Thrombosis Research 126 (2010) e1–e5

Contents lists available at ScienceDirect

Thrombosis Research

j ourna l homepage: www.e lsev ie r.com/ locate / th romres

Regular Article

Home treatment in pulmonary embolism

Remedios Otero a,⁎, Fernando Uresandi b, David Jiménez c, Miguel Ángel Cabezudo d, Mikel Oribe e,Dolores Nauffal f, Francisco Conget g, Consolación Rodríguez h, Aurelio Cayuela i

a Service of Pneumology, Hospital Universitario Virgen del Rocío, Avda Manuel Siurot s/n, 41013 Seville, Spainb Service of Pneumology, Hospital Cruces, Bizkaiac Respiratory Department, Hospital Ramón y Cajal, Madridd Service of Pneumology, Hospital Central Asturias, Oviedoe Service of Pneumology, Hospital Galdakao, Bizkaiaf Service of Pneumology, Hospital La Fe, Valenciag Service of Pneumology, Hospital Lozano Blesa, Zaragozah Section of Pneumology, Internal Medicine Department, San Juan de Dios Hospital, Bormujosi Medical Record Department, Virgen del Rocío Hospital, Seville, Spain

⁎ Corresponding author. Tel.: +34 34 955013159; faxE-mail address: rotero@separ.es (R. Otero).

0049-3848/$ – see front matter © 2009 Elsevier Ltd. Aldoi:10.1016/j.thromres.2009.09.026

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 25 May 2009Received in revised form 7 September 2009Accepted 28 September 2009Available online 24 October 2009

Keywords:Early dischargelow molecular weight heparinpulmonary embolismanticoagulantshome treatmentdisease management

Background: Limited data exist on the feasibility of providing outpatient care to patients with acute pulmonaryembolism (PE).Methods:We conducted a multicenter randomized clinical trial in acute symptomatic PE to compare the efficacyand safety of early discharge versus standard hospitalization. A clinical prediction rule was used to identify low-risk patients. All patients were followed for three months. The primary outcomes were venous thromboembolicrecurrences, major and minor bleeding, and overall mortality.Results: One hundred and thirty two low-risk patients with acute symptomatic PE were randomized to earlydischarge (n=72) or standard hospitalization (n=60). Overall mortality was 4.2% (95% CI, 0.5-8.9) in the earlydischarge group and 8.3% (95% CI, 1.1-15) in the standard hospitalization group (Relative Risk (RR) 0.5; 95%confidence interval [CI], 0.12-2.01). Non-fatal recurrenceswere 2.8% (95%CI, 1.1-6.6) in the early discharge groupand 3.3% (95% CI, 1.3-8%) in the standard hospitalization group (RR 0.8; 95% CI, 0.12-5.74). The rates of clinicallyrelevant bleedingwere 5.5% in the early discharge group and 5% in the standard hospitalization group (P=0.60).

Short-term mortality was 2.8% (95% CI, 0.8-9.6%) in the early discharge group as compared with 0% in thestandardhospitalization group. Based on the rate of short-termdeath in a carefully selectedpopulation, the studywas suspended.Conclusions: In spite of the number of complications in patients with acute symptomatic PE randomized tostandard hospitalization or early discharge did not differ significantly. The rate of short-term mortality wasunexpectedly high in a (a priori) low-risk group of patients with acute PE. The accuracy of clinical predictionscores needs to be validated in well designed clinical trials. (ClinicalTrials.gov number, NCT00214929.)

© 2009 Elsevier Ltd. All rights reserved.

Introduction

Home treatment with subcutaneous low-molecular-weight-heparin(LMWH) is a safe and effective therapy for selected patients with acutedeep vein thrombosis (DVT) [1,2]. Treatment with LMWH is as safe andeffective for inpatientmanagement of acute PE as conventional treatmentwith unfractioned heparin (UFH) [3,4]. Some professional organizationssuch as The British Thoracic Society recommend consideration ofoutpatient treatment for clinically stablepatientswithPE [5,6].Outpatienttreatment in selected patients with PE would lead to substantial costsavings [7]. However, the safety of this approach has not been adequately

: +34 34 955013168.

l rights reserved.

evaluated, and caution should be exercised as PE is a potentially life-threatening disease.

