Thromboprophylaxis for trauma patients

  • Review
  • Intervention

Authors


Abstract

Background

Trauma is a leading causes of death and disability in young people. Venous thromboembolism (VTE) is a principal cause of death. Trauma patients are at high risk of deep vein thrombosis (DVT). The incidence varies according to the method used to measure the DVT and the location of the thrombosis. Due to prolonged rest and coagulation abnormalities, trauma patients are at increased risk of thrombus formation. Thromboprohylaxis, either mechanical or pharmacological, may decrease mortality and morbidity in trauma patients who survive beyond the first day in hospital, by decreasing the risk of VTE in this population.

A previous systematic review did not find evidence of effectiveness for either pharmacological or mechanical interventions. However, this systematic review was conducted 10 years ago and most of the included studies were of poor quality. Since then new trials have been conducted. Although current guidelines recommend the use of thromboprophylaxis in trauma patients, there has not been a comprehensive and updated systematic review since the one published.

Objectives

To assess the effects of thromboprophylaxis in trauma patients on mortality and incidence of deep vein thrombosis and pulmonary embolism. To compare the effects of different thromboprophylaxis interventions and their effects according to the type of trauma.

Search methods

We searched The Cochrane Injuries Group Specialised Register (searched April 30 2009), Cochrane Central Register of Controlled Trials 2009, issue 2 (The Cochrane Library), MEDLINE (Ovid) 1950 to April (week 3) 2009, EMBASE (Ovid) 1980 to (week 17) April 2009, PubMed (searched 29 April 2009), ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 to April 2009), ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 to April 2009).

Selection criteria

Randomized controlled clinical trials involving people of any age with major trauma defined by one or more of the following criteria: physiological: penetrating or blunt trauma with more than two organs and unstable vital signs, anatomical: people with an Injury Severity Score (ISS) higher than 9, mechanism: people who are involved in a 'high energy' event with a risk for severe injury despite stable or normal vital signs. We excluded trials that only recruited outpatients, trials that recruited people with hip fractures only, or people with acute spinal injuries.  

Data collection and analysis

Four authors, in pairs (LB and CM, EF and RC), independently examined the titles and the abstracts, extracted data, assessed the risk of bias of the trials and analysed the data. PP resolved any disagreement between the authors.

Main results

Sixteen studies were included (n=3005). Four trials compared the effect of any type (mechanical and/or pharmacological) of prophylaxis versus no prophylaxis. Prophylaxis reduced the risk of DVT in people with trauma (RR 0.52; 95% CI 0.32 to 0.84). Mechanical prophylaxis reduced the risk of DVT (RR = 0.43; 95% CI 0.25 to 0.73). Pharmacological prophylaxis was more effective than mechanical methods at reducing the risk of DVT (RR 0.48; 95% CI 0.25 to 0.95). LMWH appeared to reduce the risk of DVT compared to UH (RR 0.68; 95% CI 0.50 to 0.94). People who received both mechanical and pharmacological prophylaxis had a lower risk of DVT (RR 0.34; 95% CI 0.19 to 0.60)

Authors' conclusions

We did not find evidence that thromboprophylaxis reduces mortality or PE in any of the comparisons assessed. However, we found some evidence that thromboprophylaxis prevents DVT. Although the strength of the evidence was not high, taking into account existing information from other related conditions such as surgery, we recommend the use of any DVT prophylactic method for people with severe trauma.

Résumé scientifique

La thromboprophylaxie pour les patients présentant un traumatisme

Contexte

Les traumatismes sont une des principales causes de décès et d'invalidité chez les jeunes. La thromboembolie veineuse (TEV) est une cause majeure de décès. Les personnes présentant un traumatisme ont un risque élevé de thrombose veineuse profonde (TVP). L'incidence varie en fonction de la méthode utilisée pour mesurer la TVP et de l'emplacement de la thrombose. Le repos prolongé et les anomalies de la coagulation exposent les patients présentant un traumatisme à un risque accru de formation de thrombus. En réduisant le risque de TEV, la thromboprophylaxie, qu'elle soit mécanique ou pharmacologique, est susceptible d'abaisser la mortalité et la morbidité chez les patients présentant un traumatisme qui survivent au-delà de la première journée d'hospitalisation.

Une revue systématique précédente n'avait pas trouvé de preuves de l'efficacité des interventions pharmacologiques ou mécaniques. Toutefois, cette revue systématique avait été menée il y a 10 ans et la plupart des études incluses étaient de mauvaise qualité. Depuis lors, de nouveaux essais ont été menés. Bien que les directives actuelles recommandent l'utilisation de la thromboprophylaxie chez les patients présentant un traumatisme, il n'y a pas eu de revue systématique exhaustive et actualisée depuis celle publiée.

Objectifs

Évaluer les effets de la thromboprophylaxie chez les patients présentant un traumatisme, pour ce qui concerne la mortalité et l'incidence de la thrombose veineuse profonde et de l'embolie pulmonaire. Comparer les effets de différentes interventions thromboprophylactiques ainsi que leurs effets en fonction du type de traumatisme.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre spécialisé du groupe Cochrane sur les blessures (1er décembre 2009), le registre Cochrane des essais contrôlés 2009, numéro 2 (The Cochrane Library), MEDLINE (Ovid) (de 1950 à la 3ème semaine de novembre 2009), EMBASE (Ovid) (de 1980 à fin avril 2009), PubMed (29 avril 2009), ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (de 1970 à avril 2009), ISI Web of Science : Conference Proceedings Citation Index-Science (CPCI-S) (de 1990 à avril 2009).

Critères de sélection

Des essais cliniques contrôlés randomisés impliquant des personnes de tout âge présentant un traumatisme majeur défini par un ou plusieurs des critères suivants : physiologique : traumatisme pénétrant ou contondant avec plus de deux organes touchés et signes vitaux instables, anatomique : personnes ayant un score de gravité de blessure (ISS) supérieur à 9, mécanisme : personnes impliquées dans un événement de 'haute énergie' avec un risque de blessure grave malgré des signes vitaux stables ou normaux. Nous avons exclu les essais qui n'avaient recruté que des patients ambulatoires, les essais qui n'avaient recruté que des personnes atteintes de fractures de la hanche ou des personnes souffrant de lésions médullaires aiguës.

Recueil et analyse des données

Quatre auteurs ont, indépendamment deux à deux (LB et CM, EF et RC), examiné les titres et les résumés, extrait les données, évalué le risque de biais des essais et analysé les données. PP a résolu tout désaccord entre les auteurs.

Résultats principaux

Seize études ont été incluses (n=3005). Quatre essais avaient comparé l'effet de n'importe quel type (mécanique et/ou pharmacologique) de prophylaxie par rapport à l'absence de prophylaxie. La prophylaxie avait réduit le risque de TVP chez les personnes présentant un traumatisme (RR 0,52 ; IC à 95% 0,32 à 0,84). La prophylaxie mécanique avait réduit le risque de TVP (RR 0,43 ; IC à 95% 0,25 à 0,73). La prophylaxie pharmacologique avait été plus efficace que les méthodes mécaniques pour réduire le risque de TVP (RR 0,48 ; IC à 95% 0,25 à 0,95). L'HBPM s'est avérée réduire le risque de TVP en comparaison avec l'héparine non fractionnée (RR 0,68 ; IC à 95% 0,50 à 0,94). Les personnes ayant bénéficié d'une prophylaxie à la fois mécanique et pharmacologique avaient un plus faible risque de TVP (RR 0,34 ; IC à 95% 0,19 à 0,60)

Conclusions des auteurs

Dans aucune des comparaisons évaluées nous n'avons trouvé de preuve que la thromboprophylaxie réduise la mortalité ou l'EP. Nous avons cependant trouvé des preuves que la thromboprophylaxie prévient la TVP. Malgré la solidité limitée des preuves, la prise en compte d'informations provenant de situations connexes, telle que la chirurgie, nous amène à recommander l'utilisation d'une méthode quelconque de prophylaxie de la TVP chez les personnes présentant un traumatisme grave.

Plain language summary

Preventing death from blood clots, the formation of blood clots and blood clots in the lungs in people who have had physical trauma

Thromboembolism (unwanted clotting of the blood) is a frequent complication in people who have experienced physical trauma and is also an important cause of death. The type of trauma, association with vascular injuries, and prolonged hospital bed rest are known risk factors for the development of deep vein thrombus (clot in veins of lower extremities) that can travel (embolize) to the lungs and cause death. Because of this it is usually recommended that people who have had major trauma are given mechanical or pharmacological treatments to prevent their blood forming unwanted blood clots. Mechanical interventions can include compression stockings, an air-filled plastic tube that presses around the leg, a metal blood clot filter placed inside a vein; pharmaceutical drugs include unfractionated heparin, low weight molecular heparin, anticoagulants (e.g. warfarin), antiplatelet drugs (e.g. aspirin) and others. Sixteen studies involving 3,005 people are included in this review. We did not find strong evidence that either mechanical or pharmacological interventions reduce death or clots travelling to the lungs, but we found some evidence that they can prevent clots from forming in the legs.

Résumé simplifié

Prévention de la mort causée par un caillot de sang, de la formation de caillots de sang et des caillots de sang dans les poumons chez les personnes ayant subi un traumatisme physique

La thromboembolie (coagulation indésirable du sang) est une complication fréquente chez les personnes ayant subi un traumatisme physique et elle est également une importante cause de décès. Le type de traumatisme, l'association avec des lésions vasculaires et le repos prolongé sur un lit d'hôpital sont des facteurs connus de risque pour le développement de thrombus dans les veines profondes (caillots dans les veines des membres inférieurs) qui peuvent se déplacer (emboliser) dans les poumons et causer la mort. Pour cette raison, il est généralement recommandé que les personnes ayant subi un traumatisme majeur reçoivent un traitement mécanique ou pharmacologique visant à prévenir la formation de caillots dans leur sang. Les interventions mécaniques pourront être des bas de contention, un tube en matière plastique rempli d'air qui appuie autour de la jambe ou un filtre de caillots de sang en métal placé à l'intérieur d'une veine ; les médicaments sont notamment l'héparine non fractionnée, l'héparine de bas poids moléculaire, les anticoagulants (p.ex. warfarine), les antiagrégants plaquettaires (p.ex. l'aspirine) et d'autres. Seize études totalisant 3 005 personnes sont incluses dans cette revue. Nous n'avons pas trouvé de preuves solides que les interventions mécaniques ou pharmacologiques réduisent la mortalité ou les caillots se déplaçant dans les poumons, mais nous disposons de preuves qu'elles peuvent prévenir la formation de caillots dans les jambes.

Notes de traduction

Traduit par: French Cochrane Centre 22nd March, 2013
Traduction financée par: Instituts de Recherche en Sant� du Canada, Minist�re de la Sant� et des Services Sociaux du Qu�bec, Fonds de recherche du Qu�bec-Sant� et Institut National d'Excellence en Sant� et en Services Sociaux pour la France: Minist�re en charge de la Sant�

Background

 

Description of the condition

Trauma is one of the leading causes of death and disability in young people (Evans 2003). Worldwide, about five million people die as a result of trauma every year (WHO 2008). For patients who reach hospital within one hour of trauma (called the 'golden hour'), blood loss and traumatic brain injury are the main causes of death (Sauaia 1995). For patients who survive beyond the first day, multiple organ failure, central nervous system (CNS) injury and venous thromboembolism (VTE) are the principal causes of death (Acosta 1998).

Trauma patients are at known risk of entering into a hypercoagulable state (intrinsic alterations in the nature of the blood itself). Mechanisms of hypercoagulability in the trauma setting include stasis, vessel wall dysfunction and alterations in clotting mechanisms (Virchow's triad). Injured patients are often immobilized after high-energy trauma. Being in a static position causes a reduction in venous blood returns and a decrease in the supply of oxygen and nutrients to endothelial cells. In addition, endothelial damage caused by direct trauma to the vessels causes the exposition of tissue factor bearing cells. This initiates a procoagulant factor that amplifies the coagulant response. These tissue factor bearing cells move to the cell surface of the platelets, which produces a propagation of the signal through the accumulation of thrombin, activated cofactors and more platelets, inducing thrombosis (Hoffman 2001).

On the other hand, trauma patients experience a reduction of fibrinolytic pathways that seems to result from increased plasminogen activator inhibitor (PAI) 1. PAI 1 inhibits tissue plasminogen activator (tPA) and thus decreases the production of plasmin (Rogers 1995; Kelsey 2000). Coagulation abnormalities and the reduced ability to use the muscular pump of the calf in the injured patient can produce deep venous thrombosis (DVT) in the inferior and superior extremities (Spaniolas 2008). When the thrombus extends to the proximal segments, there is an increased risk of clot migration to the lungs and a fatal outcome (Geerts 2008).

