Diagnostic Reference Levels: A Practical Guide

Diagnostic reference levels (DRLs) are benchmark dose values that flag imaging protocols delivering unusually high—or unusually low—radiation dose for a standard-sized patient undergoing a specified examination. A DRL is not a dose limit for an individual patient. It is a dose-management tool: a facility compares the median dose of a representative patient sample to the published benchmark, and a median above the DRL signals that the protocol should be reviewed and optimized.¹²³

Used correctly, DRLs are the first practical step in optimizing patient dose without sacrificing diagnostic image quality. Used incorrectly—as a per-patient ceiling, or as a target to barely satisfy—they can either alarm clinicians unnecessarily or stall genuine optimization. This guide explains where DRLs come from, how they differ from achievable doses, which dose quantities are used, and how a facility translates a benchmark comparison into a defensible optimization program.¹²⁴

Introduction

Radiation dose in medical imaging varies enormously between facilities performing the same examination. Multiple large registry studies have shown two- to threefold spreads—and sometimes wider—in CT dose for the same clinical indication, even after accounting for patient size.⁴⁵ That variation is not explained by patient anatomy or clinical need; it reflects differences in protocol design, scanner technology, and operator practice. A chest CT that is diagnostic at one site may be delivering twice the dose at another with no diagnostic benefit.

Diagnostic reference levels exist to surface that variation in a structured, comparable way. The concept, introduced by the International Commission on Radiological Protection (ICRP) in the 1990s and developed in detail in ICRP Publication 135, gives facilities a benchmark against which to compare their own typical dose for a specific protocol.³ In the United States, NCRP Report No. 172 and the ACR–AAPM–SPR practice parameter translate that framework into national survey data and practical recommendations, while the ACR Dose Index Registry supplies the large-scale CT dose data that makes meaningful benchmarking possible.¹²⁴

DRLs sit at the heart of every modern dose-optimization and accreditation program. DRPS integrates DRL benchmarking into CT physics testing, protocol review, and accreditation support for imaging facilities across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

Topic Explanation

What is a diagnostic reference level?

A diagnostic reference level is a benchmark value of an easily measured dose quantity, set at a selected percentile of a survey dose distribution, that indicates whether the typical dose for a specific imaging protocol is unusually high for a standard-sized patient.³ Three features of that definition matter in practice:

• It is statistical, not individual. A DRL applies to the median dose of a representative sample of standard-sized patients (or to a standard phantom), never to a single patient.
• It is protocol- and modality-specific. A DRL for an adult routine head CT is meaningless when applied to a CT pulmonary angiogram.
• It is a trigger, not a pass/fail line. Exceeding a DRL does not mean a facility is non-compliant; it means the protocol warrants investigation.

The companion concept is the achievable dose (AD). NCRP Report No. 172 formalized achievable doses as a more demanding optimization target—typically the median (50th percentile) of the survey distribution—against which a well-optimized facility can measure itself. Where the DRL flags the high end of the distribution, the AD describes what good practice routinely attains.²

Why "standard-sized patient" is the foundation

Dose-index quantities such as CTDIvol depend strongly on patient size. A benchmark is only meaningful if the comparison population is matched in size, which is why DRL surveys define a standard-sized adult (often expressed through a water-equivalent diameter or weight band) or use pediatric weight bands.³ ICRP Publication 135 specifically recommends weight bands for establishing pediatric DRLs for trunk examinations, because a single adult-style benchmark cannot describe the dose appropriate to a 5-kg neonate and a 70-kg adolescent at the same time.³

For the radiation-dose vocabulary that underlies DRLs—CTDIvol, DLP, and how they relate to patient dose—see our companion guide to CTDIvol, DLP, and CT dose metrics.

DRLs as the first step in optimization

A DRL comparison answers one question: "Is our typical dose for this protocol unusually high?" It does not, by itself, tell a facility how to fix a protocol or whether image quality is adequate. That is why ICRP and NCRP describe DRLs as the first step in optimization, not the whole process.²³ Once a protocol is flagged, the medical physicist investigates technique factors, automatic exposure control settings, reconstruction methods, and image-quality requirements to determine whether dose can be reduced while preserving diagnostic performance. The work of optimization happens after the benchmark comparison, drawing on tools like CT protocol optimization and tube current modulation.

Key Technical Principles

Which dose quantities are used as DRLs

DRLs use practical, measurable machine-output or dose-index quantities—not organ dose or effective dose, which cannot be measured directly at scale. The quantity chosen depends on the modality:

These quantities are the same ones collected during routine physics testing, which is what makes large-scale benchmarking feasible. For the mammography-specific quantity, see our guide to mean glandular dose in mammography.

