Electronic Alerts Aid in C diff Diagnostic Stewardship

Using electronic alerts for diagnostic stewardship for Clostridioides difficile reduced testing for the infection as well as the proportion of inappropriate testing, according to a report from Clinical Infectious Diseases.

Investigators from the Cleveland Clinic conducted a literature review in order to determine if alerts related to C diff infection diagnostic stewardship were effective at reducing inappropriate testing volume and rates of C diff among hospitalized adult patients. The study authors wrote that the optimal diagnostic strategy for C diff remains unclear, though Infectious Diseases Society of America and Society for Healthcare Epidemiology of America currently recommend choosing testing methods based on whether there are prespecified institutional criteria to limit appropriate testing. With this in mind, the investigators also aimed to determine if alerts had any bearing on patient-centered outcomes, such as rate of infection with C diff.

The investigators searched 5 databases for original studies and found 11 that met inclusion criteria. Studies were not included if they focused solely on pediatrics or if they were review articles, editorials, or conference abstracts. The studies ultimately selected by the investigators underwent review for data extraction based on the characteristics of the study, its design, the type of alert triggers used, co-interventions, and its outcomes.

All of the studies were conducted and published within the last decade in the United States. Additionally, 6 of 11 were single-hospital studies. All 11 articles studied the number of tests ordered, while 6 studied the number of “appropriate” tests, and 7 studied C diff infection rate, the study authors found.

The different alerts used in the studies were as follows:

• An alert discouraged C diff infection testing when the patient had a recent laxative use (9 studies)
• An alert discouraged C diff infection testing when there was a lack of clinical indication for testing (7 studies)
• An alert discouraged C diff infection testing when the patient had recent C diff infection testing (6 studies)
• A mechanism that prevented a physician from immediately overriding the alert (5 studies)

“Despite good evidence suggesting the efficacy of clinical decision support tools in reducing inappropriate C diff testing, there is a general lack of investigation into the unintended adverse consequences of clinical decision tools, including delayed diagnosis or alert fatigue on the part of clinicians, with the exception of a few studies,” study authors Abhishek Deshpande, MD, PhD, and Aaron Dunn told Contagion®. “Based on this observation, we note in the article a need for increased attention to unintended adverse consequences of such tools in future studies.”

There was a statistically significant decrease in the number of C diff tests ordered after the alert intervention, as demonstrated in 6 studies, the study authors found. No statistically significant measurement was found in 4 studies.

All 6 studies that examined the measure of “appropriate” testing found that there was a significant increase in the proportion of appropriate testing following the alert intervention, the investigators learned. There was about a 15% increase in the proportion of appropriate tests in 3 studies while 2 studies showed an absolute increase of <3%.

The alert intervention was responsible for a statistically significant decrease in C diff infection, at least according to 4 of the included studies. The study authors wrote that “it is challenging to summate the results of these studies or to make generalizations based on alert trigger-types” given the variations in the different types of the alerts they studied.

“The findings of our study reinforce the need for judicious testing in patients with suspected C diff infection,” Deshpande and Dunn added. “We hope that the findings of our study will encourage physicians to continue engaging with and investigating clinical decision support tools.

“Caution must be taken when using such tools, as their negative impacts remain understudied,” Deshpande and Dunn concluded.