An understanding of the prevalence of resistant Staphylococcus aureus and the risk factors for infection with resistant isolates is essential to help clinicians choose appropriate antibiotic therapy. Selection pressure due to prior and inappropriate antibiotic use of any kind seems to be the main driving force behind the increasing rates of multidrug resistance in methicillin-resistant S. aureus (MRSA) strains. Resistance to glycopeptide antibiotics in MRSA has also emerged in recent years, along with increased use of vancomycin to treat serious infections due to MRSA. Infections due to MRSA are associated with significantly increased morbidity, mortality, length of hospital stay, and costs, compared with infections due to methicillin-susceptible S. aureus, despite adjustment for disease severity and initially appropriate antibiotic treatment. Improvements in the preparation and dissemination of antibiograms, along with adequate public reporting of MRSA trends, are needed to address the challenge of choosing appropriate initial antibiotic treatment for MRSA infections.
The past decade has witnessed an alarming increase in the prevalence of antimicrobial resistance among many commonly encountered bacterial pathogens . This phenomenon has been particularly striking in hospital-acquired infections due to Staphylococcus aureus. The National Nosocomial Infections Surveillance System reported a 12% increase in methicillin-resistant S. aureus (MRSA) infections in 2003, compared with the prevalence noted from 1998 to 2002 . Strikingly, now <63% of all S. aureus isolates recovered from critically ill patients are MRSA. More importantly, methicillin resistance, first described in 1961, has not remained geographically contained to certain regions of the globe. Current information from The Surveillance Network–USA, an electronic surveillance network that reports resistance trends, shows that MRSA is prevalent in all regions of the United States (figure 1) . Additionally, MRSA now represents a worldwide problem, with the pathogen increasingly noted in Europe and Asia . Researchers believe that MRSA initially arose because of selection pressure resulting from antibiotic use and misuse in hospitals. Issues with continuing antibacterial stewardship, poor infection control, and the increasing severity of illness in hospitalized patients have contributed to the recent explosion in the prevalence of MRSA. Similar selection pressures due to physician behavior (e.g., inappropriate use of antibiotics and failure to follow hospital infection control policies) have contributed to the emergence and spread of community-associated MRSA (CA-MRSA) .
Resistance Patterns In Mrsa
At the level of the individual patient, the site of infection correlates with the likelihood of infection due to a resistant pathogen in general and due to MRSA specifically. For example, an analysis of antimicrobial surveillance data for 15,439 S. aureus—infected patients indicated that MRSA was most commonly implicated in infections in the lung (46% of lung infections), the urinary tract (38% of urinary tract infections), the bloodstream (30% of bloodstream infections), and skin/soft tissues (30% of skin/soft tissue infections) . This general distribution of the patient-centered burden of MRSA also reflects the pattern of nosocomial infections regularly encountered in hospitalized patients. In other words, hospital-acquired pneumonia is one of the most frequently diagnosed nosocomial infections, and, thus, MRSA is regularly recovered from subjects with hospital-acquired pneumonia. One should also note that MRSA is one of the few pathogens routinely implicated in nearly every type of nosocomial infection, ranging from pneumonia to bacteremia. Notably, in nosocomial infections, biofilm provides an excellent and relatively protected growth medium for MRSA. Hence, because both endotracheal and tracheostomy tubes are often covered with biofilm, the lung can serve as an important reservoir of MRSA . Similarly, MRSA is a leading pathogen in catheter-related bloodstream infection, again because these devices quickly become coated with biofilm. MRSA itself contributes to and promotes the formation of biofilm, which, in turn, facilitates the transfer of genetic material conferring resistance between species .
In addition to variability based on site of infection, the epidemiological profile of MRSA varies by point of acquisition. A retrospective cohort study of 4543 hospitalized patients with pneumonia from 59 US hospitals found significant differences in the occurrence of MRSA infections as a percentage of S. aureus infections, based on whether the infection was classified as community versus health care acquired . Specifically, MRSA was isolated relatively infrequently in hospitalized patients with community-acquired pneumonia. In patients with hospital-acquired pneumonia and ventilator-associated pneumonia, however, MRSA was a leading pathogen.
