Socioeconomic and Psychosocial Predictors of Magnetic Resonance Imaging After Cervical and Thoracic Spine Trauma in the United States

Abstract

Socioeconomic factors are known to influence outcomes after spinal trauma, but it is unclear how these factors affect health care utilization in acute care settings. We aimed to elucidate if sociodemographic and psychosocial factors are associated with obtaining magnetic resonance imaging (MRI), a costly imaging modality, after cervical or thoracic spine fracture. Data from the 2012-2016 American College of Surgeons National Trauma Data Bank were used. We assessed the relationship between receipt of MRI and patient-level sociodemographic and psychosocial factors as well as hospital characteristics while correcting for injury-specific characteristics. Multiple logistic regression was performed to assess for associations between these variables and MRI after spine trauma. A total of 213,071 patients met the inclusion criteria, of whom 13.0% had an MRI (n= 27,757). After adjusting for confounders in multivariate regression, patients had increased odds of MRI if they were Hispanic (odds ratio [OR], 1.09; P= 0.001) or black (OR, 1.14; P < 0.001) or were diagnosed with major psychiatric disorder (OR, 1.06; P= 0.009), alcohol use disorder (OR, 1.05; P < 0.001), or substance use disorder (OR, 1.10; P < 0.001). Patients with Medicare (OR, 0.88; P < 0.001) or Medicaid (OR, 0.94; P < 0.011) were less likely to have an MRI than were those with private insurance, whereas patients treated in the Northeast (OR, 1.48; P < 0.001) or at for-profit hospitals (OR, 1.12; P < 0.001) were more likely. After adjusting for injury severity and spinal cord injury diagnosis, psychosocial comorbidities and for-profit hospital status were associated with higher odds of MRI, whereas public insurance was associated with lower odds. Results highlight potential biases in the provision of MRI as a costly imaging modality.