• Keaton Piper, Ian DeAndrea-Lazarus, Hanna Algattas, Kristopher T Kimmell, James Towner, Yan M Li, Kevin Walter, and George E Vates.
• Department of Neurosurgery, University of Rochester Medical Center, Rochester, New York, USA. Electronic address: keaton_piper@urmc.rochester.edu.
• World Neurosurg. 2018 Feb 1; 110: e627-e635.

ObjectiveReoperation and readmission are often avoidable, costly, and difficult to predict. We sought to identify risk factors for readmission and reoperation after spine surgery and to use these factors to develop a scoring system predictive of readmission and reoperation.MethodsThe National Surgical Quality Improvement Project database for years 2012 to 2014 was reviewed for patients undergoing spinal surgery, and 68 perioperative characteristics were analyzed.ResultsA total of 111,892 patients who underwent spinal surgery were identified. The rate of reoperation was 3.1%, the rate of readmission was 5.2%, and the occurrence of either was 6.6%. Multivariate analysis found 20 perioperative factors significantly associated with both readmission and reoperation. Preoperative and operative factors found significant included age >60 years, African-American race, recent weight loss, chronic steroid use, on dialysis, blood transfusion required, American Society of Anesthesiologists classification ≥3, contaminated wound, >10% probability of experiencing morbidity, and operative time >3 hours. Postoperative associations identified included urinary tract infection, stroke, dehiscence, pulmonary embolism, sepsis, septic shock, deep and superficial surgical site infection, reintubation, and failure to wean from ventilator. An unweighted and weighted risk score were generated that yielded receiver operating characteristic curves with areas under the curve of 0.707 (95% confidence interval [CI]: 0.701-0.713) and 0.743 (95% CI: 0.736-0.749) 0.708 (95% CI: 0.702-0.715), respectively.ConclusionsPatients with an unweighted score ≥7 had a more than 20-fold increased risk of reoperation or readmission and a more than 1000-fold increased risk of mortality than did patients with a score of 0.Copyright © 2017 Elsevier Inc. All rights reserved.

24 keywords...

hide…