Effective-Dose Estimates and Diagnostic Reference Levels

Effective dose and diagnostic reference levels

Effective dose is a tissue-weighted reference-population quantity used in radiation protection and broad exposure comparison. It differs from SSDE, which is expressed in mGy and concerns dose in the scanned region. Effective dose is expressed in mSv and should not be presented as a precise probability of cancer for an individual patient.

A practical approximate method multiplies DLP by a regional conversion coefficient, often written k. The coefficient's units cancel mGy·cm to leave mSv. Its applicability depends on anatomy, patient age or model, reference phantom and the underlying dosimetry and tissue-weighting system. There is no universal k for every CT study.

Arithmetic using explicitly stated coefficients

For an illustrative head study with CTDIvol 56 mGy and length 18 cm, DLP = 56 × 18 = 1008 mGy·cm. If the exercise supplies a historical adult head coefficient of 0.0021 mSv/(mGy·cm), the estimated effective dose is 1008 × 0.0021 = 2.1168 mSv, rounded to 2.12 mSv. That is a calculation under the supplied model, not a precise measurement of that patient's risk.

For a chest example, CTDIvol 11.5 mGy over 32 cm gives 368 mGy·cm. Using a supplied historical coefficient of 0.014 gives 5.152 mSv, rounded to 5.15 mSv. The different coefficients reflect the region and the model, not a rule that a lower DLP always means a lower effective dose across different anatomy.

These coefficients are historical AAPM Report 96 examples. More appropriate contemporary or indication-specific methods may give different estimates. Do not mix an older coefficient with newer tissue-weighting factors and imply that the combination is a single validated model. Use the stated inputs for exam arithmetic and an appropriately selected model for actual dose assessment.

Multiphase and mixed-region studies

Additional acquisitions add exposure. For separate series in the same appropriately modeled region, sum their relevant DLP contributions before applying the matching coefficient when that approach is valid. For a study spanning substantially different regions, one coefficient may be a poor approximation; a physicist or validated organ-dose method can provide a more suitable assessment.

A diagnostic chest CT and an attenuation-correction CT in a hybrid examination may serve different tasks and use different settings. The CT dose record must include the relevant events. The radiopharmaceutical contribution is assessed separately; it is not embedded in CTDIvol or DLP. Likewise, CT fluoroscopy's repeated local acquisitions should not be hidden by reporting only the final volumetric series.

Diagnostic reference levels are investigation benchmarks

A diagnostic reference level (DRL) helps identify unusually high typical exposures for a defined clinical task and patient group. It is commonly set at the 75th percentile of a relevant distribution of facility median dose indices. An achievable dose is often the median benchmark in a specified dataset. Definitions and datasets should be identified rather than treated as universal numerical limits.

A DRL is not a legal ceiling for an individual patient's examination and does not guarantee that every result below it is optimized. A large patient or complex task may reasonably need more exposure. Conversely, a very low value can accompany inadequate image quality. Review clinical task, size and quality together before changing a protocol.

Match the comparison before judging the result

Compare the same indication, age or size group, region, phantom reference and dose-index definition. A pediatric chest reference cannot be applied directly to an adult brain scan. An accreditation pass/fail dose criterion is also a different type of requirement from a DRL. Clearly label which comparison is being made.

If a facility's typical values exceed a matched DRL, investigate phase count, range, exposure settings, positioning and reconstruction with the responsible team. A single high value does not establish malpractice or mandate repeating the examination. Documentation should identify the reason and any protocol correction.

Communicate estimates without false precision

Natural-background comparisons can provide context, but exposure varies geographically and the biological distribution differs from a CT examination. Avoid claiming that a scan is exactly equivalent to a fixed number of days for every patient. Explain that the figures are approximate and relate them to why the examination was needed.

A useful report preserves the actual indices, range, phase count and calculation method. That record permits a later reassessment with a better model. An unlabeled effective-dose number copied from a generic table cannot provide the same evidence.

References: AAPM Report 96, FDA DRL and optimization explanation.