Background:

The incidence of new or recurrent acute pulmonary embolism in patients on warfarin with an INR of = 2 is unknown. Also, the utility of using a modified Wells score in diagnosing acute pulmonary embolism in these patients has not been assessed.

Methods:

We performed a retrospective observational study on patients aged 18–89 presenting to the emergency department with an INR of = 2 on warfarin who then underwent a contrast enhanced chest computed tomography for a clinical concern of acute pulmonary embolism (CT–PE). We calculated a modified Wells score on patients that met our inclusion criteria and then constructed a 2×2 table to determine the diagnostic characteristics of using a modified Wells score as a tool to exclude pulmonary embolism. Fisher’s exact test was used to determine the statistical significance of the 2×2 table analysis.

Results:

135 patients met inclusion criteria over a period of six years, with six (4%) patients diagnosed with an acute pulmonary embolism on CT–PE (Table 1). The mean INR ([pm] SD*) and the mean modified Wells score ([pm] SD) were 3.4 ([pm]2) and 2.8 ([pm]2.2) respectively. The diagnostic characteristics of a using a modified Wells score of = 4 to exclude pulmonary embolism are listed in Table 2. The odds in favor of a positive CT–PE for a patient with a modified Wells score of = 4 were 12.9 times the odds in favor of a positive CT–PE for a patient with a modified Wells score < 4 (p < 0.05). All the patients who had a positive CT–PE had an underlying diagnosis of either a malignancy or a thrombophilia.

Conclusions:

The incidence of new or recurrent acute pulmonary embolism in patients on warfarin with an INR of = 2 is low and a modified Wells score of < 4 safely excludes pulmonary embolism in these patients.

Table 12×2 Analysis of CT–PE and Modified Wells Score

Table 2Diagnostic Characteristics of a Modified Wells Score >= 4