Antibiotic Usage and Resistance: Title and subTitle BreakGaining or Losing Ground on Infections in Critically Ill Patients?

In this issue of JAMA, Vincent and colleagues¹ report the results of a remarkable point prevalence survey of infections in intensive care units (ICUs) worldwide and the association of these infections with outcomes of critically ill patients. The study included 13 796 patients present on a single day (May 8, 2007) in more than 1200 ICUs from 75 countries around the world. Known as EPIC II (Extended Prevalence of Infection in the ICU), the study is a 15-year follow-up to another point prevalence investigation, EPIC (European Prevalence of Infection in the ICU),² which was conducted in 1995 and included 10 038 patients, primarily from ICUs in western Europe; many of the same European institutions participated in both studies. The scope and magnitude of EPIC II, the largest of any ICU infection prevalence study, reveals several noteworthy insights into the current practice patterns of antibiotic use and infection risks in ICU patients.

The burden of infection among critically ill patients is striking, especially given the marked efforts in recent years to decrease ICU infections. For instance, in EPIC II, 51% of ICU patients were considered infected and 71% were receiving antimicrobial agents on the study day, with some antibiotic exposure for prophylaxis, and the majority of patients were receiving 2 or more antibiotics.¹ In the 1995 EPIC study, 44.8% of ICU patients were considered infected and 62.3% were receiving antimicrobial agents. In EPIC II, the risk of infection increased with disease severity and length of ICU stay,¹ infection was independently associated with hospital mortality, and infection with multidrug-resistant organisms such as Acinetobacter and Pseudomonas species and fungal pathogens was statistically correlated with excess mortality rates.¹

Some concerning trends are evident when comparing the microbiology data from 15 years ago with those from the present study. Gram-negative bacterial infections, previously thought to be on the wane,³ now outnumber gram-positive infections in ICU patients, accounting for 63% of infections in 2007 vs 39.1% of infections in 1995. This is not a favorable trend, because resistance among gram-negative bacteria is increasing^{4 - 5} and the number of therapeutic alternatives to treat these infections is diminishing.⁶ The developmental pipeline for new classes of antimicrobial agents against gram-negative bacteria has virtually run dry, and the prospects for new drugs against these pathogens in the immediate future are not good.

The proportion of ICU infections caused by Staphylococcus aureus decreased from 30.1% in EPIC to 20.5% in EPIC II, and prevalence of methicillin resistance among these isolates decreased from 60% to 50%.⁷ However, resistance trends are regional. The majority of S aureus isolates from North America were resistant to methicillin,¹ and S aureus is still the single most commonly recognized microbial pathogen accounting for ICU infection. Fungal infections increased (from a prevalence of 17% in 1995 to 19% in 2007), although severe viral infections in ICU patients have remained relatively rare (<1%).^{1 - 2} This situation will likely change radically over the next year as pandemic 2009 influenza A(H1N1) continues to cause critical illness worldwide.⁸

Bacterial pathogens have a seemingly unlimited capacity to develop resistance to environmental toxins such as antimicrobial agents.⁹ Evolutionary forces have outfitted bacteria with sufficient genetic variability and mechanisms for genetic exchange to rapidly defend themselves against antimicrobial agents. The widespread use of multiple classes of antibiotics within the ICU setting makes critical care areas the epicenter for the acquisition and dissemination of antibiotic resistance in bacterial pathogens. Selection pressures created by intense antibiotic exposure favor the selection and expression of antibiotic resistance genes among bacterial populations. Pathogens with the capacity to express and exchange these resistance genes flourish in environments with heavy antibiotic use.

The prevalence of intrinsically multidrug-resistant, environmental bacteria (eg, Pseudomonas, Acinetobacter, Stenotrophomonas species) has continued to be substantial in ICU patients, with these species combined accounting for a substantial proportion of gram-negative infections in both studies. Moreover, among commensal microorganisms, those with antibiotic resistance genes, such as enterococci and enteric gram-negative bacteria, are selected for over their antibiotic-susceptible counterparts in an ICU setting. Resistance genes are often located on gene cassettes known as integrons that can be mobilized by transposable elements and spread by multiresistant R plasmids to other bacterial pathogens.⁹ Dissemination of antibiotic-resistant bacteria within the ICU environment is an ongoing risk. Vulnerable patients with multiple catheters and instruments and with various degrees of immune dysregulation from medications and underlying disease processes are at risk for colonization and infection by antibiotic-resistant bacteria.

The critical care clinician faces a therapeutic dilemma on a daily basis. Despite published guidelines,¹⁰ the uncertain interaction at the host-pathogen interface makes it difficult to distinguish between colonization vs early infection in the ICU patient. Early intervention with appropriate antibiotics is lifesaving in patients with severe infection, yet the profligate use of antimicrobial agents contributes to progressive antimicrobial resistance.¹¹ Quality-of-care indicators now penalize physicians for delayed antibiotic use in specific situations; no such imperatives are used to limit extended and unnecessary antibiotic use. With few alternatives available, it is understandable why intensivists opt for liberal antibiotic use and rely heavily on these therapeutic agents to carry patients through critical illness to recovery.

In light of the current therapeutic quandary that ICU clinicians face, what does the future hold for antimicrobial therapy and emerging antibiotic resistance? The increasing prevalence of antibiotic resistance genes among bacteria in the community and the continued exposure of commensal bacteria to antibiotics ensures that antibiotic resistance will persist and likely worsen in the future. While contamination from the environment and cross-infection from the hands and devices of health care workers account for some infections, many infections are caused by the intrinsic microorganisms residing within or on the patient at the time of arrival in the ICU. Critically ill patients regularly exposed to invasive procedures will continue to experience morbidity from bacterial and fungal infections. Without some radical new intervention (such as antibacterial vaccines, biotherapy, immunotherapy, or genome-based therapies), trends in antibiotic resistance will continue to emerge, and therapeutic options will become increasingly limited.

With few new antibiotic classes in development, the functional antibacterial activity of existing antimicrobial agents must be preserved. Use of infection control measures that prevent cross-contamination from other patients or the ICU environment is a primary safety issue, but these measures will not eliminate the risk of infection or antibiotic resistance.⁶ Good antimicrobial stewardship using pharmacokinetic or pharmacodynamic principles to optimize the benefit and minimize the risk of antibiotics should be the norm in ICU patients.¹¹ Limiting use of antibiotics to patients with clear evidence of infection rather than colonization is essential, and discontinuation of antibiotics when their possible benefits have been obtained is also critical. New initiatives such as the use of biomarkers to aid clinicians in the decision to discontinue unnecessary antibiotic therapy should be encouraged.¹² Immunotherapies and reduced reliance on invasive diagnostic and hemodynamic monitoring techniques might also be useful in the future. Development of novel classes of antimicrobial agents is sadly lacking and needs to be a major research priority.⁶ New drugs are needed to replace the increasingly obsolete classes of antibiotics that currently exist. A “postantibiotic era” is difficult to contemplate but might become a reality unless the threat of progressive antibiotic resistance is taken seriously.