Mortality rates in patientswith objectively confirmed acute PE duringthefirst 3 months of anticoagulation vary greatly from1.4% to 17.4% [8,9].Differences among studies might be explained through variation in thestudy design and selection of patients. Thus, identification of patients atlow risk for death and other adverse effects is mandatory for outpatienttreatment of clinically stable patients with PE. Several prognostic toolshave been used to risk stratify patients with acute PE, including clinicalprediction rules [10], cardiac biomarkers [11], and imaging testing [12].Clinical prediction rules that accurately classify patients with PE intoincreasing risk categories have been derived and validated [10,13,14].However, before low-risk patients with PE based on these models can betreated as outpatients, the safety and clinical usefulness of this approachmust be tested in a clinical trial.

e2 R. Otero et al. / Thrombosis Research 126 (2010) e1–e5

We designed a multicenter prospective randomized trial tocompare the efficacy and safety of an early discharge program withstandard hospitalization for acute management of PE.

Materials and methods

Study Organization

We conducted a randomized clinical trial to compare the efficacyand safety of home treatment with those of hospitalized therapy inpatients with acute symptomatic pulmonary embolism.

The steering committee had final responsibility for the studydesign, protocol, statistical analysis, study oversight, verification ofthe data, and data analysis. The protocol was approved by theinstitutional review board at each of the participating centers. Anindependent data and safety monitoring board (DSMB) periodicallyreviewed the study's outcomes and advised the steering committee.The DSMBwas composed of threemembers of the ethics committee ofVirgen del Rocio Hospital, Seville, Spain, and had the advice of twointernational experts on venous thromboembolic disease (Prof. Dr H.Bounameaux and Prof. Dr P. Wells).

Patients

Consecutive patients over 18 years of age who presented with acutesymptomatic PEwere eligible. Criteria for PEwere an intraluminalfillingdefect in subsegmental or more proximal pulmonary arteries on spiralcomputed tomography (CT), a high probability finding on a ventilation-perfusion lung scan, or a nondiagnostic finding with documented deepvein thrombosis [15,16]. A standardized clinical prediction rule wasused to identify patients with acute PE and low risk of death and short-term adverse events [17].

Patients were ineligible if they met one or more of the followingcriteria: a clinical score >2 points; hemodynamic instability atenrolment (defined as cardiogenic shock, systolic blood pressure<90 mmHg, or need of inotropic drugs support); T-troponin concen-trations of ≥0.1 ng mL-1; oxygen saturation <93%; need of hospitali-zation for other comorbidities; dyspnea (New York Heart Association[NYHA] III/IV); severe chronic obstructive pulmonary disease (FEV1<50% of predicted), severe asthma; active bleeding or high risk ofbleeding (subjectively assessed by the attending physician); recentsurgery (in the lastfifteendays); pregnancy;morbid obesity (bodymassindex [BMI] >30 Kg m-2); right ventricular dysfunction assessed bytransthoracic echocardiography (TTE) and defined when either the Acriterion or two of the B criteria below were fulfilled.

The A criterion: dilatation of the right ventricular cavity (rightventricle appearing larger than the left ventricle) from the apical,subcostal, or four chamber view, or a right ventricular end diastolicdiameter greater than 30 mm from the left precordial view at the levelof the mitral valve apparatus.

The B criteria: 1) tricuspid regurgitation jet velocity over 2-8 m/s asassessed by Doppler echocardiography or 2) over 2-5 m/s in theabsence of inspiratory collapse of the inferior vena cava; 3) dilatationof the right pulmonary artery (>12 mm/m2 body surface area) on thesuprasternal echocardiogram12; 4) right ventricular wall thicknessover 5 mm; 5) loss of inspiratory collapse of the inferior vena cava.Patients were also ineligible if they had a life expectancy of less than3 months.

