The tipping point: patients predisposed to Clostridium difficile infection and a hospital antimicrobial stewardship programme.

BACKGROUND

The incidence and severity of Clostridium difficile infection (CDI) have increased in recent years. Predictive models may help to identify at-risk patients before the onset of infection. Early identification of high-risk patients could help antimicrobial stewardship (AMS) programmes and other initiatives to better prevent C. difficile in these patients.

AIM

To develop a predictive model that identifies patients at high risk for CDI at the time of hospitalization. This approach to early identification was evaluated to determine if it could improve upon a pre-existing AMS programme.

METHODS

Logistic regression and receiver operating characteristic (ROC) curve analyses were used to develop an analytic model to predict risk for CDI at the time of hospitalization in a retrospective cohort of inpatients. The model was validated in a prospective cohort. Concurrence between the model's risk predictions and a pre-existing AMS programme was assessed.

FINDINGS

The model identified 55% of patients who later tested positive as being at high risk for CDI at the time of admission. One in every 32 high-risk patients with potentially modifiable antimicrobial risk factors tested positive for CDI. Half (53%) tested positive before meeting the risk criteria for the hospital's AMS programme.

CONCLUSION

Analytic models can identify most patients prospectively at the time of admission who later test positive for C. difficile. This approach to early identification may help AMS programmes to pursue susceptibility testing and modifications to antimicrobial therapies at an earlier stage in order to better prevent CDI.