Pulmonary Embolus Outpatient Treatment

Treat a patient with PE as an outpatient? Insanity! But where is the evidence that says we’re doing the right treatment with all of the PEs we diagnose? The original publications that launched our current way of thinking about PEs are not what I would call practice changing, and if published today, would probably be interesting anecdotes rather than news makers.

The question: Do we need to hospitalize all patients diagnosed with pulmonary embolism in the ED? Can some be discharged and managed as outpatients?

Listen to the show

Direct Download 

Read the Review

Currently, the diagnosis of PE in the United States means admission to the hospital. This is much easier than setting up outpatient anticoagulation and risk stratifying patients. Setting up a patient to go home on low molecular weight heparin (LMWH) and warfarin, getting lab rechecks, connecting them with follow up, etc, takes time. In the case of PE, is that going to be time well spent? Read on.

When we’re talking about the outpatient treatment of  PE, there are several factors to consider.

 1.Is this safe for our patient?

The corollary to that is: Is the patient going to benefit from hospitalization? If they are unstable, they may need critical or at least very closely monitored care. But for the majority of PEs, that’s not the case. What else do they get? They get definite medication. In this case LMWH and warfarin and perhaps connection with follow-up. Although I suspect that many patients are discharged from the hospital before their INR is therapeutic, so in a way, they are also being treated as outpatients.

2. Is outpatient treatment inferior to inpatient  and are there systems in place to make outpatient treatment work?

If you’re talking about a treatment  that doesn’t require monitoring, like antibiotics for pneumonia, there’s no need get doxycycline levels checked. But with PE, close and continued monitoring is critical.

3.Are we going to cause any harm? 

Patients are going to get the same anticoagulants whether they’re in the hospital or at home. This was the magic of LMWH in treating DVT-we no longer had to admit patients to give them heparin.

The main complication/iatrogenic effect we worry about is bleeding. An argument in favor of hospitalization could be that some hemorrhage risk factor was missed and we can observe for serious hemorrhage. Also, hospitalized patients will have their blood monitored frequently. As we’ll see, this could mean a shorter time on dual anticoagulation (shots and pills) and subsequently lower risk of bleeding. But no matter where where patients are dispositioned, bleeding is going to happen.  If you thin enough people’s blood, some of them are going to bleed.

What are the key points to tease out from the literature?

1) Is outpatient treatment inferior to inpatient?

2) Will we cause our patients harm?

3) What are the inclusion and exclusion criteria we can replicate them in our clinical practice?

4) Is the treatment strategy feasible and something we can and should be doing?

Outpatient versus inpatient treatment for patients with acute pulmonary embolism: an international, open label, randomized, non inferiority trial. Lancet July 2011.

Design: 344 patients with acute, symptomatic pulmonary embolism and low risk of death as determined by a PESI (pulmonary embolism risk severity index) score of 1 or 2, randomized to either outpatient or 5 days of inpatient treatment. Primary outcomes:  symptomatic, recurrent venous thromboembolism (VTE) within 90 days. Safety outcomes: mortality within 90 days and major bleeding.

 1) Is outpatient treatment inferior to inpatient? There was one recurrent VTE in the outpatient group versus zero for the inpatients. Outpatient therapy was deemed non inferior to inpatient.

 2) Will we cause our patients harm? Outpatients were treated with LMWH for 11.5 days versus 8.9 days for inpatients. Could longer treatment with dual anticoagulation lead to increased bleeding? It surely doesn’t help. Two outpatients and no inpatients had major bleeding with 14 days. At 90 days, 3 outpatients and no inpatients had major bleeding. There was one death in each group.

 3) What are the inclusion/exclusion criteria and can we replicate them in our clinical practice?  PESI is an 11 point scoring system used to predict 30 day mortality in patients with PE. No special testing beyond a good H and P and a little bit of diagnostic data is needed to populate the pieces of PESI. Patients were deemed low risk and appropriate for outpatient management if they were PESI class 1 or 2. There were also several exclusion criteria that are easy to determine in the ED. This can be replicated in clinical practice.

The Hestia study. Outpatient treatment in patients with acute pulmonary embolism. Journal of Thrombosis and Hemostasis 2011.

A single treatment arm study with 297 patients managed as outpatients. There was no control group.

1) Is outpatient treatment inferior to inpatient? Since there was no control arm in this study, non-inferiority cannot be directly determined. The authors used a 7% recurrence rate of VTE as the efficacy cutoff (based on previous publications). Six patients (2%, 95% CI 0.8-4.3) had recurrent VTE. Based on a goal efficacy of 7%, outpatient therapy was deemed non-inferior to inpatient.

2 ) Will we cause our patients harm? Three patients died in the three month follow up period, but none of the deaths were felt to be caused by PE. One patient had  a fatal intracranial hemorrhage (0.3%) and two had major bleeding events.

