The impact of antibiotic prescription rates on the incidence of MRSA bloodstream infections: A county-level, US-wide analysis

• Published in: International journal of antimicrobial agents 2018-08, Vol.52 (2), p.195-200
• Main Authors: Andreatos, Nikolaos, Shehadeh, Fadi, Pliakos, Elina Eleftheria, Mylonakis, Eleftherios
• Format: Article
• Language: English
• Subjects: Anti-Bacterial Agents - adverse effects; antibiotics; Bacteremia - drug therapy; Bacteremia - epidemiology; Bacteremia - microbiology; county; Databases, Factual; Glycopeptides - adverse effects; Humans; Incidence; infection; Lincosamides - adverse effects; Medicare Part D; Methicillin-Resistant Staphylococcus aureus - drug effects; Methicillin-Resistant Staphylococcus aureus - growth & development; MRSA; Multivariate Analysis; poverty; Prescriptions - statistics & numerical data; Socioeconomic Factors; Staphylococcal Infections - drug therapy; Staphylococcal Infections - epidemiology; Staphylococcal Infections - microbiology; Sulfonamides - adverse effects; United States - epidemiology
• ISSN: 0924-8579; 1872-7913
• DOI: 10.1016/j.ijantimicag.2018.04.003

•Antibiotic use was the most powerful predictor of MRSA bloodstream infection rates at the county level.•Indicators of socioeconomic disadvantage had a secondary impact on MRSA incidence.•Combined hospital and community-based antimicrobial stewardship programs may help contain the spread of MRSA. To investigate the association of methicillin-resistant Staphylococcus aureus (MRSA) bloodstream infection with socioeconomic factors and antibiotic prescriptions at the county level. MRSA bloodstream infection rates were extracted from the Medicare Hospital Compare database. Data on socioeconomic factors and antibiotic prescriptions were obtained from the US Census Bureau and the Medicare Part D database, respectively. In multivariate analysis, antibiotic prescriptions demonstrated a powerful positive association with MRSA bloodstream infection rates [Coefficient (Coeff): 0.432, 95% Confidence Interval (CI): 0.389, 0.474, P < 0.001], which was largely attributable to lincosamides (Coeff: 0.257, 95% CI: 0.177, 0.336, P < 0.001), glycopeptides (Coeff: 0.223, 95% CI: 0.175, 0.272, P < 0.001), and sulfonamides (Coeff: 0.166, 95% CI: 0.082, 0.249, P < 0.001). Sociodemographic factors, such as poverty (Coeff: 0.094, 95% CI: 0.034, 0.155, P=0.002) exerted a secondary positive impact on MRSA bloodstream infection. Conversely, college education (Coeff: −0.037, 95% CI: −0.068, −0.005, P=0.024), a larger median room number per house (Coeff: −0.107, 95% CI: −0.134, −0.081, P < 0.001), and an income above the poverty line (100% < income < 150% of the poverty line) (Coeff: −0.257, 95% CI: −0.314, −0.199, P < 0.001) were negatively associated with MRSA incidence rates. A multivariate model that incorporated socioeconomic data and antibiotic prescription rates predicted 39.1% of the observed variation in MRSA bloodstream infection rates (Pmodel < 0.001). MRSA bloodstream infection rates were strongly associated with county-level antibiotic use and socioeconomic factors. If the causality of these associations is confirmed, antimicrobial stewardship programs that extend outside acute healthcare facilities would likely prove instrumental in arresting the spread of MRSA.