Morbidity and mortality of major adult spinal surgery. A prospective cohort analysis of 942 consecutive patients

BACKGROUND CONTEXT: To date, most reports on the incidence of adverse events (AEs) in spine surgery have been retrospective and dependent on data abstraction from hospital-based administrative databases. To our knowledge, there have been no previous rigorously performed prospective analysis of all AEs occurring in the entire population of patients presenting to an academic quaternary referral center. PURPOSE: To determine the mortality and true incidence and severity of morbidity (major and minor, medical and surgical) in adults undergoing complex spinal surgery, both trauma and elective, in a quaternary referral center. To examine the influence of the introduction of a dedicated weekly multidisciplinary rounds, and a formal abstraction tool, on the recording of this prospective perioperative morbidity data. To examine the validity and inter-and intraobserver reliability of a dedicated Spine AdVerse Events Severity system, version 2 (SAVES V2) AE abstraction tool. STUDY DESIGN: Ours is an academic quaternary referral center serving a population of 4.5 million people. Beginning in April 2008, a spine-specific AE-recording instrument, entitled SAVES V2, was introduced at our center for reporting, categorization, and classification of AEs. The use of this system remains an ongoing prospective study. PATIENT SAMPLE: All adult patients admitted to the spine service of a quaternary referral center for a 12-month period. OUTCOME MEASURES: Avalidity and an inter-and intraobserver reliability examination of the SAVES V2 system, as used at our institution. Morbidity and inhospital deaths, unplanned second surgeries during index admission, wound infections requiring reoperation, and readmissions during the same calendar year. We also examined in detail all intraoperative and nonsurgical postoperative AEs, as well as hospital length of stay (LOS). METHODS: Data on all patients undergoing surgery over a 12-month period were prospectively collected using a perioperative morbidity abstraction tool at weekly dedicated mortality and morbidity rounds. This tool allows identification of each specific AE and grades the severity. Before the introduction of this system, and using the hospital inpatient database, our documented perioperative morbidity rate (major and minor, medical and surgical) was 23%. Diagnosis, operative data, hospital data, major and minor complications both medical and surgical, and deaths were recorded. RESULTS: One hundred percent of all patients discharged from the unit had complete data available for analysis. Nine hundred forty-two patients with an age range of 16 to 90 years (mean, 54 years; mode, 38 years) were identified. There were 552 males and 390 females. Around 58.5% of patients had undergone elective surgery. Thirty percent of patients were American Spinal Injury Association class D or worse on admission. The average LOS was 13.5 days (range, 1-221 days). Eight hundred twenty-two (87%) patients had at least one documented complication. Thirty-nine percent of these adversely affected hospital LOS. There were 14 mortalities during the study period. The rate of intraoperative surgical complication was 10.5% (4.5% incidental durotomy and 1.9% hardware malposition requiring revision and 2.2% blood loss >2 L). The incidence of postoperative complication was 73.5% (wound complications, 13.5%; delerium, 8%; pneumonia, 7%; neuropathic pain, 5%; dysphagia, 4.5%; and neurological deterioration, 3%). CONCLUSIONS: Major spinal surgery in the adult is associated with a high incidence of intra-and postoperative complications. We identified a very high rate of previously unrecognized postoperative complications, which adversely affect LOS. Without strict adherence to a prospective data collection system, the true complexity of this surgery may be greatly underestimated. (C) 2012 Elsevier Inc. All rights reserved.