DI +3.0 on a Chest Image and a Wrong-Patient Image Already in PACS
A DR chest room has a target exposure index of 320 for PA chests at 125 kVp. A 45-year-old woman's PA chest returns an EI of 640 and a deviation index of +3.0. On the monitor the image looks acceptable, because the technologist has already adjusted the window and level. Earlier that morning, an urgent portable was performed in resus; the technologist left the previous patient's demographics on the console, acquired the image, realised the error, corrected the name at the modality and re-sent it. The original had already been dispatched to PACS, which now holds the same image under two patient names, and a report on one of them is due within the hour.
1. What do the EI and DI values tell you, should the chest be repeated, and what would you change?
Read the worked answer
What the numbers mean: DI of +3.0 corresponds to 10^(3/10) ≈ 2, so the receptor received about twice the exposure needed to reach the target EI. The image is not under-exposed; it is over-exposed. It looks 'fine' on the monitor only because display brightness in DR is set by window and level, not by mAs - and that is exactly why windowing must never be used to judge exposure.
Repeat? No. There is no diagnostic deficiency. Repeating would double an already excessive dose for no information gain.
What to change:
- Do not adjust mAs to alter image brightness; in DR that only changes dose and noise.
- Check that the correct AEC chambers were selected (for a PA chest, the two outer chambers, not the central spine chamber).
- Check centring and collimation over the chambers - off-centre positioning over a denser or thinner part of the patient is a common cause of a high DI.
- Confirm the EI target is appropriate for the room's calibration and that the room's EI calibration is current.
- Consider a higher kVp with lower mAs if penetration is marginal.
- Log the DI and trend it: three consecutive readings above +2 on the same room is a QA trigger for an AEC/calibration service check, not a technologist problem.
2. Describe, in order, what you do about the image now held in PACS under two patient names, and why each step matters.
Read the worked answer
- Stop and identify. Confirm both patients using two identifiers (name plus MRN or date of birth) against the request and the worklist. Do not rely on the exam description.
- Do not delete anything. Deletion destroys the audit trail and may leave an orphaned report. Images are re-linked or moved to a rejected/hold status, not destroyed.
- Prevent the report. Immediately ask the PACS/RIS administrator to place the incorrect accession on hold and to unlink the images from the wrong patient's record, so the reporting radiologist cannot issue a report against the wrong name.
- Notify the radiologist and, if a report has already been verified, the referring clinician directly - a report on the wrong patient is a potential wrong-patient treatment event and must be corrected verbally, then in writing, not corrected quietly in the system.
- Re-link the image to the correct accession once the correct order exists, and confirm it appears in the correct patient's folder and nowhere else.
- Incident report and root cause: demographics were not cleared between patients and there was no two-identifier check at the console. Fix the process - a hard stop before exposure, and a 'clear previous patient' step built into the workflow - not just the individual.
3. What quality-assurance follow-up would you expect from this morning's two events, and how would you present it at the next departmental meeting?
Read the worked answer
Treat them as two different failure modes and report them separately but side by side, because both are system problems disguised as individual errors.
Exposure trend: present DI as a run chart for the room, not a single value. Show the proportion of chests with DI outside ±2 over the last three months, the AEC chamber used for each, and whether the deviation clusters by technologist, by patient size, or by time of day. A cluster by patient size points to technique charts and AEC chamber choice; a cluster in time points to a calibration drift or a busy period where centring is rushed.
Wrong-patient error: report it as a process audit - how many portables in the last month were acquired without a two-identifier check, how many had demographics carried over, and what the corrected workflow now is. Use it to justify a pre-exposure checklist and a training refresher.
Reject analysis: combine both into a monthly reject/QA review with a clear owner, a due date, and a re-audit date. Present the numbers, the cause, the change, and the evidence the change worked - that is what turns an incident into a quality improvement rather than a reprimand.