Trauma patients are at high risk for DVT, with an incidence of 11.8% to 65% (Sevitt 1961; Geerts 1994; Velmahos 2000). The incidence varies according to the method used to measure the DVT and the location of the thrombosis. Incidence of thrombosis in the thigh (proximal DVT) is estimated at 18% (Geerts 1994). The incidence of pulmonary embolism (PE) is estimated between 1.5% and 20% (Shackford 1988; O'Malley 1990; Velmahos 2000). Many risk factors for DVT and PE in trauma patients have been identified such as spinal cord injury, lower extremity and pelvic fractures, need for surgical procedures, increasing age, femoral venous line insertion or surgical repair of venous injuries, prolonged immobility, long duration of hospital stay, severity of the trauma, and mechanism of injury (Geerts 1994; Knudson 1994; Frezza 1996; Velmahos 2000; Cipolle 2002; Rogers 2002; Meissner 2003).

Description of the intervention

Thromboprophylaxis describes any intervention used to prevent the development of VTE, and can be categorized into mechanical and pharmacological interventions.

External mechanical devices such as graded compression devices or intermittent pneumatic compression (IPC) have been shown to be effective in preventing DVT, but they cannot be used in patients with lower extremity trauma (Fisher 1995; Elliott 1999; Velmahos 2000). Internal mechanical devices are used to prevent the migration of thrombus from DVT to the lungs, thus preventing PE. One such device is the inferior vena cava filter (IVCF) which may be particularly useful in trauma patients because of the risk of ongoing bleeding at injured sites (McMurty 1999).

Pharmacological thromboprophylaxis was first described in the 1940s by Bauer 1944, and since then a number of interventions have been proposed. The anticoagulant effect of unfractionated heparin (UH) is initiated by the activation of antithrombin III (ATIII). The ATIII/heparin complex inactivates the thrombin factor IIa, and factors Xa, IXa, XIa and XIIa. However, UH is associated with a number of adverse events, such as thrombocytopenia. More recently, alternatives such as low molecular weight heparin (LMWH), a derivative of UH, have been proposed. LMWH acts in the same way as UH, but its low molecular weight fragments reduce the binding to other cells and proteins (and it also has a major affinity to factor Xa) (Hirsh 2004). These drugs have potential as effective prophylactic interventions for trauma, although there is concern due to the associated increased risk of bleeding (Geerts 1996; Haentjens 1996; Knudson 1996; Cohn 1999). Other methods of thromboprophylaxis, such as anticoagulants (warfarin) or antiplatelets (aspirin), seem less practical for use in critically ill patients, because of their delayed action and oral presentation. Pentassacharides (a new class of synthetic selective factor Xa inhibitor, with parenteral presentation which does not bind to platelets, other cells or proteins) have been studied as prophylaxis in surgical orthopedic patients and have been shown to be as effective as UH and LMWH (Nijkeuter 2004).

How the intervention might work

Due to prolonged rest and coagulation abnormalities, trauma patients are at increased risk of thrombus formation. Thromboprohylaxis, either mechanical or pharmacological, may decrease the mortality and morbidity in trauma patients who survive beyond the first day in hospital, by decreasing the risk of DVT and PE in this population. A previous Cochrane review focusing on high-risk patients indicated that combined methods (pharmacologic and mechanical interventions) decreased the incidence of DVT (Kakkos 2008). However, this systematic review did not examine the effects in the subgroup of trauma patients.

Why it is important to do this review

Trauma patients are at an increased risk of VTE, and thromboprophylaxis has the potential to be effective in this population. However, trauma patients are at an increased risk of bleeding, which is one of the adverse events associated with pharmacological interventions. For some trauma patients with injured extremities, the use of mechanical interventions (e.g. external mechanical compression) is not feasible. A previous systematic review (Velmahos 2000) did not find evidence of effectiveness for either pharmacological or mechanical interventions. However, this systematic review was conducted 10 years ago and most of the included studies were of poor quality. Since then new trials have been conducted. Although current guidelines (Rogers 2002; Geerts 2008) recommend the use of thromboprophylaxis in trauma patients, there has not been a comprehensive and updated systematic review since the one published by Velmahos et al. Furthermore, there are still uncertainties about the relative benefit of interventions for different subgroups of trauma patients. Therefore it is necessary to conduct a systematic review to establish whether the effect of different thromboprophylaxis interventions varies according to the type of trauma, location of the trauma, severity of trauma and type of management (surgical or medical management).

Objectives

To assess the effects of thromboprophylaxis in trauma patients on mortality and incidence of DVT and PE.

To compare the effects of different thromboprophylaxis interventions and their effects according to the type of trauma.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled clinical trials.

Types of participants

People of any age with major trauma defined by one or more of the following criteria;

  • physiological: penetrating or blunt trauma with more than two organs and unstable vital signs,

  • anatomical: patients with an Injury Severity Score (ISS) higher than 9,

  • mechanism: patients who are involved in a 'high energy' event with a risk for severe injury despite stable or normal vital signs.

We excluded trials that only recruited outpatients, trials than recruited patients with hip fractures only, or patients with only acute spinal injuries.  

Types of interventions

We included trials investigating any of the following interventions;

  1. Unfractionated heparin (UH),

  2. Low weight molecular heparin (LWMH),

  3. Mechanical methods: graded compression stocking, and sequential compression devices,

  4. Oral anticoagulants (e.g. warfarin),

  5. Antiplatelet drugs (e.g. aspirin),

  6. Pentassacharides,

  7. Pulmonary embolism prophylaxis (e.g. inferior vena cava filter (IVCF)).

We compared the effects of any intervention with placebo, and any two interventions (e.g. LMWH versus UH) or combination of interventions (UH plus mechanical methods versus UH).

Types of outcome measures

Primary outcomes

The primary outcome was mortality.

Secondary outcomes

The secondary outcomes were the incidence of DVT, PE and adverse events, such as:

  • bleeding (major and minor);

  • whether the adverse event (bleeding of the injured site, intracranial bleeding, gastrointestinal bleeding, epistaxis, etc.) required transfusion or any procedure to control it;

  • and other adverse events as defined by the trial authors.

Search methods for identification of studies

We did not restrict searches by date, language or publication status.

Electronic searches

We searched the following electronic databases:

  • Cochrane Injuries Group’s Specialised Register (searched April 30 2009);

  • Cochrane Central Register of Controlled Trials 2009, issue 2 (The Cochrane Library);

  • MEDLINE (Ovid) 1950 to April (week 3) 2009;

  • EMBASE (Ovid) 1980 to (week 17) April 2009;

  • PubMed [http://www.ncbi.nlm.nih.gov/sites/entrez] (searched 29 April 2009);

  • ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 to April 2009);

  • ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 to April 2009).

Search strategies are listed in full in Appendix 1.

Searching other resources

We searched the bibliographies of all included studies and other relevant papers for further potentially eligible trials. We searched the Internet using the Google (www.google.com) search engine with selected terms from the strategy to identify any further unpublished or grey literature.

Data collection and analysis

The Injuries Group Trials Search Co-ordinator ran the electronic database searches, collated the search results, removed duplicates then sent the remaining records to the authors for screening

Selection of studies

Four authors, in pairs (LB and CM, EF and RC), independently examined titles, abstracts, and keywords of citations from electronic databases for eligibility. We obtained the full texts of all potentially relevant records and two authors (EF and CM) independently assessed whether each met the pre-defined inclusion criteria. We resolved any disagreement through discussion with a third author (PP).

Data extraction and management

Four authors, in pairs (LB and CM, EF and RC) extracted data independently, using a standardized data extraction form. LB entered the extracted information into Review Manager (RevMan 2008) for analysis. We extracted data on the following:

  1. General Information: title, authors, source of publication, country, published or not, language and year of publication.

  2. Trial characteristics: study design and information that meets the Cochrane Collaboration's tool for assessing risk of bias.

  3. Participants: sample size, inclusion criteria, exclusion criteria, location of trauma (brain, chest, abdomen, pelvis, extremity, polytrauma), severity of trauma (ISS, RTS, or according to the scale used by the trialists), type of injury (blunt or penetrating), and type of surgical procedure (non-operative or surgical management).

  4. Intervention: type and dose of thromboprophylaxis used, type and dose of control or placebo used.

  5. Outcomes: incidence of mortality, incidence of DVT (symptomatic or asymptomatic) and diagnostic test used, incidence of PE and diagnostic test used. Incidence of adverse events as follows: any bleeding, major bleeding defined as use of transfusion or any procedure to control bleeding (bleeding from the injured site, gastrointestinal bleeding, brain bleeding, epistaxis, etc.) and minor bleeding. Other outcomes recorded by the authors.

  6. Results: number of patients in each group, missing patients.

  7. Subgroup characteristics: number of patients by localization of trauma, by severity, by type (blunt or penetrating), by type of management (surgical or non-surgical).

  8. Other information: funding source.

Assessment of risk of bias in included studies

Four authors in pairs (LB and CM, EF and RC) assessed the risk of bias of each included trial using the Cochrane Collaboration's tool for assessing risk of bias presented in Higgins 2008. We assessed the following domains: sequence generation; allocation concealment; blinding; incomplete outcome data; and selective outcome reporting. We completed risk of bias tables for each trial, incorporating a description of the trial's performance against each domain and our overall judgment of the risk of bias for each entry as follows: 'Yes' for low risk of bias; 'No' for high risk of bias, or 'Unclear'. We resolved disagreements by consulting a third author (PP).

Measures of treatment effect

For dichotomous data we calculated risk ratios (RR) and 95% confidence intervals (CIs). We also calculated number needed to treat (NNT) and number needed to harm (NNH).

For continuous data we calculated the mean difference (MD) and 95% CIs when the same scale was used in a similar manner across studies. If results for continuous outcomes were reported using different scales or different versions of the same scale, we calculated the standardised mean difference (SMD) and 95% CIs.

Dealing with missing data

We did not contact the trial authors for missing information.

Assessment of heterogeneity

We examined trial characteristics in terms of participants, interventions and outcomes for evidence of clinical heterogeneity. We examined statistical heterogeneity by both the I2 statistic and Chi2 test. The I2 statistic describes the percentage of total variation across studies due to heterogeneity rather than chance. A value of 0% indicates no observed heterogeneity, and larger values show increasing heterogeneity; substantial heterogeneity is considered to exist when I2 is larger than 50%. For the Chi2 test, we used a p value of less than 0.10 to indicate the presence of statistically significant heterogeneity.

Assessment of reporting biases

We planned to investigate the presence of reporting (publication) bias using funnel plots, however there were too few included trials contributing data to each outcome to enable meaningful analysis.

Data synthesis

We judged that the trials were sufficiently homogenous, both clinically and statistically, to pool the outcome data. Dichotomous data were pooled using the Mantel-Haenszel fixed-effect method and continuous data were pooled using the fixed-effect inverse-variance method.

Because different effects were expected according to the intervention, we performed data synthesis separately for each type (e.g. UH, LWMH or mechanical devices).

Subgroup analysis and investigation of heterogeneity

There were insufficient data to perform the following planned subgroup analyses:

  • type of trauma (blunt, penetrating);

  • location of the trauma (brain, chest, abdominal, pelvis, extremity or polytrauma);

  • severity of trauma defined with ISS or other similar scores;

  • management (surgical or medical management);

  • diagnostic method.

Sensitivity analysis

We performed a sensitivity analysis to investigate whether the results were robust. We examined the effect of excluding certain studies according to their risk of bias. We reported the data synthesis for all the included studies and repeated the calculations after excluding studies judged as having a high risk of bias for allocation concealment. We also examined the effect of using a different effect measure (odds ratio) for the dichotomous outcomes.

Results

Description of studies

 

Results of the search

The combined search strategy identified 2858 citations, of which 38 were judged to be potentially eligible based on title or abstract, or both, and the full texts were obtained. After a full text review, 16 trials were judged to be eligible and were included in the review, one was reported as an abstract.

Included studies

We included 3005 people of any age who had major trauma. 1898 people had blunt trauma and 225 people had penetrating trauma. Four trials did not report the injury mechanism (Fuchs 2005; Ginzburg 2003; Knudson 1996; Yanar 2007). Most of the studies reported the ISS. The average of the ISS reported by the trials varies between 13 and 30. Five studies did not report the ISS value. Six studies reported a specific management. 816 with operative treatment and 173 with non operative treatment.

The Knudson 1994 All groups study was divided in four groups: Knudson 1994 All groups, Knudson 1994 group I (UH vs SCD vs Placebo), Knudson 1994 group II (UH vs Placebo) and Knudson 1994 group III (SCD vs Placebo).

Four trials compared a method of prophylaxis with no prophylaxis (Dennis 1993; Fisher 1995; Knudson 1994 All groups; Velmahos 2005).

