Computer modeling research shows that healthcare facilities and public health departments can greatly increase patient safety by working together in coordinated networks and sharing information about CRE patients, the CDC reports.
Some of the case count projections cited by CDC and others at the early stages of the Ebola outbreak turned out to be greatly overestimated, contributing to the perception by some that ratcheting up the fear level was designed to get the full support needed for the outbreak response. In this case, the CDC has collaborated with other researchers and used modeling methods already developed and time tested.
Two independently developed and complementary mathematical simulation models were used to measure the impact of a coordinated approach to prevent the spread of antibiotic-resistant organisms within a group of healthcare facilities interconnected through patient sharing (i.e., a network), using CRE as a test case.1
“CDC analysis done in collaboration with the Johns Hopkins Bloomberg School of Public Health, the University of Utah and the University of California Irvine School of Medicine clearly shows that we could see many fewer antibiotic-resistant infections and much less C. difficile if healthcare facilities and health staff, public health professionals, work together as a team,” Tom Frieden, MD, director of the CDC said at an Aug. 4 press conference.
The models are computer simulations that represent hospitalized patients as “agents” and track their dynamic interactions with other patients and CRE status throughout the healthcare system.
The model first projects out five years after CRE enters an area where 10 facilities share patients with no network or interlinked communications. The approach, which is basically the status quo for most healthcare facilities today, results in 2,000 patients getting CRE, impacting 12% of patients.
No hospital is an island
When a facility acts alone to enhance their infection control practices, the situation improves, but the lack of communication and networking with other facilities undermines the effort to some degree. The independent effort results in 1,500 patients getting CRE, impacting 8% of the total patients.
With a coordinated approach where facilities work together to prevent infections and notify each other of CRE issues before transferring patients, the modeling shows far fewer patients at risk. Four hundred patients are predicted to get CRE, impacting only 2% of the total patients.
The coordinated approach allowed for healthcare facilities to share CRE test results with a central public health authority, which used that information to strategically target prevention activity across multiple facilities. Notification of patient status as CRE-colonized or CRE-infected to facilities receiving a patient upon inter-facility transfer varied by model, and increased in frequency from independent efforts to coordinated approaches.
The models simulated the movement of patients within and between different healthcare facilities and transmission of CRE in a healthcare network based upon key parameter estimates that included inter-facility patient movement, the proportion of colonized patients recognized by routine clinical tests, and effectiveness of barrier precautions in preventing transmission. Further, models representing both large and small networks of interconnected healthcare facilities illustrate that a coordinated approach to interrupting transmission is more effective than traditional approaches that have relied on individual hospital efforts to independently identify and implement interventions.
“Facilities which go it alone can’t effectively protect their own patients,” Frieden said. “Findings from [this report] offer specific actions to turn the antibiotic resistance epidemic around.”
Given the limited options to gather this kind of data, models are a useful tool to base projections on, said one of the principal investigators, John Jernigan, MD, director of the CDC office of HAI Prevention Research and Evaluation in the Division of Healthcare Quality Promotion.
“The use of modeling is an important tool. Sometimes we can’t get the answers we need through traditional epidemiologic study techniques,” he said. “To do so would be impossible or potentially not feasible. Even if we did attempt it, there are so many complexities and weaknesses that it might be hard to interpret those results. In those cases such as this, we have choices. We can take no action. We can take action based upon opinions or guesses or use modeling as a tool to provide additional information that we think is valuable in guiding our actions.”
Centers for Disease Control and Prevention. Vital Signs: Estimated Effects of a Coordinated Approach for Action to Reduce Antibiotic-Resistant Infections in Health Care Facilities — United States. MMWR 2015;64(30);826-831