After giving written informed consent, patients were randomlyassigned to early discharge or hospitalization. Allocation by randomsampling numbers was used in each center. Sealed, opaque envelopesarranged in a computer-generated random order were prepared bythe data coordinating center and distributed to each participatinginstitution.

Study interventions

Patients who were assigned to the early discharge group weredischarged in the third day after diagnosis (after TTE had ruled outright ventricular dysfunction), or in the fifth day (if TTE could not beperformed).

All patients received standard therapy with weight-adjusted dosesof LMWH. Vitamin K antagonist therapy was started on day 10 afterrandomization. After an initial “overlap” treatment period, patientswere continued on dose-adjusted acenocoumarol; [target INR of 2.5(therapeutic range 2.0-3.0)]. The INR was usually monitored dailyuntil the therapeutic range had been achieved, then twice or threetimes weekly for the first weeks, and then once a week to once amonth, depending on the stability of the results.

Surveillance and follow-up

Patients in the early discharge group were contacted by telephoneeach day up to 14 days. For all patients, two additional visits werescheduled at one and three months after recruitment. At each contact,patients were evaluated for symptoms and signs of recurrent venousthromboembolism and bleeding. Patients were instructed to report tothe study center immediately if any of these conditions occurred. Theprotocol required objective testing in cases of suspected recurrentpulmonary embolism or deep vein thrombosis.

Outcome assessment

The primary efficacy outcome was symptomatic recurrent venousthromboembolism, defined as objective assessment of recurrentpulmonary embolism, deep vein thrombosis or death attributed topulmonary embolism up to three months. A diagnosis of recurrent PEwas established by a new perfusion defect involving 75% or more of alung segment; or by the presence of a new intraluminal filling defector an extension of a previous filling defect on helical CT [16]. New orrecurrent DVT was diagnosed by the appearance of a new noncom-pressible vein segment, or a 4-mm ormore increase in the diameter ofa thrombus on complete lower limb ultrasound testing (CCUS) [18].

The main safety outcome was clinically relevant bleeding and deathfrom all causes. Bleeding was classified as major or minor. Major bleedswere defined as: 1) overt bleeding causing a fall in haemoglobin con-centration of >2 g/dL; 2) requirement for transfusion of two or moreunits of blood; 3) retroperitoneal or intracranial bleeding, or 4) bleedinginto a major prosthetic joint.

Deathwas classified as due to pulmonary embolism, bleeding or otherestablished diagnoses. Pulmonary embolismwas considered the cause ofdeath if there was objective documentation or if the cause of death wasunexplained and pulmonary embolism could not be confidently ruledout.

Statistical methods

On the basis of previous data [19], the rate of early complications(deaths, thromboembolic recurrences and bleeding) in patients withPE was 6.3%. The postest probability for a negative result in the clinicalpredictive score was 2.7% (95% CI, 1.3-5.4) [17]. Moreover we usedother exclusion criteria . Thus, we assumed a early complication rateless of 1%. It was estimated that at least 671 patients per group wouldbe required to show non-inferiority in absolute risk for early discharge(80% power; two-sided α=0.05). All analyses were performed on anintention-to-treat principle. The primary analysis of survival wasbased on the time from random assignment to death. Patients alive atthe end of follow-up were censored. Survival estimates werecalculated according to the Kaplan-Meier method. Ninety-five percentconfidence intervals were calculated when appropriate.

Fig. 1. Flow diagram of the study.

e3R. Otero et al. / Thrombosis Research 126 (2010) e1–e5

Categorical variables are presented by frequency counts. Normallydistributed data in groups were expressed as mean±standarddeviation (SD). Categorical baseline characteristics, adverse eventincidence rates and other categorical outcomes were comparedbetween treatments with Fisher's exact test or the chi-square test, as

Table 1Baseline characteristics of included and excluded patients.

