Implementation of Modified CDI Algorithm to Improve Diagnosis of Clostridium difficileInfection at Hahnemann University Hospital

Abstract

Introduction: Numbers and severity of C. difficile iatrogenic outbreaks have increased due to hypervirulent strains, increased use of antibiotics, and increased exposure in health care facilities. At HUH, an increased volume of C. difficile tests was observed on patients taking laxatives, and repeat testing was performed in known C. difficile infected patients. A change for early identification and isolation algorithm for C. difficile was implemented on May 2015. The modification discontinued C. difficile testing on samples from patients receiving laxatives and on samples within 5 days of a previous test. A quality improvement study was performed to detect the number of C. difficile diagnosis 6 months prior (PRE) and 6 months post (POST) intervention. Methods: A database was retrospectively collected of patients that were all tested for CDI from November 1, 2015 to November 1, 2016. The data of the CDIFF QUIK CHEK COMPLETE, Cepheid PCR test, and clinical data, such as laxative use was collected into REDCap for analysis. Results: A total of 1768 patients were reviewed, 919 (52%) in PRE and 848 (48%) in POST. Positive tests were 233, with 155 in PRE and 81 in POST (p<0.05). Negative tests were 1033, with 550 in PRE and 483 in POST. Canceled tests were 498, with 214 in PRE and 284 in POST (p<0.05). Laxative use in patients was seen in 627 cases, 366 in PRE and 261 in POST (p<0.05). Median for C. difficile tests ordered per day had p=0.15, processed per day had p<0.01, canceled tests per day had p=0.01, and ordered per day for patients on laxatives had p<0.01. Delay in diagnosis was based off of 291 patients who had more than one C. difficile test ordered. Of these, 15 patients had a delay in diagnosis, defined as a canceled test preceding a positive test with a p=0.02. Conclusion: The intervention was not associated with a lower number of C. difficile tests being ordered after the intervention. It showed a lower number of tests being processed due to a greater number of canceled tests. Among the patients on laxatives, the intervention showed a decrease in C. difficile tests being ordered and processed. The intervention also showed a lower number of patient encounters having a delay in diagnosis. False positives for hospitals can lead to costly contact precautions and unnecessary treatment for incorrect CDI diagnosis. Re-education on when to order C. difficile tests should be based on certain CDI clinical patient risk criteria and discontinuing testing on patients receiving laxatives.