Recurring outbreaks of cholera coupled with lack of laboratory diagnostic capacity in low resource settings fuels clinicians’ reliance on clinical case definitions and highlights the importance of accurate diagnostic guidelines. While “rice-water” stool color is the hallmark predictor of cholera, few have examined the diagnostic accuracy of this assessment. This study assesses the sensitivity, specificity and positive and negative predictive value (PPV; NPV) of classifying stool color as “rice”, “clear” (i.e. watery) or “rice or clear” stool by either the patient or nurse for diagnosing cholera.

From March 2019-2020, a random sample of patients presenting to the International Centre for Diarrhoeal Diseases Research, Bangladesh with acute diarrhea who had a stool sample obtained were included in this analysis (N=2135).

Of the 1198 (56.1%) of patients that had culture growth, 641 (53.5%) were positive for Vibrio cholerae. “Rice” stool was reported by 518 (23.8%) patients and 640 (29.5%) nurses, while “clear” stool was reported by 1081 (49.8%) patients and 353 (16.3%) nurses. When observed by nurses, both “rice” (76%) and “clear” (85%) stool were reasonably specific but not very sensitive for cholera (44% and 20%, respectively). The combined “rice or clear” colored stool had the best balance of sensitivity (65%) and specificity (61%) with a PPV of 42% and NPV of 80%. When reported by patients, “rice” stool had high specificity for cholera (76%) but low sensitivity (25%), while “clear” stool had both poor sensitivity (54%) and specificity (51%).

Current international guidance that recommends classifying watery (clear) stool as cholera in outbreaks may still miss many patients with culture confirmed cholera even when the stool color is observed by trained health professionals and is likely not useful at all self-reported by patients. The combination of “rice or clear” diarrhea may provide somewhat more accurate assessments.