Nearly everyone with a pulmonary embolism is treated with an anticoagulant, and for most people that is the whole treatment. The arguments are about the minority whose heart is under strain: whether to add something that clears the clot faster, and how much bleeding risk to accept for it. In 2026 the answer changed in the United States and did not change in the UK. This chapter gives both, with the numbers.
1. First, how serious is it: the tiers
Severity is graded on four things: blood pressure, a clinical score, blood markers of heart strain, and whether the right ventricle looks enlarged on CT or echo.
The clinical score most used in the UK is the simplified PESI (sPESI). One point for each of: age over 80, cancer, chronic heart or lung disease, pulse 110 or more, systolic blood pressure under 100, oxygen saturation under 90%. A score of zero is low risk and the 30-day death rate is about 1%. The Hestia criteria are an alternative list of practical reasons a person cannot be sent home.
| UK / European tier | Definition | US 2026 category | What it means |
|---|---|---|---|
| Low risk | Normal blood pressure, sPESI 0, no strain markers | A (found incidentally or without symptoms) and B (symptomatic, low score) | Consider treatment at home |
| Intermediate-low | Normal blood pressure, sPESI 1 or more, at most one of raised troponin/BNP or an enlarged right ventricle | C1 (neither) and C2 (one of the two) | Admit, anticoagulate, watch |
| Intermediate-high | Normal blood pressure, sPESI 1 or more, both a raised marker and an enlarged right ventricle | C3 | Admit to a monitored bed. The 2026 catheter trials were done here |
| — | (Not separately named in Europe) | D1: transient low blood pressure that recovers. D2: normal pressure but signs of poor perfusion — raised lactate, falling urine output, confusion, kidney injury ("normotensive shock") | New in 2026. Reperfusion should be considered |
| High risk | Systolic blood pressure under 90 for 15 minutes, or needing drugs to hold it up, or cardiac arrest | E1: persistent shock. E2: refractory shock or cardiac arrest | Clot-busting treatment now |
The respiratory modifier. The US scheme adds an "R" to categories C, D and E when the person is hypoxic, breathing fast or needs oxygen (C3R, D2R), because those patients deteriorate more often than the bare category suggests. There is no UK equivalent; a UK clinician would say the same thing in words.
2. Who goes home the same day
NICE says consider outpatient treatment for low-risk PE, chosen using a validated tool such as sPESI or Hestia. Outpatient treatment means: the anticoagulant is started, the person goes home, and they are given written information on symptoms to watch, a direct contact number for a thrombosis team, and out-of-hours details. If you are sent home without those three things, the discharge is incomplete by NICE's own standard. The US guideline agrees that categories A and B can be managed without admission, and asks for a structured plan for the observation of borderline cases.
3. The anticoagulant
UK (NG158): apixaban or rivaroxaban, first line, started the same day. If neither is suitable: low-molecular-weight heparin (LMWH) injections for at least five days followed by dabigatran or edoxaban, or LMWH overlapping with warfarin. Unfractionated heparin by infusion is reserved for people with kidney failure, high bleeding risk, or an unstable PE that may need clot-busting treatment.
US (2026): the same DOAC-first policy for categories A and B. For the sicker categories C to E the guideline prefers LMWH over heparin infusion for the initial days, on the evidence that it reaches a therapeutic level more reliably and bleeds less. The one exception is the very sickest patients who may need ECMO, where heparin infusion's short action is an advantage.
| Situation | UK first choice |
|---|---|
| Most people | Apixaban or rivaroxaban |
| Kidney function reduced (clearance 15–50) | Apixaban or rivaroxaban with dose care, or LMWH then edoxaban (dabigatran only if clearance 30 or above), or LMWH/heparin with warfarin |
| Kidney failure (clearance under 15) | LMWH or heparin, alone or bridging to warfarin |
| Active cancer | Consider a DOAC; LMWH if a DOAC is unsuitable (stomach or bowel cancers bleed more on DOACs) |
| Triple-positive antiphospholipid syndrome | LMWH then warfarin. Not a DOAC — an MHRA alert followed higher recurrence rates |
| Under 50 kg or over 120 kg | Consider drug-level monitoring; follow the product literature |
| Blood pressure collapsing | Heparin infusion, and consider thrombolysis |
The tablets do not dissolve the clot. They stop it extending and stop new ones forming; the body's own clot-clearing system does the rest over weeks to months. That is why "the anticoagulant isn't working" is rarely the right reading of persistent breathlessness in the first fortnight.
4. Clot-busting drugs (systemic thrombolysis)
Alteplase given into a vein dissolves clot throughout the body. It saves lives when the heart is failing and it causes serious bleeding in a predictable minority, so the question is always who crosses the line.
Both guidelines agree: high-risk PE gets it. Persistent low blood pressure or cardiac arrest from PE is treated with systemic thrombolysis unless there is an absolute contraindication such as recent brain surgery or a recent haemorrhagic stroke.
Both agree the intermediate group does not get full-dose systemic lysis. The PEITHO trial in 2014 gave full-dose tenecteplase to 1,006 intermediate-high-risk patients. It cut early deterioration from 5.0% to 1.6%, but haemorrhagic stroke rose from 0.2% to 2.0% and major bleeding outside the brain from 1.2% to 6.3%, and it did not reduce deaths. NICE's wording is direct: do not offer systemic thrombolysis to people with PE who are haemodynamically stable, with or without right ventricular dysfunction. The US guideline says the same for C1 and C2. Both allow rescue lysis if a stable patient deteriorates.
5. Catheter treatment: where 2026 changed things
The alternative to flooding the whole body with alteplase is to thread a catheter into the lung artery and either drip a small dose directly onto the clot (catheter-directed thrombolysis, CDT) or pull the clot out mechanically (mechanical thrombectomy, MT). Until this year the evidence was registries and surrogate endpoints.
| Trial | Who | Compared | Result |
|---|---|---|---|
| HI-PEITHO, NEJM March 2026 | 544 intermediate-high-risk patients with extra signs of severity, nine countries including the UK | Ultrasound-assisted CDT plus anticoagulation vs anticoagulation alone | 7-day death, decompensation or recurrence: 4.0% vs 10.3%, driven by fewer decompensations. No brain bleeds. No difference in deaths or major bleeding |
| PRAGUE-26, NEJM September 2026 | 558 intermediate-high-risk patients, 11 Czech centres | Conventional CDT (no ultrasound catheter) plus anticoagulation vs anticoagulation alone | 7-day composite 0.7% vs 6.8%. Bleeding 4.6% vs 5.0%. Both brain bleeds in the trial were in the catheter arm. Deaths at 7 days 0 vs 4 |
| PEERLESS, Circulation 2025 | 550 intermediate-risk patients | Large-bore mechanical thrombectomy vs CDT | Similar safety; fewer ICU stays and less deterioration with thrombectomy, in a composite that weighted those. No anticoagulation-only arm |
Read carefully, the two 2026 trials show that catheter thrombolysis reduces early deterioration in the sickest patients with normal blood pressure, without the bleeding penalty of full-dose systemic lysis. They do not show it saves lives; deaths were too few to tell. Both were open-label, meaning clinicians knew who had the procedure, and the deterioration endpoint involves judgement. HI-PEITHO was funded by the catheter manufacturer, and around 87% of screened patients were excluded, so the result applies to a narrow, carefully selected group. PRAGUE-26 used an ordinary catheter, which matters for the NHS, because the ultrasound device is expensive and the benefit looks the same without it.
Where the guidelines now stand. The US 2026 guideline, written before HI-PEITHO reported, puts catheter therapy at "may be considered" for category C3 and treats all reperfusion options as equivalent for D and E. Its own authors have said since that HI-PEITHO is the first hard evidence for the C3 recommendation. NICE NG158 was last reviewed in May 2026 with no change: no thrombolysis in stable patients, and a separate interventional-procedures notice covering ultrasound-enhanced CDT that permits it with special arrangements. In practice, catheter treatment in the NHS is available in a minority of tertiary centres and is decided case by case.
What to ask if you or a relative are in the intermediate-high group. "Is the right ventricle enlarged and is troponin raised?" (that is the group). "Is there a pulmonary embolism response team, or a centre we could be discussed with?" "Are there signs of low perfusion even though the blood pressure is normal?" The honest answer in most UK hospitals will be that anticoagulation with close monitoring in a high-dependency bed is the plan, with lysis held in reserve. That is defensible. Asking makes sure the case has been thought about.
6. Surgery, ECMO and response teams
Surgical embolectomy — opening the chest and removing the clot — is for high-risk PE when thrombolysis is contraindicated or has failed, and needs a cardiac surgical centre. ECMO, a machine that takes over the heart and lungs, is a bridge for patients in refractory shock or after cardiac arrest while a definitive treatment is arranged. Both are rare. The US guideline recommends that hospitals run a pulmonary embolism response team (PERT): a rapid multidisciplinary discussion for every intermediate-high and high-risk patient. Some UK centres have one; most do not, and the equivalent is a phone call to the regional centre.
7. Filters
An inferior vena cava filter is a small metal cage placed in the main vein of the abdomen to catch clots before they reach the lung. It does nothing to the clot already there, and it does not replace anticoagulation. NICE restricts filters to two situations: anticoagulation is contraindicated (usually active bleeding), or a PE has happened despite anticoagulation after adherence and dosing have been checked. Every filter should be placed with a documented plan to remove it, and removed as soon as anticoagulation is established. Filters left in cause their own clots.
8. What monitoring should look like
Intermediate-risk patients should be on a ward that can watch for deterioration in the first 48 to 72 hours, when it happens if it is going to. The sign that matters is falling blood pressure, but the earlier ones are a rising heart rate, a rising early-warning score, a rising lactate and a falling urine output. Anyone in the intermediate-high group who develops those is a rescue thrombolysis decision, and both guidelines say so.
9. Where UK and US guidance differ: summary
- Catheter treatment for the stable-but-strained group. US: may be considered, with two 2026 trials now behind it. UK: not in NG158; case by case in a few centres.
- The middle tier is subdivided. US separates transient hypotension and "normotensive shock" (D) from overt shock (E) and treats D as a reperfusion decision. UK: those patients are "intermediate-high" and watched.
- Initial injection for sicker patients. US prefers LMWH; UK uses heparin infusion for the unstable and LMWH or a DOAC for the rest.
- Response teams. Recommended in the US; patchy in the UK.
- Everything else is the same: DOAC first line, home treatment for low risk, systemic lysis only for shock, filters only when anticoagulation cannot be given.