MedSysEvidence-graded reference

MedSys / Pulmonary embolism / For doctors

Pulmonary embolism · Chapter 7 of 8

For doctors: the pathway on one page

NG158 and the 2026 AHA/ACC multisociety guideline in one table, with the doses and the special populations.

Written for the clinician who meets a PE a few times a year and wants the two guidelines on one page: NICE NG158 (2020, reviewed unchanged May 2026) and the 2026 AHA/ACC multisociety guideline (Circulation, March 2026), with the ESC 2019 document where NG158 is silent. Doses are for adults with normal renal function unless stated, and should be confirmed against the current BNF.

1. Diagnosis: the pathway as NG158 writes it

  1. History, examination, chest X-ray, ECG (1.1.15)

    Low suspicion on gestalt plus a feasible alternative: consider PERC (1.1.16). All eight negative → stop.

  2. Two-level PE Wells (1.1.17)

    >4 likely → CTPA immediately, or V/Q SPECT (planar if unavailable) for contrast allergy, CrCl <30, or high radiation risk. Cannot image immediately → interim anticoagulation (1.1.18). ≤4 unlikely → D-dimer within 4 h in any setting; if not achievable, interim anticoagulation while waiting (1.1.21). Age-adjusted threshold over 50 (1.1.14); quantitative POCT acceptable (1.1.13).

  3. Interim anticoagulation (1.3.2–1.3.4)

    Pick the agent you would continue. FBC, U&E, LFT, PT, APTT before the first dose; do not wait for results; review within 24 h.

  4. PE confirmed (1.1.19)

    Anticoagulate, or mechanical intervention if anticoagulation is contraindicated. Haemodynamic instability → UFH infusion and consider thrombolysis (1.3.12, 1.6.2). Not identified and DVT suspected → proximal leg ultrasound (1.1.20).

2. Risk stratification: two schemes side by side

ESC 2019 / UK usageCriteriaAHA/ACC 2026Criteria
LowHaemodynamically stable, sPESI 0 (or PESI I–II), no RV dysfunction, biomarkers normalAIncidental / asymptomatic
BSymptomatic, low severity score (PESI ≤85, sPESI 0 or Bova ≤4)
Intermediate-lowStable, sPESI ≥1, one or neither of RV dysfunction / raised troponin or NPC1 / C2Elevated score with neither (C1) or one (C2) of abnormal biomarker and abnormal RV
Intermediate-highStable, sPESI ≥1, both RV dysfunction and raised biomarkerC3Elevated score with both
(High, transient)D1Transient hypotension: SBP <90 or a fall >40 mmHg for <15 min or fluid-responsive
(not named)D2Normotensive shock: any of lactate >2, AKI, urine output <0.5 mL/kg/h, altered mental state, CI <2.2, MAP <60, rising shock score
HighSBP <90 for ≥15 min or vasopressors, or obstructive shock, or arrestE1 / E2E1 persistent shock; E2 refractory shock or cardiac arrest

Respiratory modifier R (C, D, E): hypoxaemia, tachypnoea or oxygen requirement. sPESI: age >80, cancer, chronic cardiopulmonary disease, HR ≥110, SBP <100, SpO2 <90%; one point each; 0 = low. Bova: SBP 90–100 (2), raised troponin (2), RV dysfunction (2), HR ≥110 (1). RV dysfunction on CT: RV/LV ratio ≥1.0 on axial images is the usual threshold; on echo, RV dilatation, RV/LV >1, TAPSE <16 mm, or septal flattening. A MAP above 80 mmHg in a C3 patient identifies a group at low short-term risk of decompensation in the Italian registry analysis the US guideline cites.

3. Treatment: NG158 and AHA/ACC 2026 compared

QuestionNICE NG158AHA/ACC 2026Comment
Outpatient careConsider for low-risk PE using a validated tool; written information, direct contact, out-of-hours details mandatory (1.2)Categories A and B without admission; structured observation for borderlineSame
First-line anticoagulantApixaban or rivaroxaban; else LMWH ≥5 d then dabigatran/edoxaban, or LMWH+VKA (1.3.8)DOAC over VKA in essentially all eligible, including obesity, non-severe CKD, brain tumourSame
Initial parenteral agent in sicker patientsUFH infusion for haemodynamic instability (1.3.12); UFH not routine otherwise (1.3.9)LMWH over UFH for C, D and E; UFH only where VA-ECMO is contemplated (E2)Differs. The US position rests on the Cochrane comparison and PE-specific observational data; NICE reasons from controllability in shock
Systemic thrombolysisConsider for haemodynamic instability (1.6.2). Do not offer if stable, with or without RV dysfunction (1.6.3)E2: systemic lysis over other reperfusion. E1 and D: all reperfusion options equal weight. C1, C2: not recommended (harm)Same for shock and for the stable low tiers. Differs for D
Catheter-directed lysis / mechanical thrombectomyNot in NG158; IPG524 covers ultrasound-enhanced CDT under special arrangementsC3: may be considered (2b). D, E1: equal to other reperfusion options. C1–C2: unclear benefit. Free-floating RA/RV thrombus in C3–E: reasonableDiffers. HI-PEITHO (4.0% v 10.3%) and PRAGUE-26 (0.7% v 6.8%) post-date both documents
Surgical embolectomy / ECMONot addressedEmbolectomy where lysis is contraindicated or failed; VA-ECMO bridge in E2
PERTNot addressedRecommended for hospitals managing PE
IVC filterOnly if anticoagulation contraindicated, or PE on treatment after the treatment-failure steps; documented removal plan (1.7.1–1.7.4)Same indications; retrievable, remove when anticoagulation establishedSame
Cancer3–6 months then review; consider a DOAC; LMWH if unsuitable (1.3.15–1.3.18)DOAC preferred; LMWH for luminal GI/GU tumours or interactionsSame in substance
Triple-positive APSLMWH then VKA, not a DOAC (1.3.20)VKASame
Duration≥3 months; stop after provoked if trigger gone; consider continuing after unprovoked; HAS-BLED ≥4 prompts a stopping discussion; annual review (1.4)Same structure; extended reduced-dose DOAC named as an optionSame
Cancer investigation after unprovoked VTEHistory, examination, baseline bloods; nothing further without symptoms (1.8)Same; no extensive screeningSame (SOME trial)
Thrombophilia testingNot if continuing; not after provoked; aPL if unprovoked and stopping; heritable if unprovoked, first-degree relative and stopping; not relatives (1.9)Same, via ASHSame
Follow-up for CTEPHNot addressedStructured review at 3–6 months; echo if symptomatic; V/Q if echo suggests PH; PH centre referralESC 2019 says the same. Worth building into the local pathway

