Venous thrombotic, thromboembolic, and mechanical complications after retrievable inferior vena cava filters for major trauma

Transcription

1 British Journal of Anaesthesia 114 (1): 63 9 (2015) Advance Access publication 30 June doi: /bja/aeu195 CLINICAL PRACTICE Venous thrombotic, thromboembolic, and mechanical complications after retrievable inferior vena cava filters for major trauma K. M. Ho 1,4 *,J.A.Tan 1, M. Burrell 2, S. Rao 2 and P. Misur 3 1 Department of Intensive Care Medicine, 2 Western Australian State Trauma Service and 3 Department of Radiology, Royal Perth Hospital, Perth, Australia 4 School of Population Health, University of Western Australia, Perth, Australia * Corresponding author. kwok.ho@health.wa.gov.au Editor s key points Inferior vena cava (IVC) filters are used for prevention of venous thromboembolism when prophylactic anticoagulation is contraindicated. Risk factors for complications from retrievable IVC filters were studied in a retrospective single-centre analysis of trauma patients. High injury severity score, tibial/fibular fractures, delayed pharmacological thromboprophylaxis, and filters left in situ for.50 days were significant risk factors for complications. Background. The ideal thromboprophylaxis in patients at risk of bleeding is uncertain. This retrospective cohort study assessed the risk factors for complications after using retrievable inferior vena cava (IVC) filters for primary or secondary thromboembolism prophylaxis in patients after major trauma. Methods. Using data from radiology, trauma and death registries, the incidence of and risk factors for subsequent deep venous thrombosis (DVT), venous thromboembolism (VTE), and mechanical complications related to retrievable IVC filters in patients, admitted between 2007 and 2012, were assessed in a single trauma centre. Results. Of the 2940 major trauma patients admitted during the study period, a retrievable IVC filter was used in 223 patients(7.6%). Thirty-six patients(16%) developed DVTor VTE subsequent to placement of IVC filters (median 20 days, interquartile range 9 33), including 27 with lower limb (DVT), 8 upper limb DVT, and 4 pulmonary embolism. A high Injury Severity Score, tibial/ fibular fractures, and a delay in initiating pharmacological thromboprophylaxis after insertion of the filters (14 vs 7days, P¼0.001) were significant risk factors. Thirty patients were lost to follow-up (13%) and their filters were not retrieved. Mechanical complications including filters adherent to the wall of IVC (4.9%), IVC thrombus (4.0%), and displaced or tilted filters (2.2%) were common when the filters were left in situ for.50 days. Conclusions. A delay in initiating pharmacological thromboprophylaxis or filter removal were associated with an increased risk of subsequent DVT, VTE, and mechanical complications of retrievable IVC filters in patients after major trauma. Keywords: complications; deep vein thrombosis; prevention; pulmonary embolism; traumatic injury Accepted for publication: 2 March 2014 Venous thromboembolism (VTE) is very common in critically injured patients and is associated with substantial morbidity and mortality. 12 Evidence suggests that thromboprophylaxis is of paramount importance in reducing mortality and morbidity of VTE, but recent studies showed that many critically ill, including patients with major trauma, remained untreated with thromboprophylaxis in a timely fashion. 3 As bleeding from pharmacological thromboprophylaxis is a predictor of death, 4 mechanical thromboprophylaxis including inferior vena cava (IVC) filters is increasingly used in patients with major trauma who have an increased risk of bleeding. 5 Although IVC filters have been used for thromboprophylaxis for many years, there is only one randomized clinical trial evaluating the benefitsandrisksof usingpermanent IVC filters forsecondary prophylaxis in patients with proven proximal deep vein thrombosis (DVT). 6 This randomized controlled study showed that permanent IVC filters did not reduce the long-term risk of pulmonary embolism (PE) or mortality, but their long-term use was associated with an increase in risk of recurrent lower limb DVT. Temporary and retrievable IVC filters can be removed when the risk of VTE or bleeding has subsided and have largely replaced permanent IVC filters in the acute trauma care An abstract of this study was presented in the Annual Scientific Meeting of the Australian and New Zealand Intensive Care Society on October 19, & The Author Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please journals.permissions@oup.com

