In Reply: Novel Barium-Enhanced 3-Dimensional-Printed Spine Model for Pedicle Screw Training: A Cost-Effective Solution and Educational Validation
Makaleler > Kısa Makale · 2025
Yazarlar: Bilal Bahadır AKBULUT, Mustafa Serdar BÖLÜK, Taşkın Yurtseven, Hüseyin BİÇEROĞLU
Operative Neurosurgery, Uluslararası, Wos, 2025, cilt 29, ss 606-607
DOI: 10.1227/ons.0000000000001737
Özet
To the Editor: We appreciate the interest shown by Daungsupawong and Wiwanitkit1 in our recently published paper “Novel Barium-Enhanced 3-Dimensional-Printed Spine Model for Pedicle Screw Training: A Cost-Effective Solution and Educational Validation.”2 Their letter raises several methodological considerations that merit discussion, and we welcome this opportunity to provide additional context and clarity regarding our work. Although our sample of 10 neurosurgical residents may seem modest, it represents all of our institutional trainees who are eligible for the study. Of our 20 total residents, 7 were excluded as junior trainees without sufficient observational experience, and 3 senior residents were excluded due to previous hands-on experience with pedicle screw placement. For educational interventions in surgical subspecialties, this sample size is comparable with or exceeds many similar studies in the literature.3 Furthermore, our study design yielded 120 screw placements (4 screws × 3 sessions × 10 residents), providing sufficient data points for meaningful statistical analysis. Regarding skill retention, we have now completed a 6-month follow-up assessment using identical methodology with all 10 residents. Statistical analysis showed significant reductions in surgical duration from 20.73 ± 4.53 min to 12.33 ± 3.55 min (Wald statistic = 50.45, df = 3, P < .0001) and fluoroscopy usage from 8 (range: 5-18) images to 5 (range: 4-8) images (Wald statistic = 11.64, df = 3, P = .0087). Visual assessment scores showed improvement trends from 0.12 (range: 0-3) to 0 (range: 0-2), although this did not reach statistical significance at the 6-month follow-up (Wald statistic = 3.91, df = 3, P = .272). Most residents maintained improved accuracy levels, whereas others showed significant skill deterioration, including 2 postgraduate year 3 residents (both over 40 years of age) with multiple pedicle breaches and 1 junior resident with a single malposition. Postassessment interviews revealed that deteriorating residents had not engaged with spinal anatomy materials during the interval and demonstrated difficulty explaining basic anatomical landmarks. This suggests that while our model improves motor skills, long-term retention requires continued engagement with foundational knowledge.4 Rather than undermining our model's value, these results highlight its potential as both a training tool and an assessment instrument for identifying residents who require additional mentorship. The clear performance differentiation at 6 months provides valuable data for program directors in resource allocation and individualized training planning. The authors correctly note that direct comparison with cadaveric specimens or commercial simulators would be valuable. However, our study was conducted with a total budget of less than $500, insufficient to purchase a single commercial simulator. This economic reality underscores the purpose of our study: to develop accessible training for resource-limited settings where commercial alternatives exceed $1000. Our statistical approach used the Brunner-Langer LD F1 model, which tracks individual performance trajectories longitudinally rather than comparing group means. This nonparametric method is designed for small-sample longitudinal studies and makes no distributional assumptions. The model's ability to demonstrate statistically significant improvements while handling conservative nonparametric testing supports the robustness of our findings. Regarding assessment reliability, while using 2 raters might seem limited, the exceptional inter-rater correlation (ICC = 0.978) demonstrates robust measurement consistency. If simplified to binary “screw in/out” classification, inter-rater disagreement would be virtually eliminated. The letter raises questions about tactile feedback simulation. Literature review reveals few studies examining bone feedback characteristics during pedicle instrumentation,5-7 and no commercial models provide validation data for their tactile properties. Moreover, bone density and tactile feedback vary significantly between patients. Thus, there is no “standard” bone feel to replicate.8 Our printing parameters were optimized through senior surgeon evaluation to replicate compact and cancellous bone characteristics. While perfect biological replication remains challenging for any synthetic model, our results demonstrate sufficient tactile fidelity for effective skill development. A formal cost-effectiveness analysis seems unnecessary given the 200-fold cost difference between our model ($1.61/session) and commercial alternatives ($200-$400/session). This work is part of our broader SpineForge initiative (https://spineforge.org), which has developed multiple open-source models for different spinal procedures, all designed for reproduction using standard FDM printers available globally. We welcome the authors' suggestions for future research directions. Force sensor integration represents a separate research endeavor that requires significant resources, cadaveric specimens, and specialized equipment, which falls well beyond the scope of our study's funding. Regarding expanding to various spinal levels, we have already addressed this through SpineForge, having developed models for multiple spinal regions,9,10 all freely available at “https://www.printables.com/@AkbulutBb_2153171/collections/2294005.” In conclusion, our study demonstrates clear educational value through rigorous longitudinal analysis, supported by 6-month follow-up data. The variable retention outcomes highlight the importance of regular practice in maintaining surgical skills, a universal challenge in surgical education. Rather than invalidating our approach, these findings emphasize the need for accessible training tools that enable frequent practice sessions. Rather than requiring expensive validation against commercial alternatives, our open-source approach enables global replication and improvement. While direct comparison would be valuable, such validation was impossible within our scope and would contradict our fundamental mission of democratizing surgical education. We welcome and will support any researchers interested in conducting comparative studies using our freely available models. We thank the editor for this opportunity to clarify our work and remain committed to democratizing surgical education through continued open-source innovation.