Universal molecular Clostridioides difficile screening and overtreatment in solid organ transplant recipients

Background: Screening for Clostridioides difficile (CD) colonization can be performed using molecular testing to identify the presence of microbial DNA of the toxin gene. Colonization rates for hospitalized patients are as high as 20% and may be considerably higher in solid organ transplant (SOT) recipients. Treatment for CD should be based on clinical disease and not colonization, yet clinicians may misinterpret a positive CD screen resulting in overtreatment. Objectives: The objective of this analysis is to determine how often positive CD screens resulted in inappropriate treatment with oral vancomycin. Methods: Clostridioides difficile screens were performed using the Xpert C difficile assay (Cepheid), a nucleic acid amplification testing method utilizing polymerase chain reaction (PCR), on peri-rectal swabs for newly admitted patients. This was a single-center cohort study of adult patients with CD screens hospitalized between July 2015 and November 2018. The primary outcome was the rate of inappropriate oral vancomycin treatment in all patients and in SOT recipients, defined as therapy in the absence of diarrhea. Results: Of the 47 076 total CD screens reviewed, 1,921 were positive. In the SOT cohort, 58 of 329 screens were positive (4.1% vs 17.9%, P <.01). Of all patients with a positive CD screen, 20.1% (386/1921) were treated with oral vancomycin within 48 hours of swab collection. In the SOT cohort, 39.6% (23/58) with positive CD screens were treated with oral vancomycin within 48 hours. Of the SOT patients who received oral vancomycin, 39% (9/23) did not have true CD infection. Conclusion: Solid organ transplant recipients were more likely to have CD colonization detected by peri-rectal screening than the general inpatient population. SOT and non-SOT patients were treated with oral vancomycin at similar rates in response to the positive screen. Nearly half of the oral vancomycin use in SOT recipients was likely overtreatment, but this finding is limited by the low number of patients in this cohort.