2 BJA Ho et al. setting when an IVC filter is indicated. However, recent evidence suggested that many temporary IVC filters were not retrieved in due course, and VTE including PE still occurred despite the presence of an IVC filter. 7 Although the complications of many older permanent IVC filters have been well described, the safety of the newer retrievable IVC filters including the incidence of and risk factors for subsequent VTE and mechanical complications remains unclear. 8 9 In this retrospective cohort study, we assessed the incidence and risk factors of DVT and VTE and mechanical complications subsequent to placement of retrievable IVC filters in a cohort of patients with major trauma. Methods After obtaining approval from the Royal Perth Hospital Ethics Committee (EC 2012/115), the Western Australian Trauma Registry data of all major trauma patients who had a retrievable IVC filter for VTE prophylaxis between 1 January 2007 and 31 December 2012 were linked to the radiology database and death registry and the data were cross checked by two investigators. Royal Perth Hospital, a university teaching hospital, is the state of Western Australia s largest trauma centre, and admits about adult trauma patients including burns, cardiothoracic, spinal and head trauma, per annum. The standard thromboprophylaxis for all injured patients included elastic stockings, intermittent pneumatic compression devices to the lower limbs when there was no physical injury to the legs, and unfractionated heparin 5000 units twice or thrice daily if the patients were not at risk of bleeding. Low-molecular-weight heparin (subcutaneous enoxaparin 40 mg daily) was also used occasionally in patients who were deemed to be not at riskof bleeding. Retrievable IVC filters (Bard w G2 and G2X, C.R. Bard, Inc. New Jersey, USA) were used as primary VTE prophylaxis if pharmacological thromboprophylaxis was contraindicated within the first 3 to 7 days of injury because of bleeding risk and intermittent pneumatic compression to the lower limbs was not possible because of lower limb injuries, or when using intermittent pneumatic compression alone for a prolonged period of time was considered inadequate to prevent fatal PE (e.g. a patient with pelvic fractures, abdominal visceral organ injury, and severe brain injury requiring multiple operations within the first 7 10 days after hospital admission). IVC filters were used for secondary VTE prophylaxis in patients who had documented symptomatic VTE but with contraindications to systemic anticoagulation. In this study, only trauma patients with an Injury Severity Score (ISS) 15 were analysed. 10 All data were collected for clinical reasons and were retrospectively retrieved from the medical records and radiology database. Risk factors for DVT and VTE subsequent to insertion of IVC filters analysed in this study included age, sex, ISS, regions of injury, body weight, interventions such as intracranial pressure monitor and mechanical ventilation, use of the IVC filter as primary or secondary VTE prophylaxis, and time to initiate pharmacological thromboprophylaxis (or systemic anticoagulation for those with VTE before pharmacological thromboprophylaxis was initiated) after insertion of the IVC filters. The latter information was retrieved from the anticoagulation chart of each patient. Outcomes included symptomatic lower limb and upper limb DVT, PE, and mechanical complications noted when attempts to remove the IVC filters were made. The mechanical complications analysed in this study included filter adherent to or penetrating the IVC wall, IVC thrombosis occluding the filter requiring systemic anticoagulation for 8 10 weeks before a second attempt at removal, and filters that were displaced or tilted within the IVC reducing its effectiveness in trapping emboli from the lower limbs. Outcome events included in this study were symptomatic DVTand PE confirmed by colour Doppler compression ultrasound and CT pulmonary angiography, respectively. Outcome events, mechanical complications related to the IVC filters, and mortality were assessed on 14th May Because this was a retrospective study, radiologists who reported DVT or