A model of vancomycin-resistant enterococci infection patterns found that an increase in infection prevalence in a single hospital could have significant effects on other hospitals in the same county.
Changes in the prevalence of drug-resistant bacteria in one hospital can affect the prevalence in other hospitals in the same region that share patients, suggest the results of a study
published in the August 2013 issue of the American Journal of Infection Control
The number of vancomycin-resistant enterococci (VRE) infections in patients discharged from United States hospitals has been on the rise. To simulate changes in VRE colonization prevalence in each hospital and calculate how much and how quickly these changes would affect the other hospitals, the researchers used an agent-based model of all 29 adult acute care hospitals in Orange County, California. The researchers developed the model using probabilities based on patient admission and transfer data from all area hospitals in 2006 and 2007.
The model indicated that a moderate 10% increase in VRE colonization prevalence in any single hospital would cause an average increase in prevalence of 2.8% in other area hospitals, translating to 11 VRE-colonized patients for every 388 yearly admissions, or 898 VRE-colonized patients in a hospital with 32,082 yearly admissions. However, this moderate increase in VRE colonization prevalence at a single hospital could cause up to a 61.9% increase in colonization levels at area hospitals, translating to 240 VRE-colonized patients for every 388 yearly admissions or 19,859 VRE-colonized patients for 32,082 annual admissions. A much larger, 50%, increase in VRE colonization prevalence at 1 hospital would cause an average increase in colonization prevalence of 10.4% to 11.1% in other regional hospitals. According to the model, most hospitals would not be fully affected by the infection’s spread until at least 1.5 years after the original increase in prevalence.
The study’s results also indicated that increases in infection prevalence at a hospital had a larger effect in hospitals located nearby. When VRE colonization prevalence was raised by 15% in a hospital located near the City of Orange, infection prevalence in other hospitals around the city increased by an average of 9.83%. In addition, prevalence changes in larger hospitals had the greatest influence on all other regional hospitals. The model predicted that the overall highest change in VRE colonization prevalence would be produced when patients frequently transferred between 2 long-term acute care facilities that were both experiencing infection outbreaks. Under these circumstances, the VRE prevalence would rise by 57.5% in 1 hospital and 61.9% in the other.
To better control the effects of regional VRE colonization prevalence increases, the researchers recommend that hospitals in the same area improve communication and cooperation with each other. They also suggest that control programs for VRE should consider regional effects, as the study’s results indicate that hospitals at the opposite ends of a large county can impact each other due to patient sharing.
“Because financial and operational alliances among hospitals can drive patient sharing, further understanding these alliances and other coordinated efforts between facilities who share patients can improve our understanding of VRE spread,” the researchers write. “Knowing a hospital's connections with other health care facilities via patient sharing can help determine which hospitals to include in a surveillance or control program.”