Development of a model to predict the probability of incurring a complication during spine surgery

Abstract

Purpose Predictive models in spine surgery are of use in shared decision-making. This study sought to develop multivariable models to predict the probability of general and surgical perioperative complications of spinal surgery for lumbar degenerative diseases. Methods Data came from EUROSPINE's Spine Tango Registry (1.2012–12.2017). Separate prediction models were built for surgical and general complications. Potential predictors included age, gender, previous spine surgery, additional pathology, BMI, smoking status, morbidity, prophylaxis, technology used, and the modified Mirza invasiveness index score. Complete case multiple logistic regression was used. Discrimination was assessed using area under the receiver operating characteristic curve (AUC) with 95% confidence intervals (CI). Plots were used to assess the calibration of the models. Results Overall, 23′714/68′111 patients (54.6%) were available for complete case analysis: 763 (3.2%) had a general complication, with ASA score being strongly predictive (ASA-2 OR 1.6, 95% CI 1.20–2.12; ASA-3 OR 2.98, 95% CI 2.19–4.07; ASA-4 OR 5.62, 95% CI 3.04–10.41), while 2534 (10.7%) had a surgical complication, with previous surgery at the same level being an important predictor (OR 1.9, 95%CI 1.71–2.12). Respectively, model AUCs were 0.74 (95% CI, 0.72–0.76) and 0.64 (95% CI, 0.62–0.65), and calibration was good up to predicted probabilities of 0.30 and 0.25, respectively. Conclusion We developed two models to predict complications associated with spinal surgery. Surgical complications were predicted with less discriminative ability than general complications. Reoperation at the same level was strongly predictive of surgical complications and a higher ASA score, of general complications. A web-based prediction tool was developed at https://sst.webauthor.com/go/fx/run.cfm?fx=SSTCalculator.

Performance of hybrid artificial intelligence in determining candidacy for lumbar stenosis surgery

“…However, we also found as important predictors certain specific clinical symptoms including motor deficit reported by the doctor, back pain and leg weakness as reported from the patient. These variables differ greatly from previous studies aiming to predict surgery complications which found the health insurance provider [13] or the ASA score (physical status score) as the best predictors [14].…”

“…In a previous application of AI for a preoperative risk prediction model, a machine learning approach based on lasso logistic regression was used to predict complications after spinal surgery depending on patient variables with AUROC ranging from 0.7 to 0.76 [13]. Another regression model obtained similar results [14]. Machine learning models were also proposed to predict pain and functional outcomes after surgery [15][16][17][18].…”

“…Here, we propose a novel AI model to compute the probability to recommend spinal surgery for LSS that is in concordance with surgeon decisions. Compared to previously published models predicting complication risks [13,14], our model has the potential to directly compute the probability to admit a patient to spinal surgery, thus representing an effective augmentation tool to the medical decision making process. The model consists of a hybrid approach combining (i) a random forest model to accurately estimate model parameters from medical vignette data reviewed by surgeons, with (ii) an expert Bayesian network model implementing surgical recommendations from peer-reviewed published literature together with the expert opinions of a multidisciplinary team in the fields of spinal surgery, rehabilitation medicine, interventional and diagnostic radiology.…”

Risk factors for reoperation after surgical treatment for degenerative spinal disease in Poland: a nationwide retrospective study of 38,953 hospitalisations

“…This may involve surgical intervention at the same site, at a different site but due to the same condition, or repair of complications resulting from the initial surgical procedure [8]. Reoperation per se is strongly predictive of surgical complications in spinal surgery [9]. An additional burden of other health problems (multimorbidity) may carry an increased risk of surgical failure in DSD patients, including the need for reoperation.…”

3D visualization and morphometric analysis of spinal motion segments and vascular networks: A synchrotron radiation‐based micro‐CT study in mice

“…When surgery is conducted near the spinal cord, complications will occur that could be very serious (Nasser et al, 2010). During spine surgery, the proximity of neural and vessel structures within the spinal cord requires high accuracy from surgeons (Zehnder et al, 2021). In patients with severe spinal deformity or congenital scoliosis, the spinal anatomy can be severely altered or extremely dysmorphic (Perez‐Machado et al, 2020).…”