VTE events and mechanical complications related to IVC filters were unaware of the initiation or design of this study. Statistical analysis A sample size of over 200 was needed to assess up to five independent risk factors subsequent to the insertion of the IVC filters if the incidence of DVTor VTE of the major trauma patients with an IVC filter was 25%. Categorical variables were analysed using a x 2 test, and continuous variables with skewed distributions were analysed using a Mann Whitney test. We used Cox proportional hazards regression to assess plausible risk factors, including the pattern and severity of injury and use of clinical interventions, associated with the time to first subsequent DVT or VTE after filter insertion. Variables entered into the multivariate analysis were eliminated, in a stepwise fashion, if their associated P-value was.0.25 to improve precision of the results of the final model. A Kaplan Meier hazard curve was used to assess whether major mechanical complications related to IVC filters were related to duration of the filters left in situ. A P-value of,0.05 was taken as significant, and all tests were two-tailed and performed using SPSS for Windows (version 19.0, IBM, Version 19.0, Armonk, NY, USA, 2011). Results DVT or VTE subsequent to insertion of IVC filters A total of 2940 major trauma (ISS 15) and minor trauma (ISS,15) patients were admitted to the study centre between 2007 and Of the 2940 major trauma patients, a retrievable IVC filter was used in 223 patients (7.6%) with 201 filters used for primary VTE prophylaxis because of the presence of contraindications to other forms of thromboprophylaxis and 22 filters used for secondary VTE prophylaxis after symptomatic DVT or VTE was diagnosed and systemic anticoagulation was considered contraindicated (Fig. 1). Thirty-six of the 223 patients [16%; 2.3 VTE events per 1000 patient-days, 95% confidence interval (CI) ] developed VTE subsequent to the placement of the IVC filters [median 20 days, interquartile range (IQR) 9 33], including 27 DVT, 8 upper limb DVT, and 4 PE. Three patients had both upper and lower limb DVT subsequent to filter placement. A high ISS, low admission Glasgow Coma Score, use of intracranial pressure monitor, and a delay 64

3 Thromboembolic and mechanical complications BJA 2940 major trauma [defined as Injury Severity Score (ISS) 15] and minor trauma (ISS<15) patients admitted to the study centre between 2007 and consecutive patients (7.6%) who had an IVC filter inserted after major trauma Patients who did not have an IVC filter (n = 2717, 92.4%); the incidence of symptomatic venous thromboembolism was 2.8% among the subgroup of patients who had received pharmacological thromboprophylaxis in the years 2007 and 2008 IVC filter as primary prophylaxis due to contraindications to pharmacological thromboprophylaxis (n=201) A total of 34 patients (17%) developed subsequent VTE (median time 22 days after filter placement, interquartile range: 9 34): - upper limb DVT (n=7,3.5%) - lower limb DVT (n=25,12.5%) - pulmonary embolism (n=3, 1.5%) IVC filter as secondary prophylaxis because of contraindications to systemic anticoagulation (n = 22): - upper limb DVT = 2 (9%), - lower limb DVT = 7 (32%), - pulmonary embolism = 18 (82%), - 2 patients had all forms of DVT and VTE - 1 patient had both lower limb DVT and pulmonary embolism Two patients (9.1%) developed subsequent DVT or VTE (median time 6 days after the placement of the IVC filter, interquartile range: 3 9): - upper limb DVT (n = 1, 4.5%) - lower limb DVT (n = 2, 9.1%) - pulmonary embolism (n = 1, 4.5%) Fig 1 Flow chart showing patient inclusion and their outcomes. in initiating pharmacological thromboprophylaxis after insertion of DVTor IVC filter (median 14 vs 7days,P¼0.001) were significantly associated with occurrence of subsequent DVTor VTE (Table 1). Cox regression analysis showed that only three factors time to initiate pharmacological thromboprophylaxis, ISS, and tibial or fibular fractures were significantlyassociated with an increased risk of subsequent DVT or VTE after filter placement (Table 2). Use of low-molecular-weight heparin against different doses of unfractionated heparin (5000 units twice or thrice daily) was not a significant factor. Further stratification of the time to initiate pharmacological thromboprophylaxis after filter insertion into 7, 8 14, or.21 days