Intervention Review

You have free access to this content

Heparin for the prevention of venous thromboembolism in acutely ill medical patients (excluding stroke and myocardial infarction)

  1. Raza Alikhan1,*,
  2. Rachel Forster2,
  3. Alexander T Cohen3

Editorial Group: Cochrane Vascular Group

Published Online: 7 MAY 2014

Assessed as up-to-date: 12 NOV 2013

DOI: 10.1002/14651858.CD003747.pub4


How to Cite

Alikhan R, Forster R, Cohen AT. Heparin for the prevention of venous thromboembolism in acutely ill medical patients (excluding stroke and myocardial infarction). Cochrane Database of Systematic Reviews 2014, Issue 5. Art. No.: CD003747. DOI: 10.1002/14651858.CD003747.pub4.

Author Information

  1. 1

    University Hospital of Wales, Haemophilia and Thrombosis Centre, Cardiff, UK

  2. 2

    University of Edinburgh, Usher Institute of Population Health Sciences and Informatics, Edinburgh, UK

  3. 3

    Kings College Hospital, Department of Vascular Surgery, London, UK

*Raza Alikhan, Haemophilia and Thrombosis Centre, University Hospital of Wales, Heath Park, Cardiff, CF14 4XW, UK. raza@doctors.org.uk.

Publication History

  1. Publication Status: Edited (no change to conclusions), comment added to review
  2. Published Online: 7 MAY 2014

SEARCH

 
Characteristics of included studies [ordered by study ID]

MethodsStudy design: single centre, randomised controlled trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: Scotland, UK


ParticipantsNumber: 100 (UFH n = 50; control n = 50)
Age: (mean years) UFH 66.6; control 65.0
Sex: (M/F) UFH 35/15; control 34/16
Inclusion criteria: patients hospitalised with heart failure or chest infection, or both
Exclusion criteria: iodine sensitivity, risk of bleeding, DVT or PE on admission, or confined to bed > 2 days prior to admission

Reason for hospitalisation (n (%)): heart failure (UFH n = 21 (42.0), control n = 17 (34.0)); chest infection (UFH n = 18 (36.0), control n = 19 (38)); both (UFH n = 11 (22), control n = 14 (28))


InterventionsTreatment: UFH 5000 IU, subcutaneous injection, three times daily
Control: no specific prophylaxis (no placebo injections)
Duration: until fully mobile


OutcomesDVT, PE, major and minor haemorrhage

Haemorrhage definition: (major) not specified; (minor) not reported


NotesIsotopically detected DVT (FUT at entry and every other day)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo matched placebo used

Blinding of outcome assessment (detection bias)
All outcomes
Low riskFUT scanning performed by a person unaware of treatment allocation

Incomplete outcome data (attrition bias)
All outcomes
Low riskAnalysis performed with all randomised participants, although no indication of loss to follow-up

Selective reporting (reporting bias)Low riskReport includes all expected outcomes, although no study protocol is available

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: multi-centre, randomised, double-blind controlled trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: France


ParticipantsNumber: 2474 (nadroparin n = 1230; control n = 1244)
Age (mean years): 76
Sex: not stated
Inclusion criteria: patients hospitalised with acute medical illness and confined to bed
Exclusion criteria: not stated

Reason for hospitalisation: not described


InterventionsTreatment: LMWH (nadroparin) 7500 antiXa units, subcutaneous injection, once daily
Control: placebo
Duration: 21 days or discharge


OutcomesPE, mortality


NotesAutopsy documented PE

Data on all-cause mortality was calculated from percentages provided in study paper.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
Low riskStudy described as double-blind and control patients received matched placebo

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo indication of blinding of assessors

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data, although no indication of loss to follow-up

Selective reporting (reporting bias)Unclear riskNo description of the protocol; it is unclear whether mortality reported is all-cause or specifically due to thromboembolism

Other biasLow riskStudy appears to be free from other sources of bias


MethodsStudy design: multi-centre, randomised open-label, active-controlled trial

Intention-to-treat analysis: efficacy outcomes no; safety outcomes yes

Country: Germany


ParticipantsNumber: 337 (certoparin n = 163; UFH n = 174)

Age: (mean years ± SD) total 70.6 ± 12.3; certoparin 70.2 ± 12.2; UFH 71.0 ± 12.4

Sex: M/F total 160/177; certoparin 77/86; UFH 83/91

Inclusion criteria: patient > 40 years, hospitalisation due to an acute non-surgical disease and significant recent decrease in mobility

Exclusion criteria: women of child bearing age unless post-menopausal or using a highly effective method of birth control, pregnancy or lactation, indication for anticoagulant / thrombolysis, major surgery or invasive procedure within 4 weeks prior to randomisation or expected within 2 weeks, immobilisation due to fracture of lower limb, acute ischaemic/haemorrhagic stroke, uncontrolled hypertension, active bleeding, active peptic ulcer, bleeding diathesis, pancreatic disease / malignant tumour with high risk of bleeding, endocarditis, thrombocytopaenia, severe renal / liver disease, history of addictive disorder, other investigational drug within 30 days or 5 half-lives before enrolment

Reason for hospitalisation (n (%)): not indicated which acute illnesses were included. Disease profile: diabetes (certoparin n = 52 (31.9), UFH n = 55 (31.6)); COPD (certoparin n = 42 (25.8), UFH n = 50 (28.7)); coronary artery disease (certoparin n = 26 (16.0); UFH n = 33 (19))


InterventionsTreatment 1: Certoparin 3000 IU once daily

Treatment 2: UFH 7500 IU twice daily

Duration: 10 ± 2 days


OutcomesEfficacy: composite of asymptomatic or symptomatic proximal / distal DVT, symptomatic PE or death related to VTE

Safety: Major and minor bleeding during the study as well as during the 3 month follow-up

