DEXA Protocol (rev. 2/3/2014)
1) Every patient must have a fully completed DXA questionnaire.
2) Know your DXA machines maximum weight capacity at your site and do not exceed it.
4) Remove all artifacts, if possible, prior to scanning (i.e. naval rings, bra clips, coins, buttons/snaps, etc)
5) Each patient shall have their weight and height taken by the technologist/nurse prior to scanning and entered into the DXA unit.
6) Make certain that the patient’s age, sex, ethnicity and historical data are accurate and entered as appropriate into the DXA system.
9) THREE valid/acceptable sites are required (if possible) for each DXA scan.
10) The spine and both hips (if able/valid) are the primary sites to be evaluated on every patient.
11) Lateral vertebral morphology should be performed on each patient every time.
non-dominant forearm must be performed to allow for a total of THREE valid sites to be evaluated.
15) If an artifact is present involving the spine at one vertebral level and it can’t be removed (i.e. naval ring, surgical screw/plate etc.) then that vertebral level should be eliminated from the scan and the remaining vertebrae should be analyzed and averaged together for interpretation. In this case, the non-dominant forearm should also be scanned and submitted for interpretation.
in addition to our standard three primary site protocol.
17) If the spine or hip(s) can’t be scanned for any reason (patient can’t hold still, severe scoliosis, significant DJD, prior surgery, metal objects, patient overweight for table etc.) a note must be provided to explain this in Synapse under the “Notes” tab and/or call the reading PA.
area scanned before.
19) FRAX analysis should be performed when applicable on each scan.
20) Assure that all paperwork is filled out and this info is scanned into the system properly.
21) End and send all DXA scans efficiently.
22) Direct comparison of DXA scans from centers that have different DXA models, software and/or have not been cross calibrated should not be performed, as the precision error is too great for comparative data to be meaningful.