Background: Numerous tests are available to assess patient risk for bleeding or thrombosis. Appropriate use of these tests must involve consideration of the clinical setting, disease prevalence, performance characteristics of the tests, cost, and consequences of false-positive and false-negative results.
Purpose: To summarize information about coagulation testing in three common clinical settings: nonsurgical hospitalized patients, surgical patients, and patients having a first venous thromboembolic event.
Data sources: All English-language studies identified in searches of MEDLINE (1966 to April 2002) and reference lists of key articles.
Study selection: All published studies of blood coagulation testing as routine diagnostic tests or in the preoperative care of patients reporting postoperative bleeding complications, and all published studies of patients with the factor V Leiden mutation reporting venous thromboembolic outcomes.
Data extraction: 5 observational studies of routine coagulation testing in nonsurgical hospitalized patients and 12 observational studies of preoperative coagulation testing, from which both sensitivity and specificity could be calculated.
Data synthesis: Test performance characteristics for the partial thromboplastin time in predicting postoperative hemorrhage were pooled by type of surgery. Likelihood ratios for positive and negative results were calculated for each group; 95% confidence intervals were calculated. Patients with prolonged partial thromboplastin times did not have a statistically significantly increased risk for postoperative complications.
Conclusion: For nonsurgical and surgical patients without synthetic liver dysfunction or a history of oral anticoagulant use, routine testing has no benefit in assessment of bleeding risk. Routine testing after a first episode of venous thromboembolism is not recommended for most patients.