3 ) What are the inclusion/exclusion criteria and can we  replicate them in our clinical practice? The exclusion criteria are an 11 point yes/no checklist. This can be replicated in clinical practice. There is nothing here that requires special testing beyond what is regularly asked or evaluated in an emergency department chest pain workup.

The Hestia Checklist

  1. Is the patient hemodynamically unstable?
  2. Is thrombolysis or embolectomy necessary?
  3. Active bleeding or high risk of bleeding?
  4. More than 24 h of oxygen supply to maintain oxygen saturation > 90%?
  5. Is pulmonary embolism diagnosed during anticoagulant treatment?
  6. Severe pain needing intravenous pain medication for more than 24 h?
  7. Medical or social reason for treatment in the hospital for more than 24 h (infection, malignancy, no support system)?
  8. Does the patient have a creatinine clearance of < 30 mL min)?
  9. Does the patient have severe liver impairment?
  10. Is the patient pregnant?
  11. Does the patient have a documented history of heparin-induced thrombocytopenia?


The Big Question

Should we manage select ED patients with PE as outpatients?

All pulmonary emboli are not created equal. The elderly patient with a saddle embolus and a history of heart failure is different from the otherwise healthy 30 year old with a small peripheral clot, no medical problems and stable vital signs. The irony is that, beyond thrombolysis in the hemodynamically unstable, all PEs are treated as equals. Evidence is mounting that we should think of PEs along a spectrum of risk and treat accordingly. Do all patients with PE need to be admitted to the hospital? Probably not. But there are a few qualifiers if you are thinking of discharging your patient from the ED…

 Easy access to outpatient resources and clearly defined follow up are essential to making this work. In the Lancet study, outpatients were on dual anticoagulants 2.5 days longer, and this was in a setting where it can be inferred that there was close monitoring and follow up.  Whether the 2.5 days of extra treatment was significant is unclear, but it stresses the importance of close monitoring.

Risk stratify to select the appropriate patients. The criteria used in these two studies do not require any special testing.

Shared decision making: Like most of the decisions made in medicine, the final call is up to the patient. Our job is to diagnose, educate and treat- with emphasis on educate.

Decide in your group/hospital/community if this is going to be the standard practice. Are systems in place to assure close follow-up and monitoring?


Emergency Medicine Literature of Note

Basics of Emergency Medicine Pocket Guide by Joe Habboushe



  1. Ken Grauer, MD

    Excellent podcast!

    To me- the 2 BIG Questions are: 1) Why is out patient treatment not inferior to in-patient treatment? -- and 2) Do these low risk “PE” patients need any treatment at all?

    You make the point superbly in your podcast that “PE” is not the disease that it used to be… Studies show PE prevalence today is greatly reduced among patients who present to an ED (and get a chest CT) -- compared to patients who used to present. Clearly, we have cast our “net” wider -- so as not to miss (and send home) patients who might have a PE. Our wider net is now cast with better and better tools (like chest CT) -- that pick up smaller and smaller PEs …

    The key question is -- Are all of these small subsegmental PEs in patients with low PESI scores clinically significant? OR- Do we all have “small PEs” at some frequency that resolve on their own  not causing problem- and are now coincidentally catching many of them when we CT a much less sick group with less clear symptoms? 

    I don’t pretend to have a “magic answer” -- other than strive to select better (not easy!) which patients to CT… -- but it is important to appreciate that the downside to doing so many CTs potentially extends way beyond immediate financial cost and radiation risks IF the above holds any truth -- as patients then are committed to at least 6 months of anticoagulation (which has its longterm risks) -- and whenever they return to the ED with similar vague symptoms, the same cycle may recur when in reality, at least some of these patients may not need to be “treated” at all…

  2. Javier Benitez

    Thanks for the podcast, as usual top of the line.

    Did you come across any literature considering the etiology of pulmonary embolism? I think that’s a major factor to consider admission for a patient.


    1. Rob Orman

      Hi Javier,

      Thanks for the comment. Regarding need to admit a patient with PE….a history of cancer, lung disease and CHF will give a higher PESI score and therefore a higher risk of all-cause mortality. So in that case, I think it would make a difference. In general, the more cormorbidities, the less the patient is going to tolerate a PE. Is that what were you were referring to or did you mean something else?

  3. Pingback: Pulmonary Embolus Outpatient Treatment

  4. Pingback: Lecture for May 9th, 2013: Do we need to admit all patients with pulmonary embolism? | It's All About the Evidence

  5. Pingback: Lecture for April 11th, 2013: Is my patient on low dose aspirin at increased risk for late bleeding after mild head injury? | It's All About the Evidence

  6. Pingback: The LITFL Review 063 - LITFL

Awesome article, I know - please share your erudite thoughts...