We undertook a departmental audit of the reporting of radiographs for paediatric scoliosis and found considerable variability in the content and accuracy of the reports.
Following discussion with our paediatric spinal surgeons,
we established standard criteria that should be included in the radiologist’s report of scoliosis radiographs.
Particular problems may arise if the radiologist reports a significantly different Cobb angle to the paediatric surgeon.
Radiologists should report the side of the scoliosis,
region of the spine affected,
apical vertebra,
correctly measured Cobb angle,
spinal developmental anomalies and Risser stage.
An audit of scoliosis radiograph reporting was performed of radiographs of paediatric patients having whole spine radiographs between January and August 2016,
with first imaging at the tertiary referral centre from January 2015.
The content of the first radiograph report was recorded,
including side,
region,
apical vertebra,
Cobb angle,
developmental anomalies and Risser stage.
The Cobb angle measured by the Paediatric Spinal Surgeon was also obtained from the related clinic letter.
RESULTS:
195 reports were audited,
of which 71% were female patients,
29% male patients.
The mean age at first imaging at our tertiary referral centre was 13.4 years,
however 53% had external imaging uploaded from external hospitals.
As expected,
the majority of patients had idiopathic scoliosis (82%),
while a small percentage had scoliosis due to neuromuscular disorders (13%),
developmental vertebral anomalies (3%),
neoplasia (1%) and mesodermal disorders (1%) ( Fig. 18 ).
47% patients had a single PA radiograph and 45% had both a PA and a lateral radiograph ( Fig. 19 ).
7% had PA,
lateral and bending radiographs.
1% had PA and bending radiographs.
In addition to plain radiographs,
56% patients had imaging with both CT and MRI; 3% had MRI alone and 1% had CT alone ( Fig. 20 ).
Radiograph reports:
The majority (51%) of radiographs were reported by Consultant Radiologists,
with registrars and reporting radiographers reporting the remainder ( Fig. 21 ).
One radiograph was autoreported in error.
Although the side and location of the curve and the Cobb angle were included in more than 80% reports,
only around 60% included the level of the apex of the curve and a comment on whether there were any developmental vertebral or rib anomalies ( Fig. 22 ).
Only 6% reports included the Risser stage.
Of the reports including the side of the scoliosis,
20% incorrectly reported the scoliosis as the concave side.
The side should be reported as the deviation from the midline,
i.e.
the side of the convexity of the curve.
Agreement of reported Cobb angle with that measured by the Paediatric Spinal Surgeon:
The mean difference in the Cobb angle measured by the Radiology reporter and the Paediatric Spinal Surgeon was 5.4 degrees,
with a range of 0-42 degrees ( Fig. 23 ).
9% of radiology-reported Cobb angles differed by more than 10o compared to the Paediatric Spinal Surgeon’s Cobb angle.
Follow-up imaging:
Surgery
21% of the patients had surgery within the follow-up period,
which was internal fixation for spinal fusion in all cases.
A survey of all the paediatric scoliosis patients with radiograph review over a single month found that 38% had previous surgical intervention,
of which 13% had extendable rods and 87% had spinal fusion.
Complications from surgery:
None of the 41 audited patients who had surgery developed complications within the follow-up period.