolated and the clinical picture allows it. Oversight of antibiotic prescriptions by ICU pharmacists and Microbiologists may prevent inadequate dosing regimens and reduce inappropriate prescriptions.

Downloaded from http://ceaccp.oxfordjournals.org/ at Primary & Community Health Services Library on September 29, 2012

References
1. Greenwood D, Finch R, Davey P, Wilcox M. Antimicrobial Chemotherapy, 5th Edn. Oxford: Oxford University Press 2. Dellinger RP, Levy MM, Carlet JM et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008; 36: 296 327 3. Niederman M. Impact of antibiotic resistance on clinical outcomes and the cost of care. Critical Care Medicine 2001; 29 (4, Suppl.): N114 20 4. Department of Health. http://www.dh.gov.uk/en/Publicationsandstatistics/ Publications/PublicationsPolicyAndGuidance/DH_4085649 5. Blunt MD, Viira DJ, Brown N et al. The implications of methicillin resistant Staphlococcus aureus (MRSA) in the general intensive care unit. Lancet 1998; ii1197 6. Gemmell CG, Edwards DI, Fraise PA et al., on behalf of the Joint Working Party of the British Society for Antimicrobial Chemotherapy, Hospital Infection Society Infection Control Nurses Association. Guidelines for the prophylaxis and treatment of methicillin-resistant Staphylococcus aureus infections in the UK. J Antimicrob Chemother 2006; 57: 589608 7. van Saene HKF, Peters AJ, Ramsay G et al. All great truths are iconoclastic: selective decontamination of the digestive tract moves from heresy to level one truth. Int Care Med 2003; 29: 677 90 8. Masterton RG, Galloway A, French G et al. Guidelines for the management of hospital-acquired pneumonia in the UK: report of the Working Party on Hospital-Acquired Pneumonia of the British Society for Antimicrobial Chemotherapy. J Antimicrob Chemother 2008; 62: 5 34 9. Raymond DP Pelletier SJ, Crabtree TD et al. Impact of a rotating empiric , antibiotic schedule on infectious mortality in an intensive care unit. Crit Care Med 2001; 29: 11018 10. Fagon JY, Chastre J, Wolf M et al. Invasive and non-invasive strategies for management of suspected ventilator-associated pneumonia: a randomised trial. Ann Intern Med 2000; 132: 62130 ffken G, Niederman MS. Treatment of Pneumonia in the ICU: the 11. Ho importance of a de-escalating strategy for antibiotic nosocomial Pneumonia. Chest 2002; 122: 218396

The importance of correct dosing


Recent work has sought to characterize optimal dosing strategies through the use of pharmacokinetic data. Put simply it means that adequate doses are important, and underdosing should be avoided.

Clostridium difcile
The choice of antibiotic is further complicated by increasing concerns about C. difcile. Recent large-scale outbreaks, particularly among elderly medical inpatients, have raised awareness and an increase in incidence and severity of disease has been noted. Possible explanations for this are antibiotic misuse (particularly cephalosporin and quinolones), the development and dissemination of strains with increased virulence, and failures in basic infection control strategies. Management includes immediate isolation and conrmation of cases with stool sample toxin assays (24 48 h). Treatment is preferably with oral antibiotics (metronidazole or vancomycin). If the oral route is unavailable, i.v. administration is a viable although less efcacious alternative.

Practical measures
As discussed, the development of antimicrobial resistance in an intensive care setting has multiple causes. Strategies to counter these include: New admissions to ICU should be screened for target organisms and wherever possible isolated to prevent the introduction of resistant strains.

Please see multiple choice questions 10 14

118

Continuing Education in Anaesthesia, Critical Care & Pain j Volume 9 Number 4 2009

You might also like