Intramedullary screw fixation for simple displaced olecranon fractures

While tension-band wiring and ORIF with plate fixation are both effective treatments for olecranon fractures, we propose that fixation of simple olecranon fractures using an intramedullary screw is also a reasonable option based on the results of this series. Previous studies have suggested that fixation using intramedullary cannulated screws is technically challenging and unreliable [11, 15‐18], but we found that this technique lead to acceptable surgical results, good patient-reported outcomes, and low rates of implant removal with reasonable operative times (average 45 min).

When surgical treatment is indicated, several treatment options, such as tension-band wiring, plate fixation, intramedullary screw fixation and intramedullary nail fixation have been described [1‐5, 7‐11]. These techniques differ in their method of fixation and have their own advantages and disadvantages, but a recent Cochrane review found that none was definitively preferable based on available evidence [2]. Tension-band wiring has shown excellent results with regard to union rates [2, 3, 9] and remains the first choice of treatment for many due to the low complexity and costs of the [8, 9]. Yet, implant removal is often necessary due to hardware irritation or k-wire migration in up to 82% of patients [3, 7]. Another technique often used is plate fixation [5, 8, 9, 25]. This technique is especially indicated in comminuted fractures and with osteoporotic bone, but can also be used for simple two-part fractures [5, 8]. In a recent prospective trial comparing plate fixation and tension-band wiring, there was no difference in union, DASH, function or costs, and plate fixation was associated with a decreased rate of implant removal [3]. However, this was offset by the increased number of infections and revisions in the plate fixation group [3]. Since neither of these options is optimal for all settings, we propose that intramedullary screw fixation should also be considered for simple olecranon fractures based on our results.

Many reasons exist for differences between our study and existing literature [11, 15‐17]. The prior studies are small and often include a wide variety of fractures treated with intramedullary screws, and frequently in settings where plate fixation would have been more appropriate given our greater understanding of proximal ulna fracture morphology [19‐21]. For example, in the series published by Helm et al., 79% of treated fractures were comminuted [17]. In the present day, with more accurate radiologic diagnostics including improved X-rays and CT scans, identifying appropriate cases (e.g., simple olecranon fractures) is easier and likely to lead to better outcomes [22]. Moreover, using intramedullary screw fixation for simple Mayo type 1 and 2A fractures is somewhat intuitive since these fracture types are most similar to olecranon osteotomies which are known to heal well with screw fixation. And in many cases, use of a screw instead of tension-band wiring also led to decreased rates of implant removal [18].

Except for one elderly frail patient with an asymptomatic non-union, all patients healed their fracture. Mechanically, the partially threaded intramedullary screw works as a lag-screw. When inserted across the fracture, the threads of the screw tip engage in the cancellous bone of the metaphyseal area, causing compression of the fracture fragments upon tightening. In osteoporotic bone the screw may not have enough purchase for a stable fixation and the proximal part may be pulled away by the force of the triceps muscle. In our series, both the malunion and the non-union patient were geriatric patients with osteoporotic bone (Table 2). Based on these results, this fixation method may not be suitable for this patient group although this requires further study.

A prior study demonstrated that the dominant factor driving re-operation for isolated olecranon fractures was the type of fixation. In that study, tension-band wiring had over twice the rate of implant removal as ORIF using plates (46.5% vs. 18.7%, respectively), and rates using intramedullary implants were not studied [12]. Based on our results, intramedullary screw fixation has rates of removal that are in-between these two fixation methods. Other, earlier studies [3‐5] have shown higher removal rates for tension-band wiring and plate fixation up to 68–82%. In addition, we did not observe any superficial or deep infections and this may have been due to the fact that, compared to tension-band wiring or plate fixation, a smaller incision could be used with minimal soft-tissue dissection or trauma. However, further studies with larger number of patients comparing between fixation options are needed to better answer these questions.

Finally, during the last 20 years several intramedullary nails have been described [10, 26, 27]. The nails have the same advantage as the intramedullary screw as they are low profile and have almost no protruding hardware. However, intramedullary nails are more complex and often designed for use with more complex fractures. For simple Mayo IA or 2A fractures, these extra design features are likely not necessary and simple screw fixation is likely to be sufficient and probably less expensive. We did not observe complications related to the insertion of a straight rigid intramedullary screw into a bone with a complex curved morphology. We believe this is because the varus curvature of the proximal ulna begins at approximately 8.2 cm from the olecranon tip and the anterior angulation starts at approximately 8.6 cm, while the inserted screws were only between 9 and 10 cm [28]. These anatomic characteristics highlight the importance of the central–central insertion point in facilitating the optimal direction of the screw [23].