Ultrasonographic and clinical predictors of intussusception

Abstract

Objective: The objective of this study was to determine the positive and negative clinical predictors of intussusception and the correlation of ultrasonography and air enema in establishing this diagnosis. Study Design: This was a prospective descriptive cohort study. Setting: This study was performed in a tertiary care pediatric emergency department. Participants: Eighty-eight of 245 candidates were assessed for clinical predictors of intussusception. All 245 cases were examined for correlation between ultrasonography and air enema. Interventions: A questionnaire, ultrasonography, and air enema were used. Results: Thirty-five of the 88 patients assessed for clinical predictors were positive for intussusception. Significant positive predictors were right upper quadrant abdominal mass (positive predictive value [PPV] 94%), gross blood in stool (PPV 80%), blood on rectal examination (PPV 78%), the triad of intermittent abdominal pain, vomiting, and right upper quadrant abdominal mass (PPV 93%, p = 0.0001), and the triad with occult or gross blood per rectum (PPV 100%, p = not significant). Significant negative predictors were a combination of ≥3 of 10 clinically significant negative features (negative predictive value 77%, p = 0.035). Of the total 245 cases, intussusception (as confirmed by doughnut, target, or pseudokidney sign) was ruled out by ultrasonography in 97.4%. Alternate ultrasound findings comprised 27% of negative cases. Conclusions: Excellent positive predictors of intussusception were identified prospectively. Although no reliable negative predictors were found, patients at low risk may be screened by ultrasonography. (J Pediatr 1998;132:836-9)