Haemorrhage definition: (major) fatal bleeding, clinically overt bleeding associated with a fall of haemoglobin concentration > 20 g/L compared with baseline, requiring transfusion of 2 or more units of packed red cells or whole blood, retroperitoneal or intracranial bleeding or bleeding warranting treatment cessation; (minor) bleeding events not meeting major bleeding definition


NotesDVT assessed with the use of compression ultrasound


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomisation list produced using a validated system. After confirming patient fulfilled inclusion /exclusion criteria, investigator telephoned a randomisation hotline

Allocation concealment (selection bias)Low riskCentral allocation of assignment, participants and investigators could not foresee assignment

Blinding of participants and personnel (performance bias)
All outcomes
High riskOpen-label administration of certoparin daily versus UFH twice daily

Blinding of outcome assessment (detection bias)
All outcomes
Low riskAll outcomes were verified by a blinded adjudication committee

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskEfficacy outcomes missing 58 patients not included in reasons for exclusion; Safety outcomes are at low risk of attrition bias

Selective reporting (reporting bias)Low riskStudy protocol is not available but detailed inclusion and exclusion criteria presented in paper and it is clear that the published report includes all expected outcomes

Other biasUnclear riskFunding of study supported by Novartis


MethodsStudy design: multi-centre, randomised, double-blind study, controlled trial with 2 parallel groups

Intention-to-treat analysis: yes

Country: Germany


ParticipantsNumber: 3241 (certoparin n = 1626; UFH n = 1615)

Age: (mean years ± SD) total 78.8 ± 6.3; certoparin 79.0 ± 6.2; UFH 78.7 ± 6.3

Sex: (M/F) total 1324/1915; certoparin 669/955; UFH 655/960

Inclusion criteria: hospitalised medical patients, aged at least 70 years, with an acute medical illness with a significant decrease in mobility (bedridden or only able to walk short distances) expected for at least 4 days

Exclusion criteria: immobilised for > 3 days prior to randomisation, those immobilised due to fracture or cast, expected to undergo major surgery or invasive procedure within 3 weeks of randomisation, severe sepsis or need for intubated ventilatory support, received LMWH > 48 hours in 5 days prior to randomisation, indications for anticoagulation or thrombolysis, life-expectancy < 6 months, presence of DVT / PE, non-haemorrhagic stroke < 3months, haemorrhagic stroke < 12 months, intracranial disease, high-risk of GI bleeding, spinal/epidural anaesthesia < 12 hr, lumbar puncture < 12 hr, uncontrolled hypertension, severe liver or renal disease, acute endocarditis, active retinopathy, intravitreal / intraocular bleeding

Reason for hospitalisation (n (%)): infections and infestations (certoparin n = 435 (26.8), UFH n = 458 (28.4)); cardiac disorders (certoparin n = 357 (22.0), UFH n = 362 (22.4)); respiratory, thoracic and mediastinal disorders (certoparin n = 280 (17.2), UFH n = 279 (17.3)); nervous system disorders (certoparin n = 115 (7.1), UFH n = 99 (6.1)); gastrointestinal disorders (certoparin n = 118 (7.3), UFH n = 95 (5.9)); vascular disorders (certoparin n = 92 (5.7), UFH n = 95 (5.9))


InterventionsTreatment 1: LMWH (certoparin) 3000 U Anti-Xa, subcutaneous injection, once daily, plus placebo twice daily

Treatment 2: unfractionated heparin 5000 IU, subcutaneous injection, three times daily
Duration: 8 - 20 days


OutcomesProximal DVT, symptomatic non-fatal PE, VTE-related mortality, all cause mortality, major and minor haemorrhage, thrombocytopaenia

Haemorrhage definition: (major) fatal bleeding, clinically overt bleeding associated with a fall in haemoglobin of more than 20 g/L from baseline, requiring a transfusion of tow or more units of packed red cells or whole blood, bleeding in a critical area or organ; (minor) bleeding events that did not meet the major bleeding events


NotesDVT assessed by compression ultrasound

For the purposes of this analysis data collected during the treatment period was used


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomisation list was generated using a validated system that automates the random assignment of treatment arms to randomisation numbers in the specified ratio. Included patients were given the lowest number on the randomisation list

Allocation concealment (selection bias)Low riskTreatment assignment was concealed from patients and investigators

Blinding of participants and personnel (performance bias)
All outcomes
Low riskCertoparin was given once daily and UFH three times daily. Patients in the certoparin group received 2 additional placebo injections during the day, to match the regimen in the UFH group and to allow for treatment blinding

Blinding of outcome assessment (detection bias)
All outcomes
Low riskAll parts of the endpoint were adjudicated by a blinded central adjudication and compression ultrasound reading centre

Incomplete outcome data (attrition bias)
All outcomes
Low riskPatient flow fully described, including discontinuations and losses to follow up. All patients who received trial medication were included in the safety evaluation

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes

Other biasUnclear riskFunding of study supported by Novartis


MethodsStudy design: single centre, randomised, double blind, controlled trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: France


ParticipantsNumber: 270 (enoxaparin n = 135; placebo n = 135)
Age: (mean years ± SD) enoxaparin 79.9 ± 6.8; placebo 80.1 ± 6.9
Sex: (M/F) enoxaparin 84/51; placebo 83/52.
Inclusion criteria: patients > 65 years, hospitalised with acute medical illness.
Exclusion criteria: ongoing anticoagulant or antiplatelet therapy; presence of active bleeding; presence of coagulation disorders; predictable short-term hospitalisation (< 7 days); iodine allergy.

