A Failing DR Room and a 138 kg Inpatient
You are the quality-control radiographer in a five-room hospital department. Room 3, a DR room installed four years ago, has produced three consecutive daily flat-field images with a faint vertical band in the left third of the detector. The service engineer cleaned the detector and reloaded the calibration; the band remains. This week's AEC test shows chamber 2 needs about 30% more mAs than chamber 1 to produce the same signal, and a kVp meter reads 78.5 kVp when the console is set to 80. The radiologists say lateral lumbar spine images from Room 3 look grainier than those from Room 5. Room 3 is the only table rated for a 140 kg patient, and six inpatients, including a 138 kg man with a suspected bowel obstruction, are booked this afternoon. The service contract guarantees an engineer within 48 hours. Department acceptance limits are: kVp within ±5% of set, AEC reproducibility within ±10% between chambers, and flat-field uniformity within ±5%.
1. Which findings pass or fail the department's limits, and what does each one tell you about the cause?
Read the worked answer
kVp: 78.5 against a set 80 is −1.9%, inside ±5% (76–84 kVp). Generator output is acceptable; do not chase it.
AEC: chamber 2 needing 30% more mAs is outside ±10%. Possible causes include chamber selection or position, calibration drift, phantom positioning, or detector gain. The consequence is that exposures using chamber 2 are unpredictable: either underexposed and noisy, or compensated and overexposed. That is a patient-dose and image-quality problem, not just a number on a log.
Uniformity: "faint band" is not a measurement. You cannot pass or fail it until it is quantified, for example by comparing region-of-interest signals across the band. The fact that it persists after cleaning and recalibration points to a detector or column defect or a gain-calibration problem, not surface dirt.
Graininess on lateral lumbar: consistent with a detector or calibration problem rather than the generator, since kVp passes. It supports taking the image-quality concern seriously and ties the radiologists' observation to the QC data.
2. Do you remove the room from service? What alternatives do you have for the 138 kg inpatient?
Read the worked answer
Decision framework. Remove the room from diagnostic use if an artifact could be mistaken for or hide pathology, or if exposure is unpredictable.
- A uniformity band crossing the spine or chest could simulate or mask a fracture or line; that alone justifies restriction until it is quantified.
- The AEC failure means chamber 2 cannot be used. Manual technique or chambers 1 and 3 only, with radiologist agreement, may allow limited work.
- kVp passes, so the generator is not the reason to stop.
The 138 kg inpatient. Options in order:
- Ask the radiologist whether a portable supine abdomen is clinically adequate.
- Check whether another table can take the weight with a bariatric extension.
- Escalate the service call from 48 hours to same-day or next-day.
- If none works, arrange transfer or another room, with a documented rationale.
Do not proceed on a defective detector simply because the table is convenient. Document the decision, who approved it, and the limitation of any image produced.
3. What is your documentation and follow-up plan?
Read the worked answer
Document: date and time, test type, equipment ID, meter used, measured values, acceptance limits, pass/fail, room status, who was notified, service reference number, and any limited-use authorisation with the approving radiologist's name.
Follow-up:
- Repeat the full QC after service, before returning the room to full use; confirm with a phantom and a clinical image.
- Trend the room's QC against the other rooms to detect drift early.
- Request a physicist's evaluation if the band persists after the next service visit.
- Report to the QC or radiation safety committee with actions and deadlines.
- If the defect may be longstanding, ask the radiologist whether earlier images from the room need review, and add the room to a closer monitoring schedule.