Suspected ectopic pregnancy. Can it be predicted by history and examination?

value of histor y and physical examination in patients with suspected ectopic pregnancy. Are there signs or symptoms on history and physical examination that could predict ectopic pregnancy (EP) in patients with abdominal pain or vaginal bleeding? Prospective observational study of consecutive female patients 18 years and older with abdominal pain or vaginal bleeding and positive results of a serum β human chorion-ic gonadotropin (βhCG) test, who presented to the emergency department of an urban academic medical centre. In most places, family physicians are responsible for the total care of pregnant patients and are faced with the possibility of EP in their offices and in emergency departments. Ectopic pregnancy is common, affecting up to 1.7% of gestations. 1,2 Up to 15% of patients presenting to emergency depar tments with first-trimester abdominal pain or bleeding have EPs. 3 As many as 45% of EPs are misdiagnosed on initial presentation. 4 Ectopic pregnancy remains the leading cause of death during the first trimester and accounts for 10% of all maternal deaths. 5 Increasing importance has been placed on ultrasound and laboratory testing to aid in diagnosis, but these services are not always readily available. Identification of reliable risk factors from the history and physical examination could be used to stratify risk and guide patient management and disposition. On weekdays during daytime hours, all patients with symptoms had a history taken, were given a physical examination, and supplied a serum βhCG sample. All patients with positive pregnancy test results had an ultrasound (US) examination, and their quantitative serum βhCG was determined. On weekends and during nights, only patients with quantitative serum βhCG levels of > 1000 µIU/mL were examined by US. Patients with indeterminate US results or βhCG values < 1000 µIU/mL (without US) were admitted for further evaluation. Patients were excluded from the study if they had had previous US examination or were lost to follow up. Final diagnosis was categorized as normal intrauter-ine pregnancy (IUP), abnormal IUP, or EP. All EPs were diagnosed by laparoscopy. A data collection form recording 42 separate variables was completed by the most senior clinician involved in each patient's care. Each variable was dichotomized and compared with the final diagnosis of each patient. Data were analyzed using χ 2 or Fisher's exact tests. Univariate analysis evaluated possible predictive factors for EP. A Classification and Regression Tree (CART) analysis, described in detail by Dart, 11 was also …