Background

Osteosynthesis using an intramedullary nail for humeral shaft fractures provides solid stability and good load sharing, while minimizing soft tissue damage and preserving the periosteal blood supply at the fracture site [1, 2], resulting in satisfactory postoperative bone union rates [3,4,5,6,7,8,9,10,11]. Contrarily, antegrade intramedullary nailing has been reported to cause more shoulder complications, such as shoulder pain and decreased range of shoulder motion or muscle strength, compared with plate fixation [5, 7, 9, 11, 4, 13,14,15, 27]. Although placing the proximal end of the nail to anchor in the zone of dense subchondral bone is of critical importance in countering varus displacing force [28, 29], this study suggests that the nail should be inserted into the humeral head to avoid protrusion above the level of the articular cartilage. In cases where proximal nail protrusion remains at the time of bone union, nail removal has been reported to improve symptoms [7, 18] and can be a treatment option in such cases.

This study also showed that older age was significantly associated with poor mid-term shoulder functional outcomes in the univariate analysis. A previous study reported that all patients with unsatisfactory shoulder functional outcomes within two years after intramedullary nail fixation were older than 78 years [17], which concurs with the results of the present study. In addition, another study, including patients who underwent conservative therapy, plate fixation, or intramedullary nailing demonstrated an association between older age and poor shoulder function at 26 weeks and 52 weeks following injury [30]. However, multivariate analysis showed no significant association between older age and poor outcomes, possibly because of the confounding effect between nail protrusion and older age due to the effect of rotator cuff degeneration on the healing of the rotator cuff damaged during nail insertion and the osteoporotic loss of cortical bone in older age [18].

Unlike in a previous report [4], preoperative radial nerve injury was not significantly associated with clinical failure in this study. This disparity was attributed to the differences in the postoperative follow-up time. In this study, all patients with traumatic radial nerve injury improved over time, with only mild sensory deficits remaining at more than five years postoperatively, suggesting that preoperative radial nerve injury may affect the short-term shoulder functional outcomes and not the mid-term outcomes.

In addition, our results are different from those of previous reports showing an association between nonunion at six months after surgery and poor postoperative functional outcomes [13, 16]. The differences in the postoperative follow-up time can partly explain this discrepancy. All patients in this study who did not achieve union six months after surgery eventually achieved union within two years after surgery. This raises the possibility that the final bone union had little effect on the mid-term outcomes.

The strength of this study is that it evaluated mid-term shoulder functional outcomes following osteosynthesis for humeral shaft fractures. Previous studies on the postoperative outcomes of humeral shaft fractures have evaluated shoulder function at 1–3 years postoperatively [1, 2, 12]; till date, no study has assessed shoulder functional outcomes at more than five years postoperatively.

However, this study had some limitations. First, the cohort of patients available for analysis was small; therefore, our results may have included the effect of β-error. However, the sample size of most reported studies has been 20–40 patients [1, 2, 12], and our sample size is comparable to these studies. Second, due to the study’s observational nature, biases from unobserved differences may have affected the outcomes. For instance, although six surgeons performed the operations in this study, their skill levels were not taken into consideration. Moreover, the fact that the choice of implants depends on the surgeon’s preference could be a limitation of this study. Third, 16 patients were excluded owing to loss to follow-up, which may decrease the generalizability of the study results.

Conclusion

This study provides new information on predictive factors affecting mid-term outcomes following osteosynthesis using an antegrade nail. Our findings demonstrated that proximal protrusion of the nail was significantly associated with poor mid-term shoulder functional outcomes. Therefore, particularly in older adults, it is essential to place the proximal end of the intramedullary nail below the level of the articular cartilage.