• 
• 

Three Part Question

Is it safe [in an adult over 18 years of age] presenting to [an Emergency Department with a suspected pulmonary-embolism] to be [treated as an outpatient with robust follow up]?

Clinical Scenario

A young male was self-presented to the ED with sudden onset left-sided pleuritic chest pain. His physiological observations were within normal limits with an unremarkable physical examination, ECG and chest x-ray. The calculated two-level PE Well’s score was ‘0’ and the D-Dimer was reported as above the 99th percentile.
He was clinically risk-stratified by sPESI scoring as low-risk mortality PE presentation. A CTPA could not be done until the following day and he wished to go home. The BTS guidelines 2018 recommends that patients with suspected PE who are deemed as low-risk are eligible for OP care with an alternative strategy of anticoagulation. Hence, can this patient with suspected PE be treated as OP safely pending radiological investigation?

Search Strategy

MEDLINE, PubMed, EMBASE, CINAHL, Cochrane, Google Scholar, NICE, BTS guidelines and bibliographies.
({[Pulmonary embolism OR PE] AND [outpatient or ambulatory management OR emergency department] and [low risk)} LIMIT to [Adult and English]).

The initial search generated a total of 124 papers. 31 papers were selected after removing duplicates and review of titles and abstracts. 12 high-quality relevant papers were included in the final analysis after a full review of the literature. References from the final papers and BTS guidelines 2018 were examined to ensure all relevant articles had been included.

Search Outcome

Our analysis was aimed at finding whether low-risk suspected PE patient can safely be treated as out-patient. The analyses of 12 studies (as per Oxford centre for evidence-based medicine - Levels of Evidence) are shown in the table:

Keywords: d-dimer, high-risk, low-risk, mortality, risk-stratification;
Abbreviations:
CTPA: Computerised tomography pulmonary angiogram
DOAC: Direct oral anticoagulants
ED: Emergency department
INT: Intention-to-treat analysis
OP: out-patient
PE: Pulmonary-embolism
sPESI: Simplified pulmonary embolus severity index
VTE: Venous thrombo-embolism
UCL: Upper Confidence Limit

