Evaluating new paralysis, mortality, and readmission among subgroups of patients with spinal epidural abscess: A latent class analysis

Abstract

Background Spinal epidural abscess (SEA) is increasing in incidence; this not-to-miss diagnosis can cause significant morbidity and mortality, particularly if diagnoses are delayed. While some risk factors for SEA and subsequent mortality have been identified, the SEA patient population is clinically heterogeneous and sub-populations have not yet been characterized in the literature. The primary objective of this project was to identify characteristics of subgroups of patients with SEA. The secondary objective was to identify associations between subgroups and three clinical outcomes: new onset paralysis, in-hospital mortality, and 180-day readmission. Methods Demographics and comorbid diagnoses were collected for patients diagnosed with SEA at an academic health center between 2015 and 2019. Latent class analysis was used to identify clinical subgroups. Chi-squared tests were used to compare identified subgroups with clinical outcomes. Results We identified two subgroups of patients in our analysis. Group 1 had a high rate of medical comorbidities causing immunosuppression, requiring vascular access, or both. Group 2 was characterized by a high proportion of people with substance use disorders. Patients in Group 2 were more likely to be readmitted within 6 months than patients in Group 1 (p = 0.03). There was no difference between groups in new paralysis or mortality. Discussion While prior studies have examined the SEA patient population as a whole, our research indicates that there are at least two distinct subgroups of patients with SEA. Patients who are younger, with substance use disorder diagnoses, may have longer hospital courses and are at higher risk of readmission within six months. Future research should explore how to best support patients in both groups, and additional implications for subgroup classification on health outcomes, including engagement in care.