The publication on “Augmented reality-guided pedicle screw fixation: an experimental study” is interesting [1]. This study has substantial statistical limitations even though it demonstrated great accuracy of pedicle screw placement utilizing the augmented reality-assisted pedicle screw fixation (ARPSF) technique. Just five pig cadaveric spines and 50 screws were used in the investigation. Statistical analysis is unable to differentiate between-subject variability (intra-specimen vs. inter-specimen variability), even though this number of screws seems adequate. Additionally, only one highly skilled surgeon carried out the procedure, making it impossible to evaluate the impact of experience, skill, or surgeon variability.

The study results are at risk of being influenced by several confounding variables, such as differences in swine and human anatomy, which may limit the translation of the results to clinical practice. Furthermore, bleeding-free surgical conditions, patient movement, or complex pathology (e.g., osteoporosis, spinal deformity) were not included in the model. Furthermore, using point-pair matching with only 15 landmarks may introduce systematic bias that is not statistically analyzed, impacting the accuracy of image registration.

Reanalysis of the results revealed that the average entrance point deviation of only 0.55 mm and the average deviation angle of 2.04° are highly effective. However, publishing only average data may mask the “outlier effect” of misplaced screws. Although only 6% of cases were classed as grade B breaches, in real-world scenarios, even minor breaches can be harmful if they occur in key places such as the cervical spine or near the spinal cord. As a result, interpretations should emphasize the “clinical safety margin” rather than relying merely on average accuracy numbers [2,3]. Furthermore, the average screw time of 2.2 minutes may seem short, but in a complex operating room, preparation time, image registration, and integration workflows must be taken into account, which may be significantly extended.

Based on this study, other research concerns might be posed, such as: (1) Will the ARPSF system maintain its accuracy when applied to real-world patients with deformities or osteoporosis? (2) Will there be a noticeable learning curve when used by surgeons with varying levels of experience? (3) How will accuracy change or decrease as the patient moves during surgery? (4) What extra information will be gained by comparing computed tomography-based or robotic-assisted fixation navigation systems? Furthermore, questions regarding cost-effectiveness and long-term clinical outcomes should also be addressed. Asking these questions will aid in the development of the ARPSF, both as a proof-of-concept and for future standardized use.