Systemwide Algorithmic Approach to C. diff Testing

Pathologists and quality improvement staff at AMITA Health Saint Francis Hospital in Evanston, Illinois, had a problem. The number of Clostridium difficile (more commonly known as C. difficile or C. diff) test-positive cases was higher than the guidelines set by the Centers for Medicare & Medicaid Services (CMS) with a standardized infection ratio (SIR) of 1.2—anything greater than 1 indicates more infections than predicted given the baseline. Hospital staff suspected the problem was caused by a combination of factors, primary of which was the test method being used—a highly sensitive polymerase chain reaction (PCR) test—as well as the patients who were being tested, not all of whom met strict criteria for the test.

When pathologists and their colleagues at AMITA Health spot a problem regarding medical testing, an opportunity often presents itself in the form of a systemwide quality improvement. Janis Atkinson, MD, FCAP, system vice president of medical affairs and medical laboratory director for AMITA Health Saint Francis Hospital; Mary Haak, RN, the hospital’s director of quality and patient safety; and their team of laboratory professionals, quality specialists, and hospital leaders were able to diagnose and triage the problem through a collaborative, systemwide initiative that would ultimately provide better patient outcomes at decreased costs.

“Implementing a highly sensitive test can be an advantage in some situations, but a disadvantage in others,” said Dr. Atkinson. “Proper patient selection makes the difference.”

Two-Pronged Approach

Working through the health system’s core laboratory, Alverno, and using best practice guidelines, Dr. Atkinson and the rest of the utilization management committee laid out a systemwide initiative to improve C. diff testing both by adjusting the tests being used and the patients being testing. A well-rounded representation of invested staff made up the committee, including infectious disease clinicians, pathologists, laboratory specialists, administration, quality leaders, infection preventionists, and others.

“It really does help to seek input from throughout the hospital when you begin an improvement process like this one,” said Ms. Haak, director of quality and patient safety. “It’s a large undertaking that requires the understanding and support from a variety of staff.”

The initiative required both a high-level review from the utilization committee and the hands-on implementation from local pathology medical directors with support from their individual chief medical officers.

“I serve as an officer for the system to help drive lab stewardship initiatives,” said Dr. Atkinson, “but the success of the initiatives depends largely on implementation efforts by the on-site medical directors.”

On-site laboratory medical directors and quality experts were essential in educating physicians, nurses, and laboratory staff on the reasoning and processes of the new initiative, providing guidance through various communication channels:

• staff memorandums
• open door policies
• leadership safety and unit safety huddles
• resident meetings
• medical executive committees
• department meetings
• town halls

Dr. Atkinson said, “We want the local medical directors to take ownership of these projects because they are the trusted laboratory professional at their hospital.”

Based on the findings of the utilization management committee, AMITA decided this far-reaching information cascade would be the best way to implement their two-pronged approach to both improve the C. diff testing platform and improve the collection process.

“There are a variety of different lab tests that can be used to diagnose the infection,” said Dr. Atkinson. “And the quality of test results depends on two main factors: One is the type of test or test combinations that you use, and the other is your patient selection.”

Patient Selection and Sample Collection

While the C. diff tests available weren’t being adequately utilized for optimal results, the clinicians at AMITA Health Saint Francis Hospital first wanted to make sure they were testing the right patients. Their C. diff collection process improvement was initiated in September 2018, beginning with a “ticket to lab,” a required form that accompanied the specimen to help ensure patients both showed C. diff symptoms and hadn’t recently been tested, improving pretest probability and reducing false positives. The ticket also included a visual representation of the appropriate, and more importantly, the inappropriate type of specimen.

This first iteration of the ticket to lab was quite literally that—a paper form that clinicians would fill out and attach to specimens sent to the laboratory. The ticket had three simple boxes that had to be checked for the laboratory to accept the testing request:

• Yes, there are more than three watery stools in the past 24 hours. (Stool must be type 7 on the Bristol Stool Scale.)
• Yes, in the past two weeks during this admission, there haven’t been any positive C. diff tests.
• Yes, the patient has not had stool softeners, laxatives, or bowel preps in the past 48 hours.

These questions helped confirm that patients’ symptoms were likely consistent with those expected for C. difficile infection and not a side effect from another medication or laxatives. The screening ticket also helped ensure that repeat tests weren’t being used to “test patients for cure,” which is not an appropriate use of the test, explained Dr. Atkinson.

Ms. Haak agreed with the success of the paper ticket system. “It was a good first step to make everyone aware of the C. diff screening process we were implementing,” she said. The new process was communicated across the organization at daily leadership and unit-based safety huddles.

"Daily compliance reports from lab leadership at the safety huddle was key in hardwiring the process,” Ms. Haak said. “While it took some reminding and collaboration with clinicians, it made them more aware of the initiative before it was integrated in the electronic medical record [EMR].”

