Assessing hard and loose “endpoints”: comparison of patient and expert Bristol Stool Scale scoring of 2280 fecal samples

Background Stool consistency is an important outcome measure to evaluate in the investigation of several gastrointestinal diseases. The Bristol Stool Scale (BSS) is one of the most commonly used tools for evaluation of stool consistency. BSS ranges from 1-7 and each score is assigned to a given consistency of the feces. Self-reported characterizations can differ from an expert evaluation, and the reliability of BSS is unclear. We aimed to evaluate the reliability of BSS by comparing patient scores with expert scores. Methods Patients with inflammatory bowel disease collected stool samples throughout a 3-year follow-up. The stool´s consistency was evaluated with BSS by the patients and matched with an expert score. Agreement between patient and expert scores was assessed using Cohen’s kappa. Results BSS scores from 2280 fecal samples collected from 992 patients at up to five time points were included. When all samples were compared, there was good to substantial agreement between patient and expert scores (Cohen’s weighted kappa: 0.66-0.72). When the BSS scores were simplified and categorized as 1 (scores 1-2), 2 (scores 3-5) or 3 (scores 6-7), the agreement improved slightly (Cohen’s weighted kappa: 0.73-0.77). When the scores from the first sample per patient were compared, the experts were more likely to assign higher scores compared to the patient. The proportion of the lowest assigned scores (1-2) was 12.1% for patients and 8.1% for experts. Conclusions The agreement between patient and expert BSS scores is good to substantial, especially when the BSS scores are simplified into three categories.