Development Of An Ecological Model To Predict Risk For Acquisition Of Clostridium difficile-Associated Diarrhea During Acute Care Hospitalization

Clostridium difficile-associated diarrhea (CDAD) is an infectious disease of special concern to WOC nurses due to its association with both skin breakdown and pseudomembranous colitis requiring ileostomy.  Traditional models of infection control have failed to stop the global spread of this disease, and increasing virulence has been identified.  An alternative model for infection control based on ecological principles may be more useful to nurses working in clinical settings, since the emphasis in such models is on identification of the most susceptible hosts and creating a protective environment.

The purpose of this study was to develop and test an ecological model for predicting susceptibility of individuals to CDAD during acute care hospitalization.  The study used a case-control design comparing 66 CDAD cases to controls matched for hospital admission date and primary inpatient unit of care.  Cases and controls were compared on demographic variable of age, and on microenvironmental variables of antibiotic burden, laxative or bowel preparation burden, nutritional status, gastric acid suppression therapy, enteral feeding, and severity of illness.

Five variables were significantly associated with CDAD status on univariate analysis:  age, severity of illness, serum albumin levels, length of exposure, and proton pump inhibitor drug burden.  Following multivariate analysis using logistic regression, severity of illness, length of exposure to the hospital environment, and a decrease in antibiotic drug burden were significantly associated with hospital-acquired CDAD.  ROC curves identified severity of illness and exposure to the hospital environment as strong predictors, but decrease in antibiotic burden to be a weak predictor. 

This study supports the use of an ecological model to identify individuals susceptible to CDAD in clinical practice and future research.