Reason for hospitalisation (n (%)): not given, but diagnoses of included patients were given: heart failure (LMWH n = 24 (17.8), control n = 25 (18.5)); respiratory diseases (LMWH n = 33 (24.4), control n = 24 (17.8)); ischaemic stroke (LMWH n = 19 (14.1), control n = 27 (20.0)); malignant diseases (LMWH n = 25 (18.5), control n = 10 (7.4)); diabetes (LMWH n = 4 (3.0), control n = 8 (5.9)); depression (LMWH n = 4 (3.0), control n = 6 (4.4)); syncope (LMWH n = 5 (3.7), control n = 8 (5.9)); infection (including chest) (LMWH n = 6 (4.4), control n = 5 (3.7)); neurologic diseases (LMWH n = 3 (2.2), control n = 4 (3.0)); joint diseases (LMWH n = 4 (3.0), control n = 7 (5.2)); hepatic or biliary diseases (LMWH n = 2 (1.5), control n = 2 (1.5)); miscellaneous (LMWH n = 3 (2.2), control n = 5 (3.7))


InterventionsTreatment: LMWH (enoxaparin) 60 mg in 0.3 ml, subcutaneous injection, once daily
Control: placebo 0.3 ml, subcutaneous injection, once daily
Duration: 10 days


OutcomesDVT, PE, mortality, major haemorrhage

Haemorrhage definition: (major) "diffuse hemorrhage", not fully defined; (minor) not reported


NotesIsotopically detected DVT (I fibrinogen scanning at entry and every day or every other day) or documented symptomatic DVT or PE


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
Low riskStudy described as double-blind and control patients received matched placebo

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo indication of blinding of assessor

Incomplete outcome data (attrition bias)
All outcomes
Low riskExclusions were explained and were similar between groups

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: multi-centre, randomised, double blind, controlled trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: France


ParticipantsNumber: 442 (enoxaparin n = 217; UFH n = 225)
Age: (years ± SEM) enoxaparin 83.8 ± 0.51; UFH 82.6 ± 0.46
Sex: M/F ratio enoxaparin 29%/71%; UFH 27%/73%
Inclusion criteria: patients hospitalised with acute medical illness leading to reduced mobility
Exclusion criteria: contraindication to anticoagulation; ongoing venous, arterial or cardiac disease requiring anticoagulation; history or allergy or thrombocytopaenia induced by UFH or LMWH; abnormal platelet count; any contraindication to isotopic or venographic investigations; renal disorders; local disorders of the lower limb that would interfere with FUT

Reason for hospitalisation (n (%)): heart failure (enoxaparin n = 40 (18.5), UFH n = 46 (20.6)); bronchopulmonary infections (enoxaparin n = 49 (22.7), UFH n = 58 (26.0)); ischaemic stroke (enoxaparin n = 18 (8.3), UFH n = 20 (9.0)); cancer (enoxaparin n = 11 (5.1), UFH n = 19 (8.5)); malnutrition (enoxaparin n = 15 (6.9), UFH n = 18 (8.1)); dehydration (enoxaparin n = 39 (18.1), UFH n = 41 (18.4)); systemic infection (enoxaparin n = 8 (3.7), UFH n = 7 (3.1)); other (enoxaparin n = 125 (57.9), UFH n = 121 (54.3))


InterventionsTreatment 1: LMWH (enoxaparin) 20 mg, subcutaneous injection, once daily, plus placebo (mannitol) once daily
Treatment 2: UFH 5000 IU, subcutaneous injection, twice daily
Duration: 10 days


OutcomesDVT, PE, mortality, major haemorrhage

Haemorrhage definition: (major) clinical and associated with either a fall in haemoglobin 20 g/L, need for transfusion of two or more blood units, or if bleeding was retroperitoneal or intracranial; (minor) not reported


NotesIsotopically detected DVT (FUT at entry and every day or every other day) or documented symptomatic DVT or PE


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer-generated randomisation schedule

Allocation concealment (selection bias)Low riskAllocation concealment performed using a computer-generated randomisation schedule, and the code was unbroken

Blinding of participants and personnel (performance bias)
All outcomes
Low risk"Patients, nurses and doctors were unaware of the treatment assigned to each patient...". Study treatment provided in pre-filled syringes

Blinding of outcome assessment (detection bias)
All outcomes
Low riskEvaluation by blinded centralised interpreters

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing data accounted for and unlikely to affect results

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: multi-centre, randomised, open trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: France


ParticipantsNumber: 295 (nadroparin n = 146; UFH n = 149)
Age: (mean years ± SD) nadroparin 82.8 ± 0.5; UFH 83.8 ± 0.6
Sex: (M/F) nadroparin 36/110; UFH 38/111
Inclusion criteria: patients hospitalised with acute medical illness for an estimated minimum duration of 4 weeks; recent transient reduced mobility; absence of a recent DVT confirmed by echo-Doppler 48 hours prior to inclusion in the study
Exclusion criteria: previous reduced mobility; existence of echo-Doppler confirmed DVT prior to entry into study; previous DVT (within the last 6 months) or PE (within the last year); surgical intervention (non-orthopaedic 8 days; orthopaedic 28 days); cerebral haemorrhage within the previous 6 months or any neurosurgical intervention; renal insufficiency; sever liver disease, severe arterial hypertension; contraindication to heparin, or venographic or angiographic investigations; patients receiving aspirin, ticlopidine, anti-inflammatories, heparinoid or vitamin K

Reason for hospitalisation (n (%)): all participants hospitalised with transient locomotor disability, no further description


InterventionsTreatment 1: LMWH (nadroparin) 3075 antiXa units, subcutaneous injection, once daily
Treatment 2: UFH 5000 to 7500 IU, subcutaneous injection, three times daily
Duration: 28 days


OutcomesDVT, PE, mortality, major and minor haemorrhage, thrombocytopaenia

Haemorrhage definition: (major) requiring the cessation of treatment, such as retroperitoneal, digestive track bleeding with haematemesis or meleana; (minor) events not requiring cessation of treatment, such as haematoma at injection site microscopic haematuria or benign epistaxis


NotesDocumented symptomatic DVT or PE by Doppler ultrasound


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomisation by drawing of lots

Allocation concealment (selection bias)Low riskAllocation using sealed envelopes

Blinding of participants and personnel (performance bias)
All outcomes
High riskDescribed and an open study