showed that there were progressive increases in risk of subsequent DVT or VTE with delays in initiating pharmacological thromboprophylaxis [.21 vs 7 days: hazard ratio (HR) ¼22.0, 95% CI , P¼0.001; vs 7 days: HR¼13.3, 95% CI , P¼0.001; 8 14 vs 7 days: HR¼4.5, 95% CI , P¼0.019] (Fig. 2); this risk factor fulfilled the proportionality assumption of the Cox regression. Although the mortality of those with DVT or VTE subsequent to the insertion of IVC filters appeared to be higher than those without, this difference was not statistically different both with and withoutadjustingforiss (Table1). Fatal PEwas notobserved in the 14 patients who had post-mortem examination. Mechanical complications related to retrievable IVC filters Of the 36 patients who had subsequent DVTor VTE after filter insertion, 13 (36%) also had at least one form of mechanical complication or were lost to follow-up. A total of 50 patients had at least one mechanical complication or were lost to follow-up, and 13 of these (26%) also had symptomatic DVTor VTE subsequent to filter insertion (Table 3). Together, a total of 73 patients 65

4 BJA Ho et al. Table 1 Difference in characteristics of the patients with and without subsequent DVTor VTE after placement of a retrievable IVC filter after major trauma (n¼223). CT, computed tomography; GCS, Glasgow Coma Score; IQR, interquartile range; *mortality censored on May 14, P-values generated by either Chi-square or Mann-Whitney test. Continuous variables are presented in median and interquartile range Variable DVT or VTE, yes (n536) DVT or VTE, no (n5187) P-value Age, yr (IQR) 27 (21 46) 33 (22 49) Male, n (%) 28 (78) 142 (76) Weight, kg (IQR) 82 (70 100) 80 (69 93) Time between injury and IVC filter insertion, days (IQR) 1 (1 3) 1 (1 3) Time to initiate pharmacologic thromboprophylaxis, days (IQR) 14 (11 21) 7 (5 10) Injury Severity Score, (IQR) 32 (27 43) 29 (24 38) Chest injury, n (%) 21 (58) 115 (61) Requiring mechanical ventilation, n (%) 34 (94) 156 (83) Head injury, n (%) 27 (75) 134 (72) Intracranial pressure monitor, n (%) 24 (67) 86 (46) Admission GCS (IQR) 6 (3 13) 11 (6 14) CT brain appearance, n (%) Subarachnoid haemorrhage 18 (50) 69 (37) Subdural haematoma 9 (25) 47 (25) Epidural haematoma 4 (11) 26 (14) Intracerebral contusion or haematoma 21 (58) 88 (47) Effaced basal cistern 6 (17) 21 (11) Midline shift 6 (17) 25 (13) Marshall CT brain grading I 11 (31) 67 (36) II 15 (41) 75 (40) III 0 (0) 10 (5) IV 0 (0) 2 (1) V 6 (17) 26 (14) VI 4 (11) 7 (4) Spinal fractures, n (%) 23 (64) 99 (53) Neurological deficits from spinal injury, n (%) 7 (19) 30 (16) Abdominal visceral organ injury, n (%) 11 (31) 63 (34) Pelvic fractures, n (%) 14 (39) 76 (41) Fracture femur, n (%) 5 (14) 43 (23) Fracture tibia/fibula, n 17 (47) 64 (34) IVC filter used as primary prophylaxis, n (%) 34 (94) 167 (89) Mortality,* n (%) 5 (14) 14 (8) Length of hospital stay, days (IQR) 45 (20 81) 34 (21 51) out of 223 (33%) had either a new DVTor VTE, mechanical complication, or were lost to follow-up subsequent to filter insertion for primary or secondary VTE prophylaxis. Of the 223 included in the study, 30 patients were lost to follow-up (13%) and their IVC filters were not retrieved. After excluding these 30 patients, the median time to filter retrieval was 79 days (IQR days). Mechanical complications, including adherence to the wall of IVC (4.9%; 0.72 per 1000 patient-days, 95% CI ), IVC thrombus occluding the filter (4.0%; 0.59 per 1000 patient-days, 95% CI ), and displaced or tilted filter within the IVC (2.2%), at the time of first attempt to remove the IVC filters were common. The risk of these mechanical complications increased substantially when filters were left in situ for.50 days (Figs 3 and 4). Filters that were adherent to or penetrated through the wall of the IVC filters were not retrieved. Those patients who had IVC thrombosis received systemic anticoagulation for 2 months, and all these filters were successfully retrieved subsequently. Discussion Our results showed that venous thrombotic, thromboembolic, and mechanical complications after using retrievable IVC filters in patients with major trauma were common (33%). A delay in initiating pharmacological thromboprophylaxis after filter insertion was an important preventable risk factor for subsequent DVT or VTE, and the risk of mechanical complications increased substantially with duration of filters left in situ. The reason why many patients who had an IVC filter for primary VTE prophylaxis did not receive pharmacological thromboprophylaxis earlier was likely because of the common misconception that an IVC filter was sufficient in preventing 66

