20,000 Deaths in a Year: Where the Fight Against Staph Bloodstream Infections Stalled

In 2017, Staphylococcus aureus got into the bloodstream of an estimated 119,247 people in the United States, and 19,832 of them died. Over 2012–2017, 18% of hospital patients with it died.

For a decade before that, the headline had been good news: hospitals drove down MRSA, the drug-resistant form, with better infection control. CDC's 2019 Vital Signs report found that story had run out of momentum. The fall in hospital-onset MRSA slowed after 2012; MRSA picked up outside hospitals barely fell; and staph that ordinary antibiotics can still treat was slowly rising in the community.

Two kinds of staph, three places they start

S. aureus causes everything from skin and soft-tissue infections to invasive disease, sepsis and death. The report tracks it two ways:

• MRSA (methicillin-resistant) was long seen as a hospital germ, until community-associated MRSA, mostly causing skin and soft-tissue infections, emerged in the 1990s.
• MSSA (methicillin-susceptible) is the kind first-line drugs still work on — and it causes about half of all health care–associated staph infections. Nationally, it had been barely studied.

Each infection is sorted by where it began:

What the numbers show

Two sources were used: CDC's Emerging Infections Program (EIP), active surveillance of MRSA in six sites covering 13 million people (three California counties, statewide Connecticut, eight Georgia counties and one county each in Minnesota, New York and Tennessee), and electronic health records from the Premier and Cerner databases — 447 hospitals, an average of 325 a year, weighted to national totals.

Adjusted MRSA bloodstream infection rates, six EIP sites, 2005–2016. CDC.

Hospital-onset and community-onset MRSA and MSSA rates, relative to 2012, Premier and Cerner hospitals. CDC.

Who died. In-hospital mortality — deaths plus discharges to hospice — was 18% overall and did not change from 2012 to 2017. It was worse for infections that began in hospital:

Why progress stalled, and where the next gains are

The earlier gains came from preventing device- and procedure-associated infections — central-line bloodstream infections fell sharply in 2001–2009, especially those caused by staph — and from stopping MRSA spreading between patients, which Veterans Affairs data suggest also helped. The fall was mostly in USA100, a strain spread in health care, and less in USA300, a strain that emerged in the community and later moved into hospitals.

The country was not on track for the national action plan's goal of cutting hospital-onset MRSA bloodstream infections 50% from the 2015 baseline by 2020. CDC pointed to several places to push:

• Concentrated hospitals. About 200 hospitals account for slightly over half of the hospital-onset MRSA above the 2020 goal (unpublished NHSN data), making them the obvious priority.
• After discharge. Most MRSA bloodstream infections are health care–associated but begin in the community, and most of those occur within three months of leaving hospital. A recent study suggests that a course of decolonization — suppressing the staph a patient carries — prescribed at discharge could significantly reduce later infections. Decolonization during high-risk periods (devices in place, high-risk units, some surgeries) is promising, but the evidence does not support it as a replacement for contact precautions.
• Injection drug use. People who inject drugs have a 16-fold risk of invasive MRSA and accounted for 9.2% of cases in 2016; the opioid epidemic may be keeping community infections up. CDC called for preventing opioid misuse, linking people to medication-assisted treatment, access to sterile injecting equipment, teaching safer injection and early signs of infection, and connecting infected people to care.
• Poverty. Community-associated staph falls disproportionately on people with lower incomes, which shapes how prevention has to work.
• MSSA itself, which needs systematic study and prevention strategies of its own, along with research on a vaccine and on reducing the amount of staph people carry.

Mortality, meanwhile, depends on care: prompt diagnosis and treatment guided by susceptibility testing.

Limits

Health records could not show whether a community-onset patient had earlier health care exposure, so those cases could not be split the way EIP's were, and hospitals vary in how they record data. Its strengths were several data sources, detailed EIP case information, and clinical criteria from health records, which do not drift with changes in billing codes.

Sources

Based on Kourtis AP, Hatfield K, Baggs J, et al., "Vital Signs: Epidemiology and Recent Trends in Methicillin-Resistant and in Methicillin-Susceptible Staphylococcus aureus Bloodstream Infections — United States," MMWR Morbidity and Mortality Weekly Report 2019;68(9), Centers for Disease Control and Prevention, released early on March 5, 2019; a work of the United States government in the public domain. The imported copy's results section was truncated; the missing figures were taken from the same report's full text in PubMed Central (PMC6421967).