With respect to CA-MRSA, skin and soft tissue remain the predominant sites of infection (77%), followed by much lower prevalences in wound infections (10%), urinary tract infections (4%), and bacteremia (3%) . The lung is presently a relatively rare site of CA-MRSA infection (2%). A comparison of CA-MRSA versus health care—associated MRSA (HA-MRSA) conducted by Naimi et al.  confirmed these observations and reported that skin/soft-tissue and otitis media infections were disproportionately due to CA-MRSA (P < .001), compared with respiratory tract and urinary tract infections. These infections, as one would predict, were due to MRSA, caused by traditional HA-MRSA strains (P < .001). It should be noted, however, that pneumonia due to CA-MRSA is increasing. Several recent case series indicate that CA-MRSA can cause an aggressive necrotizing pneumonia in previously healthy subjects. Because of this trend, the most recent community-acquired pneumonia guidelines from the Infectious Diseases Society of America and the American Thoracic Society suggest early consideration of CA-MRSA in individuals with a severe necrotizing pneumonia .
Changing epidemiological profile of MRSA. As is indicated above, there has been an extremely rapid increase in CA-MRSA in the past few years. A 15-year study of MRSA at the Naval Medical Center, San Diego, showed a relatively steady rate of MRSA isolation during the years 1990 to 2001, followed by a steep increase since 2002 in the number of MRSA isolates, including both CA-MRSA and HA-MRSA . Highlighting this changing epidemiological profile, Moran et al.  studied temporal trends in the prevalence of MRSA infection among patients presenting to emergency departments with skin/soft-tissue infections. They noted that the prevalence of MRSA varied across the country but that it was found in all regions. They also reported that MRSA accounted for nearly 60% of all infections studied. of concern, genotyping revealed that 97% of these isolates were of the staphylococcal chromosome cassette SCCmec type IV and, thus, were representative of community acquisition. Moran et al.  also found that 57% of patients received inappropriate antibiotic treatment of the index MRSA infection. Underscoring both that selection pressure may drive the trends noted in the prevalence of CA-MRSA and that clinicians must now take an antibiotic exposure history, use of any antibiotic within the past month (odds ratio [OR], 2.4) was found to be a leading risk factor for eventual CA-MRSA infection. Other variables associated with the recovery of CA-MRSA included presence of an abscess or “spider bite” lesion (OR, 3.0), history of MRSA infection (OR, 3.4), or close contact with someone with a similar skin infection (OR, 3.8) .
Antibiotic use as a risk factor for MRSA. Antibiotic use at both the patient and institutional level correlates with the risk for MRSA colonization and infection. Multiple analyses reveal that, in hospitalized patients, prior antibiotic exposure (of almost any kind) is strongly linked to subsequent infection with MRSA [4, 9, 12]. Some of these reports have been confounded by the fact that patients with MRSA infections are more seriously ill and, thus, have been in the hospital longer than those not as critically ill. Therefore, severity of illness may not be appropriately taken into consideration when the nexus between prior antibiotic use and MRSA is evaluated. However, even when this issue is controlled for, prior antibiotic exposure is a risk factor for MRSA. On an institutional level, quinolone utilization may drive selection pressure and increase rates of MRSA emergence. For example, MacDougall et al.  have demonstrated a direct association between fluoroquinolone use and emergence of resistant organisms. These authors evaluated the relationship between use of fluoroquinolones and the percentage of resistant Escherichia coli and S. aureus, using isolates from 17 US hospitals and their surrounding communities. Prescription data obtained from IMS Health was used to determine the rate of fluoroquinolone use within a 16-km radius from each hospital. Hospital antibiograms were employed to determine the percentage of resistant isolates, and linear regression was utilized to determine the relationship between fluoroquinolone use and resistance. The authors reported a statistically significant association between total fluoroquinolone use within hospitals and the prevalence of MRSA (r = 0.77; P = .0003) . This and similar associative studies strongly implicate quinolone overuse as a culprit in increasing rates of MRSA emergence. Nonetheless, these epidemiological analyses do not consider other secular trends, such as changes in infection control policies and alterations in the variability of the severity of illness when admitted to the hospitals in question. Additionally, no randomized study exists showing higher rates of superinfection with quinolone-based therapies, compared with alternative antimicrobial agents.
