Outcomes of cervical spine surgery in teaching and non-teaching hospitals

Abstract

Study design: Retrospective national database analysis.

Objective: A national population-based database was analyzed to characterize cervical spine procedures performed at teaching and nonteaching hospitals with regards to patient demographics, clinical outcomes/complications, resource use, and costs.

Summary of background data: There are mixed reports in the literature regarding the quality and costs of health care provided by teaching hospitals in the United States. However, outcomes of cervical spine surgery based upon teaching status remains largely unknown. METHODS.: Data from the Nationwide Inpatient Sample were obtained from 2002-2009. Patients undergoing elective anterior or posterior cervical fusion, or posterior cervical decompression (i.e., laminoforaminotomy, laminectomy, laminoplasty) for a diagnosis of cervical myelopathy and/or radiculopathy were identified and separated into 2 cohorts (teaching and nonteaching hospitals). Patient demographics, comorbidities, complications, length of hospitalization, costs, and mortality were compared for both groups. Regression analysis was performed to assess independent predictors of mortality.

Results: A total of 212,385 cervical procedures were identified from 2002-2009 in the United States, with 54.6% performed at teaching hospitals. More multilevel fusions and posterior approaches were performed in teaching hospitals (P < 0.0005). Patients treated in teaching hospitals trended toward male sex, increased costs, and hospitalizations. Overall, procedure-related complications and inhospital mortality were increased in teaching hospitals. Regression analysis revealed that significant predictors of mortality were age 65 years or more (odds ratio = 3.0) and multiple comorbidities. Teaching status was not a significant predictor of mortality (P = 0.07).

Conclusion: Patients treated in teaching hospitals for cervical spine surgery demonstrated longer hospitalizations, increased costs, and mortality compared with patients treated in nonteaching hospitals. Incidences of postoperative complications were identified to be higher in teaching hospitals. Possible explanations for these findings are an increased complexity of procedures performed at teaching hospitals. Older age and presence of comorbidities were more significant predictors of inhospital mortality than teaching status. Future studies should identify long-term complications and costs beyond an inpatient setting to assess if differences extend beyond the perioperative period.

Level of evidence: 4.