Episodes of care for abdominal pain in a primary care practice

Objectives: To explore the usefulness of episodes of care in describing the clinical epidemiology of abdominal pain in the primary care setting and to develop methods to analyze clinician decision-making strategies during abdominal pain episodes. Design: Complete episodes of care for nonpregnant adults with nonspecific abdominal pain from an established episode-based clinical information system were supplemented and valida ted by medical record review. Utilization decisions during episodes were quantified by summing the costs of all visits, services, tests, and referrals ordered or performed by the clinician. A decision model was used to analyze significant influences on utilization decisions. Setting: An established faculty practice site of a Mid-western academic family practice department. Subjects: Two hundred ten nonpregnant adults who had nonspecific abdominal pain. Main Outcome Measures: Utilization and costs generated during the episode of care. Results: The average abdominal pain episode required 1.32 visits and cost $123.36. In more than half of all episodes (51%), a specific diagnosis was not reached. The most common specific diagnoses were gastritis and gastroesophageal reflux disease (5% each). Bivariate analyses showed that two variables, clinician uncertainty about diagnosis and a nonspecific diagnosis, were significantly associated with episode cost. Patient age, gender, comorbidity, and the presence or absence of specific clinical findings were not associated with episode cost. Stepwise regression modeling resulted in a two-factor model. Clinician uncertainty and complexity explained only 9% of the variance in episode cost. Conclusions: Episodes of abdominal pain most often remained undiagnosed. The decision model did not predict episode cost. Utilization decisions did nor seem to be driven by commonly cited clinical risk factors, but by diagnostic uncertainty or individualized decision rules.