Patients excluded or not meetin(n=572)

Clinical characteristicsFemale, n (%) 320 (56)Age, years 69±15.5

Risk factors for VTECancer, n (%) 140 (24.5)Surgery in the last 2 months, n (%) 62 (10.8)Immobilization longer than 4 days, n (%) 102 (17.8)

Renal functionCrCl<30 mL/min 116 (20.4)

Clinical presentationDyspnea, n (%) 445 (77.8)Chest pain, n (%) 230 (40.2)Hemoptysis, n (%) 28 (4.9)Syncope, n (%) 121 (21.3)DVT signs and symptoms, n (%) 139 (24.3)Shock index⁎≥1, n (%) 105 (18.4)Heart rate>100 bpm, n (%) 161 (28.2)SBP<90 mmHg, n (%) 48 (8.4)PaO2<60 mmHg, n (%) 224 (56.7)

Data are expressed as means±SD or counts and percentage, as appropriate.Abbreviations: VTE, venous thromboembolic diseases; DVT: deep vein thrombosis; CrCl: cr⁎Shock index: heart rate divided by systolic blood pressure.

appropriate. For continuous variables, the differences between twogroups were evaluated with an unpaired t test. Survival curves wereestimated using Kaplan-Meier survival methodology. SPSS version13.0 (Chicago, IL, USA) was used for data analysis; treatment effectswere tested at the two-sided alpha level of 0.05.

g inclusion criteria Patients enrolled in study(n=132)

P value

67 (51) 0.2860±16.7 0.03

6 (4.5) <0.0118 (13.6) 0.3613 (9.8) 0.02

7 (5.3) <0.01

100 (75.8) 0.6194 (71.2) <0.0111 (8.3) 0.1212 (9.2) <0.0134 (25.8) 0.735 (3.8) <0.0127 (20.5) 0.071 (0.8) <0.0120 (21.5) <0.01

eatinine clearance; bpm: beats per minute; SBP, systolic blood pressure.

Table 2Baseline characteristics of the patients.

Standard hospitalization[n=60]

Early discharge⁎[n=72]

P value

Clinical characteristicsFemale, n (%) 31 (52) 36 (50) 0.85Age, yrs 61.5±16 60.0±17.5 0.61

Risk factors for VTECancer, n (%) 3 (5) 3 (4.2) 0.82Surgery in the last 2 months,n (%)

10 (17) 8 (11) 0.35

Immobilization longer than4 days, n (%)

5 (8.3) 8 (11.1) 0.59

Renal functionCrCl<30 mL/min 3 (5) 4 (5.6) 0.60

Clinical presentationDyspnea, n (%) 48 (80) 52 (72) 0.30Chest pain, n (%) 43 (72) 51 (71) 0.54Hemoptysis, n (%) 5 (8.3) 6 (8.3) 0.63Syncope, n (%) 4 (6.8) 8 (11) 0.30DVT signs and symptoms, n (%) 14 (23) 17 (24) 0.57Shock index≥1, n (%) 3 (5) 2 (2.8) 0.41Heart rate >100 bpm, n (%) 13 (22) 14 (19) 0.46SBP<90 mmHg, n (%) 1 (1.7) 0 (0) 0.45PaO2<60 mmHg, n (%) 11 (27.5) 9 (17) 0.23

Data are expressed as means±SD or counts and percentage, as appropriate.Abbreviations: VTE, venous thromboembolic diseases; DVT: deep vein thrombosis;CrCl: creatinine clearance; bpm: beats per minute; SBP, systolic blood pressure.⁎ 44 of 72 patients were discharged on 3 day and 28 on 5 day.⁎Shock index: heart rate divided by systolic blood pressure.

Fig. 2. Kaplan-Meier survival curves (3-month follow-up) for PTE patients randomizedto conventional hospital care or early discharge.

e4 R. Otero et al. / Thrombosis Research 126 (2010) e1–e5

Results

Between February 2005 and April 2007, a total of 1016 patients withacute symptomatic PEwere screened and 132were randomly assigned.A diagram of the enrolment of patients and randomization is shown inFig. 1. No patients were lost to follow-up. Compared to patients whowere included, excluded patients were significantly older, had moreoften a history of respiratory disease, cancer and immobilization. Chestpain, syncope, hypotension, and renal insufficiency at admission weremore frequent in the patients who were excluded (Table 1).