5 Thromboembolic and mechanical complications BJA Table 2 Factors associated with time to develop subsequent DVT or VTE after insertion of a retrievable IVC filter for primary VTE prophylaxis in patients with major trauma (n¼201). NB: use of intracranial pressure monitor, Glasgow Coma Score, pelvic and spinal vertebral bone fractures, Marshall CT brain grading, spinal neurological deficits, chest and abdominal injuries and use of low-molecular-weight heparin (against different doses of unfractionated heparin) were initially entered into the Cox proportional hazards regression and were removed in a stepwise fashion because their associated P-values were.0.25 Variables Hazard ratio (95% confidence interval) P-value Injury Severity Score 1.03 ( ) (per score increment) Days to start pharmacologic thromboprophylaxis after the insertion of IVC filter 1.03 ( ) (per day increment) Tibial/fibular bone fractures 2.21 ( ) Femoral bone fractures 0.48 ( ) Cumulative hazard Time to start pharmacological thromboprophylaxis (P<0.001) <7 days 8 14 days days >21 days Time to first VTE with IVC filter left in situ (days) Fig 2 Hazard curves for developing venous thromboembolism (VTE) after insertion of a retrievable IVC stratified by time to initiate pharmacological thromboprophylaxis (7 days or less, 8 14 days, days, and.21 days), after adjustment for ISS, tibial/ fibular fractures, and femoral fractures. Cumulative hazard on the Y-axis is equivalent to negative log of the probability of not experiencing new VTE. all forms of VTE in patients at risk. Indeed, our results are consistent with recent observational studies on retrievable IVC filters that VTE including PE could still occur despite the presence of an IVC filter. 79 The possible mechanisms for PE occurrence despite the presence of an IVC filter may include clots originating from the upper limbs, de novo thrombosis formation in the pulmonary circulation, 11 and ineffective trapping of embolus from the lower limbs if the IVC filter is tilted or displaced. Our results suggested that while IVC filters might be very effective in preventing fatal PE, they are inadequate alone to prevent all DVT and VTE events and should best be used as a bridging measure with pharmacological thromboprophylaxis initiated as soon as possible. Practically, most IVC filters are used for primary VTE prophylaxis within the first 7 10 days after major trauma when patients are at highest risk of bleeding. In the later phase of major trauma, evidence suggests that many patients become prothrombotic As such, initiating pharmacological thromboprophylaxis as soon as bleeding risk has subsided, possibly within 7 10 days after filter insertion, should be considered to minimize the risk of DVT and VTE complications. Although retrievable IVC filters can theoretically be removed when bleeding risk has subsided and pharmacological thromboprophylaxis can be commenced, we noted that the median time to retrieve filters was much longer than the time to initiate pharmacological thromboprophylaxis for most patients. The clinical imperative for leaving an IVC filter in situ beyond the time without pharmacological thromboprophylaxis is to confer additional protection against PE for patients remaining at high risk of VTE. This hypothetical benefit must be balanced with the increased risk of mechanical complications when filters are left in situ for too long. Our study showed that the risk of filters adherent to or penetrating through the wall of IVC, and also IVC thrombosis, increased substantially with time, especially after 50 days, suggesting that mechanical complications might be substantially reduced if retrievable IVC filters are removed within 50 days or soon after pharmacological thromboprophylaxis has commenced. 