Becker's Clinical Quality & Infection Control

8 INFECTION CONTROL & PATIENT SAFETY How machine learning models are rapidly predicting C. diff infections By Megan Knowles R esearchers at Boston-based Massachusetts General Hospital, Ann Arbor-based University of Michigan and Cambridge-based Mas- sachusetts Institute of Technology are developing hospital-specific machine learning models that predict patients' risk of Clostridium difficile infections much sooner than current diagnostic methods allow, according to a study published in Infection Control & Epidemiology. "Despite substantial efforts to prevent C. diff infection and to institute early treatment upon diagnosis, rates of infection continue to increase," co-se- nior study author Erica Shenoy, MD, PhD, said in a press release. "We need better tools to identify the highest risk patients so that we can target both prevention and treatment interventions to reduce further transmission and improve patient outcomes." e study authors noted most previous approaches to C. diff infection risk were limited in usefulness since they were not hospital-specific and were developed as "one-size-fits-all" models that only included a few risk factors. erefore, to predict a patient's C. diff risk throughout the course of their hospital stay, the researchers took a "big data" approach that analyzed the entire EHR. is method allows for institution-specific models that could be tailored to different patient populations, different EHR systems and factors specific to each facility. "When data are simply pooled into a one-size-fits-all model, institutional differences in patient populations, hospital layouts, testing and treatment protocols, or even in the way staff interact with the EHR can lead to differ- ences in the underlying data distributions and ultimately to poor perfor- mance of such a model," said co-senior study author Jenna Wiens, PhD. "To mitigate these issues, we take a hospital-specific approach, training a model tailored to each institution." With this machine learning-based model, the researchers looked at de-identified data, which included individual patient demographics and medical history, details on admissions and daily hospitalization, and the likelihood of C. diff exposure. e data was gathered from the EHRs of roughly 257,000 patients admitted to either MGH or to Michigan Medicine over two-year and six-year periods, respectively. e models proved highly successful at predicting which patients would ultimately be diagnosed with C. diff. In half of these infected patients, ac- curate predictions could have been made at least five days before collecting diagnostic samples, which would allow hospitals to focus on antimicrobial interventions on the highest-risk patients. e study's risk prediction score could guide early screening for C. diff if validated in subsequent studies. For patients who receive an earlier diagno- sis, treatment initiation could curb illness severity, and patients with con- firmed C. diff could be isolated to prevent transmission to other patients. e algorithm code is freely available for hospital leaders to review and adapt for their institutions. However, Dr. Shenoy notes facilities looking to apply similar algorithms to their own institutions must assemble the appropriate local subject-matter experts and validate the performance of the models in their institutions. n ECRI Institute ranks 10 top patient safety concerns for 2018 By Mackenzie Bean D iagnostic errors earned the No. 1 spot on ECRI Institute's 2018 list of Top 10 Patient Safety Concerns for Healthcare Organiza- tions in 2018. ECRI Institute compiled the list based on an assessment of more than 2 million patient safety events collected in the ECRI Institute PSO database since 2009, along with expert opinions from a panel of internal and external patient safety leaders. "The list does not necessarily represent the issues that occur most frequently or are most severe," ECRI wrote in the report. "Most or- ganizations already know what their high-fre- quency, high-severity challenges are. Rather, this list identifies concerns that might be high priorities for other reasons, such as new risks, existing concerns that are changing because of new technology or care delivery models, and persistent issues that need focused attention or pose new opportunities for intervention." ECRI Institute suggested healthcare providers use this list as a starting point for launching patient safety discussions and setting priorities at their own facilities. Here are the 10 top patient safety issues for 2018, as listed by ECRI Institute. 1. Diagnostic errors 2. Opioid safety across the continuum of care 3. Internal care coordination 4. Workarounds 5. Incorporating health IT into patient safety programs 6. Management of behavioral health needs in acute care settings 7. All-hazards emergency preparedness 8. Device cleaning, disinfection and sterilization 9. Patient engagement and health literacy 10. Leadership engagement in patient safety n

• Cover