Size correction with the size-specific dose estimate

Because CTDIvol is referenced to a standard 16-cm or 32-cm phantom rather than the actual patient, comparing CTDIvol across patients of different sizes can mislead. AAPM Report No. 204 introduced the size-specific dose estimate (SSDE), which scales CTDIvol by a size-dependent conversion factor derived from the patient's effective or water-equivalent diameter:⁶

where $f_{size}$ is the conversion factor tabulated against patient size. As a worked illustration, suppose an abdomen CT reports a ${\text{CTDI}}_{vol}$ of 12 mGy referenced to the 32-cm body phantom, and the patient has a water-equivalent diameter of 26 cm. Using an AAPM Report 204 conversion factor of approximately 1.18 for that diameter:⁶

The patient-relevant dose estimate is meaningfully higher than the displayed CTDIvol because the patient is smaller than the 32-cm reference phantom. Size correction is the reason modern CT DRL work increasingly reports SSDE alongside CTDIvol and DLP, and why registry studies stratify benchmarks by patient size.⁴⁶ (The conversion factor value here is illustrative; the actual factor is read from the patient-specific diameter using the published tables.)

Comparing a facility median to a DRL

The benchmarking comparison itself is straightforward. A facility collects the chosen dose quantity for a representative sample of standard-sized patients undergoing a specific protocol, computes the sample median, and compares it to the national DRL:

If, for example, a facility's median routine-abdomen CTDIvol is 18 mGy and the applicable national DRL is 16 mGy, the protocol sits above the benchmark and is a candidate for review. If the same facility's median were 9 mGy—below both the DRL and a 12-mGy achievable dose—the protocol is already well optimized, and attention is better spent verifying that image quality remains diagnostic at that dose.²⁴ The percentile logic is what makes DRLs robust: a benchmark anchored at the 75th percentile is, by construction, exceeded by roughly a quarter of facilities, and those are precisely the protocols optimization should target first.

The role of large registries

National DRLs require large, representative datasets. The ACR Dose Index Registry (DIR) aggregates CT dose-index data from hundreds of facilities, enabling DRLs and achievable doses to be derived as a function of patient size and clinical indication.⁴⁷ Landmark analyses of U.S. registry data established DRLs and achievable doses for the ten most common adult CT examinations stratified by patient size, and later work extended the approach to clinical-indication-based benchmarks and to pediatric body CT.⁴⁵⁸ Registry participation also lets a facility see where its own protocols sit within the national distribution in near-real time, which is far more actionable than a static published table.

Clinical Impact

DRLs convert an abstract concern—"are we using too much dose?"—into a concrete, comparable, and actionable signal. Their clinical value shows up in several ways.

First, DRLs catch outlier protocols that would otherwise go unnoticed. A protocol cloned years ago, never revised after a scanner upgrade, can quietly deliver double the necessary dose. Benchmarking against a DRL surfaces that protocol so the physics and radiology teams can act on it.⁴⁵

Second, DRLs support equity of dose across a practice. Multisite groups frequently discover that the "same" examination is performed very differently across locations. Standardizing toward the achievable dose narrows that spread, so a patient receives a comparable, optimized dose regardless of which site they visit.⁴⁸

Third, DRLs anchor the optimization conversation in data rather than anecdote. When a protocol is flagged, the discussion shifts from "this feels like a lot of dose" to "our median is above the national 75th percentile—let us review the technique." That framing is far more persuasive with referring physicians and equipment vendors.

Finally, DRLs matter for pediatric imaging, where dose stewardship is especially important because children are more radiosensitive and have a longer lifetime for effects to manifest. Weight-banded pediatric DRLs and registry analyses of pediatric body CT give children's hospitals and general facilities a concrete target for size-appropriate technique.³⁸ The same size-sensitive philosophy underlies our guidance on pediatric CT dose optimization.

Practical Optimization Tips

Build benchmarking into routine workflow

• Automate dose-data capture. Use a dose-management system or registry feed that ingests the radiation dose structured report (RDSR) from each exam, rather than transcribing values by hand. Manual collection does not scale and discourages regular review.
• Stratify by size. Compare like with like. Report SSDE or stratify CTDIvol/DLP by water-equivalent diameter or weight band so the comparison is meaningful.³⁶
• Use both the DRL and the achievable dose. Treat the DRL (75th percentile) as the "investigate now" trigger and the achievable dose (median) as the optimization target.²

Investigate flagged protocols methodically

When a protocol exceeds its DRL, work through the dose-and-quality chain rather than simply lowering technique:

1. Confirm the data—exclude mis-protocoled exams, multiphase studies counted as single-phase, and size outliers.
2. Review automatic exposure control settings, noise index or quality reference targets, and kV selection.
3. Evaluate iterative or deep-learning reconstruction, which can preserve image quality at lower dose.
4. Verify that image quality remains diagnostic at the proposed lower dose—optimization is never dose reduction alone.

