This is a controversial issue which warrants a look at risk stratification tools. The primary one used is the validated Pulmonary Embolism Severity Index (PESI) score. In Lancet 2011, the authors looked at whether PESI class I and II (low risk) patients could be managed safely as outpatients. It turns out in their study, regardless of whether their PESI class I and II patients were treated as outpatients and inpatients, all fared equally well from a complications standpoint (recurrent clot, bleeding from anticoagulation).
I like the validated PESI scoring system to risk-stratify patients as low vs high risk for complications. I, however, do caution people to look closely at the exclusion criteria for this study before applying this to all ED patients.
The exclusion filter was so strict that they likely have captured a very narrow and unrealistic scope of patients to be widely applicable. It makes sense from a research standpoint to have these criteria to achieve internal validity but the question is external validity. Two exclusion criteria that struck me as awfully strict were: (1) needing parenteral opioids or (2) active alcohol or drug abuse.
Bottom line: For me, this study alone seems not have enough external validity to decide about the decision to treat PE patients as inpatient vs outpatient. Although I think that ultimately some can be managed as outpatients, I'd like to see more studies.
Feel free to download this card and print on a 4'' x 6'' index card.
Lots of Twitter comments in the first 24 hours of the post, summarized on Storify: