Prediction of RecurrentClostridium DifficileInfection Using Comprehensive Electronic Medical Records in an Integrated Healthcare Delivery System

Abstract

BACKGROUND Predicting recurrent Clostridium difficile infection (rCDI) remains difficult. METHODS. We employed a retrospective cohort design. Granular electronic medical record (EMR) data had been collected from patients hospitalized at 21 Kaiser Permanente Northern California hospitals. The derivation dataset (2007–2013) included data from 9,386 patients who experienced incident CDI (iCDI) and 1,311 who experienced their first CDI recurrences (rCDI). The validation dataset (2014) included data from 1,865 patients who experienced incident CDI and 144 who experienced rCDI. Using multiple techniques, including machine learning, we evaluated more than 150 potential predictors. Our final analyses evaluated 3 models with varying degrees of complexity and 1 previously published model. RESULTS Despite having a large multicenter cohort and access to granular EMR data (eg, vital signs, and laboratory test results), none of the models discriminated well (c statistics, 0.591–0.605), had good calibration, or had good explanatory power. CONCLUSIONS Our ability to predict rCDI remains limited. Given currently available EMR technology, improvements in prediction will require incorporating new variables because currently available data elements lack adequate explanatory power.

Healthcare resource utilization and direct medical costs associated with index and recurrent Clostridioides difficile infection: a real-world data analysis

“…Therefore, it seems that preventing any recurrence is the key objective, which should be an imperative for all CDI care. Though there are published analyses reporting risk factors for any CDI recurrence, the risk factors for multiple rCDI vs. 1 rCDI have not been extensively studied before 8,[18][19][20][21] .…”

Recommendations for the diagnosis and treatment of Clostridioides difficile infection: An official clinical practice guideline of the Spanish Society of Chemotherapy (SEQ), Spaish Society of Internal Medicine (SEMI) and the working group of Postoperative Infection of the Spanish Society of Anesthesia and Reanimation (SEDAR)

“…As the microorganism is concern, it seems clear that strains with high toxin production, as is the case with many of those grouped as 027, are associated with an increased risk of recurrence [186][187][188][189]. Despite all these data, risk scores for predicting recurrences, based on the association of clinical signs or symptoms, have not functioned adequately on most studies [190][191][192][193][194][195][196] and only in some works are they attributed a certain orientative value [197,198]. Some authors have used also toxin production, through what we might consider a surrogate marker, that would be the amplification cycle of PCR curves.…”

An Observational Cohort Study of Clostridium difficile Ribotype 027 and Recurrent Infection

“…We utilized PCR ribotyping to distinguish strains, but approaches such as whole-genome sequencing may reveal other genome-derived biomarkers of adverse outcomes such as severity and recurrence. Evaluations of the performance of published models have not yet shown an AUC value above 90%, the level where reliable allocations of expensive and/or invasive treatments such as those employing fidaxomicin, monoclonal antibodies, or FMT can be made ( 25 , 26 ). Furthermore, ribotyping may be too coarse a typing method to lead one to general conclusions, as strains that are categorized as ribotype 027 strains can have variable in vitro characteristics, such as sporulation ( 27 ).…”