Objective: To review diagnostic barriers, emergency interventions and maternal outcomes in pregnancy-associated pulmonary embolism arising during or soon after prolonged, obstructed or difficult labour and to distinguish thrombotic pulmonary embolism from amniotic fluid embolism. Study design: Systematic review with structured narrative synthesis of diagnostic studies, registry cohorts, case series and emergency-treatment evidence. Place and duration of the study: Global published evidence was reviewed from 1 May to 15 July 2026 with searches current to 30 June 2026. Methodology: MEDLINE/PubMed was searched for pulmonary embolism in pregnancy, labour and the puerperium. Eligible publications reported diagnostic pathways, confirmed thrombotic PE, emergency reperfusion or maternal survival. Data were retained in their published form because case selection and treatment severity differed. Evidence addressing amniotic fluid embolism without objectively confirmed thrombosis was excluded from the outcome synthesis. Results: The broad search identified 5,276 records and 28 focused publications were included. In the prospective pregnancy-adapted YEARS study 510 women were screened, 12 (2.4%) were excluded and 20 (4.0%) had PE at baseline. Imaging was avoided in 195 (39%, 95% CI 35–44) and one follow-up deep-vein thrombosis occurred (0.21%, 95% CI 0.04–1.2) with no missed PE. A systematic review of 127 severe cases found maternal survival of 94% (95% CI 86–98) after thrombolysis and 86.1% (95% CI 71–95) after surgical thrombectomy. Major bleeding after thrombolysis was 17.5% during pregnancy and 58.3% postpartum. Conclusion: Maternal survival depends on early suspicion, objective diagnosis and rapid multidisciplinary treatment. Pregnancy-adapted diagnostic algorithms can safely reduce imaging in stable patients. Unstable PE requires immediate anticoagulation and individualized reperfusion with particular caution about postpartum bleeding.
Pulmonary embolism (PE) is a major preventable cause of maternal death. Pregnancy creates hypercoagulability, venous stasis and endothelial stress. Labour, operative delivery, infection, haemorrhage, immobility and the early puerperium can add risk. Obstructed labour may combine prolonged immobility, dehydration, tissue injury, emergency caesarean delivery and sepsis. Yet the literature rarely treats obstruction as an isolated causal exposure. The clinically defensible question is therefore how PE is recognized and managed when it occurs during or after prolonged or difficult labour [1].
Diagnosis is difficult because dyspnoea, tachycardia, leg swelling, chest discomfort and reduced exercise tolerance overlap with normal pregnancy or labour. Hypotension, hypoxaemia and collapse require simultaneous resuscitation and differential diagnosis. Haemorrhage, sepsis, eclampsia, peripartum cardiomyopathy, acute coronary disease, anaesthetic complications and amniotic fluid embolism can resemble PE [2]. Thrombotic PE should be confirmed by vascular or pulmonary imaging when the patient is stable enough. Amniotic fluid embolism remains a clinical syndrome associated with abrupt cardiopulmonary collapse and coagulopathy and must not be relabelled as thrombotic PE without evidence [3].
Diagnostic hesitation may delay anticoagulation or reperfusion. Conversely, indiscriminate imaging exposes many women without PE to radiation and contrast. Prospective pregnancy-adapted algorithms now combine clinical assessment, D-dimer thresholds and compression ultrasonography to reduce avoidable chest imaging [4]. Treatment is also context-dependent. Low-molecular-weight heparin is preferred for stable PE. Massive PE with shock may require systemic thrombolysis, catheter-directed therapy, surgical embolectomy or extracorporeal membrane oxygenation (ECMO). Recent delivery and tissue trauma increase haemorrhagic risk [5].
Evidence from Pakistan is limited despite delayed presentation, emergency operative delivery and uneven access to advanced imaging or multidisciplinary PE response teams. A focused synthesis can support practical escalation pathways without inventing a local case series. This review evaluated diagnostic strategies, emergency interventions and maternal survival in pregnancy-associated PE during labour or the puerperium with particular attention to prolonged or obstructed labour contexts [6-13].
