Table 1– Study and patient characteristics, and risk of bias assessment
First author [Ref.]DesignRisk of bias (exposure, patient selection, consecutive, follow-up, outcome)Definition of outpatient treatment or early dischargeOutcome measures and methodsPatients nAge yearsMalesMalignancies
Agterof [4]Prospective cohortYes, yes, yes, yes, yesDischarged immediately from ED or within 24 h after admissionRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, V′/Q′, autopsy or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: independent steering committee
152 outpatients53±1474 (49)20 (13)
Aujesky [5]RCTYes, yes, yes, yes, yesDischarged from ED or within 24 h of randomisationRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, autopsy or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: independent steering committee
171 outpatients
168 inpatients
47±1684 (49)1 (1)
Beer [20]Prospective cohortUnclear, yes, no, unclear, unclearUnclearNot described43 outpatients
54 inpatients
Davies [21]Prospective cohortNo, yes, unclear, yes, yesDiagnosis of PE confirmed within 72 h of initial assessmentThromboembolic complications (with objective confirmation)157 early discharge5886 (55)
Erkens [6]Retrospective cohortYes, yes, yes, yes, yesSent home from EDRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, V′/Q′, autopsy or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: Consensus of two investigators based on clinical records
260 outpatients55±17132 (51)83 (32)
Kovacs [22]Prospective cohortYes, yes, unclear, unclear, yesUnclearRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, V′/Q′, autopsy or extension of DVT on CUS
Major bleeding: according previous reported criteria [31]
81 outpatients
27 early discharge
5725 (23)
Kovacs [7]Retrospective cohortUnclear, yes, unclear, yes, yesUnclearRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, V′/Q′, autopsy or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: not described
314 outpatients54±18130 (41)62 (20)
Lui [23]Retrospective cohortYes, yes, yes, yes, unclearSent to “hospital in the home” within 24 h of arrivalDeath, unplanned return to hospital, unplanned staff callout, complications (recurrent PE, bleeding episode or other); methods not described21 outpatients569 (43)1 (5)
Olsson [24]Prospective cohortYes, yes, no, yes, yesUnclearRecurrent thromboembolism: V′/Q′ scan102 outpatients6345 (44)
Ong [25]Retrospective cohortYes, yes, no, yes, yesAdmitted directly into ambulant care programme via GP, specialist or EDRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, V′/Q′, autopsy or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: clinical records
60 outpatients
70 inpatients
Otero [26]RCTYes, yes, no, yes, yesPatients were randomised to hospitalisation or early discharge
Early discharge patients were discharged on day 3 (with TTE) or on day 5 (if TTE was not available)
Recurrent VTE: new intraluminal filling defect on CT or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: clinical records
132 inpatients60±1765 (49)6 (5)
Rodriguez-Cerrillo [27]Prospective cohort studyYes, yes, no, unclear, unclearUnclearRecurrent VTE: unclear how diagnosis was established
Major bleeding: according to ISTH criteria [30]
Mortality: methods not described
30 outpatients
31 inpatients
6726 (42)7 (12)
Siragusa [28]#Prospective cohortNo, yes, yes, unclear, yesUnclearRecurrent DVT: extension of thrombus on CUS or venography
Recurrent PE: new defect in V′/Q′ or CT lung scan, worsening of signs or symptoms, along with deterioration shown via chest radiograph, blood gases or ECG, or leg swelling with a positive CUS
Major bleeding: according to ISTH criteria [30]
Mortality: methods not described
36 outpatients6267/127 (53)36 (100)
Wells [29]#RCTYes, yes, yes, yes, yesUnclearRecurrent DVT: extension of thrombus on CUS; in doubt serial testing or venography was used
Recurrent PE: new defect on V′/Q′, angiography or CT lung scan according to PIOPED criteria. Patients who did not have high probability on V′/Q′ scan underwent further investigations (CUS leg, venography or angiography)
Major bleeding: according to ISTH criteria [30]
Mortality: methods not described; probably clinical records reviewed by independent committee
90 outpatients58±17273/505 (54)113/505
Zondag [8]Prospective cohortYes, yes, yes, yes, yesSent home from ED or within 24 h after admissionRecurrent VTE: new intraluminal filling defect on CT, pulmonary angiogram, V′/Q′, autopsy or extension of DVT on CUS
Major bleeding: according to ISTH criteria [30]
Mortality: clinical record or autopsy report reviewed by independent committee
297 outpatients55±15172 (58)28 (9)
  • Data are presented as n, mean±sd or n (%), unless otherwise stated. ED: emergency department; VTE: venous thromboembolism; CT: computed tomography; V′/Q′: ventilation/perfusion ratio; DVT: deep vein thrombosis; CUS: compression ultrasonography; ISTH: International Society on Thrombosis and Haemostasis; RCT: randomised controlled trial; PE: pulmonary embolism; GP: general practitioner; TTE: transthoracic echocardiography; PIOPED: Prospective Investigation of Pulmonary Embolism Diagnosis. #: baseline characteristics (age, male sex, malignancies) described for a mixed group of patients with DVT and PE together, not reported separately for patients with PE; : ambulatory care programme.