Failure of a Humerus Shaft Fixation

    When non-operative management for a fractured humerus shaft fails, an active 75-year-old farmer agrees to undergo surgery with plate and screw fixation. What goes wrong?


    William W. Cross III, MD


    The author has no disclosures relevant to this article.

    Case Presentation

    An active 75-year-old male presents with pain in his right shoulder and arm following a fall from a tractor at his farm.

    He is relatively healthy, with no history of tobacco product use or chronic illness. He has mild right-sided hemiparesis from a closed head injury 40 years ago.

    It is suspected that he has a humerus shaft fracture, which is confirmed with radiographs (Figure 1).

    Figure 1. Preoperative radiographs showing the humerus shaft fracture.

    Initial Treatment

    An initial trial of non-operative treatment is appropriate for most patients with this type of fracture, and that is what I recommended for this patient. This included a coaptation splint until post-injury day 8 and then transition to a functional brace.

    I typically re-evaluate a patient who is being treated non-operatively at 6 to 8 weeks and will be very critical of the imaging results. If I don’t see signs of healing and if the fracture remains mobile, I will counsel the patient about the potential for a non-union. I will often offer surgery at that point.

    Outcome of Initial Treatment

    At 10 weeks post-injury, this patient has some early evidence of healing with callous formation. However, physical examination reveals a completely mobile fracture.

    I counseled him extensively on the risks and benefits of surgical intervention and he chose to proceed with surgery.

    Surgical Procedure

    Preoperatively, I evaluated the patient’s endocrine status, as endocrine deficiencies – particularly vitamin D deficiency – can compromise healing in trauma patients. This patient’s levels were within normal limits.

    I did not evaluate him for an infection, as he had no previous surgical intervention.

    The surgical plan included the following:

    • Debride the non-union site to identify the fracture ends.
    • Use a microsagittal saw to flatten the bone ends, which optimizes surface contact (Figure 2).
    • Maximize compression with anatomic reduction in the coronal and sagittal axes, which is accomplished with both extrinsic compression (outside of the plate using clamps) and intrinsic compression (within the plate using off-axis screw placement) (Figures 3-4).
    • Add to the local biology of the non-union site using local autograft from the callous.

    Figure 2. In this intraoperative image, a modified, pointed bone reduction clamp is utilized to afford an anatomic reduction across the fracture site. The bone ends have been cut to maximize surface contact.

    Figure 3. The plate is placed and contoured by “pre-bending” the plate to ensure to gaping on the far cortex with maximal compression. The first screw is placed in neutral position and subsequent screws are placed into compression mode.

    Figure 4. This construct has 8 cortices of fixation proximal and distal to the fracture site. Anatomic reduction is confirmed.

    Radiographs were repeated in the immediate postoperative period. They demonstrated good alignment and compression (Figure 5).

    Figure 5. Immediate postoperative radiographs. The lateral view shows 3 of 4 distal screws are placed eccentrically which shows they were placed in compression mode.

    Outcome of Surgical Intervention

    The patient was allowed to do some home exercises after surgery, including range-of-motion exercises for his right elbow and shoulder. He continued to drive and do light work around his farm.

    At 12 days following surgery, he saw his primary care provider due to serosanguinous drainage from the incision. He was placed on cephalexin at that time.

    He returned to the orthopedic clinic for his first follow-up appointment 3 weeks postoperatively. New imaging was obtained (Figure 6). Fixation failure was obvious, and the reason for continued drainage was clear: The continued motion of the fracture site had mobilized hematoma and caused local soft tissue irritation.

    Figure 6. Radiographs from the first follow-up visit show fixation failure.

    The patient was counseled on the possible treatment options, including:

    • Staged management with hardware removal, antibiotic bead placement, and subsequent revision fixation with plates and screws
    • Staged management with hardware removal, antibiotic bead placement, and subsequent revision fixation with humeral medullary nail
    • Immediate irrigation and debridement (I&D) with revision fixation with plates/screws
    • Immediate I&D with revision fixation with a humeral medullary nail
    • Non-operative management with immobilization and a short course of antibiotics
    • I&D, hardware removal, external fixation until the wound healed, and delayed reconstruction

    The patient opted for non-operative management, which included reapplication of the functional brace and planned bracing for 8 to 12 weeks based on exam and radiographic evidence of healing. He completed a 14-day course of antibiotics with resolution of all drainage. Inflammatory markers returned to normal.

    Ultimately, the fracture healed and the patient returned to full function (Figure 7-10).

    Figures 7-9. With non-operative management, including a 14-day course of antibiotics, the fracture healed (top and middle) and the patient returned to full function (bottom).

    Figure 10. At final follow-up, radiographs show complete union. The patient remains asymptomatic and is active at his farm.

    Key Takeaway Points

    Critical evaluation of the original treatment raises concerns in a few areas and offers ways in which surgeons can better counsel their patients.

      • The working length of this construct was likely too short (Figure 11). Increasing the length of the plate decreases the force across the plate and the screws at the fracture site. Ideally, a plate with 10 to 14 holes should have been selected, instead of one with 8 holes. Not every hole would need to be filled; rather, screws would be strategically and evenly distributed across the construct.
      • Patients with hemiparesis may have stiffer joints and contractures and therapists should be cautioned to treat these fracture patients differently than they would patients without hemiparesis. In this case, I believe the treating team, including myself, did not understand the patient’s true preoperative limitation and range of motion and, thus, we placed no restrictions on motion therapy. Close attention must be paid to the patient’s recollection of pre-injury function, which can guide us in the postoperative care regimen.
      • Revision surgery must be individualized. In this particular case, after a careful and thoughtful discussion with the patient, he chose to proceed with non-operative management. He was simply not interested in pursuing another surgery and chose to see if the fracture would heal this time. I could not argue with his resistance to further intervention and so I supported his wishes.
      • Many patients are anxious to return to activities that they have been unable to do for quite some time. In this particular case, the patient had an immobile extremity for 2.5 months as he was waiting patiently for healing in his functional brace. Once the fracture stabilized, he felt immediately better and began to do more than he likely should have been doing.

    Figure 11. A plate with 10 to 14 holes might have been more appropriate than this 8-hole construct.

    Author Information

    William W. Cross III, MD, is an orthopaedic traumatologist and an Assistant Professor of Orthopaedics at Mayo Clinic, Rochester, Minnesota.