A systematic review with structured narrative synthesis was conducted. The population comprised pregnant or postpartum women with suspected or confirmed thrombotic PE. The exposure context included labour, prolonged or obstructed labour, emergency or operative delivery and the first six postpartum weeks. Diagnostic outcomes included confirmed PE, imaging avoidance and venous thromboembolism during follow-up. Treatment outcomes included maternal survival, major bleeding, fetal survival and need for rescue support. Work was undertaken from 1 May to 15 July 2026. MEDLINE/PubMed was searched from inception to 30 June 2026. The main terms combined ‘pulmonary embolism’ with pregnancy, pregnant, labour, labor, postpartum or puerperium and were expanded with diagnosis, D-dimer, computed tomographic pulmonary angiography, ventilation-perfusion imaging, thrombolysis, thrombectomy, embolectomy, ECMO and survival. The broad query returned 5,276 records. Reference lists of recent reviews and guidelines were checked for prospective diagnostic management studies and severe-PE treatment series. Prospective or retrospective diagnostic studies, national or institutional cohorts, systematic reviews of emergency treatment, case series and informative cases were eligible. Confirmed thrombotic PE required CTPA, ventilation-perfusion imaging, pulmonary angiography, objective deep-vein thrombosis in an accepted diagnostic pathway or a clearly documented intraoperative clot. Reports limited to amniotic fluid embolism were excluded from thrombotic outcomes. General pregnancy VTE guidance was retained for interpretation but not counted as a patient cohort. Publications that mentioned obstructed labour and PE as unrelated causes of maternal death without patient-level linkage were excluded. A standardized extraction framework recorded study design, setting, pregnancy stage, number assessed, confirmed PE, diagnostic algorithm, imaging use, intervention, survival, bleeding and fetal outcomes. Obstructed labour was defined as failure of descent despite adequate uterine activity because of mechanical obstruction. Prolonged labour was treated as a related but non-equivalent context. Major PE was categorized by the source report as massive or high risk when haemodynamic instability, cardiac arrest or shock was present. Major bleeding was retained according to each report’s definition. Risk of bias was assessed qualitatively. Diagnostic management studies were examined for consecutive enrolment, objective reference standards and three-month follow-up. Registry cohorts were assessed for coding validity and case ascertainment. Intervention evidence was assessed for severity imbalance, selective publication and incomplete outcome reporting. No meta-analysis was performed because emergency intervention groups differed markedly in haemodynamic severity, timing relative to delivery and availability of rescue technology. Frequencies, percentages and confidence intervals were reproduced as published. Ethical approval was not required because only published aggregate data were used. Patient confidentiality was not applicable. Statistical significance was accepted at p<0.05 when used by the original study [14].
The broad database search identified 5,276 records. Focused assessment yielded 28 publications addressing pregnancy-related PE diagnosis, risk, acute management or outcome. Exact evidence specifically naming obstructed labour was scarce. Most reports described antepartum, intrapartum, post-caesarean or early postpartum PE and identified prolonged immobility, infection, haemorrhage and operative delivery as clinically relevant coexisting factors. Figure 1 presents the selection logic and shows why amniotic fluid embolism-only reports were excluded.
The strongest diagnostic evidence came from prospective management studies. In the pregnancy-adapted YEARS study 510 women were screened and 12 (2.4%) were excluded. PE was diagnosed in 20 (4.0%) at baseline. CTPA was avoided in 195 women (39%, 95% CI 35–44). During three months one popliteal DVT occurred (0.21%, 95% CI 0.04–1.2) and no PE occurred among women in whom the algorithm excluded PE. Imaging was avoided in 65% during the first trimester and 32% during the third trimester. A retrospective US validation included 74 pregnant patients with a PE prevalence of 2.7% (2 patients). Thirty-six did not require imaging by the algorithm and no patient without initial CTPA developed PE or DVT within 30 days.