4. Doses

AgentTreatmentExtended (after 6 months)Renal
Apixaban10 mg bd for 7 days, then 5 mg bd2.5 mg bdNo dose change to CrCl 15; not recommended <15
Rivaroxaban15 mg bd for 21 days, then 20 mg od with food10 mg od (20 mg if high recurrence risk)Consider 15 mg od if CrCl 15–49 and bleeding risk high; avoid <15
Edoxaban60 mg od after ≥5 days LMWH60 mg od30 mg if CrCl 15–50, weight ≤60 kg, or strong P-gp inhibitor
Dabigatran150 mg bd after ≥5 days LMWH150 mg bd (110 mg bd if ≥80 or on verapamil)Not if CrCl <30
Enoxaparin1.5 mg/kg od or 1 mg/kg bd—1 mg/kg od if CrCl <30; anti-Xa monitoring at extremes of weight
WarfarinStart with LMWH; overlap ≥5 days and until INR ≥2.0 twiceINR 2–3—
Alteplase (PE)100 mg over 2 h; in cardiac arrest 0.6 mg/kg (max 50 mg) over 15 min, or 50 mg bolus. Stop UFH during infusion, restart when APTT <2× control
Aspirin (declined anticoagulation)75 or 150 mg od (off-label, NG158 1.4.11)

Thrombolysis contraindications (absolute): haemorrhagic stroke or stroke of unknown origin at any time; ischaemic stroke within 6 months; CNS neoplasm; major trauma, surgery or head injury within 3 weeks; bleeding diathesis; active bleeding. Relative: TIA within 6 months, oral anticoagulation, pregnancy or the first postpartum week, non-compressible punctures, traumatic resuscitation, refractory hypertension (SBP >180), advanced liver disease, infective endocarditis, active peptic ulcer. In E2 the absolute list shrinks to what would kill faster than the PE.

5. Special populations

  • Extremes of weight (<50 kg or >120 kg): consider level monitoring; apixaban and rivaroxaban have the most reassuring obesity data (NG158 1.3.11).
  • Pregnancy: outside NG158. RCOG Green-top 37b: LMWH weight-based throughout; no D-dimer; imaging as for non-pregnant with CXR first; DOACs contraindicated; continue ≥6 weeks postpartum and ≥3 months total.
  • Cancer: apixaban (Caravaggio) and edoxaban (Hokusai VTE Cancer) against dalteparin; more GI bleeding with edoxaban and rivaroxaban in luminal GI tumours; check interactions with chemotherapy (strong CYP3A4/P-gp effects).
  • Antiphospholipid syndrome: DOACs failed against warfarin in triple-positive disease (TRAPS, stopped early for arterial events). Warfarin target INR 2–3.
  • Heparin-induced thrombocytopenia: argatroban or fondaparinux; avoid all heparins.
  • Chronic kidney disease: see the dose table; UFH for CrCl <15 with dialysis-team input.
  • People who inject drugs: NG158 makes a research recommendation only; adherence is the problem, and a DOAC with pharmacy support beats LMWH in practice.

6. Triggers for escalation, and the PERT conversation

Any of the following in a C3 (intermediate-high) patient is a call to the on-call cardiology or interventional centre, or the PERT if one exists: SBP falling below 100, a fluid-responsive hypotensive episode, lactate above 2, new AKI or oliguria, rising NEWS2, new confusion, escalating oxygen requirement, or RV/LV ratio above 1.5 with a raised troponin. The question to ask the centre is specific: "Is this patient a candidate for catheter-directed thrombolysis or rescue systemic lysis, and at what trigger?" Document the answer and the trigger. Free-floating right heart thrombus on echo is an escalation on its own.

7. Discharge and follow-up checklist

The printable version is at /pe/checklist/. The items:

  • Provoked or unprovoked, documented with the reasoning.
  • Anticoagulant, dose, planned duration and the date of the 3-month review, with the responsible clinician named.
  • Written information (NG158 1.5.1), alert card (1.5.2), direct contact and out-of-hours details (1.2.4, 1.4.2).
  • Renal function and weight on the discharge letter, because both decide the dose.
  • Hormonal contraception or HRT stopped and replaced.
  • Cancer: examination documented; screening status; no further tests unless symptomatic (1.8).
  • Thrombophilia: not tested unless stopping is planned and criteria in 1.9 are met; if tested, timing and interpretation by haematology.
  • 3–6-month functional review with echo if breathless; V/Q and PH referral if echo abnormal.
  • Filter: removal plan documented.
  • Future surgery and admissions: previous VTE flagged for prophylaxis.

Reference only. Local protocols, the SmPC and the BNF take precedence for any dosing decision; the trial figures above are from the papers named in the reference list and should be read with their inclusion criteria.