Importance of Early Insertion of Inferior Vena Cava Filters in Unstable Patients with Acute Pulmonary Embolism.

Background: Immortal time bias is a possible confounding factor in cohort studies. In this investigation, we assessed mortality with inferior vena cava (IVC) filters in unstable patients with pulmonary embolism using a design to control for immortal time bias.

Methods: Data were from the Premier Healthcare Database, 2010-2014. International Classification of Diseases-Ninth Revision-Clinical Modification codes were used. Unstable patients with pulmonary embolism and an admitting diagnosis of pulmonary embolism, as well as a primary diagnosis of pulmonary embolism, were included. A time-dependent analysis was used according to the day of insertion of the IVC filter to control for immortal time bias.

Results: Among all unstable patients, irrespective of the use of thrombolytic therapy, in-hospital all-cause mortality was 35 of 180 (19.4%) in those who received an IVC filter vs 122 of 299 (40.8%) with no filter (P < .0001). Mortality was lower in patients in whom the IVC filter was inserted on days 1 or 2 (on day 1, 21.4% compared with 40.8%, P = .017, and on day 2, 14.8% compared with 29.2%, P = .023), but it was not lower in those in whom the filter was inserted on subsequent days.

Conclusions: Mortality in unstable patients with pulmonary embolism appeared to be reduced with IVC filters only when the filter was inserted on the first or second day of admission. The design used for these analyses controlled for immortal time bias as a cause of the lower mortality with IVC filters.