Are abdominal x-rays a reliable way to assess for constipation?

ABSTRACT

PURPOSE: Currently to our knowledge no validated reliable tools are available to evaluate constipation in children. Abdominal x-rays are often done in clinical practice to evaluate patients with lower urinary tract symptoms. Although 3 previously published rating tools exist to score constipation based on x-ray, there is little information on their merits. We assessed these 3 tools for reliability among multiple practitioners. MATERIALS AND METHODS: We retrospectively analyzed abdominal x-rays in a cohort of 80 patients between ages 4 and 12 years. X-rays were independently assessed by each of us using the previously published Barr, Leech and Blethyn scoring tools. Scores were analyzed for reliability using standard statistical methods. RESULTS: The range of weighted κ score, indicating reliability, were 0.0491 to 0.4809 for the Barr, 0.1195 to 0.2730 for the Leech and 0.0454 to 0.4514 for the Blethyn method. Guidelines for κ scores are greater than 0.75-excellent, 0.4 to 0.75-good and 0 to 0.4-marginal reproducibility. ICC, another reliability measure, was 0.02577 for the Barr, 0.3313 for the Leech and 0.201 for the Blethyn method. ICC interpretations are greater than 0.75-excellent, 0.4 to 0.75-good and 0 to 0.4-poor. There was a trend toward good interrater reliability between more experienced urology practitioners with the Barr and Blethyn tools (0.48 and 0.45, respectively) but not between less experienced raters or with the Leech tool. CONCLUSIONS: Currently available scoring tools to evaluate constipation by x-ray do not have good reliability among multiple examiners. Further research is needed to develop an alternate tool to increase the reliability of x-ray to assess constipation between multiple raters.