Two trials compared pharmacological methods (LMWH with UH) (Cohn 1999; Geerts 1996).

Three trials compared mechanical methods (thigh calf sequential compression device with a foot-calf pump) (Anglen 1998; Elliot 1999; Stannard 2001).

Five trials compared pharmacological with mechanical methods, of which three compared LMWH with Intermittent Pneumatic Compression (IPC) devices (Ginzburg 2003; Kurtoglu 2004; Yanar 2007) and two compared UH with Sequential Compression Devices (SCD) (Knudson 1992; Knudson 1994 group I).

Four trials compared a combination of pharmacological and mechanical interventions. One trial allocated patients to three groups to compare IPC, IPC plus LMWH and LMWH alone (Yanar 2007). The other three trials compared pharmacological plus mechanical methods with pharmacological prophylactics (Fuchs 2005; Stannard 2006; Yanar 2007).      

All 16 trials used doppler ultrasound (Duplex) to diagnose DVT. The diagnosis was confirmed with venography in three trials (Geerts 1996; Fuchs 2005; Fisher 1995), and by Duplex and MRI venography in one trial (Stannard 2001).

Sixteen trials presented data on pulmonary embolism. PE was diagnosed clinically in two trials (Anglen 1998; Fuchs 2005); by a V/Q scan and pulmonary angiography in five trials (Knudson 1992; Knudson 1994 All groups; Knudson 1996; Geerts 1996; Cohn 1999); by CT scan angiography on clinically suspected cases in two trials (Kurtoglu 2004; Yanar 2007); by clinical suspicion and autopsy in one trial (Elliot 1999); by angiography and autopsy in two trials (Fisher 1995; Dennis 1993); by used ventilation/perfusion scan (V/Q scan), angiography and CT scan angiography in one trial (Ginzburg 2003); and by MRI angiography in one trial (Stannard 2006). The method used in the remaining two trials was not described (Velmahos 2005; Stannard 2001).

Further details of the individual trials are presented in the Characteristics of included studies table.

Excluded studies

Of the trials excluded from our review three were not randomised (Greenfiel 1997; Holzheimer 2004; Reilmann 1986), two included only outpatients or elective surgery patients (Haas 2003; Wolf 1992), one involved hip fracture surgery patients (Breyer 1986) and one did not measure any outcomes of interest to this review (Murakami 2003).

Details are presented in the Characteristics of excluded studies.

Risk of bias in included studies

Allocation

Eleven of the 16 studies had a low risk in the sequence generation of the treatment groups, three of them had high risk of bias and two were unclear.

In half of the studies included the allocation concealment was unclear, four of the 16 had low risk in the allocation and the other four studies had high risk of bias.

Blinding

Blinding was well conducted in seven of the 16 studies, in five there was a high risk of bias and in four the data were insufficient to establish the quality of blinding.

Incomplete outcome data

Six studies performed an intention to treat analysis, in nine there was a high risk bias because outcome data were analysed in an incomplete fashion, and just one of them did not allow us to establish the risk.

Selective reporting

We were unable to obtain the protocols for any of the trials, therefore the reporting bias was unclear for all 16 trials.

Effects of interventions

All the collected material allowed us to perform 7 comparisons, we made four comparisons as a sensitivity analysis. As established in the protocol we compared any method of prophylaxis versus no prophylaxis, between prophylaxis methods and combinations of them. The most important are highlighted in this text section. The detailed analysis can be seen in the Data and analyses section.

Prophylaxis versus no prophylaxis

Four trials involving 997 people compared the effect of any type (mechanical and/or pharmacological) of prophylaxis versus no prophylaxis. Prophylaxis reduced the risk of DVT in trauma patients (RR 0.52; 95% CI 0.32 to 0.84). There was no evidence of statistical heterogeneity between trials (I2=12%; Chi2 P=0.333). There was no evidence for an effect on PE (RR 0.65; 95% CI 0.29 to 1.43) or mortality (RR 0.59; 95% CI 0.20 to 1.70).

Three trials reported the effect on bleeding, no events were observed in any trial.

Mechanical prophylaxis versus no prophylaxis

Six trials involving 811 people compared the effect of mechanical prophylaxis with no prophylaxis.

Mechanical prophylaxis reduced the risk of DVT (RR 0.55; 95% CI 0.34 to 0.90). There was no evidence of statistical heterogeneity between trials (I2 = 22%, Chi P=0.28). There was no evidence that mechanical prophylaxis reduced the risk of PE (RR 0.77; 95% CI 0.36 to 1.66) or death (RR 0.74; 95% CI 0.27 to 2.04).

Four trials (507 patients) reported the effect on bleeding, no events were observed in any trial.

Pharmacological prophylaxis versus mechanical prophylaxis

Six trials involving 1033 people compared pharmacological prophylaxis with mechanical prophylaxis.

Pharmacological prophylaxis was more effective than mechanical methods at reducing the risk of DVT (RR 0.48; 95% CI 0.25 to 0.95). There was no evidence of heterogeneity between trials (I2=0%, Chi P=0.56). There was no evidence for a difference in effect on the risk of PE (RR 0.94; 95% CI 0.36 to 2.42) or death (RR 1.50; 95% CI 0.44 to 5.16).

Five of the trials (953 patients) reported bleeding outcome data. Pharmacological prophylaxis increased the risk of bleeding (RR 2.04; 95% CI 1.08 to 3.86) compared to mechanical methods. There was no evidence of heterogeneity (I2=0%, Chi P=0.53). Three trials (764 patients) distinguished between major and minor bleeding. There was no evidence for a difference in effect on the risk of major bleeding (RR 1.03; 95% CI 0.26 to 4.06). However, pharmacological prophylaxis increased the risk of minor bleeding (RR 2.37; 95% CI 1.13 to 4.98). There was no evidence for heterogeneity between trials (I2=0%; Chi P=0.64).

Low Molecular Weight Heparin versus Unfractionated Heparin

Two trials involving 331 patients compared low molecular weight heparin (LMWH) with unfractionated heparin (UH). LMWH appeared to reduce the risk of DVT compared to UH (RR 0.68; 95% CI 0.50 to 0.94). There was no evidence for heterogeneity between trials (I2 = 0%, Chi P=0.46).

There was no statistically significant difference in the risk of PE between the two groups (RR 3.16; 95% CI 0.13 to 76.91) and there were no deaths reported in either trial.

There was no statistically significant difference in the risk of bleeding between LMWH and UH (RR 1.63; 95% CI 0.63 to 4.22).

Mechanical plus pharmacological prophylaxis versus Pharmacological prophylaxis

Three trials involving 507 patients compared mechanical prophylaxis plus pharmacological prophylaxis with pharmacological prophylaxis alone. Patients who received both mechanical and pharmacological prophylaxis had a lower risk of DVT (RR 0.34; 95% CI 0.19 to 0.60). However, there was evidence for statistical heterogeneity between trials (I2= 69%, Chi P=0.04).

There was no evidence for a difference in effect on the risk of PE (RR 0.32; 95% CI 0.05 to 2.01) or death (RR 0.50; 95% CI 0.05 to 5.30).

One trial assessed the effect on bleeding and found no difference in the risk of bleeding between the two groups (RR 0.99; 95% CI 0.56 to 1.78).

Other comparisons

One study compared mechanical plus pharmacological prophylaxis versus mechanical prophylaxis (Yanar 2007) and did not find difference. Also three trials compared thigh-calf versus calf-foot methods (Anglen 1998; Elliot 1999; Stannard 2001) did not find either statistical difference.

Sensitivity Analyses

The sensitivity analysis included only the studies that were considered at low risk of bias for allocation concealment. We made four comparisons of the results, which showed a statistically significant difference.

The comparisons of prophylaxis vs no prophylaxis and mechanical methods vs no prophylaxis, could include only one study (Fisher 1995). It showed a tendency to be superior for prophylaxis, but without reaching a statistically significant difference (RR 0.35; 95% CI 0.11 to 1.10).

When we compare LMWH vs UH, we could include just one study (Geerts 1996) which showed strong evidence that LMWH was superior for prevention of DVT (RR 0.57; 95% CI 0.34 to 0.94).

When we compare mechanical with pharmacological prophylaxis we also could include just one study (Ginzburg 2003) but did not show any significant difference (RR 0.17; 95% CI 0.02 to 1.41).

Discussion

 

Summary of main results

This systematic review gathers all the available evidence from all randomised controlled trials which compare the use of pharmacological and mechanical thromboprophylaxis  in patients with severe trauma. We excluded trials that only recruited outpatients, patients presenting only with hip fractures, acute spinal injuries and low energy trauma. 

Among trauma patients who did not receive any prophylaxis we found that deep venous thrombosis (DVT) incidence was 8.72% (37/424) diagnosed by Dupplex. The incidence of pulmonary embolism (PE) was 3.3% (14/424) for PE diagnosed by VQ scan or angiography or autopsy (Dennis 1993; Fisher 1995; Knudson 1994 All groups; Velmahos 2005).

We did not find evidence that thromboprophylaxis reduces the primary outcome mortality or the secondary outcome PE for any of the comparisons assessed.

However, we found some evidence of effective interventions for the prevention of the secondary outcome DVT. Prophylaxis was more effective than no prophylaxis, pharmacological prophylaxis than mechanical prophylaxis, and LMWH than UH. However, these results were based on a few small trials with relatively few events and poor methodology quality. We also found some evidence that patients who received pharmacological thromboprophylaxis have a higher risk of minor bleeding compared to patients who received mechanical therapy. At the moment there is no RCT published for major trauma with dabigatran or rivaroxiban.

Although the strength of the evidence found in this review was not strong our findings are similar to previous reviews conducted in different but related conditions (Surgical non traumatized, cancer, hip fractures) (Wille-Jørgensen 2008; Bani-Hani 2011; Handoll 2008). The present review strengthens the clinical practice  guidelines recommendations from Eastern Association for the Surgery of Trauma (Rogers 2002) and American College of Chest Physicians (Kahn 2012; Geerts 2008), providing further evidence for thromboprophylaxis in patients with severe trauma.

Overall completeness and applicability of evidence

We did not find enough studies to allow any comparison between pharmacological prophylaxis against placebo. There are also not enough studies which evaluate specific pharmacological interventions (UW or LMWH) vs specific mechanical therapies (SCD or AVI).

Quality of the evidence

The quality of evidence was low as only one of the four studies included for the main comparison (prophylaxis versus no prophylaxis) had low risk of selection bias as judged by the allocation concealment process. Also the external validity could be threatened because of sponsorship by mechanical devices manufacturers (Velmahos 2005).

Potential biases in the review process

Potential biases in the review process are mainly defined by an impossibility to analyse a study due to an abstract inclusion in the meta-analysis (Yanar 2007). We tried to minimize selection bias by working, collecting and analysing all studies in pairs (LB-CM and EF-RC). All differences where resolved by consensus and also through the input of PP.

Agreements and disagreements with other studies or reviews

Previous meta-analyses (Velmahos 2000) have shown no evidence that low-dose unfractionated heparin, mechanical prophylaxis, or low-molecular weight heparin are more effective than no prophylaxis or among each other. In addition, unlike ours, this systematic review included observational studies which provide less reliable evidence of effectiveness for medical interventions.

A recent systematic review (Smith 2011) with severe skeletal trauma as its main inclusion criteria suggested that low molecular weight heparin (LMWH) may be superior to low dose heparin (LDH), and that LMWH should be used in addition to mechanical prophylaxis measures in patients following major skeletal trauma for the prevention of thromboembolic events, this findings do not differ from ours. The analysed studies were included in this meta-analysis, the difference between these systematic reviews were that they just include major skeletal trauma.

Bleeding was not analysed as an outcome in previous meta-analysis and not all of the trials considered bleeding as an outcome.

Authors' conclusions

Implications for practice

Although the strength of the evidence included in this review was not high, taking into account existing information from other related conditions such as surgery, we believe that the use thromboprophylaxis for preventing DVT in severe trauma patients is recommended.

Implications for research

Adequately powered trials should be conducted to compare different thromboprophylaxis strategies. Also we would recommend that future RCTs should have a more uniform and defined method of diagnosing both DVT and PE so as to measure the true effect thromboprophylaxis.

In addition to mortality, DVT and PE, future trials should assess the following outcomes which have important clinical implications:

1. Adverse events of pharmacological prophylaxis such as bleeding.

2. Post-discharge consequences of DVT in postraumatic patients including postphlebitic syndrome, chronic DVT or chronic PE.

Acknowledgements

We thank Karen Blackhall for designing and performing the bibliographic database searches.

