Covert Assessment of Concurrent and Construct Validity of a Chart to Characterize Fecal Output and Diarrhea in Patients Receiving Enteral Nutrition

Abstract

An accurate and convenient method for characterizing fecal output and a consistent threshold for classifying diarrhea in patients receiving enteral nutrition are required. The aim of this study is to covertly assess the construct and concurrent validity of a chart for characterizing fecal output and classifying diarrhea in patients receiving enteral nutrition. The chart was used to monitor fecal output in patients receiving enteral nutrition for a total of 280 patient days. Nurses characterized 291 fecal samples, of which 84 underwent measurement of fecal water using lyophilization and 60 underwent Clostridium difficile enterotoxin analysis using enzyme-linked immunosorbent assay. Construct and concurrent validity was assessed covertly to measure the true performance of the chart in a real-life clinical and research context. Use of the chart demonstrated higher fecal frequency (P <or= .04), heavier stools (P <or= .167), more unformed stools (P <or= .001), higher daily fecal scores (P <or= .001), and higher incidence of diarrhea (P <or= .002) on days when patients had severe hypoalbuminemia, were receiving antibiotics, or had a recent positive C difficile assay, demonstrating construct validity. The water content of samples assigned to hard and formed (62.0%), soft and formed (72.1%), loose and unformed (79.3%), and liquid (87.9%) categories was significantly different (P < .001), demonstrating concurrent validity. Under covert assessment, the chart demonstrated construct validity for characterizing fecal output, daily fecal score, and diarrhea, together with concurrent validity for characterizing fecal consistency. Use of the chart in clinical practice and research will standardize the characterization of fecal output and classification of diarrhea in patients receiving enteral nutrition.