Almost a third of adult CT scans exceeded dose thresholds in a 254-facility audit

Design

Evaluation of an automated national quality measure

Population

783,634 adult diagnostic CT examinations at 254 US facilities

Primary outcome

Percentage of examinations above dose or noise thresholds

Effect

30.7% above dose threshold; 0.06% above noise; facility range 1.8–82.4%

This evaluation covers the first 16 months of a US CMS quality measure that automatically checks every adult diagnostic CT against category-specific thresholds for size-adjusted radiation dose and image noise. It included 783,634 examinations across 16 CT categories at 254 facilities, from May 2025 to August 2026.

Overall, 30.7% of examinations exceeded the dose threshold; only 0.06% failed on noise. Facility scores ranged from 1.8% to 82.4% out of range (median 29.5%). Routine abdomen and pelvis, the highest-volume category, had a median of 43.9% out of range, and high-dose head 82.6%. Within the same facility, the gap between its best and worst category was a median of 74 percentage points.

The near-absence of noise failures suggests many protocols run higher dose than image quality requires. Because most variation sat at the level of specific protocol in specific facility, a department-wide average hides where the problem is. The useful step for any CT service, in or outside the US, is to audit size-adjusted dose by protocol category.

• Audit CT dose by protocol category, not as a single department average
• Start with routine abdomen and pelvis, the highest-volume and often highest out-of-range category
• Use size-adjusted dose metrics so larger patients are not flagged unfairly
• Check image noise alongside dose before lowering a protocol
• Involve a medical physicist in protocol revision

Why it matters

Most excess dose sits in specific protocols, which makes it fixable once found.

Don't overread it

Exceeding a threshold is a quality flag, not evidence of harm to an individual patient.

The statistics, in plain English

An intraclass correlation of 0.12 means a facility's performance in one CT category tells you little about its performance in another. That is why the authors argue each facility needs category-level results to know which protocols to fix.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.