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
David R Vinson et al. Nov. 2012 Europe, USA and Canada	N = 777 8 Studies from Europe, USA and Canada	Systematic review (2a)	Primary: 90 days All-cause mortality, VTE and haemorrhage related, nonfatal VTE and nonfatal major haemorrhage	No (0) VTE/haemorrhage related death in 7 studies with combined 741 discharged patients (upper 95% confidence limit 0.62%). Explicit Risk stratification tool (e.g. PESI) and outpatient ineligibility criteria were used to identify low-risk patients	A small number of RCT. 4 studies with non-ED settings were included in analysis. Significant risk of bias due to lack of controlling of confounding factors
David R. Vinson et al. 2018 USA	N= 1882 21 Emergency Departments of Northern California	Controlled Pragmatic trial (2b)	Primary outcome: Discharge home directly from ED or from observation unit (<24 hours) based in the ED Adverse outcome: PE-related return visit in 5 days and recurrent VTE, major haemorrhage and all-cause mortality in 30 days	Intervention group had 881 patients and control site 882 patients. Adjusted home discharge increased in intervention site (17.4% pre- to 28.0% post-intervention) without concurrent increase at control sites (15.1% pre- & 14.5% post-intervention). The difference was 11.3% (95%CI, 3.0 – 19.5; p = 0.0007). Post-intervention 30-day all-cause mortality: 0.8%, recurrent VTE: 0.8% & major haemorrhage: 0.8%	No randomization Convenience sampling Selection bias cannot be excluded The multimodal nature of intervention makes it difficult to know which element or elements contributed most to the increase in OP management
Adam J Singer et al. May 2016 USA	N=394000	Retrospective data analysis 2006 – 2010 (2c)	Primary outcome: hospital and intensive care unit admissions Secondary outcome: sPESI score and mortality during ED visit	The admission rate was 90% for PE with no significant change over time. Of all 51% of ED patients were low-risk based on sPESI scores	Retrospective study. sPESI scoring may be inaccurate due to missing data and subsequent misclassification
Jimenez D et al. August 2010 Spain	N= 995 (derivation cohort) & N=7106 (RIETE validation cohort) Emergency department Spain	Retrospective cohort (2b)	Primary outcome: all cause of mortality within 30 days after diagnosis of acute symptomatic PE	Study’s derivation cohort: 305 of 995 patients (30.7%) were stratified as low-risk by sPESI had mortality 1.0% (95%CI, 0.0%-1.5%) RIETE validation cohort: 2569 of 7106 patients (36.2%) were low-risk as per sPESI had mortality rate 1.1% (95%CI, 0.7%-1.5%)	Prospective data collection with retrospective sPESI calculation Non randomized Few patients lost follow up and not included in analysis
Philip S. Wells et al. 2001 Canada	N= 930 Four Tertiary care hospitals, Canada	Prospective cohort study (2b)	Primary outcome: Proportion of patients had VTE events during 3-month follow up in whom the diagnosis of PE had been excluded	Among 437 patients with low clinical probability and negative D-Dimer, only 1 patient developed PE during follow up. Thus, NPV was 99.5% (CI,99.1 to 100%) to rule out PE with a combination of risk stratification and D-Dimer test	Small sample size No blinding Few patients had diagnostic imaging despite of low probability and negative D-dimer D-Dimer test with higher specificity and low sensitivity used
Wendy Zo et al. 2012	N= 1657 received outpatient treatment (13 studies), N=256 early discharge group (3 studies) and N=383 inpatient group (5 studies)	Meta-analysis (2a)	Primary outcome: 3-month recurrent VTE, major bleeding and all-cause mortality	Pooled incidence of recurrent VTE /major bleeding/mortality 1.7% (95%CI 0.92-3.1%) /0.97%/1.9% in outpatients. 1.1% (0.22 -5.4%)/0.78%/2.3% in early discharge group and 1.2% (0.16 – 8.1%) /1.0%/0.74% in Inpatient group. No death due to fatal PE in outpatient group	No high quality RCT Wide CI Most studies had selection bias 4 studies were retrospective cohort Unclear definition of OP treatment and early discharge in two studies
P M G Erkens et al. August 2010 Canada	N=473 Ottawa Hospital, Canada	Retrospective cohort study with subgroup analysis (2b)	Primary outcome: Overall mortality and fatal recurrent PE 3-month	260 patients (55.0%) managed as OP (out-patient) and 213 (45.0%) managed as Inpatient. Overall mortality 5% (95%CI 2.7 – 8.4%) with 0 recurrent fatal PE in OP group and 26.7% (95%CI 20.9 - 33.2) with 5 (2.3%) fatal recurrent PE in Inpatient Group in 3-month follow-up. In subgroup analysis, excluding cancer patients, the 3-month overall mortality in OP and inpatient group was 0% (95%CI, 0-2.1) and 9.4% (95%CI, 5.0 – 15.9) respectively	Retrospective study Small sample size No blinding and Wide CI Selection bias cannot be excluded
P M Roy et al. December 2016 Canada	N= 1081 Ottawa Hospital, Canada	Retrospective Cohort Study (2b)	Primary endpoint: Combination of major bleeding, recurrent VTE or all-cause death within 14-day of diagnosis of PE	In the low-risk subgroup (n=484), the 14-day rate of combined events was 5.1% for inpatient vs 0% for OP group (p=0.0005). Similarly, the 3-month rates were 8.6% vs 1.7% (p=0.002). All-cause mortality was 2.9% (inpatient) vs 0.1% (OP)	Single centre Retrospective study Possible selection bias as admission was the physician’s opinion based The severity of underlying co-morbidities was not analysed The intermediate-risk PE was too low for subgroup analysis Use of DOAC might have influenced the result in OP group
Drahomir Aujesky et al. July 2011	N = 344 19 EDs in Switzerland, France, Belgium and USA	Prospective study Randomized clinical trial (1b)	Primary outcome: Recurrent VTE within 90 days. Safety outcome: major bleeding in 14 or 90 days and mortality in 90 days	1 (0.6%) of 171 patients in OP group developed recurrent VTE in 90 days but none in 168 inpatients (95% UCL 2.7%; p = 0.011). Only one (0.6%) died in each group. By 90 days 3(1.8%) outpatients and no inpatient had developed major bleeding (95%UCL 4.5%; p=0.086). Mean length of stay (LOS) was 0.5 days (SD 1.0) for OP and 3.9 days for inpatients (SD 3.1)	No blinding INT was not done Relatively young study population with a low incidence of cancer
M J Kovacs et al. June 2010 Canada	N = 699 London, Canada	Retrospective cohort study (3b)	Primary outcome: 3-month recurrent VTE, major bleeding and all-cause mortality	314 (49.1%) managed as OP; 3 VTE & 3 haemorrhagic events 0.95% (95%CI, 0.25,3) in 3 months. 9 deaths: 2.9% (95%CI, 1.4, 5.6), all due to underlying cancer and all occurring within initial 7 days of treatment	Retrospective study Single centre Possible selection bias in determining OP and inpatient group Inpatients were on average 9.5 years older than outpatients
Siavash Piran et al. August 20	N = 1258	Systematic review (1a)	Outcome measure: recurrent VTE 3-month, major bleeding and overall 3-month mortality	Rate of recurrent VTE in OP group was 1.47% (95%CI: 0.47 to 3.0%), Fatal PE 0.47% (95%CI: 0.16 – 1.0%), major haemorrhage 0.81% (95%CI: 0.37-1.42%), fatal intracranial haemorrhage 0.29% (95%CI: 0.06 -0.68%); The overall 3-month mortality 1.58% (95%CI: 0.71 – 2.80%)	Proportion of cancer patient varied significantly in different studies (1 to 22%) Outcome measures were not uniform in between studies Weight estimates are not derived from patient-level longitudinal data
Gregory J. Fermann et al. 2015	N = 4831 38 countries	Post Hoc analysis of open-label randomized trial (1b)	Outcome measure: Recurrent VTE, major bleeding, fatal PE and all-cause mortality at 7, 14, 30 & 90 days and at the end of full treatment period	Calculated sPESI score: 0 (53.6%), 1 (36.7%) and >1 (9.7%); Patient with sPESI 0, all-cause mortality 0.6% (14/2299), Recurrent VTE 1.5%(35/2299) and major bleeding 1.0%(24/2294) at the end of full treatment period. Patients with sPESI>1, had frequent adverse outcomes	Selection bias possible due to open-label design Intangible social factors are not part of model p-value was unadjusted