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo indication of blinding of assessors

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing data accounted for and unlikely to affect results

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: Phase III, multi-centre, randomised, double blind trial, with 2 parallel groups
Intention-to-treat analysis: yes
Country: France


ParticipantsNumber: 223 (nadroparin n = 109; placebo n = 114)
Age: (mean years ± SD) nadroparin 69.4 ± 7.7; placebo 66.8 ± 8.2
Sex: (M/F) nadroparin 87/22; placebo 87/27
Inclusion criteria: patients age 40 to 80 years; hospitalised with COPD
Exclusion criteria: confirmed DVT within previous 6 months; Doppler-confirmed DVT at inclusion to study; gastroduodenal ulcer; recent haemorrhagic CVA; severe liver failure; severe renal impairment; confirmed or uncontrolled hypertension; congential or acquired coagulation disorder; history of hypersensitivity or thrombocytopaenia to heparins of any type; contraindications to anticoagulant therapy, venography or angiography; receiving aspirin, ticlopidine or oral anticoagulants

Reason for hospitalisation (n (%)): all participants had COPD


InterventionsTreatment: LMWH (nadroparin) 3800 to 5700 antiXa units, subcutaneous injection, once daily
Control: placebo (0.9% saline), subcutaneous injection, once daily
Duration: until patient weaned from mechanical ventilation, not exceeding 21 ± 1 days


OutcomesDVT, PE, mortality, major and minor haemorrhage, thrombocytopaenia

Haemorrhage definition: (major) overt, associated with a decrease in haemoglobin of 20 g/L compared with baseline, necessitated a transfusion of two or more units of packed red cells, retroperitoneal or intracranial, when the investigator decided to end the treatment with heparin because of benefit.risk ratio; (minor) bleeding not considered major


NotesDVT detection before and every 7 days until end of study; Venography at end of treatment or documented symptomatic DVT or PE


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
Low riskStudy described as double-blind and use of a matched placebo control

Blinding of outcome assessment (detection bias)
All outcomes
Low riskDVT diagnosis confirmed by two blinded, independent radiological experts; adverse events confirmed by blinded Critical Events committee

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing data accounted for and unlikely to affect results

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes; methods intended for assessing the outcome was unclear

Other biasUnclear riskFunding of study supported by Sanofi


MethodsStudy design: single centre, randomised trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: Canada


ParticipantsNumber: 350 (UFH n = 169; control n = 181)
Age (medical patients): (mean years) UFH 65 (range 44-83); control 63 (range 43-85)

Sex: (M/F) UFH 28/10; control 31/9
Inclusion criteria: > 40 years old; admitted for elective surgery, emergency surgery after fracture of the femoral neck, and medical patients suspected of having myocardial infarction admitted to a coronary and intensive medical-care ward
Exclusion criteria: bleeding tendency; iodine allergy; history of DVT or PE within the previous year

Reason for hospitalisation (n (%)): all included participants hospitalised for heart failure


InterventionsTreatment: UFH 5000 IU, subcutaneous, three times daily
Control: no specific prophylaxis (no placebo injections)
Duration: until mobile


OutcomesDVT, PE


NotesIsotopically detected DVT (FUT at entry, daily for first 4 days, and every 2nd day after that), confirmed by venography

For the purposes of this analysis, only participants hospitalised with heart failure were considered, Number: 26 (UFH n = 11; control n = 15)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Low riskNumbered sealed envelopes

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo indication of blinding or use of placebo

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo indication of blinding of assessors

Incomplete outcome data (attrition bias)
All outcomes
Low riskAnalysis performed using all randomised participants, although no indication of loss to follow-up

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: multi-centre, randomised, unblinded, controlled trial, with 2 parallel groups

Intention to treat: yes

Country: Sweden


ParticipantsNumber: 11693 (UFH n = 5776; control n = 5917)

Age: (mean years) 75.0

Sex: (M/F) not reported

Inclusion criteria: aged ≥ 55 years; admitted to departments of infectious diseases

Exclusion criteria: pre-existing anticoagulant treatment; readmission within 60 days of randomisation; ability to be mobile; assessment of contraindications not possible; persistent haemorrhage or increased risk of bleeding complication; heparin prophylaxis judged to be indicated by the responsible doctor; severe renal failure or liver failure; HIV infection; terminal disease in which active treatment was withheld; data not available

Reason for hospitalisation (n (%)): pneumonia (UFH n = 1617 (28.0), control n = 1646 (27.8)); skin and soft-tissue infections (UFH n = 1063 (18.4), control n = 1114 (18.8)); fever/sepsis (UFH n = 767 (13.3), control n = 843 (14.2)); urinary-tract infection (UFH n = 560 (9.7), control n = 586 (9.9)); gastroenteritis (UFH n = 517 (8.9), control n = 505 (8.5)); upper-respiratory-tract infection (UFH n = 461 (8.0), control n = 436 (7.4)); bone/joint infection (UFH n = 258 (4.5), control n = 253 (4.3)); skin rash/eruption (UFH n = 144 (2.5), control n = 144 (2.4)); meningitis (UFH n = 77 (1.3), control n = 70 (1.2)); liver disease (UFH n = 22 (0.4), control n = 30 (0.5)); other infectious diease (UFH n = 185 (3.2), control n = 165 (2.8)); other non-infectious disease (UFH n = 105 (1.8), control n = 125 (2.1))


InterventionsTreatment: 5000 IU standard sodium heparin, subcutaneous injection the abdominal area, every 12 hours

Control: no prophylactic treatment (no placebo injections)

Duration: maximum 21 days


Outcomesprimary endpoint was necropsy-verified PE of a size likely to lead to death


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
High riskStudy described as unblinded

Blinding of outcome assessment (detection bias)
All outcomes
Low riskNecropsy performed by pathologist blinded to treatment group. Significance of necropsy finding assessed by investigator unaware of treatment assignment

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo discussion of control dropout, due to study design, so cannot compare attrition rates between study groups

Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all the expected outcomes

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: single centre, randomised, open trial with five parallel groups

Intention to treat: not indicated

Country: Mexico


ParticipantsNumber: 193 (heparin n = 39; GCS n = 39; EB n = 33; ASA n = 35; control n = 46)

Age: (mean years) heparin 64.2; GCS 66.5; EB 65.7; ASA 62.5; control 62.2

Sex: (M/F) heparin 22/17; GCS 20/19; EB 14/19; ASA 18/17; control 23/23

Inclusion criteria: ≥ 40 years; pulmonary disease likely to enforce bed rest for three consecutive days or more

Exclusion criteria: patient's or physician's refusal to cooperate or rejection of the preventive method; PE, DVT, medical history of PE or DVT; clinical haemorrhage from any source; cerebrovascular accident; vasculitis; pericarditis; coagulation disorders; use of antithrombotic agents; recent postoperative status; allergy to iodine

Reason for hospitalisation (n (%)): all participants with pulmonary disease (COPD: UFH n = 23 (59.0), control n = 25 (54))


InterventionsTreatment 1: heparin 5000 IU, subcutaneous injections, every 12 hours

Treatment 2: GCS worn 24 hours a day

Treatment 3: EB reapplied every 8 - 12 hours

Treatment 4: ASA 0.5 g, orally, twice daily

Control: passive exercise and massage of lower extremities every 2 hours

Duration: Until patients were ambulatory (range 3 - 44 days)


OutcomesDVT, PE, all-cause mortality, major and minor bleeding

Haemorrhage definition: (major) not specifically defined, but patients with gastrointestinal bleeding were considered major bleeding; (minor) not specifically defined, but patients with ecchymosis and petechiae were considered minor bleeding


NotesDVT was detected using venography, venous Doppler, strain gauge plethysmography, fibrinogen uptake test and diagnosed using phlebogram

PE diagnosed by perfusion and ventilation lung scans

For the purposes of this analysis only the heparin and control populations are considered. Although the study reports an initial 221 participants, only 192 are considered for analysis, as the 29 discharged patients were not indicated as randomised, and if they were, no information was given on which treatment group they were in


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskFirst 50 participants were assigned to the control group, and the remaining were assigned at random, but no description of sequence generation

Allocation concealment (selection bias)Unclear riskNo description of allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes
High riskDescribed as an open experiment with no matched placebo

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo indication of blinding assessors

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk29 participants were discharged for described reasons, but not from which treatment group

Selective reporting (reporting bias)Low riskReport includes all expected outcomes, although no protocol is available

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: multi-centre, double-blind, placebo-controlled randomised trial, with 2 parallel groups

Intention-to-treat analysis: yes

Countries: China, India, Korea, Malaysia, Mexico, Philippines, Tunisia


ParticipantsNumber: 8319 (enoxaparin n = 4174; placebo n = 4145)

Age: (mean years ± SD) enoxaparin 65.6 ± 12.0; placebo 65.3 ± 12.2

Sex: (M/F) enoxaparin 2603/1568; placebo 2608/1528

Inclusion criteria: male or female ≥ 40 years-old; hospitalised within 48 hours of randomisation with at least one of the following: acute decompensated heart failure, active cancer (diagnosis within 6 months) - unless hospitalised for chemotherapy, or severe systemic infection and at least one of chronic pulmonary disease / obesity (BMI > 30) / previous VTE / age > 60; anticipated duration of hospitalisation at least 6 days; health status: American Society of Anesthesiologists Health status score ≤ 3 , Eastern Cooperative Oncology Group performance status ≤ 2 in cancer patient; anticipated life expectancy > 1 week; signed written informed consent

Exclusion criteria: major surgery or major trauma within the previous 6 weeks; need for any ventilatory support; symptomatic VTE at enrolment; multi organ failure; evidence of an active bleeding disorder; contraindication to anticoagulation; cerebrovascular accident at inclusion and within 10 days prior study inclusion; prosthetic heart valves; confirmed cerebral metastases; known hypersensitivity to heparin or LMWH, or pork-derived products; history of documented episode of heparin or LMWH induced thrombocytopaenia and/or thrombosis (HIT, HAT, or HITTS); participating in another clinical trial within the previous 30 days; persistent renal failure; known or suspected severe anaemia of unexplained cause considered clinically relevant by investigator; spinal or epidural analgesia or lumbar puncture within the preceding 24 hours; unlikely to be compliant (e.g. alcohol, drug abuse); women of childbearing potential not protected by effective contraceptive method of birth control and/or who are unwilling to be tested for pregnancy; refusal or inability to give informed consent to participate in the study; inability to be followed-up after discharge until day 90 after randomisation

Reason for hospitalisation (n (%)): heart failure (enoxaparin n = 1280 (30.7), control n = 1297 (31.4)); severe systemic infection (enoxaparin n = 2383 (57.2) , control n = 2336 (56.5)); active cancer (enoxaparin n = 195 (4.7), control n = 170 (4.1)); heart failure and systemic infection (enoxaparin n = 253 (6.1), control n = 262 (6.3)); heart failure and active cancer (enoxaparin n = 7 (0.2), control n = 5 (0.1)); severe systemic infection and active cancer (enoxaparin n = 45 (1.1), control n = 62 (1.5)); heart failure, severe systemic infection and active cancer (enoxaparin n = 3 (0.1), control n = 2 (< 0.1)); none of the above (enoxaparin n = 24 (0.6), control n = 25 (0.6))


InterventionsTreatment: enoxaparin 40 mg, subcutaneous injection daily

Control: placebo (0.9% saline), subcutaneous injection daily

Duration: 10 ± 4 days


OutcomesEfficacy: Primary outcome was death from any cause / cardiopulmonary death / sudden death or PE between time of randomisation and day 30. Secondary outcome was death from any cause / cardiopulmonary death / sudden death or PE between time of randomisation and day 14 and time of randomisation and day 90

