CTPA — Reduced Dose, Pregnancy-Adapted
Restricted to the pulmonary arterial tree — lung apices to the costophrenic angles — with no incidental extension into the abdomen.
Timings, volumes and delays here are representative values drawn from published guidance. Scanner generation, injector, cardiac output and local preference all move them. Confirm against your department's own protocol before you rely on a number.
When to use it
- Suspected pulmonary embolism in pregnancy or the immediate postpartum period, where CTPA has been selected over perfusion scintigraphy by the local pathway.
- Suspected PE in a young patient where breast dose is a material consideration.
Technique
- Reduced tube potential with iterative reconstruction is the standard route to dose reduction; published pregnancy-adapted protocols routinely achieve effective doses below about 1 mSv.
- Fetal dose from a CTPA is very small in absolute terms — orders of magnitude below any deterministic threshold — and the dominant radiation consideration is maternal breast dose, not the conceptus.
- Bismuth breast shielding is contentious: it lowers breast dose in some phantom work, but has also been shown to increase overall patient and fetal dose by interfering with automatic exposure control if placed before the scout. Sources genuinely disagree, and many departments have abandoned shielding in favour of tube current modulation and organ-based dose reduction.
- Timing is harder than in the non-pregnant patient; the increased cardiac output shortens the arrival time and increases the risk of a mistimed bolus.
- Bolus tracking with the monitoring region of interest in the main pulmonary trunk, a 100 HU trigger as used in the published pregnancy-adapted protocol, or an SVC-positioned ROI at the same threshold, and a post-trigger diagnostic delay of about 3-4 s. The shortened transit time is the reason the delay is at the short end and the reason a fixed delay is a poor choice here.
- A test bolus of about 15-20 mL at the diagnostic rate is the alternative where the threshold approach has already failed once, and it measures the individual transit time rather than assuming it.
- Reducing tube potential to 80-100 kVp raises iodine attenuation as well as cutting dose, and is part of how a smaller contrast volume still produces a diagnostic study — the two changes are made together, not independently.
- These are typical published values for a pregnancy-adapted protocol and vary by centre and by scanner generation; a fast wide-detector platform tolerates a tighter bolus than an older one. Confirm locally.
Where it goes wrong
- Transient interruption of contrast is more frequent in pregnancy because of increased cardiac output and inferior vena caval flow — the breathing instruction matters more here than anywhere else.
- Extending coverage below the diaphragm irradiates the conceptus directly for no diagnostic gain.
- Reducing dose so far that the study is non-diagnostic is not dose optimisation: a repeat scan doubles the exposure.
Clinical questions that reach this study
Contrast
Typically 50-80 mL of high-concentration non-ionic iodinated contrast (350-370 mgI/mL) at 4-6 mL/s through a cannula of at least 20G in an antecubital vein, followed by a 30-50 mL saline chaser. The rate is at or above the top of the non-pregnant range while the volume sits at or below the bottom of it: the aim is a short, dense bolus rather than a large one.
- The reason the contrast handling differs is haemodynamic, not radiation-related. By the second and third trimesters cardiac output has risen by roughly 30-50% and plasma volume by roughly 40-50%. The first change shortens transit time, so a bolus timed for a non-pregnant patient has already moved on; the second dilutes the same iodine load in a larger circulating volume, so peak attenuation is lower. A standard fixed delay therefore gives a suboptimal study on both counts at once — late AND dilute.
- The practical answers are a higher injection rate (iodine delivery rate, not total volume, sets the peak), bolus tracking rather than a fixed delay, and careful bolus tracking rather than a fixed delay. Note that the published pregnancy-adapted protocol (Ridge et al.) kept a 100 HU trigger at the main pulmonary artery and raised the injection RATE and volume; a lowered trigger is described in some local protocols but is not what that paper reports, and this card should not attribute it to them. A published alternative moves the monitoring ROI to the superior vena cava, again at 100 HU, because waiting for a threshold the diluted bolus may never reach is a classic cause of a non-diagnostic scan.
- Non-diagnostic and indeterminate CTPA rates are reported as higher in pregnancy than in the non-pregnant population, and this physiology is the reason. Confirm the local pregnancy-adapted regimen — practice varies and several centres run a lower-kV, higher-rate variant instead of changing the threshold.
Acquisition
- Breathing
- Quiet respiration or a shallow breath-hold; deep inspiration is specifically avoided.
- Reconstruction
- Thin axial reconstructions with reformats; noise-optimised iterative or deep-learning reconstruction where available.
- Preparation
- Cannula of at least 20G, antecubital or larger vein, capable of 4 mL/s or more. Breath-hold at end-inspiration but WITHOUT a deep Valsalva, which drives unopacified inferior vena caval blood into the right heart and interrupts the bolus.
Phases
Each phase is authored once and shared across every protocol that uses it, so the physiology below is the same wherever you meet it.
- CT angiography, bolus-trackedNo fixed delay. A monitoring ROI is placed in the target vessel (commonly the aorta) and acquisition triggers at a set attenuation rise — a ~100 HU threshold is widely used — followed by a short diagnostic delay of a few seconds for table movement and breath-hold instruction. Confirm locally.
This is not a separate physiological phase so much as a technique for hitting one reliably. The transferable principle is that a fixed delay assumes an average circulation, and the patients who most need vascular imaging — the shocked, the failing, the arrhythmic, the aneurysmal — are precisely those whose circulation time is furthest from average. Tracking the bolus in the target vessel replaces that assumption with a measurement, so peak arterial opacification is achieved in the individual patient. The trade-off is that the technique commits the scan to whatever the monitoring ROI sees: a badly placed ROI, a mistimed breath-hold or a threshold reached by a contralateral vein produces a systematically mistimed study rather than a slightly degraded one.
Safety checks this protocol carries
Derived from the contrast agent and phases above, not authored here — which is why they cannot drift apart from what the protocol actually does.
- Prior contrast reaction and elective premedication· nurse pre scan
- Intravenous access adequate for the planned injection· radiographer at scan
- Metformin and iodinated contrast· radiographer at scan
- Child-sized technique and contrast dose· radiographer at scan
- Pregnancy status before an ionising exposure· radiographer at scan
- Kidney function and intravenous iodinated contrast· radiographer at scan
References
- Low-dose CT pulmonary angiography protocol for imaging pregnant patients: dose reduction without loss of image quality.
- Ridge CA, et al. Pulmonary CT angiography protocol adapted to the hemodynamic effects of pregnancy — increased cardiac output and plasma volume shorten transit time and dilute the bolus, and a high-rate, bolus-tracked injection with a reduced trigger threshold improves pulmonary arterial opacification. AJR 2011.
- Sanghavi M, Rutherford JD. Cardiovascular physiology of pregnancy. Circulation 2014 — cardiac output rises by roughly 30-50% and plasma volume by roughly 40-50% by the second and third trimesters.
- To use or not use patient shielding on pregnant women undergoing CTPA: a phantom study. Radiat Prot Dosimetry 2020.
- Prevention, diagnosis and management of PE and DVT in pregnant women. ASH Education Program 2023.