Regular articleAccuracy and usefulness of a clinical prediction rule and D-dimer testing in excluding deep vein thrombosis in cancer patients
Introduction
Patients with malignancy have a high risk of developing venous thrombosis [1], [2], [3] and the treatment of thrombosis with anticoagulants is complicated by a substantial risk of bleeding [3], [4], [5], [6]. Therefore, accurate diagnosis of deep venous thrombosis (DVT) is required to appropriately treat patients with the disease and avoid unnecessary anticoagulation in those without venous thrombosis.
Clinical prediction models combine components of the history and physical examination to categorize a patient's probability of having a disease. The Wells pretest probability (PTP) model is validated and well established for diagnosing DVT in symptomatic patients. This model originally stratified patients as having a low, moderate, or high likelihood of having DVT but subsequent analyses showed that the model could be simplified by stratifying patients into either a likely or unlikely risk group [7], [8], [9], [10]. When used in combination with D-dimer testing [10], both of these stratifications can safely manage patients with suspected DVT. The negative predictive value, or the likelihood of not having disease, when a patient has a low or unlikely PTP and a negative D-dimer result, ranges from 99.1% to 99. 6% [8], [9], [10]. It has also been demonstrated that it is safe to withhold anticoagulant therapy in patients who have a negative D-dimer result and either a low or unlikely PTP at initial presentation [8], [9], [10].
But the clinical utility of such a strategy for ruling out DVT in patients with cancer is less certain. The studies that validated PTP models did include cancer patients but the scoring does not take into account the strong risk factors for DVT that are unique to cancer patients. Also, while some studies showed that D-dimer testing had comparable negative predictive values in patients with cancer compared to patients without cancer [11], others have not [12]. Finally, the high prevalence of DVT in cancer patients may sufficiently reduce the negative predictive value of these diagnostic tools and render them less useful [12].
In a retrospective study that evaluated the application of a PTP model and D-dimer testing in cancer patients with suspected DVT, a combination of a negative D-dimer result and a low or low-moderate PTP was useful in excluding DVT [13]. However, given that only 9% of cancer patients scored a low PTP, the authors suggested that the unlikely/likely PTP stratification might be more useful in clinical practice. Also, the confidence interval of the negative predictive value of this combination was as low as 88%, suggesting that up to 12% of patients may be missed diagnosed as not having DVT.
To further investigate the utility of PTP modeling and D-dimer testing in cancer patients, we performed a combined analysis of 3 large prospective cohort studies to compare the performance of the Wells PTP model and D-dimer testing between patients with and without cancer and to examine the utility of the two PTP model classification schemes (low/moderate/high versus unlikely/likely) in excluding DVT in patients with cancer.
Section snippets
Materials and methods
We combined the databases from three prospective diagnostic studies in consecutive outpatients with suspected DVT that were performed at 7 Canadian tertiary care centers between 1997 and 2002 (see Appendix A for names of centers) [8], [9], [10]. Patients were included in these studies if they presented with symptoms of DVT at either the emergency departments or the Thromboembolism clinics. Each study evaluated patients using the Wells PTP model (Table 1) [7], [10] and performed D-dimer testing
Results
A total of 2696 patients were evaluated. The characteristics of the study participants are displayed in Table 2. Deep vein thrombosis was diagnosed in 403 (15%) patients; 13 of these were confirmed during the 3-month follow-up. A total of 200 cancer patients were included. Of these patients, 83 (41.5%) were diagnosed with DVT. A total of 463 patients did not have D-dimer performed. Thirty patients from the 3 original studies did not have sufficient information in the databases and were excluded
Discussion
We investigated the safety and utility of excluding DVT in patients with cancer using different PTP categories (low or unlikely) in combination with D-dimer testing. As in patients without cancer, low or unlikely PTP categories in combination with a negative D-dimer result have very high negative predictive values and should be able to safely rule out DVT in patients with cancer. However, because D-dimer levels are often elevated in cancer patients in the absence of DVT, the D-dimer test has
Acknowledgements
P. Wells is a recipient of a Canada Research Chair. M. Carrier is a recipient of a Canadian Institute for Health Research Fellowship.
References (15)
- et al.
Recurrent venous thromboembolism and bleeding complications during anticoagulant treatment in patients with cancer and venous thrombosis
Blood
(2002) - et al.
Accuracy of clinical assessment of deep-vein thrombosis
Lancet
(1995) - et al.
Combined use of clinical assessment and d-dimer to improve the management of patients presenting to the emergency department with suspected deep vein thrombosis (the EDITED Study)
J Thromb Haemost
(2003) - et al.
Combined use of clinical pretest probability and D-dimer test in cancer patients with clinically suspected deep venous thrombosis
J Thromb Haemost
(2006) - et al.
Influence of age on the cost-effectiveness of diagnostic strategies for suspected pulmonary embolism
J Thromb Haemost
(2007) - et al.
Malignancies, prothrombotic mutations, and the risk of venous thrombosis
JAMA
(2005) - et al.
Risk factors for deep vein thrombosis and pulmonary embolism: a population-based case-control study
Arch Intern Med
(2000)