14 Instead of leaving IVC filters in situ for patients who remain at high risk of VTE when pharmacological thromboprophylaxis alone is considered inadequate, other alternatives such as combining intermittent pneumatic compression of the lower limbs with pharmacological thromboprophylaxis or systemic anticoagulation should also be considered. Furthermore, leaving IVC filters in situ for a prolonged period of time also runs a risk of leaving filters not retrieved when patients are lost to follow-up. Since we implemented strict Western Australia state-wide guidelines and protocols on removal of temporary IVC filters before 60 days, the number of patients without filters retrieved because of lost to follow-up has substantially reduced (unpublished observations). This study has some limitations. First, the incidence of DVTor VTE was relatively low (16%), reducing the precision of the results and the statistical power of the study to identify multiple risk factors subsequent to filter insertion. This was a single-centre study and we used only one particular brand of retrievable IVC filters. As such, confirmation of our results by other centres is required. 14 Secondly, we had data only on symptomatic DVT and PE. Whether the incidence of asymptomatic DVT and PE after insertion of IVC filters for primary VTE prophylaxis in 67

6 BJA Ho et al. Table 3 Incidence of problems or complications encountered subsequent to placement of a retrievable IVC filter as primary or secondary VTE prophylaxis for patients with major trauma (n¼223). *Three patients had both upper and lower limb DVT subsequent to placement of an IVC filter and 13 (36%) of the 36 patients who had subsequent DVTE or VTE also had at least one form of mechanical complications or were lost to follow-up. Thirteen of the 50 patients (26%) who had mechanical complications or were lost to follow-up also had symptomatic thromboembolism subsequent to the insertion of an IVC filter. Together, a total of 73 patients out of 223 (33%) had either new DVTor VTE, mechanical complications or lost to follow-up subsequent to the insertion of an IVC filter for primary or secondary VTE prophylaxis Type of complications No. of patients (%) All forms of DVT or VTE* 36 (16) Lower limb DVT 27 (12) Upper limb DVT 8 (4) Pulmonary embolism 4 (2) One or more of the after complications 50 (22) IVC filter occluded with thrombosis 9 (4) IVC filter tilted within the IVC not offering full protection 5 (2) IVC filter adherent to IVC and could not be removed 11 (5) IVC bleeding during removal of the filter 1 (0.5) Narrowing of the IVC noted after removal of the IVC filter 1 (0.5) Lost to follow-up and IVC filter not removed 30 (13) Hazard function curve Censored Hazard function curve Censored Cumulative hazard Cumulative hazard Duration of IVC filter left in situ (days) Duration of IVC filter left in situ (days) Fig 3 Hazard of having a retrievable IVC filter adherent to the IVC and not able to be removed with duration of the IVC filter left in situ (n¼193 after excluding 30 patients who were lost to followup). First event occurred on day 50 after insertion of retrievable IVC filter. patients with major trauma is much higher than symptomatic DVTor VTE remains uncertain. Furthermore, there are also limitations in using Doppler ultrasound (sensitivity 85% and specificity 88% in detecting symptomatic proximal DVT) to diagnose symptomatic DVT, and hence the true incidence of DVTafter the use of an IVC filter could be higher than we reported. 17 Finally, our previous study showed that the risk of symptomatic DVT or VTE in patients after major trauma who had no contraindications to pharmacological thromboprophylaxis and not requiring an IVC filter was about 2.8%. 2 These data suggest that the presence Fig 4 Hazard of having a retrievable IVC filter occluded with thrombus with duration of the IVC filter left in situ (n¼193 after excluding 30 patients who were lost to follow-up). First event occurred on day 65 after insertion of retrievable IVC filter. of an IVC filter might have, in part, contributed to the increased risk of DVT or VTE (16%) reported in this study. Nevertheless, this comparison is not fully valid because patients with and without contraindications to pharmacological thromboprophylaxis are intrinsically different. As such, we could not confirm whether using IVC filters for primary prophylaxis would be beneficial or harmful compared with placebo in major trauma patients who have contraindications to all thromboprophylaxis. An adequately powered randomized controlled study evaluating the benefits, harms, and costeffectiveness of using IVC filters for primary thromboprophylaxis in patients who have contraindications to other forms 68

