A clinical prediction rule for identifying short-term risk of adverse events in patients with pulmonary thromboembolism

OBJECTIVE: To identify patients with a low short-term risk of complications following acute pulmonary thromboembolism. PATIENTS AND METHODS: A prospective multicenter study was conducted in 8 Spanish hospitals; 681 consecutive outpatients diagnosed with pulmonary thromboembolism were enrolled. Clinically significant variables were weighted using coefficients derived from a logistic regression model in order to optimize the diagnostic performance of a clinical prediction rule to predict the following complications within 10 days of acute pulmonary thromboembolism: death, recurrent thromboembolism, and major or minor bleeding. RESULTS: Forty-three patients (6.3%) had 51 complications. These included 33 deaths, 12 major bleeding episodes, and 6 minor bleeding episodes. The clinical variables used in the prediction rule were assigned the following scores: recent major bleeding episode and cancer with metastasis, 4 points each; creatinine levels of over 2 mg/dL, 3 points; cancer without metastasis and immobility due to a recent medical condition, 2 points each; and absence of surgery in the past 2 months and an age of over 60 years, I point each. A risk score of 2 or less, obtained by 47.8% of patients, indicated a low short-term risk of developing complications following pulmonary thromboembolism. The area under the receiver operating characteristic curve for the prediction rule was 0.75 (95% confidence interval [CI], 0.67-0.83). For this cutoff point, sensitivity was 82.9% (95% CI, 68.7-91.5) and the likelihood ratios for a positive and negative test result were 1.63 (95% CI, 1.39-1.92), and 0.35 (95% CI, 0.18-0.69), respectively. CONCLUSIONS: Our clinical prediction rule could be useful for identifying patients with a low risk of complications in the 10 days following acute pulmonary thromboembolism. Those patients would be eligible for consideration for outpatient treatment.