Epidemiological profile of glycopeptide resistance in MRSA. To date, only a few isolated cases of complete glycopeptide resistance in S. aureus have been reported . Low-level resistance, known as “vancomycin-intermediate S. aureus” (VISA) or “glycopeptide-intermediate S. aureus,” is more common . VISA isolates display a thickened cell wall, but the genetic origin of these strains is unknown. To date, all cases of glycopeptide resistance have occurred in isolates concomitantly resistant to methicillin. Staphylococci may also become partially resistant to vancomycin; these are known as “heterogeneous VISA”. These populations have a mix of susceptible (MIC, ⩽4 μg/mL) and resistant (MIC, <4 μg/mL) organisms . The Clinical and Laboratory Standards Institute (CLSI) cutoffs have recently been updated with vancomycin MICs as follows: susceptible, ⩽2 μg/mL; intermediate (VISA), 4–8 μg/mL; and vancomycin-resistant S. aureus, ⩾16 μg/mL . The true incidence of heterogeneous VISA is unclear, but estimates suggest a prevalence rate of 0%–5% in the United States. Inconsistency in descriptions of the epidemiological profile of glycopeptide-resistant S. aureus arises, in part, because of a lack of standardization in testing and classification .
In a study conducted in Turkey, where the glycopeptide teicoplanin is used in high volumes, Sancak et al.  found that the prevalence of heterogeneous VISA/MRSA increased from 1.6% in 1998 to 36% in 2001. MRSA strains from hospitalized patients were screened both with brain-heart infusion agar containing 4 mg/L vancomycin and by macro Etest (AB BIODISK). Population analysis was used to confirm isolates with intermediate susceptibility. This study confirms that widespread use can contribute to the emergence of resistant strains and is consistent with earlier observations regarding the impact of antibiotic utilization on eventual rates of antimicrobial resistance.
Consequences of Antimicrobial-Resistant Bacteria
Antimicrobial resistance has multiple implications with respect to outcomes. Persons infected with resistant organisms, such as MRSA, face higher mortality rates [18–20]. This is partially due to the fact that infection with a resistant organism is associated with the risk for initially inappropriate antibiotic therapy, a major determinant of survival in serious infections [18, 19, 21]. Additionally, the heightened risk for death may reflect that resistant pathogens have more virulence factors. Beyond mortality, resistant pathogens contribute to escalating hospital costs [19, 21, 22]. Infection with MRSA, for example, prolongs hospital and intensive care unit lengths of stay. In other words, MRSA is associated with adverse mortality and morbidity outcomes.
Underscoring the significance of MRSA's impact on mortality, Cosgrove et al.  performed a meta-analysis of the studies of S. aureus bacteremia comparing survival in those with MRSA versus MSSA. Although limited because many of the studies included in the meta-analysis did not describe initial rates of inappropriate antibiotic prescribing, MRSA bacteremia nearly doubled the probability of death. In a second study, Cosgrove et al.  reported that MRSA infections (again, bacteremia) were associated with a longer length of stay in the hospital (1.29 times longer; P = .016) and higher costs (1.36 times greater; P = .017) than MSSA bacteremias. With respect to ventilator-associated pneumonia, Shorr et al.  noted that MRSA infection prolonged intensive care unit stays significantly (median, 33 vs. 22 days; P = .047). This added morbidity and economic burden remained even after adjustment for initially appropriate antibiotic therapy . In a similar, retrospective analysis of critically ill patients with bacteremia, Blot et al.  reported that MRSA bacteremia (relative to MSSA bacteremia) correlated with higher rates of acute renal failure and hemodynamic instability, longer durations of intensive care, and higher rates of ventilator use. Furthermore, both in-hospital and 30-day mortality rates were significantly (P < .05) higher among patients with MRSA, despite adjustment for disease severity. Representative examples of adverse outcomes of MRSA infection are depicted in figure 2.
Costs are also higher for patients with MRSA infections. In studies of bacteremia, Cosgrove et al.  and Reed et al.  reported that the costs of MRSA infections were significantly (P = .008 and P = .012, respectively) higher than those with MSSA infections. The difference was even more dramatic in a study of 479 patients with surgical site infections ($29,455 for surgical control subjects with no infection, $52,791 for patients with MSSA infections, and $92,363 for patients with MRSA infections [P < .001 for all group comparisons]) . It may be emphasized that the main factor driving excess attributable hospital costs in this and like analyses is duration of hospitalization. Added charges for antibiotic use, radiologic procedures, or laboratory services rarely contribute to the major differences seen in costs between MRSA and MSSA infections.
