Patients and inclusion/ exclusion criteria
In July 2017 the imaging protocol in our hospital was changed for patients with acute wrist trauma to include a CBCT of the wrist in patients who had negative radiographs at initial evaluation and persistent clinical concern for a radiographically occult radiocarpal fracture.
This scan was performed simultaneously with radiographs at time of follow up review in the orthopaedic clinic,
usually 7-14 days after the initial trauma.
If CBCT showed a fracture which corresponded to a point of clinical tenderness,
the patient was considered to have a fracture and treated as such.
If neither radiography or CBCT showed a fracture repeat clinical examination was performed.
If no fracture was suspected at this point the patient was discharged.
If there was ongoing concern for fracture,
immobilisation with repeat examination and MRI was arranged.
A retrospective review was performed from July 2017 to February 2018 of all patients who underwent CBCT for suspected occult fracture of the radiocarpal bones.
Patients were excluded if they met any of the following criteria: less than 16 years of age,
pregnancy,
definite fracture on initial radiographs,
CBCT performed more than 14 days after trauma,
no clinical concern for fracture at follow up examination.
Imaging techniques
Radiographs - Radiographs in patients with a suspected distal radius fracture consisted of two images: Anteroposterior (AP) and lateral views of the wrist in the neutral position.
Radiographs for suspected carpal fracture consisted of four images: AP and lateral views of the wrist in the neutral position and an oblique view and scaphoid view with 20-30 degree tube angulation in ulnar deviation.
CBCT - All CBCT images were obtained using a Planmed Verity extremity CBCT scanner (Planmed Oy,
Helsinki,
Finland).
The patient’s hand was placed on a gantry within a 13cm x 16cm field of view.
The examination was conducted using a 90kV and 36mA protocol.
No iodine contrast was injected.
Slice thickness was 0.2mm.
Total scan acquisition time is approximately 36s,
with approximately 1 minute of image processing time after the scan.
A single coronal reconstruction was sent to PACS,
and assessed using Multiplanar reconstruction (MPR).
Image Analysis
Evaluation took place under standardised viewing conditions on diagnostic reporting monitors.
All images were independently reviewed by a junior radiologist (3 years experience) and senior radiologist (11 years experience).
Each radiologist determined whether a fracture was present,
if there was fracture displacement and if the fracture represented a chip or involved the body of the bone.
Fracture displacement was defined as either gapping and/or translation of 1mm or greater at the cortical surface.
A chip fracture was defined as a fracture involving cortex only,
while a fracture extending through cortex into the trabeculae of the bone was defined as involving the body.
Statistical analysis
The sensitivity,
specificity,
negative predictive value,
positive predictive value,
accuracy,
95% confidence intervals,
and kappa coefficient for inter-rater agreement were calculated using SPSS software (SPSS,
Chicago,
Ill,
USA).