Hospital Infection Control Uses EHR to Track C. Difficile

Researchers at the University of California San Francisco developed a novel methodology using electronic health record (EHR) data to map a potential Clostridium difficile (C. diff) transmission, creating an opportunity for improvement in hospital infection control practices.

During hospitalization, patients visit procedural, diagnostic and treatment areas throughout the hospital, presenting opportunities for spore contamination of surfaces and nosocomial disease transmission.

"Every day we generate hundreds of data points for each patient in the hospital," Sara Murray, MD, MAS, Medical Director of Clinical Informatics, UCSF Health, told MD Magazine. "These data are necessary for the doctors and nurses taking care of them in real time and for the hospital staff who coordinate everything that happens to a patient. With this study, we proved that these data are incredibly valuable for answering important questions facing our patients and the health system, and are valuable in ways that we couldn’t have dreamed of even 5 years ago. The most important thing is that we’re innovative and think creatively about how to use our EHR data to tackle tough problems, and there are plenty of problems left to tackle."

The health informatics team used time and location stamps entered in EHRs whenever patients underwent procedures or were moved to a different part of the hospital to map 435,000 patient changes throughout the UCSF Medical Center between 2013—2016, which included 86,648 hospitalizations.

The team used the data to construct a map of the locations where all patients with C. diff infections traveled in the hospital and exactly when they were in each place, over the course of the 3-year study period.

If a patient diagnosed with C. diff spent time in any space, that space was considered ‘contaminated’ for the following 24 hours, regardless of cleaning practices. Patients that passed through a ‘potentially contaminated’ space were considered exposed to C. diff.

Both exposed and unexposed patients were followed for 60 days to watch the development of the disease, as measured by a positive laboratory test for C. diff — obtained in either inpatient or outpatient.

The team calculated the odds of the infection for developing the disease if exposed in comparison to unexposed individuals who passed through the same location in the hospital when it wasn’t occupied by a diagnosed patient in the preceding 24 hours. This controlled for various characteristics of patients who visited the spaces, confounding the results.

Within 60 days of being exposed in the emergency department, exposure rates were significantly associated with the development of C. diff. This effect remained significant after an adjustment for age, gender, recent antibiotics, number of location changes and length of stay.

Heat maps showed potential concentration of the effect in the emergency room’s CT scanner, and was a significant source of exposure-related infections. Adjusted analyses were performed for the CT scanner, which showed a significant association between being ‘exposed’ in the location and disease acquisition.

Patients that entered the scanner within 24 hours after a C. diff positive patient were more than twice as likely to become infected themselves.

About 4% of patients who were considered exposed in the scanner contracted C. diff within 2 months. The overall rate of infection for those who passed through the CT scanner was 1.6%.

No other sites at the hospital raised concerned in the 3 years under review. This identified an opportunity to improve cleaning practices in scanner, and the hospital moved to standardize the cleaning practices to match others used in radiology suites.

While hospitals have had EHRs, the field of clinical informatics is relatively new and allows researchers to discover how to use these records to improve patient care.

The current study is only the beginning, and researchers will continue to use EHRs to improve infection control quality at other institutions and for other infectious diseases.

The study, “Using Spatial and Temporal Mapping to Identify Nosocomial Disease Transmission of Clostridium difficile," was published in The JAMA Network.