Study population
During the 3-month study period, 452 cervical spine X-ray and 153 CT cervical spine were performed within the ED, averaging six imaging referrals a day. The mean age of patients who underwent cervical spine X-ray was 56.0 (17.3 SD) years and 52.8 (21.4 SD) years for CT cervical spine, with a slight female preponderance (n = 248/452; 54.9%) for cervical spine X-ray and a male preponderance (n = 100/153; 65.4%) for CT cervical spine.
Cervical Spine Imaging (Conventional radiography and CT)
64 (14.2%) and 70 (15.5%) X-ray referrals had imaging recommendations of “Usually Appropriate” based on ACR AC and ESR iGuide respectively, while RCR iRefer did not have this category of recommendation. Likewise, based on ACR AC and ESR iGuide, there were 100 (22.1%) and 91 (20.1%) cases that were “Usually Not Appropriate”. According to RCR iRefer, the majority of the cases (n = 444/452; 98.2%) had the recommendation of “Indicated only in specific circumstances". Only a small percentage of cases (1.3–9.1%) did not meet any of the criteria.
139 (90.8%) CT referrals had imaging recommendations of “Usually Appropriate” according to ACR AC and ESR iGuide, while RCR iRefer did not have this category of recommendation. Based on RCR iRefer, 11 (7.2%) and 142 (92.8%) cases had recommendations “Indicated” and “Indicated under specific circumstances”, respectively. A small number of cases (n = 14/153; 9.2%) were classified as “Usually Not Appropriate” by ACR AC and ESR iGuide. (Refer to Table 1).
Imaging Characteristics, including positivity and negativity rates
For X-ray cervical spine using the ACR AC and ESR iGuide, the positivity rate was highest (ACR AC: 4.7% and ESR iGuide: 4.3%) in the category of “Usually Appropriate”, while the lowest positivity rate (0%) was in the category of “Usually Not Appropriate”. With respect to the RCR iRefer, the positivity rate was 1.1% for the only category of “Indicated only in specific circumstances”. For the X-ray cervical spine cohort, there was a significant difference between the imaging recommendation(s) and positivity/negativity rate for the respective individual imaging referral guidelines: ACR AC (p = 0.039; <0.05) and ESR iGuide (p = 0.049; <0.05). We did not perform any statistical analysis for RCR iRefer because it had only one imaging recommendation.
In terms of the CT cohort, the positivity rate was the highest in the category of “Usually Appropriate” in both ACR AC (7.2%) and ESR iGuide (7.2%). The "Indicated only in specific circumstances" category for RCR iRefer had the highest positivity rate (7%) and the category “Indicated” showed no positive findings. There was no significant difference between the imaging recommendation(s) and positivity/negativity rate for the CT cervical spine cohort: ACR AC (p = 0.600; >0.05), ESR iGuide (p = 0.600; >0.05), and RCR iRefer (p = 1.00; >0.05). (Refer to Table 2 and 3).
Five (1.1%) X-ray cervical spine and 10 (6.5%) CT cervical spine cases had positive significant findings (i.e., fracture or prevertebral soft tissue swelling were considered positive; all else considered negative). Fracture of the cervical spine was detected on two (0.4%) X-ray cervical spine and 10 (6.5%) CT cervical spine. Three (0.7%) X-rays of the cervical spine showed pre-vertebral swelling.
The most common non-significant findings in X-ray cervical spine were cervical spondylosis (84.5%), followed by loss of the normal cervical lordosis (33.2%) and spondylolisthesis (31.9%). For the CT cervical spine cohort, the most prevalent non-significant findings were cervical spondylosis (73.9%), followed by spondylolisthesis (26.1%). Ossification of the posterior ligament was depicted in X-ray and CT cervical spine with a larger portion observed in the CT cohort (n = 19/153; 12.4%) than the X-ray cohort (n = 3/452; 0.7%). (Refer to Table 4-7)
The majority of patients in the X-ray cervical spine cohort were either discharged home after treatment (ACR AC: n=105/446; 23.5%, ESR iGuide: n=103/411; 25.0%, and RCR iRefer: n=103/443; 23.2%) or referred to specialists (ACR AC: n=180/446; 40.4%, ESR iGuide: n=160/452; 38.9%, and RCR iRefer: n=179/443; 40.4%) for further evaluation. A subset of patients were either admitted to orthopaedics or other medical specialties for subsequent medical care. Only a small proportion of the patients were admitted to the observation ward in the emergency department for protocolised care (emergency observational medicine). In the X-ray cohort, the majority of patients with inappropriate imaging (or indicated only in specific circumstances) received home discharges or referrals to specialists for outpatient appointments.
Most of the patients in the CT cervical spine cohort were admitted to either orthopaedics or other medical specialties for further medical care. The remaining patients were treated and discharged, referred to specialists for further assessment, or admitted to the observation ward of the ED. Similar to the X-ray cervical spine group, there were also a few patients who wished to be discharged from the hospital despite the advice of the clinicians (Discharged, At Own Risk). Most patients in the CT cohort with inappropriate imaging (or indicated only in specific circumstances) had hospital admissions for conditions unrelated to cervical spine.
The respective imaging referral guidelines recommended a test as a better imaging alternative to a cervical spine X-ray in 78.9% (ACR AC), 60.8% (ESR iGuide), and 80.6% (RCR iRefer) of cases. The number of cases in which the best imaging tests were available to replace the CT cervical spine was smaller, at 0.7% and 2%, as recommended by ESR iGuide and RCR iRefer, respectively. (Refer to Table 8).