Of the consenting patients, 60 were randomized to standardhospitalization and 72 to early discharge; 44 of 72 patients weredischarged on 3 day and 28 on 5 day. The baseline characteristics ofthe 2 treatment groups demonstrated no statistical differences withrespect to age, gender, risk factors for VTE, or severity of PE (assessedby systolic blood pressure, heart rate, PaO2, or shock index) (Table 2).

Outcome events are presented in Table 3. Overall, there were8 deaths (6.1%; 95% CI, 3.1-11.5) during the 3-month follow-upperiod. Short-term mortality was 2.8% (95% CI, 0.8-9.6%) in the earlydischarge group and 0% in the standard-hospitalization group(P=0.30). TTE was not performed in one patient who died in theearly discharge group because the discharge was on the fifth day after

Table 3Outcome events.

Early discharge group(n=72)

Sta(n

Days of hospitalization 3.4±1.1 9.3Overall mortality 3 (4.2) 5 (⁎Short-term mortality 2 (2.8) 0Non-fatal recurrences 2 (2.8) 2 (⁎Short-term non-fatal recurrences 1 (1.4) 0Bleeding 4 (5.5) 3 (Major bleeding 1 (1.4) 1 (Minor bleeding 3 (4.2) 2 (

⁎Within the first 10 days after diagnosis.

PE diagnosis. Kaplan-Meier curves showed no significant difference insurvival between the 2 group (P=0.78) (Fig. 2).

The incidence of recurrent venous thromboembolismwas 2.8% in theearly discharge group and 3.3% in the standard-hospitalization group, fora relative risk of 0.83 (95% CI, 0.12-5.74). All recurrences in both groupswere due to pulmonary embolism. The incidence of major bleeding was1.4% in the early discharge group and 1.6% in the standard-hospitalizationgroup, for a relative risk of 0.83 (95% CI, 0.05-13.04) (Table 3).

After the first 132 patients were enrolled, the DSMBwere alerted bythe unsuspected high mortality rate in a carefully selected population.Due to the evaluation by the DSMB, the steering committee decided toapply caution by suspending the study.

Autopsies were performed to investigate the causes of earlydeaths. One death (female patient, 63-yrs old, weight: 60 Kg) was dueto upper gastrointestinal bleeding occurring 5 days after discharge.She was treated with 1, 5 mgr/sc/24 h of Enoxaparin (Clexane forte®90 mgr/sc/24 h). At autopsy, no abnormalities in the gastrointestinaltract were found. Normal biochemical values and TTE were obtainedbefore her discharge from hospital on day 3.

Another death (femalepatient, 28-yrs old,weight: 58 Kg)wasdue tocardiac arrest occurring 5 days after the diagnosis of acute PE. She wastreated with 1 mgr/sc/12 h of Enoxaparin (Clexane® 60 mgr/sc/12 h).In this patient, autopsy disclosed a large right heart thrombus. Normalbiochemical values were obtained but TTE was not performed. Thepatient was randomized in the early discharge group and wasdischarged on day 5. After discharge she experienced a cardiac arrestwhen she was already within the hospital. Despite immediatecardiopulmonary resuscitation, the patient died.

The remaining deaths during follow-up were unrelated to throm-boembolic disease or anticoagulant treatment. A death due tohyperglycemic coma occurred in the early discharge group. In patientsrandomized to the standard-hospitalization group, deaths were due to

ndard hospitalization group=60)

Relative risk(95% CI)

P value

±5.7 - 0.008.3) 0.50 (0.12-2.01) 0.26

- 0.303.3) 0.83 (0.12-5.74) 0.62

- 0.545.0) 1.11 (0.26-4.77) 0.601.6) 0.83 (0.05-13.04) 0.703.3) 1.25 (0.22-7.24) 0.59

e5R. Otero et al. / Thrombosis Research 126 (2010) e1–e5

malignancy (n=2), myasthenia gravis (n=1), septic shock (n=1),and respiratory failure (n=1).

Discussion

There are two principal conclusions that follow from our results.First, the rate of mortality, nonfatal recurrences, and hemorrhagiccomplications in highly selected patients with acute symptomatic PErandomized to standard hospitalization or early discharge did not differsignificantly. Second, the rate of short-termmortalitywas unexpectedlyhigh in a (a priori) low-risk group of patients with acute PE.

