Patient characteristics

A total of 452 radiography and 153 CT imaging referrals conducted on patients from the ED were identified and included. (Refer to Table 3). The number of older patients (aged 65 years or older) in the X-ray and CT cervical spine groups was 102 (22.6%) and 46 (30.1%), respectively (age range of 17 - 99 years).

Table 3: Patient characteristics

X-ray Cervical Spine

The overall appropriateness rate of X-ray cervical spine imaging referrals was 75.2%, 70.8%, and 13.3% for ACR AC, ESR iGuide, and RCR iRefer, respectively. (Refer to Table 4).

Table 4: Summary of the comparison among the imaging referral guidelines

Most of the imaging was performed on non-traumatic patients, with a smaller number of imaging referrals performed on trauma patients (30.8%–31.2%). The use of X-ray cervical spine for diagnostic investigation of headaches formed 6.2% of the total imaging referrals. A minority of imaging referrals (1.3%–9.1%) did not align with any of the specified criteria. When the imaging referrals were assessed using the ACR AC, a significant proportion of the referrals were assigned the recommendations "May Be Appropriate" (19.5%) and "May Be Appropriate (Disagreement)" (42.9%). Conversely, the ESR iGuide classified 55.3% of the referrals as "May Be Appropriate", while the RCR iRefer did not have a similar category. The RCR iRefer classifies a substantial number of referrals as "Indicated only in specific circumstances" (n = 416/452, 92.0%). Only 60 referrals (14.4%) were classified as "Indicated" after taking into account the imaging considerations outlined in the RCR iRefer. (Refer to Table 5-7).

Table 5: Evaluation of X-ray Cervical Spine imaging referrals - ACR AC

Table 6: Evaluation of X-ray Cervical Spine imaging referrals - ESR iGuide

Table 7: Evaluation of X-ray Cervical Spine imaging referrals - RCR iRefer

CT Cervical Spine

All CT cervical spine imaging was performed on traumatic patients.  The CT cervical spine's appropriateness rate was 90.8%, with all 3 imaging referral guidelines yielding the same number of indicated procedures (Table 3). Both ACR AC and ESR iGuide had the same number of recommendations categorised as "Usually Appropriate" (n=139/153, 90.8%) and "Usually Not Appropriate" (n=14/153, 9.2%). In contrast, RCR iRefer recommended “Indicated only in specific circumstances” for most of the referrals (n = 142/153, 92.8%), with only a minority “Indicated” recommendation (n = 11/153, 7.2%). After taking into account the imaging considerations mentioned in RCR iRefer, out of the referrals categorised as "Indicated only in specific circumstances", 128 imaging referrals (n = 128/142, 90.1%) were considered “Indicated”. (Refer to Table 8-10).

Table 8: Evaluation of CT Cervical Spine imaging referrals - ACR AC

Table 9: Evaluation of CT Cervical Spine imaging referrals - ESR iGuide

Table 10: Evaluation of CT Cervical Spine imaging referrals - RCR iRefer

Concordance of guidelines

Fleiss’ Kappa was run to determine if there was agreement between the three sets of imaging referral guidelines for X-ray and CT of the cervical spine [9]. Fleiss’ Kappa showed that there was slight agreement between the imaging referral guidelines for X-ray of the cervical spine, κ=0.135 (95% CI, 0.088 to 0.183), p < 0.001. For CT cervical spine, Fleiss’ Kappa showed that there was almost perfect agreement between the imaging referral guidelines, κ=1.000 (95% CI, 0.909 to 1.091), p < 0.001.

Cohen's coefficient was employed to assess the level of agreement among the various sets of imaging guidelines. For the CT cervical spine, there was almost perfect agreement across the three sets of imaging referral guidelines. ACR AC and ESR iGuide had moderate agreement, while RCR iRefer had no level of agreement. (Refer to Table 11).

Table 11: Inter-rater agreement for the imaging referral guidelines (Cohen’s kappa)