SAS/STAT (version 9.3., SAS Institute, Cary, NC) was used for statistical analyses.ĭespite its widespread use, to our knowledge, our study is the first to evaluate the BSFS for rater reproducibility in assessing stool form. Semi-interquartile ranges were calculated to assess variability and agreement of the Rome III clinically delineated categories (constipation, normal, or diarrhea) with scores greater than zero representing notable disagreement (more than half responses in different categories). Parameters of agreement and reliability were both determined to assess reproducibility.( 14) Statistical measures included determination of inter- and intra-rater reliability (i.e., reproducibility of ratings) using intraclass correlation coefficient (ICC) two-way random effects model with absolute agreement of single measures, and determination of sources of variance.( 15) ICC interpretation parameters were: <0.40=Poor 0.40–0.59=Fair 0.60–0.74=Good and ≥0.75=Excellent.( 16, 17)Īgreement (i.e., similarity of ratings for each photograph) was assessed using percent exact agreement and percent within one rating type of the most common (modal) rating chosen for each photograph. The study was approved by the Baylor College of Medicine Institutional Review Board. The individual stool type ratings were then classified based on the Rome III criteria, with stool types 1 and 2 designated as constipation, types 3, 4, and 5 as normal stool form, and types 6 and 7 as diarrhea.( 9) Twenty raters agreed to repeat the survey for intra-rater analyses a minimum of 2 weeks after the initial ratings. Each page of the survey contained an individual stool photograph with the BSFS graphic scale beneath. Thirty-four health care providers providing gastroenterology clinical care from three institutions were asked to rate the photographs using the BSFS ( Figure). Photographs were randomized into one of 6 sequences to minimize the potential effect of one photograph biasing the rating of a subsequent photograph. Only photographs that were focused, close-ups of entire bowel movements with white backgrounds were chosen. Thirty-five color, two-dimensional stool photographs ranging from liquid to hard pellets were obtained from publicly accessible areas of the internet. We determined inter- and intra-rater reliability and agreement of the BSFS by individual stool type and when classifying stools by Rome III clinical delineation (i.e. Given the wide usage of the BSFS but lack of assessment of its reproducibility, we initiated evaluation of the reliability and agreement of the BSFS when used to determine stool form by experts. The BSFS was adopted by the Rome III committee for use in categorizing IBS into one of four subtypes based on the propensity toward constipation, diarrhea or presence of both.( 9) Subsequently, children with IBS have been categorized using the same criteria.( 10, 11), and clinical trials use the BSFS to determine eligibility and evaluate outcomes. Scale reproducibility, which includes both reliability and agreement, is an important measure of any scale, particularly when measuring diagnostic criteria and to determine eligibility for therapeutic trials. It was validated as a measure of intestinal transit but not yet as a reproducible measure of stool form amongst raters.( 1) Despite this, it has been used to assess stool form in several gastrointestinal disorders including irritable bowel syndrome (IBS) and Crohn’s disease.( 2– 8) The Bristol Stool Form Scale (BSFS), a frequently used measure in gastroenterology practice and research, categorizes stools into one of seven stool types ranging from type 1 (hard lumps) to type 7 (watery diarrhea). Inter-rater and intra-rater reliability of the BSFS by Rome III clinical categorization decreased with intraclass correlations of 0.75 (95% CI: 0.69 – 0.81, P<0.001) and 0.65 (95% CI: 0.49 – 0.81, P<0.001) respectively. type 6 stools, representing the boundaries of normal vs. These 13 photographs were rated by the majority of raters as either type 2 vs. However, agreement decreased when using Rome III designations, with 13 (37%) photographs having significantly diverging classifications (semi-interquartile range=0.5). Inter-rater and intra-rater reliability of the BSFS by individual stool type was excellent with intraclass correlations of 0.88 (95% CI: 0.86–0.90), P<0.001) and 0.89 (95% CI: 0.86–0.91, P<0.001) respectively. Though only 4 photographs had absolute agreement (all Type 1 or Type 7), general agreement was high with 1132 (95.1%) of ratings being within one category type of the modal rating. 1190 individual stool type ratings were completed.

Escala de heces de bristol pdf