Retrievable inferior vena cava filters are rarely removed.

There has been an increasing nationwide trend of inferior vena cava (IVC) filter placement over the past 3 years. Most of these have been the newer, removable variety. Although these are marketed as retrievable, few are removed. The purpose of this study was to examine the practice pattern of IVC filter placement at Huntington Hospital. This study is a retrospective chart review of all IVC filter placements and removals between January 1, 2004, and December 31, 2006. The primary data points include indication for placement, major complications (migration, caval thrombosis, pulmonary embolus [PE]), attempted removal, and successful removal. Three hundred ten patients received IVC filters at our institution during this period. Eighty-four were placed in 2004, 95 in 2005, and 131 in 2006. Of those, only 12 (3.9%) were documented permanent filters, whereas the remainder (298) were removable. Of the retrievable filters placed, only 11 (3.7%) underwent successful removal. There were four (1.3%) instances in which the filter could not be removed as a result of thrombus present within the filter and two (0.67%) in which removal was aborted as a result of technical difficulty. Our use of IVC filters has increased steadily over the last 3 years. Despite the rise in use of "removable" filter devices, few are ever retrieved. Although IVC filter insertion appears an effective method of PE prevention, it comes at a cost, both physiological and monetary. It would be wise to devise more stringent criteria to identify those patients in the various populations who truly require filter placement and to be cautious in altering our indications for placement.

Outcomes of blunt head trauma without intracranial pressure monitoring.

Although guidelines exist for intracranial pressure (ICP)-guided treatment after head trauma, no conclusive data exist that support routine ICP monitoring. A retrospective case series was reviewed of all patients admitted to the intensive care unit with a diagnosis of blunt head trauma between January 1, 1999 and December 31, 2004. None of the patients in the final analysis had ICP monitoring. Data collected included age, sex, mechanism of injury, Glasgow Coma Score (GCS) at admission, injury severity score, disposition, and length of stay. One hundred thirty-one patients with a median age of 41 years were included. There were 104 men (79%). The median GCS at admission was 12. There were 22 deaths (17% mortality). Stepwise logistic regression analysis identified older age, higher injury severity score, and lower GCS to be predictors of death. The mortality rate was higher in patients with GCS < or =8 compared with GCS >8 (33% vs 8%, respectively; P < 0.001). Ten of 23 patients with a GCS of 3 died (43% mortality). The median time to death for patients with a GCS of 3 was 2 days. Although the Brain Trauma Foundation has published guidelines advocating routine ICP monitoring, no large randomized prospective studies are available to determine its effect on outcome. None of the patients in this study had ICP monitoring. Our overall survival rate of 83 per cent is relatively high. Patients with a low GCS and, specifically, those with a GCS of 3 may not benefit from ICP monitoring because of early and irreversible trauma. Variability in the use of ICP monitoring will remain until ICP monitoring can be conclusively proven to improve outcome.

Diagnosis and management of laryngotracheal trauma.

Laryngotracheal trauma is life-threatening. We identified 23 patients between 1992 and 1998 with laryngeal (12), tracheal (8), and combined injuries (3). Nineteen patients had penetrating trauma (gunshot wound, 12; stab wound, 7), and four patients had blunt injury. Flexible laryngoscopy identified the injury in 8 of 12 patients (75%), and computer tomography scan was positive in 9 of 9 patients (100%). Twelve of the 19 patients with penetrating wounds were managed by primary repair, 4 had endotracheal intubation without surgical repair, and 3 were observed. No patient with a blunt tracheal injury required repair. Two had endotracheal intubation, and two were observed. A high index of suspicion is essential to identifying laryngotracheal injury. Computer tomography scan is a sensitive diagnostic test for laryngotracheal injury, and may be indicated despite normal flexible laryngoscopy. The decision to repair injuries or observe injuries is primarily based on respiratory distress and associated injuries.