Data and analyses

Download statistical data

Comparison 1. Prophylaxis vs No prophylaxis
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Deep Venous Thrombosis (DVT)4997Risk Ratio (M-H, Fixed, 95% CI)0.52 [0.32, 0.84]
2 Pulmonary Embolism (PE)4997Risk Ratio (M-H, Fixed, 95% CI)0.65 [0.29, 1.43]
3 Mortality4997Risk Ratio (M-H, Fixed, 95% CI)0.59 [0.20, 1.70]
4 Bleeding3553Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 1.1.

Comparison 1 Prophylaxis vs No prophylaxis, Outcome 1 Deep Venous Thrombosis (DVT).

Analysis 1.2.

Comparison 1 Prophylaxis vs No prophylaxis, Outcome 2 Pulmonary Embolism (PE).

Analysis 1.3.

Comparison 1 Prophylaxis vs No prophylaxis, Outcome 3 Mortality.

Analysis 1.4.

Comparison 1 Prophylaxis vs No prophylaxis, Outcome 4 Bleeding.

Comparison 2. Mechanical prophylaxis vs No prophylaxis
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT5907Risk Ratio (M-H, Fixed, 95% CI)0.55 [0.34, 0.90]
2 PE5907Risk Ratio (M-H, Fixed, 95% CI)0.77 [0.36, 1.66]
3 Mortality5907Risk Ratio (M-H, Fixed, 95% CI)0.74 [0.27, 2.04]
4 Bleeding4603Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
Analysis 2.1.

Comparison 2 Mechanical prophylaxis vs No prophylaxis, Outcome 1 DVT.

Analysis 2.2.

Comparison 2 Mechanical prophylaxis vs No prophylaxis, Outcome 2 PE.

Analysis 2.3.

Comparison 2 Mechanical prophylaxis vs No prophylaxis, Outcome 3 Mortality.

Analysis 2.4.

Comparison 2 Mechanical prophylaxis vs No prophylaxis, Outcome 4 Bleeding.

Comparison 3. Pharmacological prophylaxis vs Mechanical prophylaxis
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT61033Risk Ratio (M-H, Fixed, 95% CI)0.48 [0.25, 0.95]
2 PE61033Risk Ratio (M-H, Fixed, 95% CI)0.94 [0.36, 2.42]
3 Mortality61033Risk Ratio (M-H, Fixed, 95% CI)1.5 [0.44, 5.16]
4 Bleeding5953Risk Ratio (M-H, Fixed, 95% CI)2.04 [1.08, 3.86]
5 Major bleeding3764Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.26, 4.06]
6 Minor bleeding3764Risk Ratio (M-H, Fixed, 95% CI)2.37 [1.13, 4.98]
Analysis 3.1.

Comparison 3 Pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 1 DVT.

Analysis 3.2.

Comparison 3 Pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 2 PE.

Analysis 3.3.

Comparison 3 Pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 3 Mortality.

Analysis 3.4.

Comparison 3 Pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 4 Bleeding.

Analysis 3.5.

Comparison 3 Pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 5 Major bleeding.

Analysis 3.6.

Comparison 3 Pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 6 Minor bleeding.

Comparison 4. LWMH vs UH
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT2331Risk Ratio (M-H, Fixed, 95% CI)0.68 [0.50, 0.94]
2 PE2331Risk Ratio (M-H, Fixed, 95% CI)3.16 [0.13, 76.91]
3 Mortality2331Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Bleeding2331Risk Ratio (M-H, Fixed, 95% CI)1.63 [0.63, 4.22]
5 Major bleeding2331Risk Ratio (M-H, Fixed, 95% CI)1.46 [0.55, 3.85]
6 Minor bleeding2331Risk Ratio (M-H, Fixed, 95% CI)3.16 [0.13, 76.91]
Analysis 4.1.

Comparison 4 LWMH vs UH, Outcome 1 DVT.

Analysis 4.2.

Comparison 4 LWMH vs UH, Outcome 2 PE.

Analysis 4.3.

Comparison 4 LWMH vs UH, Outcome 3 Mortality.

Analysis 4.4.

Comparison 4 LWMH vs UH, Outcome 4 Bleeding.

Analysis 4.5.

Comparison 4 LWMH vs UH, Outcome 5 Major bleeding.

Analysis 4.6.

Comparison 4 LWMH vs UH, Outcome 6 Minor bleeding.

Comparison 5. Mechanical + pharmacological prophylaxis vs Pharmacological
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT3507Risk Ratio (M-H, Fixed, 95% CI)0.34 [0.19, 0.60]
2 PE3507Risk Ratio (M-H, Fixed, 95% CI)0.32 [0.05, 2.01]
3 Mortality3507Risk Ratio (M-H, Fixed, 95% CI)0.5 [0.05, 5.30]
4 Bleeding1200Risk Ratio (M-H, Fixed, 95% CI)0.99 [0.56, 1.78]
Analysis 5.1.

Comparison 5 Mechanical + pharmacological prophylaxis vs Pharmacological, Outcome 1 DVT.

Analysis 5.2.

Comparison 5 Mechanical + pharmacological prophylaxis vs Pharmacological, Outcome 2 PE.

Analysis 5.3.

Comparison 5 Mechanical + pharmacological prophylaxis vs Pharmacological, Outcome 3 Mortality.

Analysis 5.4.

Comparison 5 Mechanical + pharmacological prophylaxis vs Pharmacological, Outcome 4 Bleeding.

Comparison 6. Mechanical + pharmacological prophylaxis vs Mechanical prophylaxis
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT180Risk Ratio (M-H, Fixed, 95% CI)0.5 [0.10, 2.58]
2 PE180Risk Ratio (M-H, Fixed, 95% CI)0.5 [0.05, 5.30]
3 Mortality180Risk Ratio (M-H, Fixed, 95% CI)0.5 [0.05, 5.30]
Analysis 6.1.

Comparison 6 Mechanical + pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 1 DVT.

Analysis 6.2.

Comparison 6 Mechanical + pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 2 PE.

Analysis 6.3.

Comparison 6 Mechanical + pharmacological prophylaxis vs Mechanical prophylaxis, Outcome 3 Mortality.

Comparison 7. Thigh-calf vs Calf-foot
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT3373Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.37, 1.32]
2 PE3373Risk Ratio (M-H, Fixed, 95% CI)0.79 [0.12, 5.10]
3 Mortality3373Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Bleeding3373Risk Ratio (M-H, Fixed, 95% CI)3.04 [0.13, 73.44]
Analysis 7.1.

Comparison 7 Thigh-calf vs Calf-foot, Outcome 1 DVT.

Analysis 7.2.

Comparison 7 Thigh-calf vs Calf-foot, Outcome 2 PE.

Analysis 7.3.

Comparison 7 Thigh-calf vs Calf-foot, Outcome 3 Mortality.

Analysis 7.4.

Comparison 7 Thigh-calf vs Calf-foot, Outcome 4 Bleeding.

Comparison 8. Prophylaxis vs No prophylaxis (sensitivity analysis)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT1304Risk Ratio (M-H, Fixed, 95% CI)0.37 [0.12, 1.11]
2 PE1304Risk Ratio (M-H, Fixed, 95% CI)0.73 [0.27, 2.00]
Analysis 8.1.

Comparison 8 Prophylaxis vs No prophylaxis (sensitivity analysis), Outcome 1 DVT.

Analysis 8.2.

Comparison 8 Prophylaxis vs No prophylaxis (sensitivity analysis), Outcome 2 PE.

Comparison 9. Mechanical prophylaxis vs No prophylaxis (sensitivity analysis)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT1304Risk Ratio (M-H, Fixed, 95% CI)0.37 [0.12, 1.11]
2 PE1304Risk Ratio (M-H, Fixed, 95% CI)0.73 [0.27, 2.00]
Analysis 9.1.

Comparison 9 Mechanical prophylaxis vs No prophylaxis (sensitivity analysis), Outcome 1 DVT.

Analysis 9.2.

Comparison 9 Mechanical prophylaxis vs No prophylaxis (sensitivity analysis), Outcome 2 PE.

Comparison 10. LMWH vs UH (Sensitivity analysis)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT1265Risk Ratio (M-H, Fixed, 95% CI)0.70 [0.51, 0.97]
2 PE1265Risk Ratio (M-H, Fixed, 95% CI)3.16 [0.13, 76.91]
3 Mortality1265Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Bleeding1265Risk Ratio (M-H, Fixed, 95% CI)5.27 [0.62, 44.51]
5 Major Bleeding1265Risk Ratio (M-H, Fixed, 95% CI)4.22 [0.48, 37.23]
6 Minor Bleeding1265Risk Ratio (M-H, Fixed, 95% CI)3.16 [0.13, 76.91]
Analysis 10.1.

Comparison 10 LMWH vs UH (Sensitivity analysis), Outcome 1 DVT.

Analysis 10.2.

Comparison 10 LMWH vs UH (Sensitivity analysis), Outcome 2 PE.

Analysis 10.3.

Comparison 10 LMWH vs UH (Sensitivity analysis), Outcome 3 Mortality.

Analysis 10.4.

Comparison 10 LMWH vs UH (Sensitivity analysis), Outcome 4 Bleeding.

Analysis 10.5.

Comparison 10 LMWH vs UH (Sensitivity analysis), Outcome 5 Major Bleeding.

Analysis 10.6.

Comparison 10 LMWH vs UH (Sensitivity analysis), Outcome 6 Minor Bleeding.

Comparison 11. Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 DVT1442Risk Ratio (M-H, Fixed, 95% CI)0.17 [0.02, 1.41]
2 PE1442Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.06, 16.32]
3 Mortality1442Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Bleeding1442Risk Ratio (M-H, Fixed, 95% CI)1.67 [0.71, 3.95]
5 Major bleeding1442Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.26, 4.06]
6 Minor bleeding1442Risk Ratio (M-H, Fixed, 95% CI)2.31 [0.72, 7.40]
Analysis 11.1.

Comparison 11 Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis), Outcome 1 DVT.

Analysis 11.2.

Comparison 11 Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis), Outcome 2 PE.

Analysis 11.3.

Comparison 11 Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis), Outcome 3 Mortality.

Analysis 11.4.

Comparison 11 Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis), Outcome 4 Bleeding.

Analysis 11.5.

Comparison 11 Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis), Outcome 5 Major bleeding.

Analysis 11.6.

Comparison 11 Pharmacological prophylaxis vs Mechanical prophylaxis (Sensitivity analysis), Outcome 6 Minor bleeding.

Appendices

Appendix 1. Search strategy

Cochrane Injuries Group Specialised Register (searched April 30 2009)
1. (wound* or trauma* or injur* or fracture* or burn* or stab* or shot* or shoot* or lacerat* or accident*)
2. (thrombus* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or thromboph*) or ((deep* and (vein* or ven*) and (thromb* or embol*)) or ((pulmonary or lung*) and (thromb* or embol*)) or (DVT or PE or VTE)
3. 1 and 2
4. Thromboprophylaxis or prophylactic* or prophylaxis or Heparin* or Anticoagulant* or Warfarin or Coumadin* or apo-warfarin or gen-warfarin or warfant or Coumadin or aldocumar or tedicumar or Antiplatelet* or anticoagulant* or Aspirin* or "acetylsalicylic acid" or acylpyrin or aloxiprimum or colfarit or dispril or easprin or ecotrin or endosprin or magnecyl or micristin or polopirin or polopiryna or solprin or solupsan or zorprin or acetysal or Pentassacharide* or fondaparinux or Heparin* or "vena cava filter" or "umbrella filter" anti-platelet*
5. 3 and 4

MEDLINE (Ovid) 1950 to April (Week 3) 2009
1. exp "Wounds and Injuries"/
2. (wound* or trauma* or injur* or fracture* or burn* or stab* or shot* or shoot* or lacerat* or accident*).mp. [mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer name]
3. 1 or 2
4. exp Venous Thromboembolism/
5. exp Venous Thrombosis/
6. exp Pulmonary embolism/
7. exp Thrombophlebitis/
8. (thrombus* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or thromboph*).ab,ti.
9. (deep* adj3 (vein* or ven*) adj5 (thromb* or embol*)).ab,ti.
10. ((pulmonary or lung*) adj3 (thromb* or embol*)).ab,ti.
11. (DVT or PE or VTE).ab,ti.
12. 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11
13. (thromboprophylaxis or prophylactic* or prophylaxis).ab,ti.
14. exp Heparin/
15. exp Heparin, Low-Molecular-Weight/
16. exp Heparinoids/
17. exp Stockings, Compression/
18. exp Intermittent Pneumatic Compression Devices/
19. exp Stockings, Compression/
20. exp Anticoagulants/
21. exp Warfarin/
22. exp Platelet Aggregation Inhibitors/
23. exp Aspirin/
24. Heparin*.ab,ti.
25. ((compression or impulse or pneumatic or elastic*) adj3 (device* or stocking* or hose* or dressing* or bandage*)).ab,ti.
26. (Anticoagulant* or Warfarin or Coumadin* or apo-warfarin or gen-warfarin or warfant or Coumadin or aldocumar or tedicumar).ab,ti.
27. (Antiplatelet* or (platelet* adj3 aggregation adj3 inhibit*) or ((blood or platelet*) adj3 (antagonist* or antiaggrega*))).ab,ti.
28. (Aspirin* or acetylsalicylic acid or acylpyrin or aloxiprimum or colfarit or dispril or easprin or ecotrin or endosprin or magnecyl or micristin or polopirin or polopiryna or solprin or solupsan or zorprin or acetysal).ab,ti.
29. exp Vena Cava Filters/
30. ((vena adj3 cava adj3 filter*) or (umbrella adj3 filter*)).ab,ti.
31. (Pentassacharide* or fondaparinux).ab,ti.
32. or/13-31
33. randomi?ed.ab,ti.
34. randomized controlled trial.pt.
35. controlled clinical trial.pt.
36. placebo.ab.
37. clinical trials as topic.sh.
38. randomly.ab.
39. trial.ti.
40. 33 or 34 or 35 or 36 or 37 or 38 or 39
41. (animals not (humans and animals)).sh.
42. 40 not 41
43. 3 and 12 and 32 and 42

