METHODS: From February 1991 to April 1998, the
trauma registry identified 7333 admissions. One hundred eighty-seven prophylactic
inferior vena cava filters were inserted. After the exclusion of 27
trauma-related deaths (none caused by
thromboembolism), 160 patients were eligible for the study. The eligible patients were contacted and asked to complete a survey and return for a follow-up examination to include physical examination, Doppler scan study, vena cava duplex scanning, and fluoroscopic examination. The patients' hospital charts were reviewed in detail. The indications for prophylactic
inferior vena cava filter insertion included prolonged immobilization with
multiple injuries,
closed head injury, pelvic fracture, spine
fracture, multiple long
bone fracture, and attending discretion.
RESULTS: Of the 160 eligible patients, 127 were men, the mean age was 40.3 years, and the mean injury severity score was 26.1. The mean day of insertion was hospital day 6. Seventy-five patients (47%) returned for evaluation, with a mean follow-up period of 19.4 months after implantation (range, 7 to 60 months). On survey, patients had leg swelling (n = 27), lower extremity
numbness (n = 14),
shortness of breath (n = 9),
chest pain (n = 7), and skin changes (n = 4). All the survey symptoms appeared to be attributable to patient
injuries and not related to prophylactic
inferior vena cava filter. Physical examination results revealed
edema (n = 12) and skin changes (n = 2). Ten Doppler scan studies had results that were suggestive of
venous insufficiency, nine of which had histories of
deep vein thrombosis. With duplex scanning, 93% (70 of 75) of the vena cavas were visualized, and all were patent. Only 52% (39 of 75) of the prophylactic
inferior vena cava filters were visualized with duplex scanning. All the prophylactic
inferior vena cava filters were visualized with fluoroscopy, with no evidence of filter migration. Of the total 187 patients, 24 (12.8%) had
deep vein thrombosis develop after prophylactic
inferior vena cava filter insertion, including 10 of 75 (13.3%) in the follow-up group, and one patient had a nonfatal
pulmonary embolism despite filter placement. Filter insertion complications occurred in 1.6% (three of 187) of patients and included one groin
hematoma, one
arteriovenous fistula, and one misplacement in the common iliac vein.
CONCLUSION: