Search

Background/objectives: Older adults are at risk for adverse outcomes after trauma, but little is known about post-acute survival as state and national trauma registries collect only inpatient or 30-day outcomes. This study investigates long-term, out-of-hospital mortality in geriatric trauma patients., Methods: Level I Trauma Center registry data were matched to the US Social Security Death Index (SSDI) to determine long-term and out-of-hospital outcomes of older patients. Blunt trauma patients aged ≥65 were identified from 2009 to 2015 in an American College of Surgeons Level 1 Trauma Center registry, n = 6289 patients with an age range 65-105 years, mean age 78.5 ± 8.4 years. Dates of death were queried using social security numbers and unique patient identifiers. Demographics, injury, treatments, and outcomes were compared using descriptive and univariate statistics., Results: Of 6289 geriatric trauma patients, 505 (8.0%) died as an inpatient following trauma. Fall was the most common mechanism of injury (n = 4757, 76%) with mortality rate of 46.5% at long-term follow-up; motor vehicle crash (MVC) (n = 1212, 19%) had long-term mortality of 27.6%. Overall, 24.1% of patients died within 1 year of trauma. Only 8 of 488 patients who died between 1 and 6 months post-trauma were inpatient. Mortality rate varied by discharge location: 25.1% home, 36.4% acute rehabilitation, and 51.5% skilled nursing facility, P < .0001., Conclusion: Inpatient and 30-day mortality rates in national outcome registries fail to fully capture the burden of trauma on older patients. Though 92% of geriatric trauma patients survived to discharge, almost one-quarter had died by 1 year following their injuries.

Background: The clinical significance of obtaining cardiac troponin (cTn) levels among trauma patients with new onset arrhythmias is unknown. We aimed to assess whether cTn levels actually influence clinical decision making or represent an inappropriate use of resources., Methods: Trauma patients admitted from 2013 to 2014 diagnosed with atrial fibrillation (AF) were retrospectively reviewed using the institutional trauma database. Demographics, cTn levels, and myocardial infarction (MI) diagnosis data were recorded. Standard univariate tests were used to compare data between patients with and without cTn., Results: There were 258 patients included of which 126 patients had cTn levels obtained (48.8%, TEST group). The remaining 132 patients (51.2%) were untested (noTEST group). Among TEST patients, use of echocardiography nearly doubled and cardiology consultations increased (all p < 0.05). No TEST patients suffered MI or PE., Conclusions: Obtaining cTn values in trauma patients with new-onset AF resulted in increased resource utilization without clinical utility., (Copyright © 2019 Elsevier Inc. All rights reserved.)

Background: Interfacility transfers are necessary and valuable for the trauma system, but despite regional guidelines, many patients are inappropriately transferred. We evaluated over-triage at our Level I center and identified risk factors for over-triage., Methods: Retrospective analysis at our Level I urban trauma center assessed patients transferred from regional facilities during 2017. Over-triage was defined as patients discharged <48 h without procedures. Exclusion criteria were leaving against medical advice or no outside records., Results: Overall, 2352 patients met criteria. Nine hundred thirty (39.5%) with complete hospital records were discharged in <48 h; 498 (53.5%) received no procedural intervention and 909 (97.7%) were ultimately discharged home., Conclusion: Many patients are inappropriately transferred to tertiary care centers without a definitive need for advanced services. Studies are needed to improve triage criteria without increasing under-triage., (Copyright © 2019 Elsevier Inc. All rights reserved.)

The role of prophylactic vena cava filters (pVCFs) in trauma patients remains controversial. After 20 years of data collection and experience, we reviewed our venous thromboembolism guideline for the efficacy of pVCFs in preventing pulmonary embolism (PE). A retrospective cohort study was performed using our Level I trauma center registry from January 1997 thru December 2016. This population was then divided by the presence of pVCFs. Univariate analysis was performed comparing the incidence of PEs, deep vein thrombosis, and mortality between those with and without a pVCF. There were 35,658 patients identified, of whom 2 per cent (n = 847) received pVCFs. The PE rate was 0.4 per cent in both groups. The deep vein thrombosis rate for pVCFs was 3.9 per cent compared with 0.6 per cent in the no-VCF group ( P < 0.0001). Given that there was no difference in the rates of PEs between the cohorts, the subset of patients with a PE were analyzed by their risk factors. Only ventilator days > 3 were associated with a higher risk in the no-pVCF group (0.2 vs 1.5%, P = 0.033). pVCFs did not confer benefit reducing PE rate. In addition, despite their intended purpose, pVCFs cannot eliminate PEs in high-risk trauma patients, suggesting a lack of utility for prophylaxis in this population.

Introduction: We hypothesized that a pH of <7.0 on presentation would correlate with almost universal mortality in trauma patients., Methods: A retrospective cohort study was performed on a Level I trauma center registry from 2013 to 2014. Hospital mortality was the primary outcome, which was compared by pH cohort (<7.0 or ≥7.0) using standard univariate statistics and multivariate logistic regression., Results: There were 593 patients included in the analysis: 66 in <7.0, 527 in ≥7.0. Mortality was 3× higher in the <7.0 pH cohort (62.1 vs. 20.3%; p < 0.0001), however there was no threshold for a pH below which there was 100% mortality. After controlling for these confounding variables, initial pH was found to be an independent predictor of inpatient mortality: pH < 7.0 (OR 6.33, 3.29-12.19; p < 0.0001)., Conclusion: This data indicates that while patients with severe acidosis are at increased risk for mortality, a pH < 7.0 is still recoverable in select cases., (Copyright © 2017 Elsevier Inc. All rights reserved.)

Background: Patient demographics, behavior, and injury patterns were assessed to inform preventative efforts for reduced incidence of all-terrain vehicle (ATV) trauma., Methods: ATV-related injuries treated at a Level I trauma center from 2008 to 2012 were retrospectively reviewed. Patient outcomes and incidence of traumatic brain injury (TBI) were compared by helmet use and alcohol intoxication., Results: Helmet data were available for 304 patients of 404 patients included; of these, 75 (24.7%) wore a helmet. Incidence of TBI was lower in the helmeted (8.0%) versus the unhelmeted subgroup (26.6%) (P < 0.001). Helmeted patients had lower injury severity scores, lower intensive-care unit (ICU) admission rates, and shorter ICU and hospital length of stay (LOS) (P < 0.05). Intoxicated patients had higher rates of TBI and ICU admission as well as prolonged ICU LOS (P < 0.05)., Conclusions: These data support the requirement for a greater emphasis on injury prevention among ATV users., (Copyright © 2016 Elsevier Inc. All rights reserved.)

Transfer of severely injured patients to regional trauma centers is often expedited; however, transfer of less-injured, older patients may not evoke the same urgency. We examined referring hospitals' length of stay (LOS) and compared the subsequent outcomes in less-injured transfer patients (TP) with patients presenting directly (DP) to the trauma center. We reviewed the medical records of less-injured (Injury Severity Score [ISS] 9 or less), older (age older than 60 years) patients transferred to a regional Level 1 trauma center to determine the referring facility LOS, demographics, and injury information. Outcomes of the TP were then compared with similarly injured DP using local trauma registry data. In 2011, there were 1657 transfers; the referring facility LOS averaged greater than 3 hours. In the less-injured patients (ISS 9 or less), the average referring facility LOS was 3 hours 20 minutes compared with 2 hours 24 minutes in more severely injured patients (ISS 25 or greater, P < 0.05). The mortality was significantly lower in the DP patients (5.8% TP vs 2.6% DP, P = 0.035). Delays in transfer of less-injured, older trauma patients can result in poor outcomes including increased mortality. Geographic challenges do not allow for every patient to be transported directly to a trauma center. As a result, we propose further outreach efforts to identify potential causes for delay and to promote compliance with regional referral guidelines.

Return transfer (RT) to the intensive care unit (ICU) negatively impacts patient outcomes, length of stay (LOS), and hospital costs. This study assesses the most common events necessitating RT in trauma patients. We performed a retrospective chart review of ICU RT from 2004 to 2008. Patient demographics, injuries and injury severity, reason for transfer, LOS, interventions, and outcomes data were collected. Overall, 158 patients required readmission to the ICU. Respiratory insufficiency/failure (48%) was the most common reason for RT followed by cardiac (16%) and neurological (13%) events. The most commonly associated injuries were traumatic brain injuries (TBIs) (32%), rib fractures (30%), and pulmonary contusions (20%). Initial ICU LOS was 6.6 ± 8 days (range, 1 to 44 days) with 4.4 ± 7.8 ventilator days. Mean floor time before ICU RT was 5.7 ± 6.3 days (range, 0 to 33 days). Forty-nine patients (31%) required intubation and mechanical ventilation on RT. ICU RT incurred an additional ICU LOS of 8 ± 8.5 days (range, 1 to 40 days) and 5.2 ± 7.5 ventilator days. Mortality after a single RT was 10 per cent (n = 16). RT to the ICU most often occurs as a result of respiratory compromise, and patients with TBI are particularly vulnerable. Trauma pulmonary hygiene practices should be evaluated to determine strategies that could decrease RT.

Current guidelines for the empiric treatment of ventilator-associated pneumonia (VAP) recommend that vancomycin is dosed 15 mg/kg and administered twice daily for a target trough level of 15 to 20 μg/mL. This study compared conventional intermittent vancomycin infusion (IVI) with continuous vancomycin infusion (CVI). Our prospective, randomized study compared CVI with IVI in trauma patients with suspected VAP. The primary outcome measure was a serum vancomycin level within the target level 48 hours after initiation of therapy. Treatment groups were compared using standard statistical methods. The study included 73 patients, 36 IVI and 37 CVI. Eighteen patients were withdrawn from the study as a result of discontinuation of the drug before 48 hours or failure to draw levels at the appropriate time, resulting in 27 IVI and 28 CVI study patients. There were no differences between treatment groups in gender (P = 0.97), Injury Severity Score (P = 0.70), total body weight (P = 0.36), or age (P = 0.81). The mean serum vancomycin level for the IVI group was 8.9 ± 3.9 μg/mL, and the CVI level was 19.8 ± 6.13 μg/mL (P < 0.0001). Two patients in the IVI group (7.4%) were in the therapeutic range compared with 16 (57.1%) in the CVI group (P < 0.0001). Six patients in the CVI group (21.4%) and none of the IVI patients had supratherapeutic levels. Four patients developed renal insufficiency, three IVI (11.1%) and one CVI (3.6%) (P = 0.36). The current American Trauma Society dosing recommendations for vancomycin for presumptive VAP treatment are inadequate. Continuous vancomycin infusion should be adopted as the standard dosing strategy.

Background: Aeromedical transport (AMT) is an effective but costly means of rescuing critically injured patients. Although studies have shown that it improves survival to hospital discharge compared with ground transportation, an efficient threshold or universal criteria for this mode of transport remains to be established. Herein, we examined the effect of implementing a Trauma Advisory Committee (TAC) initiative focused on reducing AMT overtriage (OT) rates., Methods: TAC outreach coordinators implemented a process improvement (PI) initiative and collected data prospectively from January 2007 to December 2011. OT was defined as patients who were airlifted from scene and later discharged from the emergency department. Serving as liaisons to surrounding counties, TAC outreach coordinators conducted quarterly PI meetings with local emergency medical service agencies. Patients were grouped into those who were airlifted from TAC counties versus counties outside TAC's jurisdiction (non-TAC). Standard statistical methods were used., Results: From 2007 to 2011, 3,349 patients were airlifted from 30 counties, 1,427 (43%) from TAC counties and 1,922 (57%) from non-TAC counties. The OT rates from TAC counties declined compared with non-TAC counties each year and reached statistical significance in 2008 (17% vs. 23%, p < 0.05), 2009 (11% vs. 17%m p < 0.05), and 2011 (6% vs. 12%, p < 0.05). The reduction in OT continued over the study duration, with improvement in TAC counties compared with previous years., Conclusion: Implementation of a regional TAC PI initiative focused on OT issues led to a more efficient use of AMT., Level of Evidence: Prognostic study, level III; therapeutic study, level IV.

Background: Mopeds have less stringent licensing laws than automobiles. Moped operators in motorized vehicle collisions (MVCs) exhibit significantly higher rates of driving while intoxicated (DWI) and higher blood alcohol levels than automobile or motorcycle operators. This study evaluates the public safety issue of DWI recidivism among moped operators., Methods: Moped operators evaluated after MVCs were identified from 2007 to 2009. Demographics, hospital data, and Department of Motor Vehicles records were reviewed., Results: Sixty-five moped operators were evaluated. Thirty-two (49%) had a positive blood alcohol level, 29 (45%) had a previous DWI, and 21 (72%) of those were repeat offenders. Twenty-five (38%) had a revoked license at the time of injury. Of these, 19 (76%) incurred multiple revocations. Twenty-two (34%) showed prior charges of driving with a revoked license (DWRL), with 15 (68%) incurring multiple DWRL charges., Conclusions: Moped operators are often intoxicated at the time of injury and represent a public safety hazard. The majority are recidivists with multiple alcohol-related traffic charges. Current laws allow repeat offenders the sustained opportunity to operate motorized vehicles. Re-evaluation of current moped laws is needed to keep habitual offenders off the road., (Copyright © 2011 Elsevier Inc. All rights reserved.)

Mopeds are not subject to the same laws and jurisdiction as cars or motorcycles, including the requirement of a driver's license. We undertook this study to examine the influence of alcohol (ETOH) on moped crashes. We retrospectively reviewed adult moped injuries compared with motor vehicle crashes (MVCs) and motorcycle crashes (MCCs) from 1995 through 2006. Demographics, severity of injury, mortality, and serum ETOH levels were recorded. Data were analyzed using the Student t test for continuous data and the χ² test for proportional data. Motor vehicle crashes accounted for 7186 admissions. MCC and moped crashes numbered 973 and 113, respectively. Although not statistically significant (P = 0.064), moped crashes yielded the highest mortality (9.7%) compared with MCCs (8.5%) and MVCs (6.7%). An increased association of blood ETOH levels with moped crashes, however, was statistically significant (P = 0.004). Serum ETOH levels above 0.05 g/dL were observed in 1681 MVCs (23.4%), 241 MCCs (24.8%), and 44 moped crashes (39%). In this study, we discovered that moped crashes demonstrate a significantly higher ETOH involvement than either MVCs or MCCs representing a previously unrecognized public safety risk.

Introduction: We previously demonstrated that utilization of erythropoietin (r-EPO) did not significantly reduce blood utilization in trauma patients. We undertook this study to analyze blood utilization 1 year after r-EPO removal from our trauma service anaemia practice management guideline., Methods: Electronic records of patients admitted to the trauma service were retrospectively reviewed for units of packed red blood cells (pRBCs) transfused and for units of r-EPO administered 12 months before the initiation of an anaemia practice guideline (PRE), 12 months during the use of an anaemia guideline (GUIDE), and 12 months following removal of r-EPO from the guideline (POST). Hospital acquisition cost was also reviewed for the respective time periods. Nominal data were analyzed using chi-squared or Fisher's exact tests, and interval data were compared using ANOVA followed by Tukey's test where appropriate. Results were considered significant for P<0.05., Results: Over the 3-year study period, 4881 patients were admitted to the trauma service and included in this study. The hospital length of stay, intensive care unit length of stay, and units of pRBC transfused were similar among all three groups. Group I (PRE) received a total of 228 doses of r-EPO at a cost of $102,600. Group II (GUIDE) received a total of 410 doses at a cost of $184,500. Group III (POST) received 28 doses of r-EPO at a cost of $12,600., Conclusion: Removal of erythropoietin from our trauma service anaemia practice management guideline did not result in increased blood utilization. However, it yielded a hospital acquisition cost savings of $171,900.

Hospital length of stay (LOS) is frequently used to evaluate the quality of trauma care but LOS may be impacted by nonmedical factors as well. We reviewed our experience with delays in patient discharge to determine its financial consequences and its impact on LOS. We performed an analysis of linked trauma registry and "delayed discharge" databases. Actual LOS (A-LOS) values were compared with calculated ideal LOS (I-LOS) values, and the per cent increase in LOS was calculated. Linear regression analysis was used to identify significant predictors of prolonged LOS. One thousand, five hundred and seventeen patients were studied, with an A-LOS of 6.54 days. Seven per cent of patients experienced discharge delays, resulting in 580 excess hospital days. Calculated I-LOS was 6.15 days, 6.34 per cent lower than A-LOS. Other I-LOS estimates were as much as 25 per cent lower than A-LOS. Estimated excess patient charges associated with delayed discharges were $4,000,000 to $15,000,000. Discharge delays are an infrequent, although costly, occurrence that has a significant impact on LOS. LOS therefore may not be an appropriate metric for assessing the quality of trauma care, and should only be used if it has been corrected for discharge delays. Concerted efforts should be directed towards identifying and correcting the factors responsible for delayed discharge in trauma patients.

The electrothermal bipolar vessel sealer (EBVS) was developed as an alternative to sutures, clips, and the ultrasonic scalpel for haemostasis during open and laparoscopic surgery. The purpose of this prospective clinical study was to objectively evaluate the performance of the EBVS during trauma surgery. Data from trauma cases, in which the EBVS was used, was collected prospectively for 19 consecutive months. Data collected included: total number of EBVS applications, need for additional haemostatic devices after application, calculated time savings or loss, and hemorrhagic complications. The EBVS was used in 23 trauma cases consisting of 13 small bowel resections, 4 ileocolectomies, 2 left hemicolectomies, 1 transverse colectomy, 1 right hemicolectomy with roux-en-Y gastrojejunostomy and duodenostomy, 1 Hartmann's procedure, and 1 splenorrhaphy with omental injury repair. A total of 631 applications of the device were used, averaging 27.4 applications per case. An additional suture ligature or clip placed for an inadequate EBVS seal (failure) was needed in only 1.5% of the total EBVS applications. All of these (10) occurred in one patient undergoing Hartmann's procedure for massive colonic injury. An additional 45 sutures or clips for non-EBVS failure were used in 5 cases due to proximity of bleeding to critical vascular, biliary, or bowel structures. In 17 trauma cases (74%) with intestinal resection no other means of hemostasis (sutures or clips) was required except the EBVS. The mean calculated time savings using the EBVS was 26.8 min (range 9.8-48) per case. There were no haemorrhagic complications. The EBVS is safe and effective for intestinal resections and haemostasis in trauma surgery. This novel energy source reliably seals major mesenteric vessels with little need for sutures or clips. Using the EBVS can substantially shorten operative time.

Background: Increasing patient volume and residents' work hour restrictions have increased the workload at trauma centers. Further, comprehensive tertiary surveys after initial stabilization and appropriate follow-up plans for incidental findings are time consuming. Midlevel providers (MLP) can help streamline this process. We initiated a care plan in which MLPs conducted all tertiary surveys and coordinated follow-ups for incidental findings., Methods: From November 2005 through May 2006, we implemented a MLP-driven initiative aimed at performing tertiary surveys within 48 hours of admission on all trauma patients admitted to our Level-1 trauma center. Tertiary surveys consisted of a complete history and physical, radiographic evaluations and appropriate consultations. Incidental findings were recorded and communicated to the trauma attending. A follow-up plan was devised, and the course of action was documented. Patients or family members were informed, and their acknowledgments were filed. Data are presented as mean +/- SE., Results: There were 1,027 patients admitted during the study period. Blunt mechanisms accounted for 81% of the injuries (primarily motor vehicle crashes and falls). Seventy-six patients had 87 incidental findings (7.4%); 53 were men. The mean age was 51.8 years +/- 2.1 years and mean injury severity score was 18.5 +/- 1.4. Incidental findings of clinical significance included 18 pulmonary nodules or neoplasms, 9 adrenal masses (>4 mm), 7 patients with lymphadenopathy, 5 benign cystic lesions, and 3 renal masses. Other neoplastic lesions included bladder (2), thyroid (2), ovary (1), breast (1), and rectum (1)., Conclusions: With prevalent medicolegal pressure and restricted residents' work hours, a MLP-initiative to streamline the tertiary survey effectively addresses incidental findings. This MLP-driven care plan can help reduce residents' workload, provides appropriate follow-up, and minimizes legal risks inherent to incidental findings on the trauma service.

Background: The arterial base deficit has been demonstrated to be a marker of shock and predictive of survival in injured patients. The venous blood, however, may better reflect tissue perfusion. Its usefulness in trauma is unknown. We compared central venous with arterial blood gas analysis to determine which was a better predictor of survival in injured patients., Methods: A prospective, nonrandomized series of acutely injured patients was investigated. Patients who had an arterial blood gas analysis for acid-base determination had a simultaneous central venous blood gas analysis and routine blood tests. Patient demographics, Injury Severity Score, and survival past 24 hours were recorded. Arterial and venous blood samples were analyzed for pH, PCO2, PO2, HCO3, hemoglobin-oxygen saturation, base deficit, and lactate., Results: One hundred patients were enrolled. There were 76 survivors and 24 nonsurvivors. Wilcoxon rank sum test and multivariate logistic regression were used for each recorded variable; only central venous base deficit was predictive of survival past 24 hours (P = .0081). Specifically, arterial base deficit was not predictive of survival past 24 hours., Conclusion: In a prospective series of acutely injured patients, central venous base deficit, not arterial base deficit, was predictive of survival past 24 hours.