Safety: major haemorrhage, clinically relevant non-major bleeding, minor bleeding, thrombocytopaenia, heparin induced thrombocytopaenia during treatment period

Haemorrhage definition: (major) overt bleeding associated with death, need for transfusion of at least 2 units of packed red cells or whole blood, a fall in haemoglobin of 20 g/L, requirement for major therapeutic intervention to control bleeding, a bleeding site that was retroperitoneal, intracranial or intraocular; (minor) overt bleeding that did not meet the criteria for major haemorrhage


NotesParticipants were also wearing elastic stocking with graduated compression

For the purposes of this analysis, data collected at day 30 after randomisation, which corresponds with the primary outcome


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskTreatment code list of random permuted blocks were generated by an independent contract research organization and stratified according to centre. Investigators assigned patients in sequential order of treatment number available at centre

Allocation concealment (selection bias)Low riskInvestigators assigned patients in sequential order of treatment number available at centre

Blinding of participants and personnel (performance bias)
All outcomes
Low riskInvestigators and patients were unaware of treatment assignments

Blinding of outcome assessment (detection bias)
All outcomes
Low riskSteering committee / data and safety monitoring members were unaware of treatment assignments

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data

Selective reporting (reporting bias)Low riskStudy protocol is available and all but one of the study's pre-specified outcomes have been reported. Thrombocytopaenia was not reported but there were no case of heparin induced thrombocytopaenia

Other biasUnclear riskFunding of study supported by Sanofi


MethodsStudy design: Phase III, multi-centre, randomised, double blind trial with 3 parallel groups
Intention-to-treat analysis: yes
Countries: Canada and various European countries


ParticipantsNumber: 1102 (enoxaparin 20 mg n = 364; enoxaparin 40 mg n = 367; placebo n = 371).
Age: (mean years ± SD) enoxaparin 20 mg 72.9 ± 10.1; enoxaparin 40 mg 73.1 ± 10.8; placebo 74.1 ±10.6
Sex: (M/F) enoxaparin 20 mg187/176; enoxaparin 40 mg 171/196; placebo 192/178
Inclusion criteria: patients > 40 years, hospitalised with an acute medical illness who were not immobilised for more than 3 days
Exclusion criteria: women of child-bearing age if pregnant, breast-feeding, or not using contraception; stroke or major surgery in the previous 3 months; contraindications to iodinated contrast media; serum creatinine concentration > 1.7 mg/dl; intubation; thrombophilia; HIV; uncontrolled arterial hypertension; active peptic ulcer; bacterial endocarditis; sensitivity to heparin or HIT, or platelet count < 100,000 x 109/L; patients requiring anticoagulation

Reason for hospitalisation (n (%)): congestive heart failure NYHA class III (enoxaparin 40 mg n = 103 (28.1), enoxaparin 20 mg n = 76 (20.9), control n = 95 (25.7)); congestive heart failure NYHA class IV (enoxaparin 40 mg n = 26 (7.1), enoxaparin 20 mg n = 44 (12.1), control n = 32 (8.6)); acute respiratory failure (enoxaparin 40 mg n = 195 (53.1), enoxaparin 20 mg n = 192 (52.9), control n = 202 (54.6)); acute infectious disease (enoxaparin 40 mg n = 197 (53.7), enoxaparin 20 mg n = 194 (53.4), control n = 193 (52.2)); acute rheumatic disorder (enoxaparin 40 mg n = 28 (7.6), enoxaparin 20 mg n = 40 (11.0), control n = 32 (8.6)); inflammatory bowel disease (enoxaparin 40 mg n = 3 (0.8), enoxaparin 20 mg n = 1 (0.3), control n = 1 (0.3))


InterventionsTreatment 1: LMWH (enoxaparin) 20 mg, subcutaneous injection, once daily
Treatment 2: LMWH (enoxaparin) 40 mg, subcutaneous injection, once daily
Control: placebo (0.2 ml isotonic saline), subcutaneous injection, once daily
Duration: 6 to 14 days


OutcomesDVT, PE, mortality, major and minor haemorrhage, thrombocytopaenia

Haemorrhage definition: (major) overt bleeding associated with the need for transfusion of two or more units of packed red cells or whole blood or with a decrease in haemoglobin of 20 g/L or more from baseline, or if bleeding was retroperitoneal, intracranial or fatal; (minor) overt bleeding events not meeting the criteria for major bleeding


NotesVenographically detected DVT or documented symptomatic DVT or PE

For the purposes of this analysis data used was for days 1 to 14, which corresponds to the primary outcome


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Low riskUse of a central location for randomisation

Blinding of participants and personnel (performance bias)
All outcomes
Low riskStudy described as double-blinded with use of a matched placebo control and prefilled syringes for treatment and placebo

Blinding of outcome assessment (detection bias)
All outcomes
Low riskOutcomes were reviewed by 2 independent committees whose members were unaware of treatment assignments

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing data accounted for and unlikely to affect results

Selective reporting (reporting bias)Low riskReport includes all expected outcomes

Other biasUnclear riskFunding of study supported by Rhône-Poulenc (Sanofi)


MethodsStudy design: Phase III, multi-centre, randomised, double blind trial with 2 parallel groups
Intention-to-treat analysis: yes
Countries: Africa, Australia, Canada, Europe, Israel, Lebanon, North and South America


ParticipantsNumber: 3706 (Treatment n = 1856; Control n = 1850)
Age: (mean years ± SD) dalteparin 68.5 (11.1); placebo 68.5 (11.7)
Sex: (M/F) dalteparin 884/964; placebo 888/945
Inclusion criteria: patients ≥ 40 years with an acute medical illness requiring a projected hospitalisation ≥ 4 days and had ≤ 3 days of prior immobilisation
Exclusion criteria: acute coronary syndrome or major surgical or invasive procedure within previous month; bacterial endocarditis; immobilised lower limb due to cast or fracture, stroke within 3 months; high risk of bleeding; platelet count < 100,000 x 109/L; contraindication to heparin; hepatic insufficiency; pregnancy or breastfeeding; or life expectancy < 1 month