Comment(s)

The sPESI appears to be a safe risk-stratification tool for triaging patients between in-patient and ambulatory care. Our data analysis derived all-cause mortality 0.76% (mean) & 0.70% (median) and combined major adverse events 1.56% (mean) & 1.70% (median) in the cohort of patients, categorised as low-risk by sPESI with discharge rate ranging from 28.0% to 55%.

Clinical Bottom Line

Based on the above findings, it would seem appropriate to conclude that properly assessed low-risk PE (suspected/confirmed) can safely be treated as an OP with robust follow-up monitoring. The sPESI can safely be utilised as a clinical risk scoring tool to identify low-risk PE presentations.

References

1. David R Vinson et al. Can selected patients with newly diagnosed PE be safely treated without hospitalization Annals of Emergency Medicine. 2012 Nov; 60 (5): 651-662
2. David R. Vinson et al. Increasing safe outpatient management of ED patients with Pulmonary Embolism Annals of Internal Medicine 2018 Dec;169(12):855-865
3. Adam J Singer et al. Admission rates for emergency department patients with venous thromboembolism and estimation of the proportion of low risk pulmonary embolism patients: a US perspective Clinical and experimental emergency medicine. 2016 Sep; 3(3):126-131
4. Jimenez D et al. Simplification of the pulmonary embolism severity index for prognostication in patients with acute symptomatic pulmonary embolism Archives of internal medicine 2010 Aug;170 (15):1383-1389
5. Philip S. Wells et al. Excluding pulmonary embolism at the bedside without diagnostic Imaging: Management of patients with suspected pulmonary embolism presenting to the Emergency Department by using simple clinical model a Ann Intern Med 2001;135:98-107
6. Wendy Zo et al. Outpatient vs inpatient treatment in patients with pulmonary embolism: a meta-analysis Eur Respir J. 2013;42:134-144
7. P M G Erkens et al. Safety of outpatient treatment in acute pulmonary embolism JTH 2010 Nov;8 (11):2412-2417
8. Roy P et al. Net clinical benefit of hospitalization versus outpatient management of patients with acute pulmonary embolism JTH 2017;15 (4):685-694
9. Drahomir Aujesky et al. Outpatient versus Inpatient Treatment for Patients with Acute Pulmonary Embolism: An International, Open-label, Randomised, Non-inferiority Trial Lancet July 2011;378:41–8
10. Kovacs J et al. Ambulatory management of pulmonary embolism: a pragmatic evaluation J Thromb Haemost. Nov 2010;8(11):2406-11
11. Siavash Piran et al. Outpatient treatment of symptomatic pulmonary embolism: A systematic review and meta-analysis Thrombosis Research 2013;132(5):515-519
12. Gregory J. Fermann et al. Treatment of pulmonary embolism with rivaroxaban: outcomes by simplified Pulmonary Embolism Severity Index score from a post hoc analysis of the EINSTEIN PE study Acad Emerg Med. Mar 2015;22(3):299-307