In view of the high risk of pulmonary embolism in patients with cancer, we tested the hypothesis that stable patients with pulmonary embolism who have cancer might be a subset of patients who would show a lower case fatality rate with vena cava filters than without filters.

Stable patients with pulmonary embolism and cancer at discharge from short-stay hospitals throughout the US from 1998-2009 were identified from the Nationwide Inpatient Sample. Patients with pulmonary embolism who had a diagnostic code for shock, ventilatory support, thrombolytic therapy, or pulmonary embolectomy were excluded because such patients have been shown to have lower case fatality rate with filters.

In-hospital all-cause case fatality rate was lower with vena cava filters in stable patients with pulmonary embolism and solid malignant tumors providing they were aged >30 years, but there was variability according to type of tumor and age of patient. On average, case fatality rate among those >30 years with filters was 7070 of 69,350 (10.2%) (95% confidence interval, 10.0-10.4) versus 36,875 of 247,125 (14.9%) (95% confidence interval, 14.8-15.1) without filters (P <.0001) (relative risk 0.68). Among stable patients with hematological malignancies, case fatality rate, except in the elderly, was higher among those with vena cava filters than those without filters.

Stable patients with pulmonary embolism and solid malignant tumors who are older than age 30 years appear to be a subset of patients with pulmonary embolism who would benefit from vena cava filters, but this needs to be tested prospectively.