Previous clinical trials have shown that outpatient treatment ofpatients with deep vein thrombosis is feasible, effective and safe [1,2].However, available data on early discharge programmes for patientswith acute symptomatic PE remain scarce, and several limitations(including small sample sizes, retrospective nature of the investiga-tion, and lack of randomization) are common to previous investiga-tions [20–23]. A recent observational study in the UK found asurprisingly low rate of short-term adverse events in selected patientswith acute PE [24]. In this study, 157 patients with PE receivedoutpatient anticoagulant therapy. There were no deaths, bleeding orrecurrent thromboembolic events during acute treatment with lowmolecular weight heparin. Unfortunately, the authors' selectioncriteria could not be validated in an independent population ofpatients with PE [25].

Determining which patients have a poor prognosis is a crucial stepfor selecting those patients who should be treated for pulmonaryembolism in the hospital. In our study, the incidence of short-termadverse events was higher than previously described [3,4]. Oneexplanation is that the prognostic model that was employed has notbeen validated and interobserver variability has not been assessed[17]. However, our selection was implemented with echocardio-graphic findings and cardiac troponin testing [11,12], suggesting thatshort-term deaths were due to chance and not to the treatmentmodality. Furthermore, both patients who died would have beenclassified as low risk according to previously established clinicalprognostic models [10,13].

A recent study shows that patients with a very short length ofhospital stay had greater postdischarge mortality relative to patientswith a typical hospital stay, suggesting that physicians mayinappropriately select patients with PE for early discharge [26]. Thatresults support our experience, we were surprised by the unexpectedrate of short-term mortality in spite of a careful selection.

Our study has some limitations and the results should beinterpreted with caution. First, because the number of enrolledpatients was low, the lack of association between early discharge andprognosis may just reflect a lack of statistical power. Second, theanticoagulant regimen was chosen according to local patterns ofpractice. Thus, our results may not be easily generalizable. Althoughthe definition of early discharge is debatable, in many countries thedischarge on day 5 can be considered early. Finally, the performanceof our prognostic model might not be optimal, which may results inthe inclusion of patients with a relatively poor prognosis.

In conclusion, the usefulness of clinical prediction rules forselecting patients with acute PE who might be suitable for outpatientmanagement deserves additional clinical evaluation. In addition,clinical trials are required to assess whether patients with acute PEmight benefit from outpatient therapy.

Funding/support

Supported by grants from the Ministry of Health and ConsumerAffairs, Instituto de Salud Carlos III (FIS: PI03/0192), and the SociedadEspañola de Neumologia y Cirugía Torácica (SEPAR).

Conflict of interest statement

None of the authors have conflict of interests.

References

[1] Levine M, Gent M, Hirsh J, et al. A comparison of low-molecular-weight heparinadministered primarily at home with unfractionated heparin administered in thehospital for proximal deep-vein thrombosis. N Engl J Med 1996;334:677–81.

[2] Koopman MM, Prandoni P, Piovella F, et al. Treatment of venous thrombosis withintravenous unfractionated heparin administered in the hospital as comparedwith subcutaneous low-molecular-weight heparin administered at home. TheTasman Study Group. N Engl J Med 1996;334:682–7.

[3] The Columbus Investigators. Low molecular weight heparin in the treatment ofpatients with venous thromboembolism. N Engl J Med 1997;337:657–62.

[4] Simonneau G, Sors H, Charbonnier B, et al. A comparison of low-molecular-weightheparin with unfractionated heparin for acute pulmonary embolism. The THESEEStudy Group. Tinzaparine ou Heparine Standard: Evaluation dans l`EmboliePulmonaire. N Engl J Med 1997;337:663–9.

[5] British Thoracic Society Guidelines for the management of suspected acutepulmonary embolism. British Thoracic Society Standards of Care CommitteePulmonary Embolism Guideline Development Group. Thorax 2003;58:470–83.