Avoid common DRL mistakes

• Do not treat the DRL as a per-patient limit. A large patient or a demanding indication may justifiably exceed it.³
• Do not "optimize to the DRL." The benchmark is a ceiling to stay under, not a target to approach; aim for the achievable dose.²
• Do not ignore unusually low doses. A median far below the achievable dose can signal inadequate image quality and should also prompt review.³
• Do not let benchmarks go stale. Reassess DRLs and your own protocols on a recurring schedule and after any equipment, software, or protocol change.

Connect benchmarking to accreditation

Accrediting bodies and many state programs expect documented dose review and protocol management. A DRL-based program—periodic benchmarking, documented investigation of outliers, and recorded protocol changes—produces exactly the evidence reviewers look for. DRPS folds DRL benchmarking into accreditation support and ongoing medical physicist consulting.

Regulatory Considerations

DRLs are guidance benchmarks, not federal radiation-dose limits, but they are increasingly embedded in accreditation requirements, professional practice parameters, and state expectations for dose management. Understanding that distinction is essential.

• ACR–AAPM–SPR Practice Parameter for Diagnostic Reference Levels and Achievable Doses in Medical X-Ray Imaging. This practice parameter, revised in 2023, defines how DRLs and achievable doses should be established and used in U.S. practice and is the primary domestic reference for facilities and physicists.¹
• NCRP Report No. 172. Provides U.S. survey-based reference levels and achievable doses across radiography, fluoroscopy, CT, and dental imaging, and formalizes the achievable-dose concept.²
• ICRP Publication 135. The international methodological basis for DRLs, including recommended quantities by modality, weight bands for pediatrics, and guidance on updating intervals.³
• Accreditation programs. ACR and other accrediting bodies expect documented protocol review and dose management; participation in the ACR DIR is one recognized mechanism for benchmarking CT dose.⁴⁷

Unlike CT and radiography output, which are regulated as radiation-producing machines under state and FDA authority, DRLs are not enforced as numeric limits. They function as the consensus standard of care for dose management, and a facility that ignores benchmarking exposes itself to findings during accreditation review even though no single dose value is "illegal." DRPS helps facilities document a defensible DRL program consistent with these references and with state radiation-control expectations across our service areas. For the broader compliance picture, see ACR accreditation physics requirements.

Frequently Asked Questions (FAQs)

What is a diagnostic reference level (DRL)?

A diagnostic reference level is a benchmark dose value—usually the 75th percentile of a dose distribution from a survey of many facilities—used to identify imaging protocols that deliver unusually high (or sometimes unusually low) radiation dose for a standard-sized patient undergoing a specified examination. It is a dose-management tool, not a dose limit for an individual patient.

Is a DRL a dose limit for a patient?

No. DRLs are not regulatory dose limits and are never applied to an individual patient. They apply to the median dose of a representative sample of standard-sized patients for a specific protocol. A single patient may justifiably receive a dose above the DRL when clinically warranted; the DRL only signals that a protocol's typical dose should be reviewed.

What is the difference between a DRL and an achievable dose?

A DRL is typically set at the 75th percentile of survey data and flags protocols at the high end of the dose distribution. An achievable dose (AD) is typically the median (50th percentile) and represents a more aggressive optimization target. Consistently operating near or below the AD indicates a well-optimized protocol, while exceeding the DRL is a trigger to investigate.

Which dose quantities are used as DRLs?

DRL quantities are easily measured machine-output or dose-index metrics rather than organ or effective dose. For CT, CTDIvol and DLP (and increasingly SSDE) are used. For radiography, incident air kerma or kerma-area product (KAP). For fluoroscopy, reference air kerma and KAP. For mammography, mean glandular dose. These quantities are practical to collect at scale.

How does a facility compare its dose to a DRL?

A facility collects dose-index data for a sample of standard-sized patients (or a standard phantom) for a given protocol, computes the median value, and compares that median to the published national DRL. If the facility median exceeds the DRL, the protocol is a candidate for review and optimization. Comparing the facility median to the achievable dose provides a tighter optimization target.

Where do U.S. diagnostic reference levels come from?

In the United States, DRLs are informed by NCRP Report No. 172, the ACR–AAPM–SPR practice parameter, and large registries such as the ACR Dose Index Registry, which aggregates CT dose-index data from hundreds of facilities. National and international guidance from ICRP Publication 135 establishes the underlying methodology.

How often should DRLs be reviewed and updated?

DRLs should be reassessed periodically—commonly every three to five years—because technology and practice evolve and dose distributions shift downward as optimization spreads. Facilities should also review their own protocols on a recurring schedule and after any equipment, software, or protocol change that could affect dose.