Need for improvement in reporting and antibiogram practices. Despite abundant evidence that infections with resistant pathogens are both common and associated with increased costs and mortality, many hospitals inadequately track and report rates of nosocomial infection and the prevalence of resistant organisms. Because of frustration with this issue, and in an effort to spur better awareness in both the medical and lay communities, Pennsylvania became the first state to publicly adopt regulations requiring hospitals to report nosocomial infections. The state's report examined year 2005 data on 1.6 million patients treated in 168 hospitals throughout the state . An average of 12.2 patients per 1000 had acquired infections while in the hospital, primarily surgical site, urinary tract, pneumonia, and bloodstream infections. The average cost for patients with acquired nosocomial infection was $185,260, compared with $31,389 for patients without acquired infection. In addition, the mortality rate was 12.9% when nosocomial infection occurred, versus 2.3% with no infection . The success of the Pennsylvania initiative at spurring interest in this area and at encouraging hospitals to take better “ownership” of this issue has prompted other states to consider similar legislation. The Pennsylvania reporting law, however, had certain limitations: there was no adjustment for severity of illness, there was wide variability in the infection tracking and reporting abilities of hospitals, and there were changes in infection categories; in addition, the hospital charge data do not reflect actual reimbursement amounts, which were significantly lower .
At the local level, effort is needed to improve the preparation and dissemination of antibiograms. Clinicians must have access to local information to ensure that they are making appropriate choices when it comes to initial antimicrobial therapy. Given the variability in prescribing patterns from hospital to hospital and often between providers within the same hospital, it is illogical to conclude that trends in resistance seen nationally apply to a particular center. Hence, without precise, local epidemiological information, physicians lack a key tool for improving their initial antibiotic selections. In that vein, the CLSI has placed emphasis on the preparation and dissemination of antibiograms . CLSI guidelines suggest that antibiogram reports should (1) be used to guide empirical therapy while culture results are pending, (2) have a critical role in monitoring patterns of antibiotic resistance, and (3) reflect patient care needs and the institution's formulary. Lautenbach and Nachamkin  found considerable variability in antibiogram approaches in an analysis of 65 US medical centers. More precisely, <50% of centers reported intensive care unit—specific susceptibility rates. of note, laboratories in teaching hospitals and those with more personnel offered significantly better antibiograms.
Furthermore, Zapantis et al.  analyzed 209 antibiograms from 149 institutions across the United States. Their goal was to identify compliance with CLSI guidelines. Although 85% of the antibiograms were in compliance with at least 5 of 10 CLSI recommendations, clinically relevant elements that were not met included annual analysis, notation of duplicate isolates, and exclusion of organisms with <10 isolates. Unusual susceptibility results were identified in 14% of antibiograms, including inconsistent β-lactam susceptibility among staphylococci, <100% susceptibility to imipenem among Escherichia coli; in addition, 37% of antibiograms reported <0% susceptibility to ampicillin among Klebsiella pneumoniae . These investigators also found that there was an absence of thorough screening before distribution. As a result, it has become evident that a multidisciplinary approach that includes physicians, infection control personnel, microbiologists, and pharmacists is needed for reviewing antibiogram data before its eventual dissemination.
Growing rates of MRSA infection, along with the increasing prevalence of resistance among other pathogens, must prompt physicians to alter their behavior. We must acknowledge that traditional approaches to antibiotic prescribing (that involve treating susceptible infections with narrow-spectrum antimicrobials) fail to address the frequency with which resistance is encountered and, thus, expose patients to the risk of ineffective therapy. Similarly, indiscriminate use of antibiotics only contributes to further resistance. We must also accept that poor choices have significant consequences in terms of both mortality and cost. Crucial to improving outcomes will be an emphasis on improved data collection. Through understanding local patterns of nosocomial infection and local susceptibility and antibiogram information, physicians can develop strategies that can effectively improve outcomes in locally hospitalized patients with serious infections.
Supplement sponsorship. This article was published as part of a supplement entitled “Bugging the Bugs: Novel Approaches in the Strategic Management of Resistant Staphylococcus aureus Infections”, jointly sponsored by the Dannemiller Memorial Educational Foundation and Emeritus Educational Sciences and supported by an educational grant from Ortho-McNeil, Inc., administered by Ortho-McNeil Janssen Scientific Affairs, LLC.
Potential conflicts of interest. A.F.S. has received research/grant support from and is a consultant for Ortho-McNeil, Pfizer, AstraZeneca, and GlaxoSmithKline; and is on the speakers' bureaus for Ortho-McNeil, Pfizer, AstraZeneca, Merck, and GlaxoSmithKline.
- staphylococcus aureus
- glycopeptide antibiotic
- drug resistance, multiple
- length of stay
- genus staphylococcus
- antibiotic therapy
- methicillin-resistant staphylococcus aureus
- methicillin-resistant staphylococcus aureus infection
- severity of illness
- infection risk