400

Abstract

Introduction: Pulmonary Embolism (PE) is a common and potentially life-threatening condition. Overall mortality ranges between 5–15% and varies substantially depending upon comorbidities and severity of presentation. Epidemiologic studies have shown an increase of inferior vena cava (IVC) filter placement in patients with acute PE and observational data suggest that IVC filter use, irrespective of anticoagulation, may improve outcomes. Therefore, we assessed the incidence and impact on outcomes of IVC filter use in patients admitted with acute PE in a large integrated health care system. Methods: We performed a retrospective cohort study of all adult patients admitted to either of two tertiary-care teaching hospitals in the Geisinger Health System from Jan 2006 through Dec 2012. Eligible patients had both an admission and discharge diagnosis of acute PE; chart review was performed in a subset of patients to confirm accuracy of administrative codes. Relevant clinical information was extracted and the Pulmonary Embolism Severity Index (PESI) score was assigned to each patient. The primary outcome was in-hospital and 30 day mortality. Results: Of a total of 1052 patients, 13.4% (n=141) received an IVC filter. The use of IVC filters did not show a clear trend over time, ranging from 6.8% to 20.4% per year during the study period. Patients who received IVC filters, compared to those who did not, were older (median 65 years, interquartile range (IQR) 54.7–78.0 vs 62 years, IQR 49.0–73.7, p=0.009), had significantly higher PESI scores (median 88, IQR 67–110 vs 79, IQR 61–98, p=0.0002), and had longer length of stay (median 5 days, IQR 3.1–7.7 vs 3 days, IQR 1.9–4.9, p<0.0001). There was a trend toward higher in-hospital mortality (6.4% vs 3.4%, OR=1.94, 95%CI 0.90–4.16, p=0.09), but no difference in 30 day mortality (7.8% vs 7.0%, OR=1.12, 95%CI 0.58–2.18, p=0.74) in those who received IVC filters compared to those who did not, respectively. When controlling for severity of presentation using PESI score, the mortality difference between the two groups decreased compared to uncontrolled analysis; however, neither outcome was statistically significant (in-hospital mortality OR=1.59, 95%CI 0.69–3.65, p=0.27 and 30 day mortality OR=0.71, 95%CI=0.33–1.52, p=0.38). Conclusions: Patients with acute PE who received IVC filters in this study were older, had longer length of stay, and had greater severity of illness compared to those patients who did not receive IVC filters. We found that IVC filter use did not impact either in-hospital or 30 day mortality, even when controlling for severity of presentation. While epidemiologic studies report that IVC filter use in those with acute PE has increased over the past few decades, we found that in our health care system the use of IVC filters was variable over time and no clear temporal trend was observed. Although multiple observational studies suggest that IVC filters reduce mortality in acute PE patients, we did not observe a benefit in our patient population. Practice guidelines currently recommend that IVC filter use be limited to those with contraindications to, or failure of, anticoagulation. The lack of prospective study evidence that IVC filters improve outcomes in acute PE patients at least partially explains the inconsistent use of IVC filters in our study. Based on observational data that suggest improved outcomes and uncertainty at the patient care level, prospective randomized trials in acute PE patients are needed to define which patients, if any, benefit from IVC filter use to reduce the risk of adverse outcomes.