Author
CHUMPITAZI, BRUNO - Baylor College Of Medicine | |
WEIDLER, ERICA - Children'S Nutrition Research Center (CNRC) | |
CZYZEWSKI, DANITA - Baylor College Of Medicine | |
LANE, MARIELLA - Baylor College Of Medicine | |
SHULMAN, RUBERT - Children'S Nutrition Research Center (CNRC) |
Submitted to: Gastroenterology
Publication Type: Abstract Only Publication Acceptance Date: 4/1/2009 Publication Date: 5/1/2009 Citation: Chumpitazi, B.P., Weidler, E., Czyzewski, D.I., Lane, M., Shulman, R.J. 2009. Validation of a modified Bristol stool scale - inter-rater reliability amongst pediatric gastroenterologists [abstract]. Gastroenterology. 136(5 Suppl.1):A506. Interpretive Summary: Technical Abstract: Stool form and changes in stool form are important criteria in both clinical practice and clinical research. However, descriptions of stool form from both patients and physicians alike may be subjective and objective measurements of stool form are not well developed. Although the Bristol stool scale has been used in adult and pediatric research trials to identify stool form, it was developed and validated to standardize documentation of stool transit time rather than form. To determine the ability of experts to agree in identifying stools as being of a particular form. Pediatric gastroenterologists within a tertiary care center were shown two dimensional color photographs of various expelled types of stools ranging in form from hard to liquid. Each gastroenterologist was asked to independently rate, using a modified Bristol scale, each stool image provided from 1 (separate, hard lumps) to 5 (all liquid). A Brennan and Prediger multi-rater kappa was calculated for each stool photograph based on the responses received, with a kappa value of 0.70 or greater considered to represent adequate inter-rater agreement. Fifteen pediatric gastroenterologists rated up to 32 stool pictures. Eight (25%) stool photographs had a kappa of 1.0 with perfect inter-rater agreement. Six (19%) stool photographs had an inter-rater kappa ranging from 0.80 to 0.89. Two (6%) stool photographs had an inter-rater kappa ranging from of 0.70 to 0.79. The remaining 16 (50%) stool photographs had an inter-rater kappa ranging from 0.14 to 0.68. Despite the use of an objective scale, stool form identification may remain subjective - even in expert hands. |