Diagnostic accuracy of perioperative electromyography in the positioning of pedicle screws in adolescent idiopathic scoliosis treatment: a cross-sectional diagnostic study

To investigate in the conventional techniques of the pedicle screws using triggered screw electromyography (t-EMG), considering different threshold cutoffs: 10, 15, 20 25 mA, for predicting pedicle screw positioning during surgery of the adolescent with idiopathic scoliosis (AIS). Sixteen patients (4 males, 12 females, average age 16.6 years) were included, with an average curve magnitude of 50 degrees and placement of 226 pedicle screws. Each screw was classified as "at risk for nerve injury" (ARNI) or "no risk for nerve injury" (NRNI) using CT and the diagnostic accuracy of EMG considering different threshold cutoffs (10,15, 20 and 25 mA) in the axial and Sagittal planes for predicting screw positions ARNI was investigated. The EMG exam accuracy, in the axial plane, 90.3% screws were considered NRNI. In the sagittal plane, 81% pedicle screws were considered NRNI. A 1-mA decrease in the EMG threshold was associated with a 12% increase in the odds of the screw position ARNI. In the axial and sagittal planes, the ORs were 1.09 and 1.12, respectively. At every threshold cutoff evaluated, the PPV of EMG for predicting screws ARNI was very low in the different threshold cutoff (10 and 15); the highest PPV was 18% with a threshold cutoff of 25 mA. The PPV was always slightly higher for predicting screws ARNI in the sagittal plane than in the axial plane. In contrast, there was a moderate to high NPV (78-93%) for every cutoff analyzed. EMG had a moderate to high accuracy for positive predicting value screws ARNI with increase threshold cutoffs of 20 and 25 mA. In addition, showed to be effective for minimizing false-negative screws ARNI in the different threshold cutoffs of the EMG in adolescent with idiopathic scoliosis (AIS).