EMBASE (Ovid) 1980 to 2009 (Week 17)
1.exp Injury/
2.(wound* or trauma* or injur* or fracture* or burn* or stab* or shot* or shoot* or lacerat* or accident*).mp. [mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer name]
3.1 or 2
4.exp Vein Thrombosis/
5.exp Lung Embolism/
6.exp venous thromboembolism/
7.exp deep vein thrombosis/
8.exp Thrombophlebitis/
9.(thrombus* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or thromboph*).ab,ti.
10.(deep* adj3 (vein* or ven*) adj5 (thromb* or embol*)).ab,ti.
11.(deep* adj3 (vein* or ven*) adj5 (thromb* or embol*)).ab,ti.
12.((pulmonary or lung*) adj3 (thromb* or embol*)).ab,ti.
13.(DVT or PE or VTE).ab,ti.
14.or/4-13
15.(thromboprophylaxis or prophylactic* or prophylaxis).ab,ti.
16.exp Heparin/
17.exp Low Molecular Weight Heparin/
18.exp Heparinoid/
19.exp compression garment/
20.exp Compression Bandage/
21.exp intermittent pneumatic compression device/
22.exp Vena Cava Filter/
23.exp Anticoagulant Agent/
24.exp Coumarin Anticoagulant/
25.exp Antithrombocytic Agent/
26.exp Acetylsalicylic Acid/
27.Heparin*.ab,ti.
28.(Anticoagulant* or Warfarin or Coumadin* or apo-warfarin or gen-warfarin or warfant or Coumadin or aldocumar or tedicumar).ab,ti.
29.(Antiplatelet* or (platelet* adj3 aggregation adj3 inhibit*) or ((blood or platelet*) adj3 (antagonist* or antiaggrega*))).ab,ti.
30.(Aspirin* or acetylsalicylic acid or acylpyrin or aloxiprimum or colfarit or dispril or easprin or ecotrin or endosprin or magnecyl or micristin or polopirin or polopiryna or solprin or solupsan or zorprin or acetysal).ab,ti.
31.((vena adj3 cava adj3 filter*) or (umbrella adj3 filter*)).ab,ti.
32.(Pentassacharide* or fondaparinux).ab,ti.
33.((compression or impulse or pneumatic or elastic*) adj3 (device* or stocking* or hose* or dressing* or bandage*)).ab,ti.
34.or/15-33
35.exp Randomized Controlled Trial/
36.exp controlled clinical trial/
37.randomi?ed.ab,ti.
38.placebo.ab.
39.*Clinical Trial/
40.randomly.ab.
41.trial.ti.
42.35 or 36 or 37 or 38 or 39 or 40 or 41
43.exp animal/ not (exp human/ and exp animal/)
44.42 not 43
45.34 and 3 and 44 and 14

CENTRAL (The Cochrane Library 2009, Issue 2)
#1 MeSH descriptor Venous Thromboembolism explode all trees
#2 MeSH descriptor Venous Thrombosis explode all trees
#3 MeSH descriptor Pulmonary Embolism explode all trees
#4 MeSH descriptor Thrombophlebitis explode all trees
#5 (thrombus* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or thromboph*)
#6 ((deep*) near3 (vein* or ven*)) near5 (thromb* or embol*)
#7 (pulmonary or lung*) near3 (thromb* or embol*)
#8 DVT or PE or VTE
#9 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8)
#10 thromboprophylaxis or prophylactic* or prophylaxis
#11 MeSH descriptor Heparin explode all trees
#12 MeSH descriptor Heparin, Low-Molecular-Weight explode all trees
#13 MeSH descriptor Heparinoids explode all trees
#14 MeSH descriptor Stockings, Compression explode all trees
#15 MeSH descriptor Intermittent Pneumatic Compression Devices explode all trees
#16 MeSH descriptor Stockings, Compression explode all trees
#17 MeSH descriptor Anticoagulants explode all trees
#18 MeSH descriptor Warfarin explode all trees
#19 MeSH descriptor Platelet Aggregation Inhibitors explode all trees
#20 MeSH descriptor Aspirin explode all trees
#21 Heparin*
#22 ((compression or impulse or pneumatic or elastic*) near3 (device* or stocking* or hose* or dressing* or bandage*))
#23 (Anticoagulant* or Warfarin or Coumadin* or apo-warfarin or gen-warfarin or warfant or Coumadin or aldocumar or tedicumar)
#24 Antiplatelet* or (platelet* near3 aggregation near3 inhibit*)
#25 (blood or platelet*) near3 (antagonist* or antiaggrega*)
#26 Aspirin* or acetylsalicylic acid or acylpyrin or aloxiprimum or colfarit or dispril or easprin or ecotrin or endosprin or magnecyl or micristin or polopirin or polopiryna or solprin or solupsan or zorprin or acetysal
#27 MeSH descriptor Vena Cava Filters explode all trees
#28 (vena near3 cava near3 filter*) or (umbrella near3 filter*)
#29 Pentassacharide* or fondaparinux
#30 (#10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29)
#31 (#9 AND #30)
#32 MeSH descriptor Wounds and Injuries explode all trees
#33 wound* or trauma* or injur* or fracture* or burn* or stab* or shot* or shoot* or lacerat* or accident*
#34 (#32 OR #33)
#35 (#31 AND #34)

ISI Web of Science: Science Citation Index Expanded (SCI-EXPANDED) (1970 to April 2009),
ISI Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (1990 to April 2009)

#1 Topic=(wound* or trauma* or injur* or fracture* or burn* or stab* or shot* or shoot* or lacerat* or accident*)
#2 Topic=(thrombus* or thrombotic* or thrombolic* or thromboemboli* or thrombos* or thromboph*) OR Topic=(deep* same (vein* or ven*) same (thromb* or embol*)) OR Topic=((pulmonary or lung*) same (thromb* or embol*)) OR Topic=(DVT or PE or VTE)
#3 #1 and #2
#4 Topic=(thromboprophylaxis or prophylactic* or prophylaxis or Heparin* or Anticoagulant* or Warfarin or Coumadin* or apo-warfarin or gen-warfarin or warfant or Coumadin or aldocumar or tedicumar or Antiplatelet* or Aspirin* or acetylsalicylic acid or acylpyrin or aloxiprimum or colfarit or dispril or easprin or ecotrin or endosprin or magnecyl or micristin or polopirin or polopiryna or solprin or solupsan or zorprin or acetysal or Pentassacharide* or fondaparinux)
#5 Topic=(platelet* same aggregation same inhibit*) OR Topic=((blood or platelet*) same (antagonist* or antiaggrega*)) OR Topic=((compression or impulse or pneumatic or elastic*) same (device* or stocking* or hose* or dressing* or bandage*)) OR Topic=(vena same cava same filter*) OR Topic=(umbrella same filter*)
#6 #4 or #5
#7 #3 and #6
#8 Topic=(((singl* OR doubl* OR trebl* OR tripl*) SAME (blind* OR mask*))) OR Topic=(randomised OR randomized OR randomly OR random order OR random sequence OR random allocation OR randomly allocated OR at random OR randomized controlled trial) OR Topic=(controlled clinical trial OR controlled trial OR clinical trial OR placebo) AND Topic=(human*)
#9 #7 and #8

PubMed [http://www.ncbi.nlm.nih.gov/sites/entrez] (searched 29 April 2009: limited to: added to PubMed in the Last 180 days)
#1 wound* OR trauma OR traumatic OR traumas OR traumatolog* OR injur* OR fracture* OR burn* OR stab OR stabbing* OR stabbed OR stabwound* OR stabs OR shot* OR shoot* OR lacerat* OR accident*
#2 thrombus* OR thrombotic* OR thrombolic* OR thromboemboli* OR thrombos* OR thromboph*
#3 DVT OR PE OR VTE
#4 deep* AND (vein* OR venous) AND (thrombos OR thromboe* OR emboli*)
#5 #2 OR #3 OR #4
#6 (hromboprophylaxis OR prophylactic* OR prophylaxis OR Heparin* OR Anticoagulant* OR Warfarin OR Coumadin* OR apo-warfarin OR gen-warfarin OR warfant OR Coumadin OR aldocumar OR tedicumar OR Antiplatelet* OR Aspirin* OR acetylsalicylic acid OR acylpyrin OR aloxiprimum OR colfarit OR dispril OR easprin OR ecotrin OR endosprin OR magnecyl OR micristin OR polopirin OR polopiryna OR solprin OR solupsan OR zorprin OR acetysal OR Pentassacharide* OR fondaparinux
#7 platelet* AND aggregation AND inhibitor*
#8 (blood or platelet*) AND (antagonist* or antiaggrega*)
#9 (compression OR impulse OR pneumatic OR elastic*) AND (device* OR stocking* OR hose* OR dressing* OR bandage*)
#10 (vena AND cava AND filter*) OR (umbrella AND filter*)
#11 #6 OR #7 OR #8 OR #9 OR #10
#12 #1 AND #5 AND #11
#13 ((randomized controlled trial[pt] OR controlled clinical trial[pt]) OR (randomized OR randomised OR randomly OR placebo[tiab]) OR (trial[ti]) OR ("Clinical Trials as Topic"[MeSH Major Topic])) NOT (("Animals"[Mesh]) NOT ("Humans"[Mesh] AND "Animals"[Mesh]))
#14 #12 AND #13

Contributions of authors

All authors contributed to this manuscript.

Declarations of interest

None known.

Differences between protocol and review

On the advice of the Cochrane Injuries Group's editors we have specified mortality as the primary outcome of the review. DVT and PE are listed as secondary outcomes.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Anglen 1998

MethodsRandomized control trial
Participants

117 participants randomised: 49 for sequential compression device (SCD) and 68 for Plantar venous pneumatic compression (P ).

Gender: SCD M 38 F 30 P M 27 F 22

Mean age: SCD 38 (17-82) P 41 (18-83).

Inclusion Criteria All Adult trauma with fracture of the pelvic ring, acetabulum or femur

Exclusion Criteria: Inabilty of give informed consent, preexisting thrombosis, active anticoagulation,inability to use the devices

Type of Injury All were blunt trauma

Location of trauma: Spine: Not reported Head: Not reported Face: Not reported Chest: Not reported Abdomen: Not reported Pelvis: SCD 26 P 24 Extremity: 36 to 23 PolItrauma (more than one): 42 to 30

Severity of trauma (ISS, RTS) not reported:

Type of management (Operatory, Non operatory, both or not reported): Operatory

Interventions

thigh-calf sequential-compression device (Kendall SCD; Kendall, Mansfield, Massachusetts)

PlexiPulse foot pumps (Kinetic Concepts, San Antonio, Texas)

Outcomes

Diagnostic method for DVT: Doppler ultrasonography

Diagnostic method for PE: Clinically

Any bleeding? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High riskNo (Medical record number).
Allocation concealment (selection bias)High riskNo (Medical record number).
Blinding (performance bias and detection bias)
All outcomes
Low riskThe scans were read by radiology in a blinded fashion.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Cohn 1999

MethodsRandomized control trial
Participants

66 participants randomised (32 for low dose unfractionated heparin (UH) and34 for LMWH)

Gender: Not reported

Mean age: 44±17 years for UH to 38±15 years for LMWH

Inclusion Criteria: Any trauma patients with at least one of the following risk factors: Age > 45, expectation of > 2 days bedrest, history of DVT or PE, coma (CGS <7), spine cord injury, pelvic fracture,lower extremity fractures, and repair of mayor extremity vein, complex wound of lower extremity, femoral venous catheter

Exclusion Criteria: Age < 18, sever blunt head injury, bleeding injuries no accessible to haemostatic control , active bleeding disorder, cardiovascular instability, nursing mothers, heparin, warfarin or heparinoids within 7 days of trauma, allergy of heparin, bisulphite or fish, history of protein C deficiency, history of heparin associated thrombocytopenia, malignant hypertension blood pressures over 250 systolic and 130 diastolic, liver failure with encephalopathy, renal failure or failure to obtain informed consent.

