Background: Healthcare readmissions are important causes of increased cost and have profound clinical impact. Thirty-day readmissions in spine surgery have been well documented. However, rates, causes, and outcomes are not well understood outside 30 d.

Objective: To analyze 30- and 90-d readmissions for a retrospective cohort of anterior cervical discectomy and fusions (ACDF) and total disc replacement (TDR) for degenerative cervical conditions.

Methods: The Nationwide Readmissions Database approximates 50% of all US hospitalizations with patient identifiers to track patients longitudinally. Patients greater than 18 yr old were identified. Rates of readmission for 30 and 90 d were calculated. Predictor variables, complications, outcomes, and costs were analyzed via univariate and multivariable analyses.

Results: Between January and September 2013, 72 688 patients were identified. The 30- and 90-d readmission rates were 2.67% and 5.97%, respectively. The most prevalent reason for 30-d readmission was complication of medical/surgical care (20.3%), whereas for 90-d readmission it was degenerative spine etiology (19.2%). Common risk factors for 30- and 90-d readmission included older age, male gender, Medicare/Medicaid, prolonged initial length of stay, and various comorbidities. Unique risk factors for 30- and 90-d readmissions included adverse discharge disposition and mechanical implant-related complications, respectively. When comparing ACDF and TDR, ACDFs were associated with increased 90-d readmissions (6.0% vs 4.3%). The TDR cohort had a shorter length of stay, lower complication rate, and fewer adverse discharge dispositions.

Conclusion: Identification of readmission causes and predictors is important to potentially allow for changes in periperative management. Decreasing readmissions would improve patient outcomes and reduce healthcare costs.