Abstract
Background
Previous surgery is known to increase the risk of complications during spine surgery, but few studies have quantified the dose-response effect using multivariate models to account for confounders. We quantified the effect of the number of prior spine surgeries on perioperative complications in patients undergoing surgery for lumbar degenerative disorders.
Methods
We included data from 4′940 patients documented in Eurospine's Spine Tango Registry from 2004 to 2015. Medical history and surgical details were documented on the Tango Surgery form, as were surgical and general medical complications arising between admission and discharge. Multiple logistic regression models were built to investigate the relationship between the number of any previous surgeries and the presence of a perioperative complication, controlling for other potential confounders (age, sex, smoking, BMI, comorbidity, number of vertebral levels affected).
Results
There were 9.4% surgical complications (most commonly dural tear, neurological sequelae, haematoma, infection) and 6.5 % general complications (most commonly cardiovascular, urinary, pulmonary, liver/GI). In the multivariable models, previous surgery significantly increased the odds of having a surgical complication (OR 1.148, 95%CI 1.051–1.253; p=0.002) and a medical complication (OR 1.185, 95%CI 1.074–1.307; p=0.001) i.e., for each previous surgery, the odds of a complication increased by 15% and 19%, respectively.
Conclusion
We demonstrated a significant dose-response effect of previous surgery on the risk of incurring a complication during subsequent spine surgery. The results can be used by the surgeon when assessing the risk/benefit ratio of further surgery, for informing and consenting the patient.
No conflicts of interest
No funding obtained
