UCSF Staff Using Electronic Health Records to Curb C. diff Infection

C. diff is a dangerous hospital-acquired infection that can lead to life-threatening inflammation of the colon. To decrease instances of C. diff, UCSF Informatics employees are using electronic health records to track the sources of the bacteria.

Controlling hospital-acquired infections is a continual struggle. Annually, half a million people suffer from Clostridium difficile (C. diff) infections, and U.S. acute care hospitals spend $4.8 billion on their care. However, UCSF Health’s experts may have found an innovative and surprising way to identify and reduce the risk of infection using electronic medical records. Keep reading to learn how big data, when properly analyzed, can improve our communities’ health.

Who Is At-Risk for Dangerous C. Diff Infections?

C. diff is one of the most common hospital-acquired infections (HAIs), and at least 500,000 people are infected each year. It is particularly dangerous to senior citizens; for C. diff patients who are over 65 years old, one-in-eleven die within one month of diagnosis. Unfortunately, these bacteria are hard to kill, and simple products like hand sanitizer will not prevent transmission.

Some people who contract C. diff won’t have any symptoms. Others can experience severe diarrhea, fever, nausea, lost appetite, and stomach pain. Sometimes, it even causes severe, life-threatening inflammation called colitis.

While anyone is at risk for a C. diff infection, certain types of patients are particularly vulnerable. For example, those who:

  • Are 65 years old or older
  • Require a long hospital stay
  • Have a weakened immune system
  • Were previously diagnosed with or exposed to C. diff
  • Take antibiotic medications

While you may be surprised to hear that antibiotic use is a C. diff risk factor, broad-spectrum medications can damage the helpful bacteria in your stomach that regulate gut health and keep harmful microbes in check.

UCSF Health Is Using Technology to Battle C. Diff

Traditionally, hospitals try to control C. diff by aggressively testing suspected victims, isolating them from the general hospital population, and deep cleaning rooms and equipment that have been exposed to the bacteria. Many normal disinfectants will not kill C. diff. Instead, you must use special, EPA-approved products that eliminate its spores.

Hospitals are also being more thoughtful about the use of antibiotics. The CDC estimates that between 30% and 50% of antibiotic prescriptions are unnecessary. And research suggests that even a modest reduction in the use of antibiotics could dramatically reduce hospital-related C. diff infections. According to the CDC, cutting unnecessary antibiotic use by 30% could lead to a 25% drop in C. diff cases. Another study found that a 10% drop in antibiotic use resulted in a 34% decrease in infections.

Recently, UCSF Health made C. diff reduction a top priority. In addition to implementing best-in-class protocols, it consulted with its informatics team. These technical experts proposed a unique project: use machine-learning algorithms to scan C. diff patients’ electronic records to identify hospital-wide risk factors and potential sites of contamination.

Using Electronic Health Records to Improve Community Health and Reduce Hospital-Based Infections

Electronic health records (EHRs) track your every move in a hospital. If you’re taken out for a diagnostic test, your electronic chart will document exactly where you went and what procedures the doctors performed. It lists your medications, the people you interact with, and much more. The UCSF informatics team used these digital “breadcrumbs” to identify each C. diff patient’s steps and activities within the hospital. Using this data, they discovered that a significant number of C. diff patients had used a specific CT machine in the hospital’s ER. The machine was deep-cleaned and sanitized, reducing the risk of further contamination.

This experiment is one of the first times EHRs have been used to improve a community’s overall health, but it’s probably not the last. Healthcare experts believe that as machine learning and artificial intelligence improve, healthcare facilities and physicians will be able to harness data to identify more effective treatment protocols, reduce population-specific risk factors, and improve patient outcomes.

“The electronic health record is a treasure trove of clinical data and insights, but we are just beginning to discover how to unlock its secrets,” said Dr. Robert Wachter, Chief of Medical Service and Chief of the Division of Hospital Medicine at UCSF Medical Center. “This study demonstrates the potential to transform patient care when innovative clinicians and technology experts join hands to tackle healthcare’s hardest problems.”


Kurtzman, L. (2017, October 23). UCSF innovators use EHRs to track hospital-acquired infection. UCSF Health. Retrieved from https://www.ucsf.edu/news/2017/10/408776/ucsf-innovators-use-ehrs-track-hospital-acquired-infection

Murray, S., Yim, J., Croci, R. (2017, December). Using spatial and temporal mapping to identify nosocomial disease transmission of Clostridium difficile. JAMA Internal Medicine. Retrieved from https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2659323

Nearly half a million Americans suffered from Clostridium difficile infections in a single year (2015, February 25). CDC Newsroom. Retrieved from https://www.cdc.gov/media/releases/2015/p0225-clostridium-difficile.html