The Pregnancy-Adapted Geneva score was derived from 395 women. PE prevalence increased from 2.3% in the low pretest probability group to 11.6% in the intermediate group and 61.5% in the high group. The area under the receiver operating characteristic curve was 0.795 compared with 0.684 for the original Geneva score. A South African retrospective assessment found that application of pregnancy-adapted YEARS could have reduced scans by 25.7% with a negative predictive value of 100.0%. These findings support structured triage but do not justify delaying imaging in unstable labouring or postpartum women.
Emergency-treatment evidence was dominated by selected severe cases. A systematic review identified 127 severe PE cases of which at least 83% were massive and 23% involved cardiac arrest. Among 83 women receiving thrombolysis, maternal survival was 94% (95% CI 86–98). Major bleeding was 17.5% during pregnancy and 58.3% postpartum. Fetal deaths possibly related to PE or treatment occurred in 12.0% of antenatal cases. Among 36 women treated with surgical thrombectomy, maternal survival was 86.1% (95% CI 71–95) and major bleeding was 20.0%. Fetal deaths possibly related to surgery occurred in 20.0%. About half of severe postpartum PE occurred within 24 hours of delivery.
ECMO evidence covered broader obstetric cardiopulmonary failure. A systematic review included 213 obstetric patients. PE accounted for 13.6% of indications and overall maternal survival was 79.3%. Live birth occurred on ECMO in 15.5% (95% CI 10.6–20.4) and off ECMO in 58.3% (95% CI 51.7–64.9). Another review of 97 patients reported maternal survival of 90.7% and neonatal survival of 83.3% with haemorrhage in 31.9%. These values cannot be compared directly with thrombolysis because indications and illness severity differed.
A United States mortality analysis covering 2003–2020 confirmed that pregnancy-related PE deaths remain clinically important and often cluster postpartum. Recent emergency reviews emphasized immediate anticoagulation for confirmed PE when bleeding risk permits. Unstable patients required bedside echocardiography, rapid exclusion of major alternative causes and multidisciplinary selection of thrombolysis, catheter therapy, surgery or ECMO. Tables 1–3 summarize diagnostic cohorts, emergency outcomes and practical factors associated with delayed recognition or treatment risk.
Figure 1. Study selection and evidence-synthesis flow.
Table 1. Principal diagnostic evidence
|
Study |
Sample |
PE outcome |
Diagnostic finding |
|
Pregnancy-adapted YEARS |
510 screened; 12 (2.4%) excluded |
20 (4.0%) baseline PE |
CTPA avoided in 195 (39%, 95% CI 35–44) |
|
PAG derivation cohort |
395 women |
PE prevalence 2.3%, 11.6% and 61.5% by group |
AUC 0.795 vs 0.684 |
|
US retrospective validation |
74 women |
2 (2.7%) PE |
36 did not require imaging by algorithm |
|
South African assessment |
Pregnant/puerperal cohort |
No missed PE reported |
Potential scan reduction 25.7%; NPV 100.0% |
Table 2. Emergency intervention outcomes in severe pregnancy-related PE
|
Intervention/evidence |
Patients |
Maternal survival |
Major bleeding/other outcome |
|
Systemic thrombolysis |
83 |
94% (95% CI 86–98) |
17.5% antenatal; 58.3% postpartum bleeding |
|
Surgical thrombectomy |
36 |
86.1% (95% CI 71–95) |
20.0% major bleeding |
|
ECMO, mixed obstetric indications |
213 |
79.3% |
PE was 13.6% of indications |
|
ECMO case-report review |
97 |
90.7% |
31.9% haemorrhage; neonatal survival 83.3% |
Table 3. Clinical factors affecting recognition and survival
|
Domain |
Challenge |
Action |
|
Symptoms |
Overlap with normal labour, sepsis and haemorrhage |
Use structured PE probability and reassessment |
|
Diagnosis |
Concern about radiation or transfer delay |
Use CUS when DVT symptoms exist and timely CTPA/VQ when required |
|
Differential |
Thrombotic PE confused with amniotic fluid embolism |
Assess timing, coagulopathy, imaging and RV findings |
|
Treatment |
Fresh surgical/uterine wound |
Balance immediate survival benefit against postpartum bleeding |
|
System |
Limited reperfusion capability |
Activate referral and multidisciplinary PE response pathway |
This review examined PE associated with difficult labour rather than creating an unsupported causal claim that obstruction alone produces PE. The evidence showed three decisive determinants of maternal survival: recognition before irreversible shock, objective diagnosis when feasible and rapid access to anticoagulation or reperfusion. Diagnostic algorithms safely reduced imaging in stable pregnant women. High-risk PE during labour or soon after delivery required a different pathway because resuscitation and treatment could not wait for prolonged sequential testing [15].