Type of Injury : UH Blunt: 31 Penetrating: 1 to :LMWH. Blunt 32 Penetrating:2.

Location of trauma: Not reported.

Severity of trauma ISS mean: UH ISS mean13+/-14 to LMWH ISS mean 10 +/- 5

Type of management (Operatory , Non operatory, both or not reported): Not reported

Interventions

UH 5000 U SC BID

LMWH 30 mg SC BID

Outcomes

Diagnostic method for DVT: Doppler ultrasonography

Diagnostic method for PE: Clinically, V/Q and angiograms .

Any bleeding? Yes.

Major Bleeding (need of transfusion, any procedure to control bleeding)

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer generated schedule.
Allocation concealment (selection bias)Low riskCentral allocation (pharmacy-controlled randomizations).
Blinding (performance bias and detection bias)
All outcomes
High riskThey don´t say if the radiologist were blinded.
Incomplete outcome data (attrition bias)
All outcomes
High risk104 patient were randomised but at the end just 66 patient were analysed. They don´t specify what happen with the other 42 patients.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Dennis 1993

MethodsRandomized clinical trial
Participants

395 participants randomised (281 for sequential compression device or low dose unfractionated heparin and 114 for no prophylaxis).

Gender: Not reported

Mean age: 28.6 years for prophylaxis to 27.4 years for no prophylaxis.

Inclusion Criteria: Any trauma

Exclusion Criteria: Less than 18 years or ISS of 9 or below.

Type of Injury (blunt , penetrated, both or not reported): Blunt: 320 Penetrating: 75. Prophylaxis: Blunt: 233 Penetrating: 48. No prophylaxis:Blunt: 87 Penetrating: 27.

Location of trauma: Spine: 39 to11 Head: 74 to 18 Face: Not reported Chest: 81 to 37 Abdomen: 68 to 24 Pelvis: Not reported Extremity: 137 to 51 PolItrauma (more than one): Not reported.

Severity of trauma (ISS, RTS) or not reported: ISS mean: Prophylaxis: 21.1 No prophylaxis: 20.5

Type of management (Operatory , Non operatory, both or not reported): Not reported

Interventions

Prophylactic method: Full length lower extremity sequential compression device (SCD) or those who have fractures or extensive soft tissue injuries of the lower extremity and in cases where SCD were not available were given low dose of subcutaneous heparin at a dose of 5000 units BID.

No prophylactic method.

Outcomes

Diagnostic method for DVT: Doppler ultrasonography

Diagnostic method for PE: Authopsy or Pulmonary angiography.

Any bleeding ? No.

Major Bleeding (need of transfusion, any procedure to control bleeding)

Minor Bleeding (no need none of the above)

NotesLess than 20% couldn't perform the duplex because of the extremity injuries. 67 where changed to prophylaxis group and analysed in the No prophylaxis group (Intervention B).
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskThey were randomised, but they don´t say how.
Allocation concealment (selection bias)Unclear riskThey didn't describe how allocate.
Blinding (performance bias and detection bias)
All outcomes
High riskIncomplete information to establish whether the vascular technician or attending vascular surgeon were blinded37% of the 181 patients in the control group were switched from no prophylaxis to a SCD at the discretion of the attending surgeon who felt the risk of DVT and PE was too high.
Incomplete outcome data (attrition bias)
All outcomes
High riskPer protocol analysis.
Selective reporting (reporting bias)Unclear riskAll the outcomes were reported. (Protocol not known)

Elliot 1999

MethodsRandomized control trial
Participants

149 participants randomised (74 for sequential compression device (Calf thigh device) 75 for sequential compression device (Plantar venous pneumatic compression)

Gender: Calf thigh device male 49 / female 25 Plantar venous pneumatic compression male 51 / female 24

Mean age: 33.9 years for Calf thigh device to 30.2 years for Plantar venous pneumatic compression

Inclusion Criteria: more than 13 years old and who had recent (within 24 hours) severe head injuries (Glasgow Coma Scale score < 9) and/or major trauma and were expected to be bedridden for more than 72 hours

Exclusion Criteria: Patients with external fixation devices or casts that precluded the use of calf-thigh sequential pneumatic compression devices on either or both legs, patients who were not expected to live more than 24 hours, and patients whose injuries occurred more than 24 hours before admission

Type of Injury (blunt, penetrated, both or not reported): Blunt: 137 Penetrating: 12.Calf thigh device Blunt: 68 Penetrating:6. Plantar venous pneumatic compression:Blunt: 69 Penetrating: 6.

Location of trauma: Spine: Not reported Head: 62 to 61 Face: 23 to 14 Chest: 39 to 44 Abdomen: 17 to 22 Pelvis: Not reported Extremity: 10 to 10 Polytrauma (more than one): Not reported.

Severity of trauma (ISS, RTS) or not reported: ISS mean: Calf thigh device: 31.0 Plantar venous pneumatic compression: 30.2

Type of management (Operatory, Non operatory, both or not reported): Calf thigh device Operatory 12/Non operatory 62 Plantar venous pneumatic compression Operatory 10 /Non operatory 65

Interventions

The calf-thigh sequential pneumatic compression devices consisted of four calf and two thigh plastic chambers that inflate sequentially to a pressure of 45 mm Hg. The calf chambers inflated sequentially from the ankle to the knee at 5-second intervals (as recommended by the manufacturer). The two thigh chambers then inflate sequentially in a proximal direction. All chambers remain inflated for 5 seconds, then deflate simultaneously.

The PlexipulseR has a single chamber that inflates for 2 seconds and cycles every 20 seconds. The chamber pressure was set to 160 mm Hg (as recommended by the manufacturer).

Outcomes

Diagnostic method for DVT: Doppler ultrasonography

Diagnostic method for PE: No

Any bleeding ? Yes

Major Bleeding (need of transfusion, any procedure to control bleeding)

Minor Bleeding (no need none of the above)

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomly assigned by computer.
Allocation concealment (selection bias)Low riskIndividual prophylaxis assignments were written on cards and placed in sealed opaque envelopes with only the order of assignment and stratification displayed.
Blinding (performance bias and detection bias)
All outcomes
Low riskThey removed the calf-thigh sequential pneumatic compression devices or plantar venous intermittent pneumatic compression devices from the patient and the patient's room to maintain an assessment that was blinded to the method of venous thromboembolism prophylaxis.
Incomplete outcome data (attrition bias)
All outcomes
Low risk149 patients compose the intent to treat group.
Selective reporting (reporting bias)Unclear riskAll the outcomes were reported. (Protocol not known)

Fisher 1995

MethodsRandomized clinical Trial
Participants

304 participants randomised:145 for Pneumatico sequential leg compression devices (PSLCD) and 159 for no prophylaxis.

Gender: Not reported

Mean age: Not reported

Inclusion Criteria: Pelvic, acetabular, femoral neck, intertrochanteric or subtrochanteric fractures.

Exclusion Criteria: Abnormal coagulation profile, current or recent use of an antiplaquet or anticoagulant medication, malignancy, severe liver disease, skin ulceration or large open wound on lower extremity, objective evidence of DVT, severe multi-trauma in which participation of another trauma service took precedence over the study protocol.

Type of Injury All were blunt :PSLCD: Blunt: 145 No prophylaxis:Blunt: 154.

Location of trauma: Pelvis:PSLCD 35 to No prophylaxis: 38 Extremity (hip fractures): PSLCD 110 to No prophylaxis: 121.

Severity of trauma: ISS mean: Not reported

Type of management : Operatory.

Interventions

PSLCD: Portable controller with a pair of thigh length sleeves. Each sleeve contains six chambers, four calf and two thigh. Sleevs were sequentially inflated to pressures of 45 mm Hg at the ankle, 35-40 mm Hg at the calf, and 25 mm mg at the thigh.The compression cycle is 71 sec, with each compression lasting 11 sec .

No prophylactic method.

Outcomes

Diagnostic method for DVT: Doppler ultrasonography. Equivocal or positive tests, when indicated, were followed by venograms or autopsy

Diagnostic method for PE: VQ scan, angiography or autopsy

Any bleeding ? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomization was by sealed enveloped and balanced in blocks of 20.
Allocation concealment (selection bias)Low riskSealed envelopes.
Blinding (performance bias and detection bias)
All outcomes
Low riskDiagnostic test were assessed by a blinding radiologist, vascular surgeon or nuclear medicine specialist.
Incomplete outcome data (attrition bias)
All outcomes
High riskReason for missing outcome data likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups.
Selective reporting (reporting bias)Unclear riskAll the outcomes were reported. (Protocol not known)

Fuchs 2005

MethodsRandomized clinical trial
Participants

227 participants randomised (111 Arthroflow plus unfractioned heparin and 116 for Unfractioned Heparin).

Gender: Arthroflow plus unfractioned heparin M 57 F 54 Unfractioned Heparin M74 F 42

Mean age: 47.1 (19.7) years for Arthroflow plus unfractioned heparin to 71.9 (19.5) years forUnfractioned Heparin

Inclusion Criteria: Above 18 and below 80 years. Bony or ligamentous trauma of spine, pelvis or extremities.

Exclusion Criteria: Polytrauma, decompensated coronary heart disease, advance peripheral arterial occlusion, severe liver failure, haemorraghic diathesis, stroke, pregnancy, malignant neoplasty, arthritis and arthrodesis of the lower limb, acute thrombosis or thrombophlebitis, pulmonary embolism, paraplegia, chronic muscular dystrophy, lack of compliance.

Type of Injury (blunt , penetrated, both or not reported): Not reported

Location of trauma: Spine: 40 to 30 Head: Pelvis: 27 to 20 extremities 44 to 66

Severity of trauma (ISS, RTS) or not reported: No reported

Type of management (Operatory , Non operatory, both or not reported): Not reported

Interventions

Arthroflow plus UH: UH 5000 ui TID and mobilization of the extremity for 30 min TID.

UH 5000 ui TID

Outcomes

Diagnostic method for DVT: Doppler ultrasonography and plethysmography as screening. Confirmed diagnosis by Venography

Diagnostic method for PE: Clinically

Any bleeding ? Not reported.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomly assigned by computer.
Allocation concealment (selection bias)Unclear riskNot reported.
Blinding (performance bias and detection bias)
All outcomes
Low riskEvaluated by blind radiologist.
Incomplete outcome data (attrition bias)
All outcomes
High riskPatients with trauma to the ankle had to be excluded from to the arthroflow group.
Selective reporting (reporting bias)Unclear riskAll the outcomes were reported. (Protocol not known).

Geerts 1996

MethodsRandomized clinical trial
Participants

265 participants randomised: 136 for low dose unfractionated heparin (LDUH) and 129 for LMWH.

Gender: LDUH male 99 / female 37 LMWH male 93 / female 36

Mean age: LDUH 37.0± 16.5 and LMWH 39.1 ± 16.8

Inclusion Criteria: Any trauma in adults.

Exclusion Criteria: ISS below 9. Survival below 7 days. Hospital stay below 7 days. Intracranial bleeding by CT. Uncontrolled bleeding after 36 hours of injury. Systemic coagulopathy (PT more 3segs above control or platelets less than 50,000). Need of therapeutic anticoagulation. Contrast allergy for venography. Renal failure. Pregnant. Don´t achieve venous access because amputation or major foot injury.

Type of Injury Blunt trauma

Location of trauma: Spine: 24 to 16 Head: 6 to 7 Face: Not reported Chest:Not reported Abdomen: Not reported Pelvis: 25 to 23 Extremity: 75 to 69 Polytrauma (more than one): Not reported.

Severity of trauma: ISS mean: LDUH 22.7 ± 9 LMWH 23.1± 8

Type of management Operatory 119 to 107 non operatory 17 to 22

Interventions

UH 5000 ui every 12 hours (0,3 ml preloaded syringes), 36 hours after the injury and continued by 14 days. Witheld single preoperative dose and resumed dose after the surgery.

LMWH (Enoxaparin) 30 mg every 12 hours (0,3 ml preloaded syringes), 36 hours after the injury and continued by 14 days. Witheld single preoperative dose and resumed dose after the surgery.

Outcomes

Diagnostic method for DVT:Clinical surveillance, Clinical suspect diagnosis with US-doppler and confirmation with Venography. If not (complete asymptomatic), Control Venography days 10 to 14 after injury.

Diagnostic method for PE: Clinical suspect with ventilation-perfusion scanning. Any doubt will be confirmed with pulmonary angiography, venous ultrasonography or contrast venography or combinations of these.

