How well does the activated partial thromboplastin time predict postoperative hemorrhage?

To determine the clinical usefulness of the activated partial thromboplastin time in screening patients before surgery, we related preoperative test results to the occurrence of hemorrhagic complications, studying all adult inpatients undergoing invasive diagnostic or therapeutic procedures at a large university hospital during a one-year period. Using clinical characteristics to divide the population into two groups according to a priori risk of hemorrhagic complications, we found that the activated partial thromboplastin time had no ability to predict the occurrence or absence of hemorrhage in the low-risk group, but that it was a predictor of modest strength in the high-risk group. Our data justify limiting preoperative coagulation screening to patients with active bleeding, known or clinically suspected bleeding disorders (including use of anticoagulants), liver disease, malabsorption, malnutrition, or other conditions associated with acquired coagulopathies and patients whose procedures may interfere with normal coagulation.