Pregnancy-adapted YEARS produced the most directly usable diagnostic evidence. Its 39% imaging avoidance with no follow-up PE balanced diagnostic safety against radiation. The lower efficiency in the third trimester reflects rising D-dimer and more frequent pregnancy-related symptoms [16]. Recent diagnostic reviews support clinical prediction rules but continue to regard CTPA and ventilation-perfusion scanning as acceptable when imaging is required [17]. In obstructed labour, bedside compression ultrasonography may establish VTE without chest imaging when DVT symptoms are present. A negative leg study does not exclude PE and must not delay thoracic imaging in a stable patient with persistent suspicion [18].
The differential diagnosis is especially important during labour. Thrombotic PE often presents with dyspnoea, hypoxaemia, pleuritic symptoms, syncope or right ventricular strain. Amniotic fluid embolism classically involves abrupt collapse, hypoxia and coagulopathy around delivery. Sepsis, haemorrhage and anaesthetic complications can coexist with obstructed labour [19]. Recent case-based pathways show that anchoring on a single obstetric diagnosis can delay PE recognition [20]. Point-of-care echocardiography may demonstrate right ventricular dilatation or pressure overload in shock. These signs support emergency decision-making but are not sufficiently specific to replace definitive imaging in stable women [21].
Anticoagulation remains the foundation for stable PE. Low-molecular-weight heparin has predictable pharmacology and does not cross the placenta [22]. Delivery planning must consider the timing of the last dose, neuraxial anaesthesia and postpartum restart. Severe haemorrhage, uterine rupture or immediate surgery may require temporary interruption and individualized use of unfractionated heparin because it can be stopped and reversed more rapidly [23]. Direct oral anticoagulants are generally avoided during pregnancy and are not preferred during breastfeeding when evidence is insufficient [24].
The high reported survival after thrombolysis is encouraging but must be interpreted with selection and publication bias. The 58.3% postpartum major bleeding rate was substantially higher than the 17.5% antenatal rate [25]. Fresh uterine and surgical wounds explain much of this difference. Systemic thrombolysis remains reasonable for life-threatening PE with refractory shock when death is otherwise likely. In the immediate postpartum period catheter-directed thrombectomy or surgical embolectomy may offer reperfusion with less systemic lytic exposure where expertise is immediately available [26]. The choice should reflect haemodynamic urgency, bleeding, delivery timing and local capability rather than a rigid hierarchy.
ECMO can bridge patients with cardiac arrest or profound right ventricular failure to recovery or definitive clot removal [27]. Its apparent survival must not be interpreted as superiority because the evidence consists mainly of case reports and mixed obstetric indications. Cannulation, anticoagulation and obstetric haemorrhage demand coordinated critical care, cardiology, obstetrics, anaesthesia, haematology and surgery. A PE response team model is particularly suitable for high-risk intrapartum events [28]. Transfer protocols are important for hospitals that lack catheter or surgical capability.
