RESUMEN
The prognostic significance of preoperative MRI findings in patients undergoing discectomy is incompletely understood. Identifying the radiological predictors of revision surgery on pre-operative MRI can guide management decisions and potentially prevent multiple surgeries. We included 181 patients who underwent primary lumbar discectomy between 2010 and 2014. All patients were contacted via a short telephone interview to determine if they had revision surgery within 5 years of their index surgery. Preoperative MRI of the lumbosacral spine was evaluated for various radiological factors including type of disc herniation, anatomical location of herniation, direction of herniation, degree of disc degeneration, end plate changes and presence of listhesis. Other potential confounders including age, gender, smoking status and index level of surgery were also recorded. Multivariate model of all radiological predictors and confounders were developed and a step-wise approach was used to remove insignificant variables in order to develop final significant multivariate model. P value of <0.05 was considered statistically significant. Patients with retrolisthesis were found to be 2.7 times more likely than the patients without listhesis to require revision surgery (p = 0.019). Patients with foraminal disc herniation were 3.45 times more likely than the patients with paramedian disc herniation to require revision surgery (p = 0.026). Other MRI predictors failed to achieve statistical significance. Based on the data presented patients with retrolisthesis and/or foraminal disc herniation should be counselled on the relatively higher risk of revision surgery when proceeding with discectomy, or alternative options should be considered.