RESUMO
OBJECTIVES: The aim of the study was to determine the test performance characteristics for point-of-care ultrasound in diagnosing constipation, through measuring the transrectal diameter (TRD). We sought to develop a sonographic numeric cutoff value for diagnosing constipation. Secondary objectives included whether certain patient characteristics would affect the TRD measurement and the accuracy of the TRD in comparison with abdominal radiographs. METHODS: We conducted a prospective, cohort study of pediatric patients between the age of 4 and 17 years, presenting with abdominal pain to a pediatric emergency department. A point-of-care ultrasound was performed with a phased array transducer (5-1 MHz). In addition to a thorough history and physical examination, each patient was administered with the Rome III questionnaire, which served as the criterion standard for the diagnosis of constipation. Radiographs and enema treatments were performed at the discretion of the treating attending physician. When enemas were administered, a postenema TRD measurement was obtained. RESULTS: Fifty subjects were "constipated" or "nonconstipated," as determined by the Rome III questionnaire. A TRD cutoff of 3.8 cm or greater correlated with the diagnoses of constipation (P < 0.001). Ultrasound-diagnosed constipation had a sensitivity of 86% (95% confidence interval, 69%-96%), specificity of 71% (95% CI, 53%-85%), negative predictive value of 0.87 (95% CI, 0.68-0.95), and positive predictive value of 0.70 (95% CI, 0.52-0.84). The TRD measurement was not affected by patient physical characteristics or bladder fullness. In 7 patients, an enema was administered. There was an overall mean (SD) decrease of 1.22 (1.62) cm; this difference was not statistically significant (P = 0.093). Abdominal radiographs were performed in 25 patients. When compared with abdominal radiographs, ultrasound had a higher specificity of 71% (95% CI, 53%-85%), but this difference was not statistically significant. Ultrasound performed similarly to abdominal radiographs with regard to sensitivity 86% (95% CI, 67%-95%), positive predictive value of 0.70 (95% CI, 0.52-0.84), and negative predictive value of 0.87 (0.68-0.95). In 22 of 25 patients who received radiographs, the ultrasound diagnosis was the same as the radiologist read of the radiographs. Potentially, 88% of radiographs could have been avoided in these patients. CONCLUSIONS: In pediatric patients with abdominal pain, there is a strong correlation of an enlarged TRD with constipation. Our results suggest that point-of-care ultrasound is a useful adjunct for diagnosing constipation and has the potential to replace the use of abdominal radiographs.