Venous thromboembolism (VTE) in trauma patients is a capricious problem that continues to plague trauma surgeons and critical care physicians alike. Pharmacologic preventions of VTE with anticoagulants are often contraindicated in the trauma patient because of risk of bleeding diathesis. Mechanical prophylaxis in the form of venous compression boots often cannot be placed because of external fixators, swelling, and so forth. Providing effective VTE prophylaxis, while at the same time providing definitive care for the trauma patient, can be a nightmare. This review will first discuss the incidence and prevalence of VTE, as well as investigate the condition's diagnosis and treatment. Solutions to frequently encountered clinical dilemmas in managing VTE in trauma patients are considered in the form of frequently asked questions. Diagnostic techniques such as magnetic resonance venography, D-dimer, and various computed tomography methods are evaluated. Recent literature on preventive pharmacologic therapies is explored. The authors also consider whether vena cava filters prevent pulmonary embolism in trauma patients.

Background: Retrieval of optional vena cava filters (VCF) has been demonstrated to be safe and feasible in injured patients in 4 recent studies. However, 2 pulmonary emboli PE were reported in these studies with mean implant durations less than 19 days. In light of these occurrences, we changed our practice for VCF retrieval when patients had recovered from their injuries and at least 30 days after their discharge, or had been stable on therapeutic anticoagulation for deep venous thrombosis (DVT) or PE for at least 2 weeks. The aim of the current study was to assess the safety of this approach., Methods: A review of prospectively collected data on optional VCF over a 16-month period. The filters were inserted prophylactically per an institutional practice guideline or for the presence of DVT or PE with a contraindication and/or complication to anticoagulation. All patients underwent duplex imaging of the lower extremities and had pre- and post- retrieval cavagrams. Demographics, duration of implantation, and complications were recorded., Results: Eighty-three patients had optional VCF inserted since the change in our clinical practice. Indications included prophylaxis for high-risk trauma patients (n = 58), DVT or PE with acute contraindication to therapeutic anticoagulation (n = 22), or complications of anticoagulation (n = 3). Two patients developed lower extremity DVT after filter insertion and 1 patient developed a vena cava thrombosis. Retrieval was successful in 47 of 54 cases (87%) attempted. Median implantation duration was 142 days (range 17-475). A filter strut fracture occurred during retrieval without further consequences. No post-insertion or post-retrieval PE occurred in this study., Conclusion: Extended intervals for retrieval of VCF are safe and may maximize protection against pulmonary embolism.

Background: Four recent reports of the retrieval of optional vena cava filters (VCF) in trauma patients had average implant durations of 10, 19, and 19 days (one not specified). Two patients in these studies had pulmonary emboli after VCF removal. No evidence-based guidelines exist on the appropriate time to remove optional VCF. The purpose of this study was to examine the timing of pulmonary emboli (PE) and determine the optimal time to remove optional VCFs., Methods: A multicenter retrospective chart review of trauma patients who had a postinjury PE between January 2001 and December 2004 was performed. We examined the demographics, prophylaxis at the time of PE (pharmacologic [unfractionated or low molecular weight heparin] or sequential compression devices [SCD]), diagnostic test used, timing of PE from the date of injury, and survival outcome., Results: In all, 146 patients were identified, mean age 45.1 (+/- 21.1 SD); Injury Severity Score 18.0 (+/- 12.1 SD). Diagnosis was obtained by spiral computed tomography (N = 93), pulmonary arteriogram (N = 18), V/Q (N = 26), autopsy (N = 6), clinical (N = 6), and unknown (N = 3). Overall mortality was 17.8% (N = 26). Pulmonary embolism was felt to contribute to or was the cause of death in 85% (N = 22) of these patients. Two late PE deaths occurred (days 21 and 43). Sixty (37%) patients had pharmacologic prophylaxis at the time of PE and 83 (50.9%) had SCDs. Average time from injury to PE was 7.9 days (+/- 8.1 SD), the longest being 43 days postinjury. Eleven percent of PE occurred after 21 days, including fatal PE., Conclusions: Clinical criteria defining the time to remove optional VCFs without exposing patients to the risk of PE by removing a filter too soon should be determined.