Any bleeding ? yes

Major Bleeding (need of transfusion, any procedure to control bleeding)

Minor Bleeding (no need none of the above)

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomly assigned by a computer.
Allocation concealment (selection bias)Low riskBlinded fashion with preloaded syringes.
Blinding (performance bias and detection bias)
All outcomes
Low riskImaging studies and episodes of bleeding were adjudicated by a panel of experts who were unaware of the clinical data.
Incomplete outcome data (attrition bias)
All outcomes
Low riskWithout losses.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Ginzburg 2003

MethodsRandomized clinical trial
Participants

442 participants randomised: 218 for LMWH and 224 for Intermitent Pneumatic Compression (IPC)..

Gender: 160 male for LMWH to 167 male for IPC

Mean age: 42 years forLMWH to 41 years for IPC.

Inclusion Criteria: Above 18 yeas old. ISS above 9. At least one leg and one arm available for the IPC. Don't need anticoagulation. Had no contraindication for LMWH.

Exclusion Criteria: Under 18 years. ISS under 9. Unlikely to survive or remain in hospital at least 7 days. Acute renal failure (creatinine above 3.4 mg/dL). Pregnant women. Unable to undergo bilateral duplex. Morbid obesity (>25kg/m2). Contraindication of coagulation: intracranial bleeding of uncontrolled bleeding from other sites for more than 24 hours after admission. Coaugulopathy (PT> 3s/ control or plat < 50.000)

Type of Injury: Not reported.

Location of trauma: Spine: 18 to 15 Head: 41 to 60 Face: 11 to 11 Chest: 80 to 92 Abdomen: 53 to 63 Pelvis: 35 to 44 Extremity: 174 to 136 PolItrauma (more than one): Not reported.

Severity of trauma (ISS, RTS) or not reported: Moderately (ISS 9-19): 294.Severe (ISS >20):148

Moderate: ISS mean: LMWH: 12.4 (3.8) to IPC: 12.7 (3.9). Severe: ISS mean: LMWH: 25.3 (4.8) to IPC 25.9 (5.8). All ISS mean: ISS mean: LMWH: 16.7 (7.3) to IPC: 17.1 (6.8).

Type of management: Not reported

Interventions

LMWH (Enoxaparin): 30 mg SC BID. Begining 24h after trauma. Witheld 12h before surgery and resumed 12h after.Two doses missed.

IPC (Calf garment)(Huntleigh Flowtron, Manalapan, New Jersey, USA) placed both legs inflated alternately to 40 mmHg every 60 s cycle. First cycle inflated 12 s and deflated 48s. Second cycle 30 s inflated and 30 s deflated. Tolerated for up to 8 h.

Outcomes

Diagnostic method for DVT: Doppler ultrasonography

Diagnostic method for PE: Clinically PE or CT scan or angiography or ventilation/perfusion scan.

Any bleeding ? Yes.

Major Bleeding (need of transfusion, any procedure to control bleeding).

Minor Bleeding (no need none of the above)

NotesThe primary endpoint, the development of significant thromboembolic complications while in hospital, served as the basis for the calculation of sample size. The protocol was designed to consider all randomised and compliant patients in the analysis. The risk of proximal DVT and pulmonary embolism in this study population was projected to be around 10 per cent in the enoxaparin group on the basis of reports from similar trials8. Enrolment of a total of 900 patients (450 in each arm) would be needed to allow the
confident detection of a clinically significant 30 per cent difference in treatment efficacy between enoxaparin and IPC (power 80 per cent, one-sided significance 0·050) with an anticipated non-compliance (drop out) rate of 5 per cent. This population size was based on an expected incidence of DVT or pulmonary embolism of 20 per cent. However, at planned interim analysis (442 patients), it was realized that the incidence was only 2·0 per cent. Based on this analysis, the decision was made to cease enrolment of subjects
because a tenfold increase in patient numbers would have been needed to reach significance. Because this protocol was designed to test equivalence and because LMWH has been recognized as a standard of thromboprophylaxis, use of a one-tailed test was justified.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated random.
Allocation concealment (selection bias)Low riskInvestigators were blinded to treatment assignment before randomisation.
Blinding (performance bias and detection bias)
All outcomes
High risk

Single technician performed all scans. He was not blind.

Also all patients were assessed daily by the investigators. Which means that investigators could see the intervention.

Incomplete outcome data (attrition bias)
All outcomes
High riskOf the 442 randomised subjects 44 (15 in the IPC group and 29 in the LMWH
group) were excluded from analysis because of compliance issues.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Knudson 1992

MethodsRandomized clinical trial
Participants

113 participants randomised (37 for low dose unfractionated heparin (LDUH) and 76 for Sequential pneumatic compression boots and elastic stockings (SCD and TED)).

Gender: Not reported.

Mean age: 37.7 ± 15.8 years for LDUH to 37.8 ± 19.4 years for SCD and TED.

Inclusion Criteria: Adults traumatized patients.

Exclusion Criteria: Pediatric trauma patients (below 17 years), pregnant trauma patients, minor injuries, hospitalisation less than 48 hours,

Type of Injury: Blunt: 91 Penetrating: 11.

Location of trauma: Spine and Pelvis: 7 to 22 Head: 12 to 24 Face: Not reported Chest: 18 to 22 Abdomen: 13 to 23 Extremity: 27 to 30 PolItrauma (more than one): Not reported.

Severity of trauma (ISS, RTS) or not reported: ISS mean: LDUH: 15.9 ± 7.9 SCD and TED: 18.5 ± 11.8

Type of management: Not reported.

Interventions

Low Dose Unfractioned Heparin: 5000 ui SC BID.

Sequential pneumatic compression boots and elastic stockings (SCD and TED).

Outcomes

Diagnostic method for DVT: Real time B mode ultrasound / 5 days for tree weeks until discharge

Diagnostic method for PE: V/Q radionuclide and or pulmonary angiograms

Any bleeding ? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High riskEven and odds.
Allocation concealment (selection bias)High riskAlternative.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskThey didn't say whether the technologist or radiologist were blinding.
Incomplete outcome data (attrition bias)
All outcomes
High riskPer protocol analysis.

Knudson 1994 All groups

MethodsRandomized clinical trial
Participants

251 participants randomised in three groups:

Gender: Male 200/ Female 51

Mean age: 38 years.

Inclusion Criteria:

Group I: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who could receive either of the two methods of prophylaxis.

Group II: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who could not wear bilateral devices.

Group III: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who had any contraindication to receive heparin.

Exclusion Criteria:

Group I: Younger than 18 year old, pregnant and prisoners.

Group II: Younger than 18 year old, pregnant and prisoners.

Group III: Younger than 18 year old, pregnant and prisoners.

Type of Injury: 140 blunt, 111 penetrating ( 65 stab wounds and 46 firearm)

Location of trauma: Not reported.

Severity of trauma: Not reported

Type of management: Not reported

Interventions

Group I: Low Dose Unfractioned Heparin: 5000 ui SC BID or Sequential pneumatic compression boots and elastic stockings (SCD and TED) vs Control.

Group II: Low Dose Unfractioned Heparin: 5000 ui SC BID vs Control.

Group III: Sequential pneumatic compression boots and elastic stockings (SCD and TED) vs Control.

Outcomes

Diagnostic method for DVT: Dupplex.

Diagnostic method for PE: V/Q and angiograms

Any bleeding ? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandom drawing.
Allocation concealment (selection bias)High riskAllocated by the possibility to wear the devices or to receive heparin.
Blinding (performance bias and detection bias)
All outcomes
High riskIt is not reported if the technologist or radiologist were blinded.
Incomplete outcome data (attrition bias)
All outcomes
High risk145 missing patients.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Knudson 1994 group I

Methods 
Participants 
Interventions 
Outcomes 
Notes 

Knudson 1994 group II

MethodsRandomized clinical trial
Participants

251 participants randomised in three groups:

Gender: Male 200/ Female 51

Mean age: 38 years.

Inclusion Criteria:

Group I: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who could receive either of the two methods of prophylaxis.

Group II: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who could not wear bilateral devices.

Group III: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who had any contraindication to receive heparin.

Exclusion Criteria:

Group I: Younger than 18 year old, pregnant and prisoners.

Group II: Younger than 18 year old, pregnant and prisoners.

Group III: Younger than 18 year old, pregnant and prisoners.

Type of Injury: 140 blunt, 111 penetrating ( 65 stab wounds and 46 firearm)

Location of trauma: Not reported.

Severity of trauma: Not reported

Type of management: Not reported

Interventions

Group I: Low Dose Unfractioned Heparin: 5000 u SC BID or Sequential pneumatic compression boots and elastic stockings (SCD and TED) vs Control.

Group II: Low Dose Unfractioned Heparin: 5000 u SC BID vs Control.

Group III: Sequential pneumatic compression boots and elastic stockings (SCD and TED) vs Control.

Outcomes

Diagnostic method for DVT: Dupplex.

Diagnostic method for PE: V/Q and angiograms

Any bleeding ? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandom drawing.
Allocation concealment (selection bias)High riskAllocated by the possibility to wear the devices or to receive heparin.
Blinding (performance bias and detection bias)
All outcomes
High riskThey didn´t say weather the technologist or radiologist were blinding.
Incomplete outcome data (attrition bias)
All outcomes
High risk145 missing patients.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Knudson 1994 group III

MethodsRandomized clinical trial
Participants

251 participants randomised in three groups:

Gender: Male 200/ Female 51

Mean age: 38 years.

Inclusion Criteria:

Group I: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who could receive either of the two methods of prophylaxis.

Group II: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who could not wear bilateral devices.

Group III: Any trauma patients seriously injured. With duplex scan negative the first day before the randomising. Who had any contraindication to receive heparin.

Exclusion Criteria:

Group I: Younger than 18 year old, pregnant and prisoners.

Group II: Younger than 18 year old, pregnant and prisoners.

Group III: Younger than 18 year old, pregnant and prisoners.

Type of Injury: 140 blunt, 111 penetrating ( 65 stab wounds and 46 firearm)

Location of trauma: Not reported.

Severity of trauma: Not reported

Type of management: Not reported

Interventions

Group I: Low Dose Unfractioned Heparin: 5000 u SC BID or Sequential pneumatic compression boots and elastic stockings (SCD and TED) vs Control.

Group II: Low Dose Unfractioned Heparin: 5000 u SC BID vs Control.

Group III: Sequential pneumatic compression boots and elastic stockings (SCD and TED) vs Control.

Outcomes

Diagnostic method for DVT: Dupplex.

Diagnostic method for PE: V/Q and angiograms

Any bleeding ? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandom drawing.
Allocation concealment (selection bias)High riskAllocated by the possibility to wear the devices or to receive heparin.
Blinding (performance bias and detection bias)
All outcomes
High riskThey didn't say weather the technologist or radiologist were blinding.
Incomplete outcome data (attrition bias)
All outcomes
High risk145 missing patients.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Knudson 1996

MethodsRandomized clinical trial
Participants

202 participants randomised (120 for low molecular weight heparin -LMWH- and 82 for SCD or arteriovenous impulse (AVI) device ).

Gender: Not reported

Mean age: 39 years for LMWH to 38 years for SCD and AVI.

Inclusion Criteria: Injury Severity Score (ISS) of > 10 or Abbreviated Injury Scale score greater than or equal to 3 in any category; head injury with Glasgow Coma Scale score less than or equal to 8; unstable spine fracture without neurologic deficit; spine fracture with deficit; major pelvic fracture (Tile Class B or C); fracture of the lower extremity above the ankle; previous history of DVT; acute venous injury; or age >50 years

Exclusion Criteria: the presence of DVT on admission, patients younger than 18 years of age, pregnant patients, prisoners, patients who could not be randomised within the first 24 hours of hospitalisation, and patients who refused consent. included discharge or transfer from the hospital before 5 days post injury, bleeding disorders possibly associated with the use of LMWH, noncompliance with the assigned prophylactic measure, or evidence of DVT on duplex examination or PE by ventilation-perfusion scan or angiography

Type of Injury Blunt trauma and penetrating trauma.

Location of trauma: Not reported.

Severity of trauma ISS mean: LMWH 14 SCD and AVI. 16

Type of management (Operatory , Non operatory, both or not reported): Not reported

Interventions

LMWH (enoxaparin sodium, Rhone Poulenc Rorer Pharmaceuticals Inc., Collegeville, Pa) 30 mg subcutaneously every 12 hours.

Sequential gradient pneumatic compression (SCD) or arteriovenous impulse (AVI) device

Outcomes

Diagnostic method for DVT: venous duplex

Diagnostic method for PE: Clinical observation, Arterial blood gases and oxygen saturation , three ventilation-perfusion scans and two pulmonary angiograms were performed for unexplained respiratory distress or hypoxic episodes

Any bleeding ? No.

