Identifying risks factors in thoracolumbar anterior fusion surgery through predictive analytics in a nationally representative inpatient sample.

Purpose: Anterior thoracolumbar (TL) surgical approaches provide more direct trajectories compared to posterior approaches. Proper patient selection is key in identifying populations that may benefit from anterior TL fusion. Here, we utilize predictive analytics to identify risk factors in anterior TL fusion in patients with trauma and deformity.

Methods: In this retrospective cohort study of patients receiving anterior TL fusion (between and including T12/L1), population-based regression models were developed to identify risk factors using the National Readmission Database 2016-2017. Readmissions were analyzed at 30- and 90-day intervals. Risk factors included hypertension, obesity, malnutrition, smoking, alcohol use, long-term opioid use, and frailty. Multivariate regression models were developed to determine the influence of each risk factor on complication rates.

Results: A total of 265 and 375 patients were identified for the scoliosis and burst fracture cohorts, respectively. In patients with scoliosis, alcohol use was found to increase the length of stay (LOS) (p = 0.00061) and all-payer inpatient cost following surgery (p = 0.014), and frailty was found to increase the inpatient LOS (p = 0.0045). In patients with burst fractures, malnutrition was found to increase the LOS (p < 0.0001) and all-payer cost (p < 0.0001), obesity was found to increase the all-payer cost (p = 0.012), and frailty was found to increase the all-payer cost (p = 0.031) and LOS (p < 0.0001).

Discussion: Patient-specific risk factors in anterior TL fusion surgery significantly influence complication rates. An understanding of relevant risk factors before surgery may facilitate preoperative patient selection and postoperative patient triage and risk categorization.