The CT Dose Index Discrepancy
A medical physicist is reviewing CT dose data after a new protocol was installed. The console reports CTDIvol = 24 mGy and DLP = 720 mGy·cm for a 30 cm phantom. The physicist's independent measurement with a 100 mm pencil chamber in the same phantom gives CTDIvol = 31 mGy. The scanner's annual survey is due, and the department has been using the console values for its dose registry. A clinician asks whether the difference matters for a patient who received 3 chest CTs in one week. The physicist must decide which value to trust, whether the registry is wrong, and what to tell the clinician.
1. Why might the console and the independent measurement differ, and which should be used for patient dose records?
Read the worked answer
Differences between console CTDIvol and an independent measurement can arise from: chamber calibration and its traceability, beam width and the 100 mm chamber's integration length, phantom material and positioning, the scanner's own CTDI phantom assumptions (16 cm vs 30 cm), and the scanner's internal calibration factors. A 24 vs 31 mGy discrepancy (~29%) is large enough to investigate. The independent measurement is generally the reference if the chamber is calibrated and the geometry matches; the console value is only valid for the conditions it was calibrated for. For patient dose records, use the value that is traceable and validated - typically the independent measurement or a validated console factor - and document the method. The judgement is not to blindly trust either number: verify chamber calibration, repeat the measurement, and check whether the console uses a different phantom or beam width. Until resolved, the registry should be flagged, not silently continued.
2. The clinician asks whether 3 chest CTs in a week is dangerous. How do you frame the answer quantitatively?
Read the worked answer
Estimate effective dose from DLP using a conversion coefficient (k ≈ 0.014–0.018 mSv/mGy·cm for chest). With DLP = 720 mGy·cm per scan, effective dose ≈ 10–13 mSv per scan, so 3 scans ≈ 30–39 mSv. Using the higher measured CTDIvol (31 mGy) the DLP and effective dose would be proportionally higher - perhaps 40–50 mSv total. Frame this against context: natural background is ~2–3 mSv/yr, and the risk at these doses is small and stochastic (no deterministic effects). The clinician should understand that the dose is not trivial and that justification and optimization matter, but the patient is not in acute danger. The answer should be quantitative, state assumptions (conversion coefficient, scan length), and avoid false reassurance or alarm. The judgement is to translate instrument units into a clinically meaningful effective dose and to flag the discrepancy's impact on that estimate.
3. What quality-control and reporting actions follow from the discrepancy?
Read the worked answer
Actions: (1) recalibrate/verify the pencil chamber and electrometer; (2) repeat the measurement with correct phantom and beam width; (3) compare with the scanner's acceptance-test baseline; (4) if the console is systematically low, apply a correction factor to registry data and re-run the dose registry for affected patients; (5) document the investigation and any protocol change; (6) review the protocol's justification and optimization (is 24–31 mGy CTDIvol appropriate for chest?). If the console under-reports, the registry and any dose-tracking alerts are unreliable, so the correction must propagate. The judgement is that a QC discrepancy is a data-integrity issue with downstream patient-record consequences, not just a physics curiosity.