This site is intended for health professionals only
Kathleen B Bamford
Reader and Honorary Consultant in Diagnostic Microbiology
Imperial College London and Hammersmith Hospitals Trust
In the UK, the profile of antimicrobial resistance in different pathogens has been raised significantly since the House of Lords report on antimicrobial resistance improved public awareness and the requirements for trusts to report methicillin-resistant Staphylococcus aureus (MRSA) bacteraemia. But what does antimicrobial resistance mean in clinical practice and how does it impact services for patients? In this article the dilemmas caused by antimicrobial resistance when treating infections at the frontline of acute care will be highlighted. While this will address antimicrobial resistance in common bacterial pathogens relevant to hospital practice, the principles also apply to other types of pathogen.
What we mean by antimicrobial resistance
When an organism is described as being resistant to an antimicrobial it usually refers to the situation where the minimum inhibitory concentration of that agent against the organism in question is above a value that is safely achievable in the tissue. Calling an organism resistant to an antibiotic implies that, when a clinical infection is caused by that organism, it is unlikely to be effectively treated by that antimicrobial when given at standard doses.
No clinician is likely to prescribe or recommend an antimicrobial to which they know the infecting organism is resistant (ie, when relevant microbiological information is available). Thus, the real issue with antimicrobial resistance arises when empirical therapy is required and antimicrobial resistance is not suspected.
What this means for the patient with an infection requiring antimicrobials is that they are unlikely to improve if the organism is resistant to the agent chosen, that there will be a delay (possibly up to 48 hours before treatment failure is apparent), and that there may be a further delay after a second choice is made before recovery can be expected. The dilemma here is that there is a gap between the information the prescriber needs when making a choice (identity and antimicrobial sensitivities of the infecting organism) and that which is available. Some particular organisms are more likely to be associated with clinically relevant antimicrobial resistance, such as Staphylococcus aureus, Streptococcus pneumoniae (pneumococus), members of the enterobacteriaceae such as Escherichia coli and Klebsiella species and, more recently, Acinetobacter baumannii.
The pneumococcus is a good organism to illustrate the dilemmas that affect antimicrobial choice in clinical practice. This organism causes pneumonia, otitis media, meningitis and bacteraemia, with the most severe cases and greatest mortality occurring at extremes of age. In the USA, it causes 40,000 deaths per annum due to invasive disease, with mortality rates in community-acquired meningitis ranging from 20 to 50% (rates are higher in less developed countries). Importantly, the pneumococcus is part of the normal flora of the nasopharynx. Acquisition resulting in colonisation as well as dissemination are highest in preschool children.(1) Thus, this is an important pathogen that is frequently present in clinical samples. Recent antibiotic therapy has been shown to be associated with increased pneumococcal colonisation. In addition, when present, pneumococcus is more likely to be resistant to penicillin. Interestingly, this effect is more marked with some antimicrobials, such as cotrimoxazole and erythromycin, rather than others such as coamoxiclav.
Evidence suggests that high-dose, short-course antimicrobials that produce higher levels in tonsillar tissue are associated with less postantimicrobial resistance in the colonising flora.(2) Does this translate into treatment failure? Not always, but it can. High levels of treatment failure are reported in infections caused by strains with the highest level of resistance and where there is more severe, or higher risk of, underlying disease.(3) The treatment dilemma here is that this is a potentially life-threatening infection, and appropriate treatment, particularly for invasive disease, is important; in the natural course of treatment, however, clinical response may not be evident immediately. Therefore, there is a clinical imperative to assume the worst and consider the possibility of treatment failure due to antibiotic resistance. This, in turn, increases the likelihood of use of additional or second-line agents.
The outcome in infection associated with methicillin-sensitive isolates in staphylococcal pneumonia associated with bacteraemia has been compared with methicillin-resistant isolates in a study setting. In this case, poor outcome was associated with septic shock, vancomycin treatment and respiratory distress. Thus, rather than resistance to antibiotics, it was underlying disease severity and treatment with a potentially less effective agent that determined the outcome.(4)
These results are in accordance with reports on the effects of methicillin resistance in S aureus on disease outcome. In this case, again, there is a treatment dilemma: the possibility of methicillin resistance means that glycopeptide therapy could be considered, but if the infection is caused by a sensitive strain, this will be a suboptimal treatment, associated with higher mortality.
In a number of studies, the effects of using ineffective initial antibiotic therapy (IIAT) has been shown to be independently linked to mortality in Gram-negative sepsis. One study concentrated on extended-spectrum β-lactamase producing E coli and Klebsiella; IIAT in the first 48 hours for infections outside the urinary tract was associated with significantly poorer outcome.(5) Another study investigating Gram-negative bacteraemia found that IIAT in the first 24 hours was associated with higher mortality.(6)
Thus, there is a body of evidence indicating the importance of using antimicrobials that are active against the infecting organism early in sepsis. This remains the crux of the problem for prescribing physicians. We must make choices to prescribe effective therapy, often without the information we need to do so. At the same time, we must balance good microbiological practice – targeting the infection appropriately – with reducing the likelihood of unwanted side-effects. This includes recognising the risk of secondary colonisation or infection with a different or a more resistant organism, as well as the effects of antimicrobial resistance in the community.
Strategies to tackle the problem
No single approach is effective in dealing with antimicrobial resistance. This is a relentless fight that requires a joined effort from clinical microbiology and infection services, with sustained expertise provided by pharmacists and infection control practitioners.
More and better guidelines that are up to date and tailored to local clinical needs will help guide empirical therapy. These guidelines need to be informed by high-quality levels of evidence, which, in turn, requires clinical trials and more systematic gathering of information. They need to be translated into local policy at practice and unit levels, with regular audits and variance analyses.
Multidisciplinary antibiotic teams with clinically active microbiologists, antibiotic pharmacists and infection specialists are best placed to carry out antibiotic and infection ward rounds with clinical teams. Crucially, these groups need current, accurate information about the infecting organism and its anti‑microbial sensitivities. The key here is better and faster diagnostics.
Microbiology is on the brink of developing and introducing a range of molecular tools that will help in this task. Today, this is limited by cost and skills, but perhaps by weighing up the costs of getting it wrong at the bedside, the cost-benefit of novel approaches will impact the way we target anti‑