Sentinel events in lumbar spine surgery are rare but potentially catastrophic life threatening or altering events. These are typically investigated on a case-by-case basis using root cause analysis, but they are so infrequent that systematic study to determine their incidence, associated risk factors, and trends over time is difficult without a large database of cases to analyze. As such, Dr. Marquez-Lara and colleagues from Chicago queried the National Inpatient Sample (NIS) database from 2002-2011 to determine the rates of wrong site surgery, bowel or peritoneal injury, vascular injury, nerve injury, and retained foreign body in over 500,000 inpatient lumbar spine surgeries. They found the overall rate of sentinel events was 0.8/1000, which remained relatively stable over time. Wrong site surgery was reported in approximately 1/3,000 cases, vascular and nerve injury in 1/5,000 cases, and retained foreign objects in about 1/10,000 cases. The only sentinel event with a significant change in incidence over time was wrong-site surgery, which had no reported cases in 2002 and a rate of 1/2,000 cases in 2011 (p=0.006). Fusion of three or more levels was almost twice as common in the sentinel event cohort, and anterior surgery was associated with a higher rates of vascular and bowel injuries. There was a trend towards an increased overall rate of sentinel events in cases involving anterior surgery compared to posterior only surgery, though this was apparently not statistically significant. Patients with a sentinel event were significantly more likely to have a host of complications following surgery, including death. Bowel injury increased the risk of death 200-fold, while the overall relative risk of mortality in the sentinel event group was increased 20-fold. The sentinel event cohort had a length of stay twice that of the non-sentinel event cohort and a hospital cost nearly twice as high.
This paper represents one of just a few studies looking at safety in spine surgery on a large scale, and it is probably the first one focused specifically on sentinel events. A large database like the NIS is needed to perform such a study, and this study design has both advantages and disadvantages that need to be considered when interpreting the results. The NIS only captures complications that were coded during the index hospitalization, so complications that were diagnosed on an outpatient basis or during a subsequent admission would not be included. Additionally, some sentinel events may not be coded, and an audit of actual cases could provide an estimate of how consistently the events are captured in this database. The nerve injury rate of 1/5,000 cases seems low compared to what is in the literature, but this complication is likely undercoded. A potentially concerning finding was the supposed increased rate of wrong site surgery over time despite efforts such as marking the site, time-outs, and surgical checklists. In 2002, no cases of wrong site surgery were reported, which also raises concern for the accuracy of the coding data. It seems possible that with increased awareness of wrong site surgery that the rate of reporting this sentinel event has increased more than the actual rate of its occurrence. This paper provides some benchmark rates of sentinel events in lumbar spine surgery and also demonstrates some of the consequences in terms of costs, hospital stays, complications, and death. While root cause analysis may give a better understanding of the underlying cause of individual sentinel events, this system-wide perspective is important as well.
Please read this article in the May 15 issue. Does this paper change how you view sentinel events in lumbar spine surgery? Let us know by leaving a comment on The Spine Blog.
Adam Pearson, MD, MS
Associate Web Editor