Intra-rater Variability of the Brussels Infants and Toddlers Stool Scale (BITSS) Using Photographed Stools.

Abstract

The Brussels Infants and Toddlers Stool Scale (BITSS) was developed for the assessment of stool consistency in non-toilet-trained children. This study aimed to (1) investigate the intra-rater reliability of the BITSS among health care professionals (HCPs) and caregivers (CGs); (2) study a potential learning curve; (3) explore the impact of photo quality on intra-rater reliability. Photos of diapers containing stool were assessed twice by 4 HCP (2432 photos) and 8 CGs (492 photos) using the BITTS. Intra-rater reliability was calculated by the percentage of exact agreement and a κ-value. A learning effect and the impact of photo quality was explored using mixed linear model and generalized estimating equations. HCPs generated 24,320 stool consistency ratings: 12.1% were scored as watery, 31.0% loose, 29.4% formed, and 27.6% hard. CGs performed 7872 ratings: 9.2% classified as watery, 34.6% loose, 28.9% formed, and 27.3% hard. Intra-rater reliability (κ) for HCPs ranged from 0.64 [95% confidence interval (CI) = 0.61-0.66] to 0.78 (95% CI = 0.76-0.80) and from 0.68 (95% CI = 0.63-0.73) to 0.94 (95% CI = 0.91-0.97) in the CG group. Both groups had <1% improvement in the odds of identical classification per 50 photos. The percentage of absolute agreement was higher in photos rated as good quality than those that were not (HCPs: 80.3% vs 69.5%, P < 0.001; CGs: 90.4% vs 86.3%, P < 0.001). The BITSS has an excellent intra-rater reliability for the stool consistency scoring of photographs of stools in diapers, but can be influenced by photo quality. A clinically meaningless learning effect was found.