In the United States, age-adjusted opioid overdose death rates increased by >200% during 1999–2015, and heroin overdose death rates increased nearly 300% during 2011–2015 (1). During 2011–2013, the rate of heroin use within the past year among U.S. residents aged ≥12 years increased 62.5% overall and 114.3% among non-Hispanic whites, compared with 2002–2004 (2). Increases in human immunodeficiency virus (HIV) and hepatitis C virus (HCV) infections related to increases in injection drug use have been recently highlighted (3,4); likewise, invasive bacterial infections, including endocarditis, osteomyelitis, and skin and soft tissue infections, have increased in areas where the opioid epidemic is expanding (5–7). To assess the effects of the opioid epidemic on invasive methicillin-resistant Staphylococcus aureus (MRSA) infections during 2005–2016, surveillance data from CDC’s Emerging Infections Program (EIP) were analyzed (8). Persons who inject drugs were estimated to be 16.3 times more likely to develop invasive MRSA infections than others. The proportion of invasive MRSA cases that occurred among persons who inject drugs increased from 4.1% in 2011 to 9.2% in 2016. Infection types were frequently those associated with nonsterile injection drug use. Continued increases in nonsterile injection drug use are likely to result in increases in invasive MRSA infections, underscoring the importance of public health measures to curb the opioid epidemic.

Active, population-, and laboratory-based surveillance data collected through the Healthcare-Associated Infections/Community Interface (HAIC) component of CDC’s EIP during 2005–2016 were analyzed to assess the effects of the opioid epidemic on invasive MRSA infection. A case was defined as the isolation of MRSA from a normally sterile site (e.g., blood, cerebrospinal fluid, or bone) from a surveillance area resident. National invasive MRSA disease prevalence (adjusted for age, race, sex, and dialysis) among persons aged ≥13 years who inject drugs and among persons aged ≥13 years who do not inject drugs were estimated for 2011 from EIP/HAIC data using a previously described method (8); invasive MRSA rates per 100,000 persons in both groups (and the corresponding rate ratio) were calculated in conjunction with a published population point estimate of the U.S. population aged ≥13 years who injected drugs in the previous year for 2011 (9). The six-site surveillance area used for the remainder of this report included California (three counties); Connecticut (statewide); Georgia (eight counties); and Minnesota, New York, and Tennessee (one county each). Demographic characteristics and clinical diagnoses of invasive MRSA cases among persons who inject drugs were compared with those among persons who do not inject drugs. The proportion of invasive MRSA cases that occurred among persons who injected drugs (among all invasive MRSA cases) was calculated overall and by site for each year; significance of trends was analyzed using linear regression. P values <0.05 were considered statistically significant.

Among invasive MRSA cases occurring in persons who inject drugs, demographics and health care–associated risk factors for cases ascertained during 2005–2010 were compared with those that occurred during 2011–2016 to describe changes over time. Health care–associated risk factors include specimen collection for culture >3 days after hospital admission; dialysis, hospitalization, surgery, or long-term care residency in the 12 months preceding culture; and/or presence of a central venous catheter ≤2 days before invasive MRSA culture collection. Cases among persons with none of these risk factors were considered community-associated. Trends in the proportion of invasive MRSA cases that occurred among persons who inject drugs also were assessed in three sites that reported data from 2005–2014 only (Colorado and Maryland [one county each]; Oregon [three counties]).

Among 39,050 invasive MRSA cases reported from six sites during 2005–2016, a total of 2,093 (5.4%) occurred in persons who injected drugs. The estimated rate of invasive MRSA among persons aged ≥13 years who injected drugs in the previous year was 472.2 per 100,000 in 2011, and the estimated rate among persons aged ≥13 who did not inject drugs in the previous year was 29.0 per 100,000 (rate ratio [RR] = 16.3; 95% confidence interval [CI] = 15.7–16.8). Overall, cases of invasive MRSA among persons who inject drugs were more likely than cases among persons who did not inject drugs to occur in persons who were younger (median age = 45 versus 63 years; p<0.05) and to be community-associated infections (odds ratio [OR] = 4.4, 95% CI = 4.0–4.8). Clinical diagnoses frequently associated with injection drug use were more common among patients with invasive MRSA who injected drugs than among those who did not (Table), including septic embolism, endocarditis, abscess (skin and internal), cellulitis, and osteomyelitis.

The proportion of invasive MRSA cases that occurred among persons who inject drugs approximately doubled in some sites (counties in Connecticut, Georgia, Minnesota, and Tennessee) after 2011. In the six-site catchment area, the percentage of invasive MRSA cases among persons who inject drugs declined from 6.4% in 2005 to 3.5% in 2010 (p<0.05), but subsequently increased steadily to 9.2% in 2016 (p<0.05) (Figure). Among invasive MRSA cases that occurred among persons who inject drugs, cases during 2011–2016 were more likely to occur in persons who were white (OR = 1.7; 95% CI = 1.4–2.0) and be community-associated (OR = 1.3; 95% CI = 1.1–1.6) than were cases during 2005–2010. In two of three sites (Colorado and Oregon) that reported data from 2005–2014 only, similar increases in the proportion of invasive MRSA cases that occurred among persons who inject drugs after an initial decrease (2005: 11.1% of cases; 2011: 10.6%; 2014: 15.2%) were observed.