Background:

Recent studies report the incidence of C. difficile infections (CDI) in children is increasing across the nation and that CDI may occur outside the traditional risk factors.Our objective: To describe clinical factors associated with CDI in the hospitalized pediatric population.

Methods:

Retrospective chart review over a 2 year period at an academic free–standing children’s hospital. 138 patients with CDI and 274 age–matched controls (2:1) with diarrheal illness were evaluated. Charts were reviewed for multiple factors in the clinical presentation and hospital course including measures of diarrhea severity and consistency, antibiotic exposure, Proton Pump Inhibitor (PPI) use, previous hospitalizations, and co–morbid conditions. Statistics were performed using Chi squared, Fisher’s Exact, and Mantel Hansel tests.

Results:

Cases (CD) and controls (CTL) were similar in gender and race. More CD had antibiotic exposure (65% vs. 20%; p<0.001) and recent hospitalization (65% vs. 19%; p <0.000001). However, 20% of CD had no antibiotic exposure or recent hospitalizations. Immunodeficiency (46% vs. 6%; p<0.00001) and PPI use (22% vs. 7%; p<0.001) was more frequent in CD. Of patients who were symptomatic on admission (Community–Acquired (CA)), more CD had bloody stools (26% vs. 13%; p<0.03) but CTL had more abdominal pain (41% vs. 23%; p<0.001). CD were more likely to have a history of GI disease (30% vs. 18%; p<0.005) and GI surgery (30% vs. 16%; p< 0.001). Although CD had a higher rate of return overall, (30% vs. 14%; p<0.001), the rate of return due to GI symptoms was similar in both groups (8%). Of the CD, 59% were hospital–acquired (HA–CD). HA–CD had more antibiotic exposure (33% vs 60%; p=0.003), less abdominal pain (8% vs 43%; p <0.0001) compared to community–acquired CD (CA–CD). History of CDI and antibiotic prophylaxis use was not different between HA–CD and CA–CD.

Conclusions:

Traditional risk factors of antibiotic exposure and recent hospitalization are not found in all CD patients, confirming concerns that CDI are becoming a problem in otherwise low–risk pediatric populations. Immunodeficiency, GI disease, GI surgery were highly associated with CD disease. Presence of bloody stools was the only useful symptom in the diagnosis of CDI in patients with CA–CD. Overall, the presence of risk factors was more helpful in predicting CDI than clinical symptoms, particularly for HA–CD.