The Pakistani context requires pragmatic adaptation. No Pakistan-specific study met the focused outcome criteria and national linked surveillance remains limited [29]. Delayed referral after obstructed labour can combine dehydration, infection, anaemia, surgery and immobility. Facilities should complete VTE risk assessment at admission, after major obstetric complications and after delivery. Mechanical prophylaxis and pharmacological prophylaxis should be used according to bleeding risk and local guidelines. Sudden postpartum dyspnoea should trigger a PE pathway rather than reassurance based on young age [30].
Maternal mortality surveillance provides a preventability lens. International data have found modifiable delays in recognition, investigation and escalation [31]. Structured handover after difficult labour should document haemorrhage, infection, immobility, thrombosis history, obesity and operative delivery. Women discharged after emergency caesarean section should receive clear advice about breathlessness, chest pain, syncope and unilateral leg swelling. Prevention does not replace diagnostic vigilance because PE may occur without a recognized major risk factor [32].
Recent reviews agree that the postpartum period carries the greatest risk and that management evidence remains weaker than in nonpregnant adults [33]. Contemporary emergency literature supports multidisciplinary reperfusion for massive PE [34]. The strongest clinical message is time-sensitive differentiation: stable suspected PE permits algorithmic exclusion and targeted imaging whereas shock or arrest requires parallel resuscitation, bedside assessment and preparation for reperfusion.
An obstructed-labour pathway should begin before collapse. Admission assessment should identify previous VTE, thrombophilia, obesity, severe infection, dehydration, prolonged immobility and anticipated operative delivery. The risk profile should be reassessed after haemorrhage, transfusion, hysterectomy, intensive care admission or prolonged surgery. Pharmacological prophylaxis must be timed against active bleeding and neuraxial procedures. When anticoagulation is temporarily unsafe, mechanical methods and frequent reassessment can reduce unprotected time. Documentation is important because responsibility often moves between labour ward, theatre, recovery and postnatal teams.
A minimum emergency bundle can be implemented without a specialist PE centre. It includes oxygen, haemodynamic support, electrocardiography, blood tests, bedside ultrasound capability and an agreed route to definitive imaging. A single call should mobilize senior obstetrics, anaesthesia, critical care and medicine. Blood products and a massive haemorrhage protocol should be available before reperfusion in the early postpartum period. If transfer is required, the referring team should communicate delivery time, operative details, bleeding, anticoagulation and right ventricular findings. Delayed transfer without stabilization can eliminate the opportunity for definitive treatment.
Communication with the patient and family is difficult during emergency care but remains important. In a stable patient the rationale for CTPA or ventilation-perfusion imaging should be explained in plain language. Radiation concern should be balanced against the maternal and fetal danger of missed PE. In shock the team may need to proceed under emergency consent principles. After recovery, counseling should cover anticoagulant duration, breastfeeding compatibility, warning symptoms, future pregnancy planning and recurrence prevention. Follow-up should assess post-thrombotic symptoms, chronic thromboembolic pulmonary hypertension and psychological consequences of maternal collapse.
Research should move beyond isolated successful cases. Prospective registries need standardized reporting of labour duration, obstruction, delivery mode, haemorrhage, infection, prophylaxis, time to diagnosis, imaging, haemodynamic class and treatment. Maternal and fetal outcomes should be reported at discharge and follow-up. Low- and middle-income centres require data on transfer delay and availability of CTPA, echocardiography, thrombolysis, catheter treatment and ECMO. Such information would allow prevention and emergency recommendations to reflect real resource constraints rather than assuming immediate access to every reperfusion modality.
Simulation and audit can improve readiness for an event that individual clinicians rarely encounter. Multidisciplinary drills should include intrapartum collapse, competing diagnoses, activation of imaging, anticoagulation decisions and postpartum haemorrhage after reperfusion. Debriefing should identify delays in senior review, blood availability, transport and communication. Audit indicators can include time from deterioration to PE consideration, time to imaging, time to anticoagulation and time to reperfusion decision. These process measures may improve before a hospital accumulates enough cases to compare mortality.