Reason for hospitalisation (n (%)): acute congestive heart failure NYHA class III or IV (dalteparin n = 965 (52.2), control n = 940 (51.3)); acute respiratory failure (dalteparin n = 561 (30.4), control n = 560 (30.6)); other acute conditions (dalteparin n = 749 (40.5), control n = 781 (42.6)): infectious disease (dalteparin n = 673 (36.4), control n = 687 (37.5)); rheumatological disorder (dalteparin n = 200 (10.8), control n = 198 (10.8)); inflammatory bowel disease (dalteparin n = 10 (0.5), control n = 8 (0.4))


InterventionsTreatment: LMWH (dalteparin) 5000 IU, subcutaneous injection, once daily
Control: placebo, subcutaneous injection, once daily
Duration: 14 days


OutcomesDVT, PE, mortality, major and minor haemorrhage, thrombocytopaenia

Haemorrhage definition: (major) intraocular, spinal/epidural, intracranial or retroperitoneal; if haemoglobin decreased by 20 g/L or more, if a transfusion of two or more units of blood were needing, if significant medical or surgical intervention was needed, or if fatal; (minor) all other bleeding events not fitting criteria of major bleeding


Notes Compression ultrasound detected proximal DVT or confirmed symptomatic DVT or PE

For the purposes of this analysis outcome data measured at day 21 was used.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
Low riskStudy described as double-blinded with use of matched placebo control

Blinding of outcome assessment (detection bias)
All outcomes
Low risk"All clinical end points were centrally adjudicated by a blinded Clinical Events Committee"

Incomplete outcome data (attrition bias)
All outcomes
Low riskExclusion were explained and similar between groups

Selective reporting (reporting bias)Low riskReport includes all expected outcomes

Other biasUnclear riskFunding of study supported by Pharmacia Corporation


MethodsStudy design: Phase III, multi-centre, randomised, double blind trial with 2 parallel groups
Intention-to-treat analysis: yes
Country: Germany and Austria


ParticipantsNumber: 959 (enoxaparin n = 477; UFH n = 482)
Age: (years) enoxaparin 74 ± 13; UFH 74 ± 13
Sex: (M/F) enoxaparin 183/294; UFH 178/304
Inclusion criteria: patients > 18 years; hospitalised with acute medical illness and at least one additional thromboembolic risk factor (e.g. > 60 years, malignancy, obesity, former thromboembolic event, cardiac insufficiency, paresis of the lower limbs, hemiplegia, paraplegia, or sever infection); expected immobilisation for more than half of the day-time for the whole study period (7 days)
Exclusion criteria: treatment with anticoagulants, antiplatelets or NSAIDs in the preceding 7 days; regional anaesthesia; pregnancy and lactation; bleeding disorder; thrombocytopaenia < 100,000 x 109/L; head trauma in the previous 6 months; haemorrhagic stroke the preceding 4 weeks; endocarditis; severe liver disease or renal insufficiency; thromboembolism; or participation in a clinical trial in the preceding 6 weeks

Reason for hospitalisation (n (%)): cardiovascular diseases (enoxaparin n = 322 (67.5), UFH n = 340 (70.5)); endocrinologic diseases (enoxaparin n = 133 (27.9), UFH n = 145 (30.1)); respiratory diseases (enoxaparin n = 116 (24.3), UFH n = 113 (23.4)); gastrointestinal and urogenital diseases (enoxaparin n = 108 (22.6), UFH n = 105 (21.8)); central nervous diseases (enoxaparin n = 75 (15.8), UFH n = 86 (17.8)); cancer (enoxaparin n = 70 (14.7), UFH n = 62 (12.9)); bone diseases (enoxaparin n = 51 (10.8), UFH n = 59 (12.2)); skin diseases (enoxaparin n = 17 (3.5), UFH n = 15 (3.1)); other (enoxaparin n = 39 (8.2), UFH n = 43 (8.9))


InterventionsTreatment 1: LMWH (enoxaparin) 40 mg in 0.2 ml, subcutaneous injection, once daily, plus twice daily placebo subcutaneous injections (0.2ml isotonic mannitol solution)
Treatment 2: UFH 5000 IU, subcutaneous injection, three times daily
Duration: 7 days


OutcomesDVT, PE, mortality, major haemorrhage

Haemorrhage definition: (major) a decrease in haemoglobin of 20 g/L or more, a transfusion of 2 or more units of blood, intracranial or retroperitoneal bleeding; (minor) not reported


NotesUltrasound detected DVT before and at end of study, confirmed by venography or documented symptomatic DVT. PE confirmed by pulmonary perfusion scan, angiography or autopsy


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod not stated

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
Low riskStudy described as double-blinded with use of a matched placebo control

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo indication of blinding of assessors

Incomplete outcome data (attrition bias)
All outcomes
Low riskDescribed treatment and control populations as being similar, although concern that there are no reasons given for exclusion

Selective reporting (reporting bias)Low riskReport includes all expected outcomes

Other biasLow riskThe study appears to be free from other sources of bias


MethodsStudy design: multi-centre, controlled, randomised, open, with 2 parallel groups

Intention-to-treat: efficacy outcomes no; safety outcomes yes

Country: Germany


ParticipantsNumber: 665 patients (enoxaparin n = 332; UFH n = 333)

Age: (mean years ± SD) enoxaparin 70 ± 14; UFH 70 ± 14

Sex: (M/F) enoxaparin 160/172; UFH 183/150

Inclusion criteria: ≥ 18 years old; hospitalised for severe respiratory disease or heart failure; confined to bed for more than two thirds of each day

