Diagnosing functional abdominal pain with the Rome II criteria: parent, child, and clinician agreement.
Product Description:
EVALUATION OF ROME II CRITERIA FOR PEDIATRIC FUNCTIONAL GASTROINTESTINAL DISORDERS: PARENT, CHILD, AND CLINICIAN AGREEMENT
Caroline Elder Danda, University of Kansas Medical Center, Craig A Friesen, Jennifer Verrill Schurman, Linda Andre, Elly Welchert, Children's Mercy Hospital, Teri Comninellis, Jose T Cocjin, Paul E Hyman, University of Kansas Medical Center
The Rome II pediatric working team proposed diagnostic criteria for functional gastrointestinal disorders in children, but validity is not established. During their first visit to a pediatric gastroenterology clinic, we interviewed 59 children aged 101-211 mo (Mage=139?30 mo) with abdominal pain duration of > 3 mo and a parent, using a questionnaire based on Rome II criteria (Walker, Caplan, Rasquin, 2000). On the parent questionnaire, 90% of children met diagnostic criteria for a functional gastrointestinal disorder. On the child questionnaire, 76% of children met criteria. Child report was more likely to not to meet criteria than parent report (24% vs. 10%, p < .03). Parent-child agreement for diagnosis was 79%, with a kappa statistic of .5 (moderate agreement). Parent and child reports were more likely to agree when child questionnaire data was consistent with dyspepsia as opposed to irritable bowel syndrome (IBS; 86% vs. 47%, p < .01). To evaluate validity of Rome II criteria, clinicians made diagnoses based on medical history (60% had IBS, 88% had dyspepsia, 31% had both IBS and dyspepsia). Parent and child questionnaire sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), percent agreement, and kappa statistics are presented in Table 1. Conclusions: Parent-child agreement on a questionnaire based on Rome II criteria was moderate, with the best agreement for children reporting dyspepsia symptoms. Compared to clinician diagnoses, the parent and child questionnaires demonstrated moderate sensitivity and strong specificity. Further research is needed to determine which history (parent, child, or clinician) provides a diagnosis most predictive of therapeutic response or which factors influence symptom reporting and agreement.
Table 1. Comparing clinician diagnoses to diagnoses based on parent and child questionnaire (Rome II criteria).
Sensitivity Specificity PPV NPV Agreement Kappa
IBS-Parent Report 52% 92% 80% 75% 76% .5
IBS-Child Report 48% 81% 61% 71% 68% .3
Dyspepsia-Parent Report 65% 86% 97% 25% 68% .3
Dyspepsia-Child Report 52% 100% 100% 22% 58% .2
Abstract before publication
