Suspected Pulmonary Embolism at 22 Weeks' Gestation at 23:00
A 33-year-old woman at 22 weeks' gestation presents to the emergency department with sudden-onset dyspnoea, pleuritic chest pain, heart rate 116 per minute, and a D-dimer of 2,100 micrograms per litre fibrinogen-equivalent units. She has no previous venous thromboembolism, no haemoptysis, and a normal chest radiograph. Her creatinine is normal and she has no contrast allergy. She is breastfeeding a 7-month-old and is anxious: "I don't want anything that could hurt the baby." The on-call radiology fellow is off-site until 08:00 and CT pulmonary angiography cannot be reported before then. Your department has a SPECT/CT gamma camera and can perform ventilation/perfusion imaging overnight. It is 23:00 and the emergency physician wants a definitive answer within the hour because the patient is tachycardic and hypoxic on ambulatory oxygen. The obstetric team will not start anticoagulation without imaging.
1. Choose the study, justify it with dose and diagnostic reasoning, and state the specific technical adjustments you would make.
Read the worked answer
With a normal chest radiograph and a high clinical probability, V/Q imaging is the preferred first test in pregnancy. A normal perfusion scan excludes pulmonary embolism, and V/Q has higher specificity than CTPA in this population, so it is less likely to produce an inconclusive result that forces a second study.
Doses, roughly: perfusion with Tc-99m MAA at about 40-100 MBq gives a fetal absorbed dose in the order of 0.1-0.2 mGy and a maternal breast dose under 1 mGy. CTPA gives a comparable or somewhat higher fetal dose and, importantly, a maternal breast dose in the order of 10-20 mGy, which matters in a 33-year-old. Neither is a large dose; undertreated pulmonary embolism is the bigger threat to mother and fetus.
Technical adjustments:
- Perfusion-only first, because the radiograph is normal. Reserve ventilation for an abnormal perfusion study.
- Halve the activity to around 40 MBq and use the minimum particle number consistent with a diagnostic scan - roughly 100,000-200,000, and well under 500,000 - because of the small risk of pulmonary hypertension.
- Aspirate and discard any blood before injection; blood contact with MAA alters biodistribution.
- Image supine to improve count rate; use SPECT for segmental localisation, and if CT is used for attenuation correction keep it to a low-dose non-contrast technique.
- Do not delay to 08:00. The clinical risk of waiting outweighs the small dose.
2. How do you counsel her about fetal and breastfeeding risk, and what do you document?
Read the worked answer
Counsel with numbers and without hedging, then document that the conversation happened.
Fetal dose: the absorbed dose is in the order of a fraction of a milligray - comparable to a few weeks of natural background radiation, and lower than or similar to CTPA. Say that explicitly, because her fear is not irrational and vague reassurance will not resolve it.
Breastfeeding: for Tc-99m MAA, current guidance from both the American College of Radiology and the Society of Nuclear Medicine and Molecular Imaging is that breastfeeding need not be interrupted - the activity excreted in milk is negligible and the infant's absorbed dose is trivial. Offer her the option of expressing and discarding one feed for her own peace of mind, but tell her clearly that this is not required. (For iodinated contrast at CTPA, current guidance is also that breastfeeding may continue.)
Document: the indication and pretest probability, that the choice of study was discussed with the emergency and obstetric teams, the consent conversation including the fetal and breastfeeding discussion, the administered activity and particle number, the time of administration, and the result with its limitations. If the perfusion study is abnormal, document that ventilation imaging was added and why.
3. The perfusion study shows a single subsegmental mismatch in the right lower lobe with normal ventilation. How do you handle this result at 01:00?
Read the worked answer
A single subsegmental mismatch in a patient with high pretest probability is not a negative study and must not be reported as one. In the SPECT era, a solitary subsegmental defect has poor positive predictive value, but in this clinical context it is indeterminate-to-suspicious, not reassuring.
The failure mode to avoid is a report that reads "low probability" and a team that then withholds anticoagulation. Call the emergency physician and the obstetric team with the finding and your interpretation, and state clearly that the study does not exclude pulmonary embolism.
Options, in order of practicality at 01:00:
- Bilateral lower-limb ultrasound now. If a deep vein thrombosis is found in a symptomatic pregnant patient, treatment is indicated and the chest imaging question becomes moot. This is fast, radiation-free, and highly specific.
- If ultrasound is negative, proceed to CTPA once the fellow is available - recognising that a negative CTPA also has limits in this setting.
- Empirical anticoagulation with a therapeutic dose of low-molecular-weight heparin while awaiting completion, in discussion with obstetrics, is defensible given the high pretest probability and is safer than a delay.
Document the discussion, the reasoning, and the decision, and make sure the result is communicated verbally, not just posted to the worklist at 01:00.