Exclusion criteria: advanced acquired immunodeficiency syndrome; evidence of contraindications for LMWH or UFH therapy; hypersensitivity to contrast media; severe pancreatic, hepatic or renal disease; severe arterial hypertension; intracranial bleeding or haemorrhagic stroke in preceding 6 months; ocular or CNS surgery in preceding 4 weeks; coagulation disorders; drug or alcohol abuse; acute signs of DVT/PE; gastrointestinal ulcer; immobilised for > 24 hours before enrolment; taking anticoagulants/platelet inhibitors or nonsteroidal anti-inflammatory drugs

Reason for hospitalisation (n (%)): respiratory disease (enoxaparin n = 168 (50.6), UFH n = 164 (49.4)); heart failure (enoxaparin n = 164 (49.2), UFH n = 169 (50.8))


InterventionsTreatment 1: enoxaparin 40 mg, subcutaneous injection, once daily

Treatment 2: UFH 5000 IU, subcutaneous injection, three times daily

Duration: 10 ± 2 days


OutcomesDVT, PE, mortality, major haemorrhage

Haemorrhage definition: (major) intracranial or retroperitoneal, overt bleeding with either a decreased in haemoglobin of 20 g/L or more, or a transfusion of at least two units of blood; (minor) not reported


NotesDVT detected by fibrin monomer/D-dimer tests and verified by venography or autopsy

Original number of randomised participants is 668, but as the three participants that withdrew do not have an explanation of which treatment group they came from this analysis can only reliably use n = 665 (enoxaparin n = 332; UFH n = 333)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomisation performed separately in blocks of 6 patients for each of the 2 disease groups to guarantee a stratified enrolment, but this does not describe the method of sequence generation

Allocation concealment (selection bias)Unclear riskMethod not stated

Blinding of participants and personnel (performance bias)
All outcomes
High riskStudy described as open label with no use of a matched placebo

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskEfficacy outcomes (low risk): blinded evaluation of efficacy endpoint by independent central reviewers

Safety outcomes (unclear risk): blinding not indicated for safety endpoints as explicitly stated for efficacy endpoints

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskEfficacy outcomes (high risk): "...more patients receiving UFH discontinued the study because of adverse effects or serious adverse events, including death."

Safety outcomes (low risk): all participants reported for safety outcomes

Selective reporting (reporting bias)Low riskReport includes all expected outcomes

Other biasUnclear riskFunding of study supported by Rhône-Poulenc (Sanofi)

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Aquino 1990Elderly patients with an acute medical disorder, however also included patients undergoing orthopaedic rehabilitation

Cade 1982aVirtually all medical patients with DVT had suffered a recent acute MI. DVT in comparison study found mainly in patients who had undergone recent vascular or cancer surgery

Cade 1982bPatients from coronary care unit post-MI and patients who had recently undergone surgery as well as general medical patients

EXCLAIMAll included patients received enoxaparin and then were randomised to placebo or continued on enoxaparin

Halkin 1982Patients randomised based on hospital number (odd or even). Patient eligibility determined after allocation to treatment arm. 411 of 669 patients (even hospital number) received heparin, 689 patients with odd number acted as control

Harenberg 1990Included patients with acute coronary and cerebrovascular disease, unable to remove from final data analysis

HESIMIncluded patients with acute coronary and cerebrovascular disease, unable to remove from final data analysis

Manciet 1990Elderly patients with an acute medical disorder, however also included patients undergoing orthopaedic rehabilitation

Mottier 199310% of acute medical diseases were stroke

Poniewierski 1988No DVT identified in either treatment group. Method used to identify DVT, thermography, not sensitive enough to detect asymptomatic DVT

PROMPTThis study of hospitalised general medical > 60 years included patients with MI and cerebrovascular disease. It was not possible to extract separate data from the analysis

 
Comparison 1. Heparin versus placebo or no treatment

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Deep vein thrombosis75511Odds Ratio (M-H, Random, 95% CI)0.41 [0.25, 0.67]

 2 Non-fatal pulmonary embolism65485Odds Ratio (M-H, Fixed, 95% CI)0.46 [0.20, 1.07]

 3 Fatal pulmonary embolism627563Odds Ratio (M-H, Fixed, 95% CI)0.71 [0.43, 1.15]

 4 Combined non-fatal and/or fatal pulmonary embolism927971Odds Ratio (M-H, Fixed, 95% CI)0.66 [0.43, 1.02]

 5 All cause mortality727786Odds Ratio (M-H, Fixed, 95% CI)0.97 [0.87, 1.08]

 6 Major bleeding713804Odds Ratio (M-H, Fixed, 95% CI)1.65 [1.01, 2.71]

 7 Minor bleeding513434Odds Ratio (M-H, Fixed, 95% CI)1.61 [1.26, 2.08]

 8 Thrombocytopaenia413349Odds Ratio (M-H, Fixed, 95% CI)1.05 [0.64, 1.74]

 
Comparison 2. Low molecular weight heparin versus unfractionated heparin

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Deep vein thrombosis65942Odds Ratio (M-H, Fixed, 95% CI)0.77 [0.62, 0.96]

 2 Non-fatal pulmonary embolism65942Odds Ratio (M-H, Fixed, 95% CI)0.93 [0.42, 2.08]

 3 Fatal pulmonary embolism23581Odds Ratio (M-H, Fixed, 95% CI)0.33 [0.01, 8.14]

 4 Combined non-fatal and/or fatal pulmonary embolism65942Odds Ratio (M-H, Fixed, 95% CI)0.86 [0.39, 1.90]

 5 All cause mortality55605Odds Ratio (M-H, Fixed, 95% CI)0.79 [0.54, 1.16]

 6 Major bleeding65942Odds Ratio (M-H, Fixed, 95% CI)0.43 [0.22, 0.83]

 7 Minor bleeding33876Odds Ratio (M-H, Fixed, 95% CI)0.70 [0.48, 1.00]

 8 Thrombocytopaenia33876Odds Ratio (M-H, Fixed, 95% CI)0.41 [0.08, 2.11]