Computer-assisted planning vs. conventional surgery for the correction of symptomatic mid-shaft clavicular nonunion...

The aim of this study was to compare the clinical and radiographic outcomes of treatment of symptomatic mal- and/or nonunion of midshaft clavicle fractures using radiographically based free-hand open reduction and internal fixation (ORIF) or computer-assisted 3D-planned, personalized corrective osteotomies performed using patient-specific instrumentation (PSI) and ORIF. The hypotheses were that (1) patients treated with computer-assisted planning and PSI would have a better clinical outcome, and (2) computer-assisted surgical planning would achieve a more accurate restoration of anatomy compared to the free-hand technique.

Between 1998 and 2020, 13 patients underwent PSI, and 34 patients underwent free-hand ORIF and/or corrective osteotomy. After application of exclusion criteria, 12/13 and 11/34 patients were included in the study. The clinical examination included measurement of the active range of motion and assessment of the absolute and relative Constant-Murley Scores and the subjective shoulder value. Subjective satisfaction with the cosmetic result was assessed on a Likert scale from 0 to 100 (subjective aesthetic value). 11/13 and 6/11 patients underwent postoperative computed tomography evaluation of both clavicles. Computed tomography scans were segmented to generate 3D surface models. After projection onto the mirrored contralateral side, displacement analysis was performed. Finally, bony union was documented. The average follow-up time was 43 months in the PSI and 50 months in the free-hand cohort.

The clinical outcomes of both groups did not differ significantly. Median subjective shoulder value was 97.5% (70; 100) in the PSI group vs. 90% (0; 100) in the free-hand group; subjective aesthetic value was 86.4% (±10.7) vs. 75% (±18.7); aCS was 82.3 (±10.3) points vs. 74.9 (±26) points; and rCS was 86.7 (±11.3) points vs. 81.9 (±28.1) points. In the free-hand group, 2/11 patients had a postoperative neurological complication. In the PSI cohort, the 3D angle deviation was significantly smaller (PSI/planned vs. free-hand/contralateral: 10.8° (3.1; 23.8) vs. 17.4° (11.6; 42.4); P = .020)). There was also a trend toward a smaller 3D shift, which was not statistically significant (PSI/planned vs. free-hand/contralateral: 6 mm (3.4; 18.3) vs. 9.3 mm (5.1; 18.1); P = .342). There were no other significant differences. A bony union was achieved in all cases.

Surgical treatment of nonunion and malunions of the clavicle was associated with very good clinical results and a 100% union rate. This study, albeit in a relatively small cohort with a follow-up of 4 years, could not document any clinically relevant advantage of 3D planning and personalized operative templating over conventional radiographic planning and free-hand surgical fixation performed by experienced surgeons.