Imaging in low back pain: a cross-sectional analysis of Australian early-career general practitioners' ordering of imaging for non-specific low back pain.

Abstract

Introduction Low back pain is a leading cause of morbidity worldwide, but its cause is often non-specific. Imaging should not take place unless there are specific 'red flag' features present, as it does not improve patient outcomes. Aim To establish prevalence and factors associated with ordering of imaging for new onset non-specific low back pain (new onset NSLBP) by early-career GPs ('registrars'). Methods Cross-sectional analysis from a subset of data from the ReCEnT study of GP registrars' clinical experiences (2010-2018). The outcome was whether imaging was ordered for a patient with new onset NSLBP. Analyses employed univariable and multivariable regression. Results A total of 2333 GP registrars (96.0% response rate) contributed data from 325,058 consultations, comprising 508,316 patient problems/diagnoses. A total of 3066 problems/diagnoses (0.6%) were new onset NSLBP. Of the 3066 new onset NSLBP problems/diagnoses, 450 (15%) had imaging ordered. In multivariable models, variables significantly associated with imaging for new onset NSLBP were patient age: 35-64 years (adjusted OR 1.55, P =0.002) or over 65 years (OR 2.32, P P =0.009), scheduling of follow-up with the registrar (OR 3.61, P P =0.013), and 'generation of learning goals' (OR 1.96, P =P P Discussion Although Australian GP registrars are taking a considered approach to imaging for NSLBP, the prevalence of imaging likely exceeds optimal levels, at significant cost and potentially poorer patient outcomes. Refinement in guidelines regarding appropriate 'red flags' for imaging may aid in the further reduction of unnecessary imaging.