RESUMEN
BACKGROUND: Virtual surgical planning (VSP) is increasingly used in maxillomandibular osseous free flap reconstruction. Non-commercial ('in-house') VSP may offer the same level of accuracy and other benefits, without the inflated costs and time delays inherent in using commercial providers. Comparisons between commercial and in-house methods are lacking. This study aims to determine the accuracy of VSP, compare in-house and commercially planned cases, and explore predictors of the reconstruction error. METHODS: Seventy-six patients who had a virtually planned maxillomandibular reconstruction between January 2012 and July 2020 were retrospectively identified. The preoperative digital plan was compared to the postoperative CT scan in terms of length of bone segments, angle between adjacent segments and intercondylar, and intergonial angle distances (mandibular reconstructions only). RESULTS: Forty-four patients fulfilled the inclusion criteria. The mean intergonial and intercondylar distances error was 1.7 ± 1.01 mm, mean segment length error was 1.3 ± 1.40 mm, and mean angles error was 1.9 ± 2.32°. The difference in error of in-house VSP compared to commercial VSP was not statistically significant for intercondylar and intergonial distance (p = 0.76), segment length (p = 0.15), or angle between segments (p = 0.92). The increased error was associated with osteoradionecrosis as the indication for surgery, greater number of segments, and secondary reconstructions. CONCLUSION: VSP is an accurate method of maxillary and mandibular reconstruction. In-house VSP may be similar in accuracy to commercial VSP options. Higher levels of inaccuracy are likely to occur in more complex reconstructions, particularly secondary reconstructions, and in the setting of osteoradionecrosis.