Pediatric Back Pain: A Scoring System to Guide Use of Magnetic Resonance Imaging

Background: The prevalence of back pain in the pediatric population is increasing, and the workup of these patients presents a clinical challenge. Many cases are selflimited, but failure to diagnose a pathology that requires clinical intervention can carry severe repercussions. Magnetic resonance imaging (MRI) carries a high cost to the patient and health care system, and may even require procedural sedation in the pediatric population. The aim of this study was to develop a scoring system based on pediatric patient factors to help determine when an MRI will change clinical management. Methods: This is a retrospective cohort analysis of consecutive pediatric patients who presented to clinic with a chief complaint of back pain between 2010 and 2018 at single orthopaedic surgery practice. Comprehensive demographic and presentation variables were collected. A predictive model of factors that influence whether MRI results in a change in management was then generated using cross-validation least absolute shrinkage and selection operator logistic regression analysis. Results: A total of 729 patients were included, with a mean age of 15.1 years (range: 3 to 20 y). Of these, 344 (47.2%) had an MRI. A predictive model was generated, with nocturnal symptoms (5 points), neurological deficit (10 points), age (0.7 points per year), lumbar pain (2 points), sudden onset of pain (3.25 points), and leg pain (3.75 points) identified as significant predictors. A combined score of greater than 9.5 points for a given patient is highly suggestive that an MRI will result in a change in clinical management (specificity: 0.93; positive predictive value: 0.92). Conclusions: A predictive model was generated to help determine when ordering an MRI may result in a change in clinical management for workup of back pain in the pediatric population. The main factors included the presence of a neurological deficit, nocturnal symptoms, sudden onset, leg pain, lumbar pain, and age. Care providers can use these findings to better determine if and when an MRI might be appropriate.