Novel drugs for Clostridium difficile (C. difficile) infections have been proven to reduce recurrent infections. Because of their high financial costs, identification of patients at high risk for recurrence is essential to provide optimal treatment. The ATLAS score’s ability to predict 90-day recurrence, disease complications and 1‑year all-cause mortality was evaluated.

144 consecutive symptomatic patients with positive stool test for C. difficile were enrolled. The ATLAS score (consisting of the variables age, temperature, leukocyte count, albumin, systemic antibiotics, serum creatinine) was calculated and patients were stratified into 4 subgroups according to their scores. A Cox regression model was used to estimate the extent to which ATLAS was associated with 90-day recurrence. Furthermore, the score was correlated with disease complications and one-year all-cause mortality.

ATLAS was unable to predict 90-day recurrence (p = 0.064, HR 1.134 [0.993;1.295]), but performed well for disease complications (D = 0.382, p < 0.001, HR 1.547 [1.266;1.889]) and mortality (p < 0.001, HR 1.374 [1.194;1.583]). Serum albumin was the only parameter able to predict 90-day recurrence (p = 0.016, HR 0.958 [0.926;0.992]) and was also a predictor of disease complications (p < 0.001, HR 0.865[0.809;0.924]) and one-year all-cause mortality (p < 0.001, HR 0.923 [0.896;0.950]). A threshold of 33.1g/L (sensitivity = 56%, specificity = 80%, AUC 0.683) and 29.2g/L (sensitivity = 75%, specificity = 70%, AUC 0.763) of serum albumin could be identified to be predictive for 90-day recurrence and one-year all-cause mortality, respectively.

Serum albumin and ATLAS are predictors of disease complications and mortality, while only serum albumin is significantly associated with 90-day disease recurrence.