Prevention after obstructed labour must be individualized. Emergency caesarean delivery and immobility increase risk but active haemorrhage may initially preclude pharmacological prophylaxis. A written plan should state when bleeding will be reassessed and who is responsible for starting prophylaxis. Early mobilization, hydration and mechanical measures are supportive but do not replace anticoagulant prophylaxis when indicated. Discharge planning should consider the duration of postpartum risk and whether the woman can obtain injections, administer them correctly and return promptly if symptoms develop.
The immediate postpartum period deserves enhanced observation after a difficult birth. Routine vital signs may identify tachycardia or hypoxaemia but trends and clinical context are more informative than isolated thresholds. Unexplained oxygen requirement, syncope, persistent tachycardia or disproportionate breathlessness should prompt senior review. A normal chest radiograph or nonspecific electrocardiogram does not exclude PE. Staff should avoid attributing symptoms solely to pain, anxiety or anaemia until serious cardiopulmonary causes have been considered. This approach can shorten diagnostic delay without subjecting every postpartum woman to imaging.
Anticoagulant continuity after discharge is another safety point. Women need written dosing instructions, a plan for missed doses and access to follow-up. Clinicians should check renal function, body weight and interacting medicines when selecting dose. Bleeding counseling should be balanced so that patients do not stop treatment for minor bruising without advice. Subsequent pregnancy requires early specialist review because recurrence risk and prophylaxis depend on whether the previous event was provoked, pregnancy-related or associated with thrombophilia. These steps extend emergency survival into long-term prevention.
Terminology should be standardized in reports and practice. ‘Obstetric embolism’ can refer to thrombotic PE, amniotic fluid embolism or air embolism although their mechanisms and treatments differ. A diagnosis of thrombotic PE should document the objective basis whenever possible. A clinical diagnosis during arrest should state the evidence and competing causes. Obstructed labour should also be defined rather than inferred from a long labour. Precise terminology improves mortality review, coding and research and prevents an apparent association from being created by grouping unrelated emergencies under one label.
This review had limitations. Direct literature linking confirmed thrombotic PE specifically to obstructed labour was sparse and most evidence concerned pregnancy, caesarean delivery or the puerperium more broadly. Case reports and treatment series were highly susceptible to publication bias, duplicate reporting and severity-based treatment selection. Diagnostic studies enrolled stable suspected cases and cannot be directly generalized to intrapartum collapse. Definitions of massive PE and major bleeding varied. Evidence from Pakistan was limited and no national PE registry was identified. The review used a focused MEDLINE/PubMed strategy and narrative synthesis rather than patient-level meta-analysis [35].
Pregnancy-associated PE during or after prolonged, obstructed or difficult labour is uncommon but rapidly fatal when recognition and reperfusion are delayed. Pregnancy-adapted algorithms can safely reduce imaging in stable patients. Sudden hypoxaemia, syncope, hypotension or right ventricular strain requires urgent objective assessment and immediate multidisciplinary escalation. Anticoagulation is central and high-risk PE may require thrombolysis, catheter therapy, surgical embolectomy or ECMO. Postpartum bleeding risk must shape reperfusion choice. Future work should establish prospective obstetric PE registries in Pakistan and other low-resource settings. Declarations Financial support and sponsorship: Nil. Conflicts of interest: There are no conflicts of interest.
Coggins AS, Gomez E, Sheffield JS. Pulmonary Embolism and Amniotic Fluid Embolism. Obstet Gynecol Clin North Am 2022;49(3):439-460. doi: https://doi.org/10.1016/j.ogc.2022.02.015. PMID: 36122978; https://pubmed.ncbi.nlm.nih.gov/36122978/