Key Takeaways

• A DRL is a benchmark, not a limit. It flags protocols whose typical dose is unusually high for a standard-sized patient and triggers review, not a pass/fail compliance line.³
• DRL versus achievable dose. The DRL is typically the 75th percentile (investigate-now trigger); the achievable dose is typically the median (optimization target).²
• Use measurable quantities. CTDIvol, DLP, and SSDE for CT; air kerma and KAP for radiography and fluoroscopy; mean glandular dose for mammography.³⁶
• Correct for size. SSDE scales CTDIvol by a size-dependent factor so the dose estimate reflects the actual patient, not a reference phantom.⁶
• Registries make benchmarking real. The ACR Dose Index Registry supplies the large, size-stratified datasets that underpin credible U.S. CT DRLs.⁴⁷
• Benchmarking is the first step, not the whole job. After a protocol is flagged, optimization preserves diagnostic image quality while reducing dose.²³

Conclusion

Diagnostic reference levels turn the wide, often invisible variation in imaging dose into a structured, comparable signal that a facility can act on. They are deliberately statistical and protocol-specific: a benchmark for the typical dose of a standard-sized population, never a ceiling for an individual patient. The DRL flags the high end of practice; the achievable dose describes where good practice routinely operates; and the gap between them is the optimization opportunity.

A mature program collects dose data automatically, stratifies by patient size, compares facility medians to national DRLs and achievable doses, investigates flagged protocols methodically, and documents the result. Done well, DRL benchmarking protects patients, standardizes dose across a practice, and produces exactly the evidence that accreditation reviewers expect—while keeping image quality firmly in the foreground.¹²³⁴

How DRPS Can Help

Diagnostic Radiation Physics Services helps imaging facilities build and document DRL-based dose-optimization programs. That work includes CT physics testing, dose-index data review and registry benchmarking, protocol optimization, SSDE and size-stratified analysis, pediatric technique review, and accreditation support prepared by board-certified medical physicists.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware. A strong DRL program is not about chasing the lowest possible number—it is about delivering consistent, optimized, diagnostic-quality imaging and being able to prove it.

Related Resources

• CTDIvol, DLP, and CT dose metrics
• CT protocol optimization
• CT tube current modulation
• Pediatric CT dose optimization
• Mean glandular dose in mammography
• ACR accreditation physics requirements
• CT physics testing
• Accreditation support
• Medical physicist consulting

References

1. American College of Radiology, American Association of Physicists in Medicine, Society for Pediatric Radiology. ACR–AAPM–SPR Practice Parameter for Diagnostic Reference Levels and Achievable Doses in Medical X-Ray Imaging. Revised 2023 (Resolution 22). acr.org
2. National Council on Radiation Protection and Measurements. Reference Levels and Achievable Doses in Medical and Dental Imaging: Recommendations for the United States. NCRP Report No. 172. Bethesda, MD: NCRP; 2012. ncrponline.org
3. International Commission on Radiological Protection. Diagnostic Reference Levels in Medical Imaging. ICRP Publication 135. Ann ICRP. 2017;46(1). icrp.org
4. Kanal KM, Butler PF, Sengupta D, Bhargavan-Chatfield M, Coombs LP, Morin RL. U.S. diagnostic reference levels and achievable doses for 10 adult CT examinations. Radiology. 2017;284(1):120-133. doi:10.1148/radiol.2017161911. doi.org
5. Bos D, Yu S, Luong J, et al. Diagnostic reference levels and median doses for common clinical indications of CT: findings from an international registry. Eur Radiol. 2021;32(3):1971-1982. doi:10.1007/s00330-021-08266-1. doi.org
6. American Association of Physicists in Medicine. Size-Specific Dose Estimates (SSDE) in Pediatric and Adult Body CT Examinations. AAPM Report No. 204. College Park, MD: AAPM; 2011. aapm.org
7. American College of Radiology. Dose Index Registry (DIR). National Radiology Data Registry. acr.org
8. Wildman-Tobriner B, Strauss KJ, Bhargavan-Chatfield M, et al. Using the American College of Radiology Dose Index Registry to evaluate practice patterns and radiation dose estimates of pediatric body CT. AJR Am J Roentgenol. 2018;210(3):641-647. doi:10.2214/AJR.17.18122. doi.org
9. Jordan DW, Becker MD, Brady S, et al. Validation of adult relative radiation levels using the ACR Dose Index Registry: report of the ACR Appropriateness Criteria Radiation Exposure Subcommittee. J Am Coll Radiol. 2019;16(2):236-239. doi:10.1016/j.jacr.2018.08.008. doi.org
10. Ahmad M, Liu X, Morani AC, et al. Oncology-specific radiation dose and image noise reference levels in adult abdominal-pelvic CT. Clin Imaging. 2022;93:52-59. doi:10.1016/j.clinimag.2022.10.016. doi.org