Vital Signs

Trends in Staphylococcus Aureus Infections in Veterans Affairs Medical Centers — United States, 2005–2017

Makoto Jones, MD; John A. Jernigan, MD; Martin E. Evans, MD; Gary A. Roselle, MD; Kelly M. Hatfield, MSPH; Matthew H. Samore, MD


Morbidity and Mortality Weekly Report. 2019;68(9):220-224. 

In This Article

Conclusions and Comment

During 2005–2017, following introduction of a system-wide, multifaceted infection control intervention that included admission screening for nasal MRSA carriage and use of contact precautions for MRSA-colonized patients, VAMCs across the United States experienced a sharp decline in S. aureus infections among hospitalized patients. Most of the reductions were explained by decreases in MRSA; reductions in MSSA rates were more modest. Although the precise relationship between the observed trends and infection control interventions are difficult to demonstrate and likely complex, a careful examination of the potential mechanisms that could explain discordant MRSA and MSSA trends provides important insights for S. aureus prevention strategies.

One potential explanation for the discordant MSSA and MRSA trends is that the observed trends represent an artifact of differential detection bias, by which MRSA-infected patients would be progressively less likely than would MSSA-infected patients to have cultures obtained over the course of the study period. There is no obvious reason that likelihood of obtaining a diagnostic culture in patients with suspected infection would differ according to a provider's clinical suspicion of MSSA versus MRSA, and there was no change in rate of diagnostic cultures obtained over the study period, nor was there any difference in diagnostic culture rate based on admission MRSA carriage status.

A second potential explanation is that shifts in S. aureus epidemiology might have influenced the observed trends. It has been suggested that downward temporal trends in community-associated infections caused by community-associated MRSA strains (e.g., USA300) might explain decreases in health care–associated MRSA..[5] Although strain data were not available for this analysis, data describing the national MRSA experience do not support this hypothesis. Population-based surveillance data from CDC's Emerging Infections Program show that although rates of health care–associated MRSA infection rates have been declining, community-associated MRSA rates have remained unchanged since 2005.[6] In addition, almost all MRSA reductions resulted from decreases in USA100, a strain associated with health care system transmission.[7] Conversely, only modest reductions were observed in USA300, a strain associated with community transmission. In the absence of replacement by other strains, this suggests that successful interruption of MRSA transmission in health care settings is an important contributor to national trends.

Infection control interventions might produce differential trends in MRSA and MSSA infection rates. Two broad approaches to preventing health care–associated infection include reducing the likelihood of invasive disease given colonization or exposure and decreasing transmission of pathogens (preventing infection by avoiding colonization or exposure in the first place). The VA system adopted both of these strategies. Similar to programs elsewhere, the VA system implemented bundled interventions designed to prevent device- and procedure-related infections (e.g., central line–associated bloodstream and surgical site infections). However, if such interventions were primarily responsible for the observed S. aureus trends, MSSA and MRSA rates would have been expected to have been affected approximately equally.

Other evidence also suggests decreased MRSA transmission as the primary mechanism for S. aureus reductions in VA hospitals. First, the discordance between MRSA and MSSA trends is consistent with mathematical modeling studies of health care transmission. Models predict that a decrease in overall transmission of bacterial pathogens in health care settings will result in disproportionately greater impact on strains having characteristics that provide a selective advantage for health care transmission, such as resistance to multiple antibiotics, including MRSA.[8,9] Thus, the VA trends are consistent with decreased S. aureus transmission as the causative mechanism, regardless of whether improvements in infection control practices specifically targeted MRSA. Second, the rate of MRSA acquisition, a direct measure of MRSA transmission, decreased markedly during the course of the study. Third, reductions in hospital-onset MRSA infection were significantly greater among patients who were not carrying MRSA at the time of admission, suggesting that practices preventing acquisition of MRSA colonization had a greater impact than practices preventing progressing to infection among colonized patients. These findings are not consistent with the hypothesis that device- and procedure-associated prevention bundles, which are designed to prevent progression from colonization to infection, were primary drivers of S. aureus reduction in VA hospitals. Finally, the striking reductions in MRSA infection rates in the early postdischarge period are consistent with decreased acquisition during inpatient stays.

The mechanisms by which transmission was prevented are difficult to determine with precision, in part because multiple interventions were occurring simultaneously. It is highly plausible that the aggressive and targeted approach to preventing MRSA transmission (i.e., screening for MRSA carriage and implementation of contact precautions for all carriers) contributed to the pronounced decrease in MRSA infections. However, the discordant MRSA/MSSA trends might also be explained by infection control practices that prevent transmission of all bacterial pathogens, but do not specifically target MRSA, such as hand hygiene. A sustained decline in gram-negative rod bloodstream infections in the VA system after implementation of the MRSA prevention program was also observed.[10] However, it is likely that contact precautions for MRSA-colonized patients contributed to this trend: another VA study showed that 31% of patients with multidrug-resistant gram-negative bacteria would have been under contact precautions because of a positive MRSA screen.[11] Changes in antibiotic use could have contributed as well. There is evidence that fluoroquinolone use is associated with increased MRSA colonization,[12] and the reduction in fluoroquinolone use could contribute to selective reduction in MRSA because it is more commonly fluoroquinolone-resistant than is MSSA. The VA did observe a substantial reduction in fluoroquinolone use, but the fluoroquinolone reductions did not begin until 2009, after substantial MRSA reductions had already occurred.

The findings in this report are subject to at least five limitations. First, the patient population in VAMCs is predominately male, although it is not clear that this characteristic would affect these findings. Second, the models used in this analysis did not include data regarding adherence to infection control practices; including such data might have provided additional insight into which components of the intervention might have had the most impact. Third, information about MSSA colonization was lacking, making it difficult to characterize MSSA transmission dynamics. Fourth, no information on MRSA or MSSA strain characteristics was available. Finally, simple exponential trends improve interpretability but might not always closely reflect trends in complex systems.

The significant reduction in S. aureus infection observed across VAMCs, driven primarily by a decrease in MRSA infection rates, offers important insights that can inform national S. aureus prevention strategy. Although the causal relationship between specific components of the VA-wide infection control intervention and the reduction in infection rates is difficult to determine with precision, it seems likely that decreased MRSA transmission played a substantial role. These data suggest that recent calls to withdraw infection control interventions[5] designed to prevent MRSA transmission, such as use of contact precautions, might be premature and inadvisable, at least until more is known about effective control of bacterial pathogen transmission in health care settings. Adherence to CDC recommendations[13] for antimicrobial stewardship, preventing device- and procedure-associated infections and interrupting transmission of health care–prevalent strains (e.g., use of contact precautions for MRSA) continue to be a mainstay of S. aureus prevention.