[6] Büller HR, Agnelli G, Hull RD, et al. Antithrombotic Therapy for VenousThromboembolic Disease: The Seventh ACCP Conference on Antithrombotic andThrombolytic Therapy. Chest 2004;126:401S–28S.

[7] Aujesky D, Smith KJ, Cornuz J, Roberts MS. Cost-effectiveness of low-molecular-weight heparin for treatment of pulmonary embolism. Chest 2005;128:1601–10.

[8] Douketis JD, Kearon C, Bates S, et al. Risk of fatal pulmonary embolism in patientswith treated venous thromboembolism. JAMA 1998;279:458–62.

[9] Goldhaber S, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes inthe International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet1999;353:1386–9.

[10] Aujesky D, Obrosky DS, Stone RA, et al. Derivation and validation of a prognosticmodel for pulmonary embolism. Am J Respir Crit Care Med 2005;172:1041–6.

[11] Giannitsis E, Müller-Bardorff M, Kurowski V, et al. Independent prognostic value ofcardiac troponin T in patients with confirmed pulmonary embolism. Circulation2000;102:211–7.

[12] Grifoni S, Olivotto I, Cecchini P, et al. Short-term clinical outcome of patients withacute pulmonary embolism, normal blood pressure and echocardiographic rightventricular dysfunction. Circulation 2000;101:2817–22.

[13] Wicki J, Perrier A, Perneger TV, et al. Predicting adverse outcome in patients withacute pulmonary embolism: a risk score. Thromb Haemost 2000;84:548–852.

[14] Jiménez D, Yusen RD, Otero R, et al. Prognosis models for selecting patients withacute pulmonary embolism for initial outpatient therapy. Chest 2007;132:24–30.

[15] The PIOPED Investigators. Value of the ventilation/perfusion scan in acutepulmonary embolism. Results of the prospective investigation of pulmonaryembolism diagnosis (PIOPED). JAMA 1990;263:2753–9.

[16] Remy-Jardin M, Remy J, Wattinne L, Giraud F. Central pulmonary thromboembo-lism: diagnosis with spiral volumetric CT with the single-breath-hold-technique-comparison with pulmonary angiography. Radiology 1992;185:381–7.

[17] Uresandi F, Otero R, Cayuela A, et al. A clinical prediction rule for identifying short-term risk of adverse events in patients with pulmonary thromboembolism. ArchBronconeumol 2007;43:617–22.

[18] Prandoni P, Cogo A, Bernardi E, et al. A simple approach for detection of recurrentproximal vein thrombosis. Circulation 1993;88:1730–5.

[19] Nauffal D, Cayuela A, Otero R, et al. Early complications in patients with pulmonaryembolism. Med Clin (Barc) 2006;127:1–4.

[20] Lui B, Tran A, Montalto M. Treatment of patients with pulmonary embolismentirely in hospital in the home. Aust Fam Physician 2007;36:381–4.

[21] Dager WE, King JH, Branch JM, et al. Tinzaparin in outpatients with pulmonaryembolism or deep vein thrombosis. Ann Pharmacother 2005;39:1182–7.

[22] Ong BS, Karr MA, Chan DKY, et al. Management of pulmonary embolism in thehome. Med J Aust 2005;183:239–42.

[23] Beer JH, Burger M, Gretener S, Bernard-Bagattini S, Bounameaux H. Outpatienttreatment of pulmonary embolism is feasible and safe in a substantial proportionof patients. J Thromb Haemost 2003;1:186–7.

[24] Davies CWH,Wimperis J, Green ES, et al. Early discharge of patients with pulmonaryembolism: a two-phase observational study. Eur Respir J 2007;30:708–14.

[25] Otero R, Jiménez D. on behalf of the Cooperative Study for theAmbulatory TreatmentofPatientswith Pulmonary EmbolismResearchStudyGroup.Pulmonaryembolismathome. Eur Respir J 2008;31:686–7.

[26] Aujesky D, Stone RA, Kim S, Crick EJ, Fine MJ. Length of Hospital Stay andPostdischarge Mortality in Patients with Pulmonary Embolism. Arch Intern Med2008;168:706–12.