RESUMO
INTRODUCTION: Clinical clearance of a child's cervical spine after trauma is often challenging due to impaired mental status or an unreliable neurologic examination. Magnetic resonance imaging (MRI) is the gold standard for excluding ligamentous injury in children but is constrained by long image acquisition times and frequent need for anesthesia. Limited-sequence MRI (LSMRI) is used in evaluating the evolution of traumatic brain injury and may also be useful for cervical spine clearance while potentially avoiding the need for anesthesia. The purpose of this study was to assess the sensitivity and negative predictive value of LSMRI as compared to gold standard full-sequence MRI as a screening tool to rule out clinically significant ligamentous cervical spine injury. METHODS: We conducted a ten-center, five-year retrospective cohort study (2017-2021) of all children (0-18y) with a cervical spine MRI after blunt trauma. MRI images were re-reviewed by a study pediatric radiologist at each site to determine if the presence of an injury could be identified on limited sequences alone. Unstable cervical spine injury was determined by study neurosurgeon review at each site. RESULTS: We identified 2,663 children less than 18 years of age who underwent an MRI of the cervical spine with 1,008 injuries detected on full-sequence studies. The sensitivity and negative predictive value of LSMRI were both >99% for detecting any injury and 100% for detecting any unstable injury. Young children (age < 5 years) were more likely to be electively intubated or sedated for cervical spine MRI. CONCLUSION: LSMRI is reliably detects clinically significant ligamentous injury in children after blunt trauma. To decrease anesthesia use and minimize MRI time, trauma centers should develop LSMRI screening protocols for children without a reliable neurologic exam. LEVEL OF EVIDENCE: 2 (Diagnostic Tests or Criteria).
RESUMO
BACKGROUND: The purpose of this study was to accurately predict pediatric choledocholithiasis with clinical data using a computational machine learning algorithm. METHODS: A multicenter retrospective cohort study was performed on children <18 years of age who underwent cholecystectomy between 2016 to 2019 at 10 pediatric institutions. Demographic data, clinical findings, laboratory, and ultrasound results were evaluated by bivariate analyses. An Extra-Trees machine learning algorithm using k-fold cross-validation was used to determine predictive factors for choledocholithiasis. Model performance was assessed using the area under the receiver operating characteristic curve on a validation dataset. RESULTS: A cohort of 1,597 patients was included, with an average age of 13.9 ± 3.2 years. Choledocholithiasis was confirmed in 301 patients (18.8%). Obesity was the most common comorbidity in all patients. Choledocholithiasis was associated with the finding of a common bile duct stone on ultrasound, increased common bile duct diameter, and higher serum concentrations of aspartate aminotransferase, alanine transaminase, lipase, and direct and peak total bilirubin. Nine features (age, body mass index, common bile duct stone on ultrasound, common bile duct diameter, aspartate aminotransferase, alanine transaminase, lipase, direct bilirubin, and peak total bilirubin) were clinically important and included in the machine learning algorithm. Our 9-feature model deployed on new patients was found to be highly predictive for choledocholithiasis, with an area under the receiver operating characteristic score of 0.935. CONCLUSION: This multicenter study uses machine learning for pediatric choledocholithiasis. Nine clinical factors were highly predictive of choledocholithiasis, and a machine learning model trained using medical and laboratory data was able to identify children at the highest risk for choledocholithiasis.