Major spinal surgery procedures have increased significantly 1) 2) for three main reasons: the increased age demographic of the general population, the introduction of minimally invasive methods including percutaneous procedures, and improved outcomes including reduced hospital stay and return to desirable lifestyle. Lumbar and cervical fusion are the main reported procedures on the spine and these numbers seems to be significantly increase because of life style variations 3) 4).
Although spine fusions are now considered minimally invasive techniques, the aggregate costs related to these surgeries has increased since the complexity of spinal involvement and number of levels to be fused have increased 5).
A metaanalysis on the effectiveness of minimally invasive techniques for lumbar spinal stenosis has revealed that there was no difference in terms of improved outcome for the most commonly used surgical techniques 6). Other important factors to be considered in complex spinal surgeries include length of the procedure and anaesthesia time, prolonged prone positioning and blood loss which can contributors to postoperative adverse events.
One of the medical fields that has intensively utilized the most advanced technologies is complex spinal surgery. A multitude of novel types of instrumentation, implants, navigation and biologics have recently become available for the use in complex spine surgery 7). However, critics point out that technologically advanced treatments may offer little or no clinical benefit compared to traditional treatment strategies 8).
Predictive clinical decision support is having an increasing impact in the field of risk stratification in complex spine surgery. Researchers are building accurate multivariate predictive models that can be applied to clinical practice in the form of decision support systems (DSS). Bekelis et al. created a statistical model to predict complications in spine surgery based on data from 13,660 patient cases. The model’s outcome variables included 30-day postoperative risk of stroke, myocardial infarction (MI), wound infection, urinary tract infection (UTI), death, deep vein thrombosis (DVT), pulmonary embolism, and unplanned return to surgery. Predictors were preoperative patient characteristics. The model was able to successfully discriminate between cases that did and did not experience complications. Areas under the receiver operating characteristics curves for each of the outcome variables ranged from moderate to high 9).