Major Bleeding (need of transfusion, any procedure to control bleeding)

Minor Bleeding (no need none of the above)

NotesFrom 372 patients were assigned to a Heparin or a No-Heparin group, depending upon the presence of injuries that would preclude the use of heparin in the early period after injury. Just 202 patient were randomised in the heparin group to receive heparin or SCD or AVI.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated random.
Allocation concealment (selection bias)Unclear riskInsufficient information.
Blinding (performance bias and detection bias)
All outcomes
High riskVenous duplex examinations were performed by vascular technician with over 10 years of experience. But the article didn't say whether the vascular technician were blinded.
Incomplete outcome data (attrition bias)
All outcomes
Low riskThe heparin group that were randomised did not have any losses.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Kurtoglu 2004

MethodsRandomized clinical trial
Participants

120 participants randomised (60 for intermittent pneumatic compressionIPC or 60 for LMWH

Gender: Male 47 Female 73

Mean age: 37.1 (R 18-76)

Inclusion Criteria: polytrauma with severe head/spinal injuries

Exclusion Criteria: younger than 14 years old were. excluded, as were individuals with hepatic or urinary dysfunction, a spinal cord injury, a history of DVT, or a high bleeding risk (platelets < 100,000/μl or INR > 1.5) and those using anticoagulants. Patients with continuing haemorrhage on control scans within 24 hours of admission or who required craniotomy were excluded from the study.

Type of Injury Not reported (Blunt?).

Location of trauma: Spine: 5 to 6 Head: 55 to 54 Face: Not reported Chest: Not reported Abdomen: Not reported Pelvis: Not reported Extremity: Not reported PolItrauma (more than one): Not reported.

Severity of trauma (ISS, RTS) or not reported: ISS mean: ISS mean Intervention A 18.3 +/- 3.2 (4-35) Intervetion B 19.5 +/- 1.7 ( 4-45)

Type of management (Operatory , Non operatory, both or not reported): Not reported

Interventions

IPC Intermitent pneumatic Compression the IPC patients were placed on below-knee IPC devices (Prophylactic D.V.T System, modelAC
550; Flowtron Excell, Bedfordshire, UK) or another IPC device (AV Impulse System, Duo; Novamedix, Andover, UK) for prophylaxis.

LMWH group patients received enoxaparin sodium 40 mg/
day (Clexane; Aventis, Strasbourg, France

Outcomes

Diagnostic method for DVT: Venous duplex colour-flow Doppler ultrasonography of the lower extremities was obtained on admission to the ICU, each week of hospitalisation, and 1 week after discharge for all patients

Diagnostic method for PE: Patients with a suspected DVT, those whose clinical status was deteriorating, and
those who displayed a sudden change in blood gas levels were subjected to spiral CT scanning. Pulmonary CT examinations were undertaken with a Somatom Plus-S scanner (Siemens, Erlangen, Germany

Any bleeding ? Yes

Major Bleeding macroscopic hematuria without renal injury, overt bleeding, and a sudden drop in the haemoglobin level ( 2 g/dl).

Minor Bleeding Microscopic hematuria, hematoma at the site of injection, and a drop in the haemoglobin level of less than 2 g/dl were considered minor bleeding complications.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High riskNo (Alternate selection)
Allocation concealment (selection bias)High riskThe patients were randomly allocated by alternate
selection to either group
Blinding (performance bias and detection bias)
All outcomes
Unclear riskUnclear (Not described)
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll patients randomised were analysed
Selective reporting (reporting bias)Unclear riskProtocol not known.

Stannard 2001

MethodsRandomized clinical trial
Participants

107 participants randomised: 54 for thigh-calf low-pressure sequential-compression device 53 for calf-foot high-pressure pulsatile-compression pump

Gender: Not reported

Mean age: Not reported

Inclusion Criteria: blunt trauma causing a pelvic or acetabular
fracture with a pattern requiring surgical fixation, an age of at
least sixteen years, and an ability and willingness to comply
with both the mechanical prophylaxis protocol and the
screening studies for deep-vein thrombosis

Exclusion Criteria: a history of venous thromboembolic disease, initiation
of mechanical compression more than seventy-two hours
following the injury, a body habitus or weight that made it difficult
for the patient to fit in the magnetic resonance imaging
scanner, or a stable injury that did not require surgical treatment.

Type of Injury: all were blunt

Location of trauma: Spine: Not reported Head: Not reported Face: Not reported Chest: Not reported Abdomen: Not reported Pelvis:All were pelvis (they don´t say how many per group Extremity: Not reported Polytrauma Not reported.

Severity of trauma (ISS, RTS) or not reported: ISS mean: thigh-calf low-pressure sequential-compression : 19.8 (9-59) calf-foot high-pressure pulsatile-compression pump: 16.1(9-50)

Type of management Operatory

Interventions

- a thigh-calf low pressure sequential- compression device (Kendall SCD; Kendall, Mansfield, Massachusetts).

- combination sequential pump that covers the calf and foot (PlexiPulse; NuTech, Kinetic Concepts,
San Antonio, Texas).

Outcomes

Diagnostic method for DVT: Doppler ultrasonography and MRI venogram

Diagnostic method for PE: not reported.

Any bleeding ? No.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated randomisation table.
Allocation concealment (selection bias)Unclear riskNot described.
Blinding (performance bias and detection bias)
All outcomes
Low riskAll studies were interpreted by radiologists blinded regarding the type of prophylaxis against deep-vein thrombosis and the result of the other study.
Incomplete outcome data (attrition bias)
All outcomes
High risk"Thirty-three patients who were initially enrolled in the study did not complete it. The reasons for withdrawal from the study included claustrophobia (six patients); death (six) (no patient died because of a thromboembolic event); refusal to undergo magnetic resonance venography (five); inadvertent initiation of anticoagulation by another service (five); discharge before the appropriate studies had been performed (three); inability of the patient to fit in the magnetic resonance scanner (two); inability of the patient to remain immobile during the magnetic resonance imaging secondary to a closed head injury (two); prior venous thromboembolic disease missed on the initial screening (two); inadvertent switching of
the pump types (one); and pregnancy (one). The patients who withdrew had a total of three deep-vein thromboses, which were not included in the statistical analysis". (Stannard 2001, page 1048)
Selective reporting (reporting bias)Unclear riskProtocol not known.

Stannard 2006

MethodsRandomized clinical trial
Participants

200 participants randomised (Group A 97 LMWH and Group B 103 for PlexiPulse foot pump and LMWH

Gender: Not reported

Mean age: Group A 38.2 (19-75) Group B 41.0 (19-80)

Inclusion Criteria: The inclusion criteria were blunt trauma and at least one of the following findings: an Abbreviated Injury Score of 3 or more and a long-bone fracture, multiple (two or more) long-bone fractures,
or an age of more than fifty-five years and a long-bone fracture. In addition, all patients were more than eighteen years of age, had no contraindications to anticoagulation, and had been admitted to our hospital less than seventy-two hours after the time of trauma or had a negative magnetic resonance venogram
prior to enrolment.

Exclusion Criteria: Exclusion criteria included renal insufficiency; severe cranial or spinal cord injury; the use of anticoagulants; any contraindication to anticoagulation, including severe active bleeding; pregnancy; a history of venous thromboembolic disease; any contraindication to magnetic resonance venography; the presence of a vena cava filter; and severe ocular trauma. Cranial and ocular trauma associated with
bleeding and a risk of increased intracranial or ocular pressure was considered severe and led to exclusion from the study.

Type of Injury Blunt trauma

Location of trauma: Spine: A 2 to B 6 Head:No Face: No Chest:No Abdomen:No Pelvis: A 89 to B 85 Extremity: A 161 To B 141 PolItrauma (more than one): No

Severity of trauma ISS mean:Group A 14.41 (8-57) Group B 14.43 (4-41)

Type of management : Not reported

Interventions

Group A LMWH 30 mg sc BID until the discharge

Group B LMWH 30 mg sc BID until the discharge and PlexiPulse foot pumps (Kinetic Concepts, San Antonio, Texas)

Outcomes

DVT Dupplex ultrasound and venous MRI before the discharge or before it was indicated

MRI

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated randomisation program.
Allocation concealment (selection bias)Unclear riskNot described in the full text article.
Blinding (performance bias and detection bias)
All outcomes
Low riskTwo radiologists (including one of the authors , who were blinded with regard to the type of deep-vein thrombosis treatment, reviewed the coronal reformatted multiple- intensity projections and source axial images. When there was a discrepancy between the radiologists with regard to the correct interpretation, the radiologists reviewed the studies in question together and came to a consensus
Incomplete outcome data (attrition bias)
All outcomes
High riskTwo hundred and twenty-four patients were enrolled, and
200 completed the entire protocol.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Velmahos 2005

MethodsRandomized clinical trial
Participants

47 participants randomised (26 for Muscle electrostimulation (MEST) and 21 for Control).

Gender: MEST: male 21/female 5 to Control: male 16/female 5

Mean age: 32 ± 15 years for MEST to 45 ± 21 years for Control.

Inclusion Criteria: trauma (Injury Severity Score higher than 9) and had contraindications for receiving prophylactic heparin (unfractionated or low molecular weight) were considered for inclusion in the study. Such
patients were those with (a) significant head injuries, (b) operations for extensive organ injuries,
(c) major retroperitoneal hematomas, (d) liver, spleen, or kidney injuries higher than Grade II,
managed nonoperatively, or (e) other injuries that at the discretion of the trauma surgeon were
deemed to be associated with a high likelihood for bleeding. Additional inclusion criteria were (1)
anticipated survival for longer than 7 days, (2) anticipated hospital stay of longer than 7 days, and
(3) randomisation within 24 hours of injury.

Exclusion Criteria: a) age less than 18 years, (b) known allergy to contrast material, precluding
the use of venography, (c) cardiac demand pacemakers or other implanted stimulators or implants
containing metal parts within the area of treatment, (d) spastic paralysis, (e) local infection at the
site of application, and (f ) history or present evidence of venous thrombosis.

Type of Injury (blunt , penetrated, both or not reported): Blunt: 34 Penetrating: 13. MEST: Blunt: 18 Penetrating: 8. Control: Blunt: 16 Penetrating: 5.

Location of trauma: Spine: 5 to1 Head: 7 to 1 Face: Not reported Chest: Not reported Abdomen: Not reported Pelvis: Not reported Extremity: 9 to 13 PolItrauma (more than one): Not reported.

Severity of trauma (ISS, RTS) or not reported: ISS mean: MEST: 20 Control: 18

Type of management (Operatory , Non operatory, both or not reported): MEST: operatory: 19 Control: operatory: 21

Interventions

Muscle electrostimulation (MEST): MEST patients received two 30-minute sessions daily, one in the morning
and one in the evening, by using the Lymphavision stimulator (Bexley Trading, San Rafael, Calif).
Electrodes were placed at the calves and medial thighs of both extremities. Voltage was applied
gradually (0-120 V) to obtain a slight visible twitch of the muscles, usually evident by movement of the
toes. Stimuli were 3 milliseconds long at a frequency of 1.75 Hz (105/minute) with inversion of
polarity every 5 seconds. Daily treatments were continued for a minimum of 7 days and a maximum
of 14 days.

Control.

Outcomes

Diagnostic method for DVT: Venography or duplex

Diagnostic method for PE: Not described

Any bleeding ? None

NotesCo-interventions:Patients of both groups were allowed to have standard prophylaxis by subcutaneous unfractionated or low-molecular-weight heparin when the contraindication for its use was no longer
present (usually 3-5 days after admission) and/or by SCD if the extremities were not injured at the site of placement.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomized by a computer generated system
Allocation concealment (selection bias)Unclear riskNot described.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNot described.
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll randomised patients were analysed.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Yanar 2007

MethodsRandomized clinical trial (Abstract)
Participants120 patients three groups. They don´t say inclusion or exclusion criteria
Interventions

Group A IPC

Group B IPC +LMWH

Group C LMWH

Outcomes

Diagnostic method for DVT: duplex ultrasound

Diagnostic method for PE: CT Scan

Any bleeding ? Not reported

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskIncomplete data in the abstract.
Allocation concealment (selection bias)Unclear riskIncomplete data in the abstract.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskIncomplete data in the abstract.
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskIncomplete data in the abstract.
Selective reporting (reporting bias)Unclear riskProtocol not known.

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Breyer 1986Insufficient information given in the abstract to consider for inclusion.
Greenfiel 1997It is not randomised.
Haas 2003It is an abstract and included patients with hip fractures and outpatients.
Holzheimer 2004It is a review.
Murakami 2003DVT is not an outcome
Reilmann 1986It is not a RCT.
Wolf 1992It was not for trauma patients.

Ancillary