7 Thromboembolic and mechanical complications BJA of thromboprophylaxis after major trauma is thus needed. Because many VTE events are clinically silent, a proactive approach to detect mildly symptomatic or asymptomatic VTE will be essential in such a randomized controlled study. In summary, DVT, VTE, and mechanical complications after using retrievable IVC filters for VTE prophylaxis in patients with major trauma were common, especially when pharmacological thromboprophylaxis was delayed or when filters were not removed within 50 days. Until data from randomized controlled trials on cost-effectiveness and safety of using retrievable IVC filters in patients with major trauma are available, initiating pharmacological thromboprophylaxis and removing retrievable IVC filters as soon as bleeding and VTE risks have subsided are recommended. Authors contributions K.M.H.: initiated the idea of the study, involved in the data collection, analysed and interpreted the data. J.A.T. and M.B.: involved in the data collection and interpretation. S.R. and P.M.: involved in the data interpretation. All authors were involved in the drafting of the manuscript and agreed on the content of the final manuscript. K.M.H. is the guarantor of the study. Declaration of interest None declared. Funding This work was solely funded by Department of Intensive Care Medicine, Royal Perth Hospital, Perth, Australia. References 1 Schultz DJ, Brasel KJ, Washington L, et al. Incidence of asymptomatic pulmonary embolism in moderately to severely injured trauma patients. J Trauma 2004; 56: ; discussion Ho KM, Burrell M, Rao S, Baker R. Incidence and risk factors for fatal pulmonary embolism after major trauma: a nested cohort study. Br J Anaesth 2010; 105: Ho KM, Chavan S, Pilcher D. Omission of early thromboprophylaxis and mortality in critically ill patients: a multicenter registry study. Chest 2011; 140: Eikelboom JW, Quinlan DJ, O Donnell M. Major bleeding, mortality, and efficacy of fondaparinux in venous thromboembolism prevention trials. Circulation 2009; 120: Shackford SR, Cook A, Rogers FB, LittenbergB, OslerT. The increasing use of vena cava filters in adult trauma victims: data from the American College of Surgeons National Trauma Data Bank. J Trauma 2007; 63: Decousus H, Leizorovicz A, Parent F, et al. A clinical trial of vena caval filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. Prévention du Risque d Embolie Pulmonaire par Interruption Cave Study Group. N Engl J Med 1998; 338: Sarosiek S, Crowther M, Sloan JM. Indications, complications, and management of inferior vena cava filters: the experience in 952 patients at an academic hospital with a level I trauma center. JAMA Intern Med 2013; 173: Datta I, Ball CG, Rudmik L, Hameed SM, Kortbeek JB. Complications related to deep venous thrombosis prophylaxis in trauma: a systematic review of the literature. J Trauma Manag Outcomes 2010; 4: 1 9 Smoot RL, Koch CA, Heller SF, et al. Inferior vena cava filters in trauma patients: efficacy, morbidity, and retrievability. J Trauma 2010; 68: Baker SP, O Neill B, Haddon W Jr, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: Velmahos GC, Spaniolas K, Tabbara M, et al. Pulmonary embolism and deep venous thrombosis in trauma: are they related? Arch Surg 2009; 144: Duff OC, Ho KM, Maybury SM. In vitro thrombotic tendency of reactive thrombocytosis in critically ill patients: a prospective case-control study. Anaesth Intensive Care 2012; 40: Ho KM, Yip CB, Duff O. Reactive thrombocytosis and risk of subsequent venous thromboembolism: a cohort study. J Thromb Haemost 2012; 10: Uberoi R, Tapping CR, Chalmers N, Allgar V. British Society of Interventional Radiology (BSIR) Inferior Vena Cava (IVC) Filter Registry. Cardiovasc Intervent Radiol 2013; 36: Ho KM, Tan JA. Stratified meta-analysis of intermittent pneumatic compression of the lower limbs to prevent venous thromboembolism in hospitalized patients. Circulation 2013; 128: Ho KM. Benefit of intermittent pneumatic compression of lower limbs in reducing venousthromboembolism in hospitalised patients: interactions between risk and effectiveness. Anaesth Intensive Care 2014; 42: Rogers FB, Cipolle MD, Velmahos G, Rozycki G, Luchette FA. Practice management guidelines for the prevention of venous thromboembolism in trauma patients: the EAST practice management guidelines work group. JTrauma2002; 53: Handling editor: H. C. Hemmings 69