Decline in Invasive MRSA Infection: Title and subTitle BreakWhere to Go From Here?

Staphylococcus aureus, unlike many virulent pathogens, is a common commensal asymptomatically colonizing the nares¹ and other body sites in approximately 30% of healthy individuals. The remarkable success of S aureus as a human pathogen is due in large part to its ability to develop resistance to antimicrobial agents. Although it might appear that the increasing prevalence of resistant S aureus, particularly methicillin-resistant S aureus (MRSA), has been continuous and uninterrupted, understanding the evolution of S aureus resistance is helpful for interpretation of data on recent trends in invasive MRSA infection, such as those reported by Kallen and colleagues² in this issue of JAMA.

Antibiotic-susceptible S aureus of the preantibiotic era has given way to waves of antibiotic-resistant strains that have emerged over the past 60 years.³ During the mid-1940s the proportion of infections caused by penicillin-resistant S aureus increased, until these strains became pandemic in the community in the 1950s and 1960s.⁴ The first published reports of what came to be called MRSA appeared in 1961.⁵ By the late 1970s, outbreaks of MRSA appeared in US hospitals and by the 1980s, MRSA was endemic in hospitals worldwide.⁶ In US intensive care units, the proportion of all S aureus isolates that were methicillin-resistant increased from 36% in 1992 to 64% in 2003.⁷ S aureus evolution has continued with the emergence of vancomycin-resistant S aureus⁸ and most recently, linezolid-resistant S aureus.⁹

The emergence of MRSA as a community-associated pathogen began in the 1990s, and since then, community-associated MRSA strains have increasingly caused hospital-onset and health care–associated, community-onset infections.¹⁰ Even the seemingly inevitable increase in community-associated MRSA rates was neither uniform nor continuous. The initial strains (USA400) that predominated before 2001 have now been replaced by the unrelated USA300 strains that currently cause the majority of community-associated MRSA infections in the United States.³

The emergence of MRSA as an important pathogen was accompanied by increasing physician and public awareness through scientific and media reports. For instance, following publication of a large-scale epidemiological investigation of MRSA,¹⁰ media coverage about MRSA increased and wound swab cultures for MRSA doubled within 1 month following the report.¹¹ In addition, numerous US states have enacted MRSA-specific public reporting and mandatory surveillance laws.¹²

Against the backdrop of a perceived inevitable increase in MRSA and increased public awareness of the pathogen and its harms, the report by Kallen et al² in this issue of JAMA may help to refocus S aureus control efforts. Using data from 2005 through 2008, investigators of the Centers for Disease Control and Prevention's (CDC’s) Emerging Infections Program/Active Bacterial Core surveillance system, which includes a population of approximately 15 million in 9 metropolitan areas, analyzed the incidence of invasive MRSA infections and MRSA bloodstream infections (BSIs). Apart from reporting that the proportion of invasive community-associated MRSA infections increased slightly during the period (17% to 19%), the analysis and discussion were limited to health care–associated invasive MRSA infections.

The most important finding of this study was documentation of a continuous decline of invasive MRSA disease, including an estimated 9.4% annual decrease in hospital onset and an estimated 5.7% annual decrease in health care–associated community-onset infections. These decreases were even greater when the analysis was limited to BSIs and were evident in the subpopulation with dialysis-associated BSIs. Similar decreases in MRSA infection have been reported outside the United States, with a recent report from England assessing the same period noting a 24% decrease in S aureus bacteremia, solely due to declines in MRSA with the proportion of all S aureus bacteremia caused by MRSA decreasing from 41% to 23% by 2008.¹³ For the first time, recent European data also demonstrate that more countries are experiencing decreases in the proportion of invasive S aureus infections caused by MRSA than are experiencing increases.¹⁴ Burton et al¹⁵ reported a similar reduction in central line–associated BSIs caused by MRSA that began in 2001 in US adult intensive care units.

There are a variety of theories as to why invasive MRSA infection rates appear to be decreasing. Kallen et al² suggest that their reported reductions complement previously published small quasi-experimental studies of MRSA-specific interventions, such as active detection, isolation, and decolonization. However, the decreases in the United States and European Union began several years prior to the wider adoption of MRSA-specific control programs^{13 ,16} and such interventions still have not been implemented in many US hospitals. Therefore, the observed declines in MRSA reported may instead be the result of general infection control efforts such as wider adoption of alcohol-based hand rubs, improved hand hygiene compliance, efforts targeted at eliminating central line–associated BSIs, and enhanced antimicrobial stewardship programs.¹⁷ Additionally, given the general decline of health care–associated MRSA reported by Kallen et al, the internal validity of any “before-after” quasi-experimental study may be questioned.¹⁸ For example, a decrease in MRSA infections observed after an initiation of an active detection and isolation strategy might be due to the background decrease in MRSA rates rather than the intervention. Well-designed studies of factors that may have contributed to the decrease in MRSA should be performed to understand the reported decrease and direct additional control efforts.

However, it may be presumptuous to assume that hospital-based infection prevention efforts have a major effect on the natural history of a pathogen that colonizes up to a third of humans.¹ Natural biological trends, including the emergence and disappearance of specific clones, are likely to override the best-laid attempts at infection control.¹⁹ These trends affect the human epidemiology of all S aureus disease, not solely MRSA. Surveillance programs that are limited to MRSA will always provide an incomplete epidemiological picture.

An important lesson from the report by Kallen et al is that current understanding of the epidemiology of S aureus colonization and disease remains relatively limited, in part because robust surveillance systems are lacking. The CDC's Emerging Infections Program/Active Bacterial Core surveillance system, which provides active population-based sampling combined with a medical record review, is an excellent model, but includes only 9 metropolitan areas and currently MRSA is its sole tracked hospital pathogen. Continued surveillance for S aureus and the other important health care–associated pathogens (eg, Enterobacteriaceae, Pseudomonas, Acinetobacter, Candida) by this system and by the CDC's National Healthcare Safety Network, is essential for informing and targeting infection prevention efforts. In addition, the Emerging Infections Program/Active Bacterial Core surveillance system should be expanded beyond the 9 metropolitan areas and include more rural areas, and eventually all 50 states. Such surveillance should move beyond host risk factors to also characterize the pathogens more fully. A recent example is the molecular-epidemiological analysis of S aureus using isolates from 28 European reference laboratories and highlighting the regional distribution of specific MRSA clones, which could be exploited for targeting future regional or health care network interventions.²⁰ A similar network and analysis should be attempted in the United States.

Prevention of S aureus disease in humans will also require a focus on the complex interplay between colonization and infection. The specific host, pathogen, and microbiome factors that determine whether S aureus carriage is maintained in the human host, and factors that predict invasive disease among carriers must be understood. Robust surveillance programs that include collection of information about both host and pathogen will be required to begin to understand these issues.

The study by Kallen et al, when combined with previous investigations, suggests that there may be an ongoing decrease in MRSA as a cause of human infection, particularly in noncommunity settings. Although the present decrease in MRSA may be used to argue for or against MRSA-specific vs general infection prevention interventions, these arguments would be missing the point. The decreases are occurring for a reason, and only by improving existing surveillance and prevention research programs can clinicians and infection control researchers begin to explain why. Such research will be essential for guiding future approaches to all S aureus prevention. Although MRSA may be in decline, it is unlikely that S aureus will follow suit.