Once the IT department built the clinical screening questions into the EMR, clinicians were provided with the indications for appropriate C. diff testing right at the time of placing the test order. This is an efficient, up-front education strategy to help ensure that the right patients are receiving the C. diff test. Now, Dr. Atkinson and her utilization management team needed to discern the appropriate test, or tests, for the job.

Three Tests, Two Steps

Before the testing intervention at AMITA Health Saint Francis Hospital, the problem didn’t lie with missed cases of C. difficile but rather with too many false positives due to the type of test being used and the types of patients being tested.

“We were using a PCR test that was very, very sensitive to the presence of C. diff bacteria,” said Dr. Atkinson, “but not very specific for identifying patients with the disease C. diff colitis.”

The hospital was primarily using the C. diff molecular PCR test, which is hypersensitive to detecting the C. diff toxin gene’s presence because it amplifies the bacterial DNA several million times. However, it’s not specific enough in its measurement of infection, leading to an increase in false positives by identifying both carriers with a trace amount of the bacteria (carriers) and those who are actually infected (patients with C. diff colitis).

Why didn’t the hospital simply switch to another test for C. diff?

As Dr. Atkinson put it, “There’s no perfect test or combination of tests for C. diff out there.”

The C. diff toxin enzyme-linked immunosorbent assay (EIA) is, conversely, specific in its measurement levels but not sensitive enough to guarantee patients with C. diff won’t slip through undetected. The third type of test available, the antigen EIA, is sensitive but lacks clinical specificity in its detection of C. diff, because it can detect non-C. diff bacteria as well.

Without a good option for single-test diagnosis of C. diff, the utilization committee turned to the newest guideline recommendations from the Infectious Disease Society of America and the Society for Health Care Epidemiology of America.

After consulting the literature on the performance and testing options, a two-step algorithm involving all three C. diff tests was put into place.

Samples from patients suffering symptoms of C. diff—and who passed the improved patient selection process—were tested both by the antigen EIA and toxin EIA, covering both the sensitive and specific qualities of each test. If both tests reported positive, then that patient was treated for C. diff. If both tests reported negative, that patient did not have the infection. These tests, conducted on site at AMITA Health Saint Francis Hospital, were able to be done more quickly and at less cost.

However, if the antigen EIA came back positive and the toxin EIA came back negative, the sample was sent to the offsite core laboratory where a molecular PCR test was conducted to confirm the presence of infection. The process moved the initial tests on site to decrease costs and reduce turnaround time allowing clinical staff to more quickly isolate and treat patients testing positive for C. diff—only about 10% of tests needed to be sent off site for confirmation by the molecular PCR test.

Results: Decreased Cost, Improved Patient Care and Satisfaction

Between 2018 and 2019, the average turnaround time for a delivered result from a C. diff sample received in the laboratory dropped from 16 hours to two hours at AMITA Health Saint Francis Hospital—a staggering 87% improvement. Not only does that allow clinicians to more quickly care for patients who are positive for the infection, but the improved turnaround time, paired with a more reliable testing platform, helps avoid a great number of other costs and overtreatments that can negatively affect a patient’s satisfaction or overall health.

When a patient is thought to have C. diff, Dr. Atkinson explains, “They are put on isolation precautions, meaning they need to be in a room by themselves and staff need to gown up to enter—that costs the hospital extra money, the nurses and doctors extra time, and negatively impacts patient satisfaction. Patients are started on expensive antibiotics, and the patients sometimes get other complications from those antibiotics. If the patient is negative, none of those things need to happen.”

As Dr. Atkinson said, “When your family comes to visit you, they don’t have to wear masks and gloves—nobody likes to be in the hospital and then have to have everyone who comes to see you be covered in masks and gloves.”

Not only did turnaround time drop, but the number of tests and C. diff positives decreased significantly. Compared to the first quarter of 2018, the hospital performed 23% fewer tests for C. diff with 40% fewer positives. Fewer cases required less treatment, resulting in an estimated drug cost savings of approximately $16,500 every quarter. While not a tremendous amount of savings on its own, when paired with length of stay, isolation, and overtreatment, the costs begin to quickly add up.

Nearly one year later, AMITA Health Saint Francis Hospital continues to report an average of two-hour turnaround times, fewer cases of C. diff, and near-100% compliance with their algorithm from laboratory and clinical staff. Dr. Atkinson and Ms.Haak both credit their success to open communication and partnership across the medical, administrative, and technical teams at AMITA Health Saint Francis Hospital, Alverno Laboratories, and systemwide at AMITA Health.

As Dr. Atkinson said, “This project was a win-win. Our patients got a more accurate test result, and our hospital improved its health care-associated infections ranking.”