Removal of broken intramedullary nails can be extremely challenging. Implant failure following intramedullary nailing has its origin in repetitive sub-threshold loading of the construct leading to fatigue failure. During the period of secondary bone healing by callus formation, the intramedullary nail is exposed to significant deforming forces and care should be taken to select an implant most likely to avoid failure before the fracture has had time to heal. Newer closed section nails are believed to be able to withstand more load cycles before failure than older generation slotted nails. Similarly, larger diameter nails are less likely to break than smaller diameter nails. When a nail breaks it indicates that there is still movement present at the fracture site. Therefore a broken nail should be considered as a sign of non-union and fracture instability and removal of the implant is usually indicated to facilitate further treatment.

Some surgeons regard broken locking bolts as a method for “auto-dynamisation” of the fracture site and beneficial for healing. In these instances there might be no urgency for the broken bolts to be removed. Dynamisation is only of benefit in simple fracture configurations where axial loading will achieve compression between healing fracture fragments. In other scenarios removal of the implants and revision fixation is usually indicated. Prior to revision surgery, the surgeon should discuss with the patient whether it is necessary or desirable to remove all fragments of broken locking bolts. Sometimes it may be simpler and safer to leave asymptomatic fragments of broken locking bolts in situ, instead of pursuing troublesome options for their removal. Once the fracture has united, broken locking bolts are usually not an independent source of pain.

Deep infection following nailing of closed fractures of the tibia is rare with an incidence around 1 % but infection rates are higher in open fractures, up to 17 % [13, 14]. Infection rates are similar or lower in femur fractures [15]. Deep infection usually mandates removal of the intramedullary nail at some stage, but it is often beneficial to delay until fracture healing has been achieved [16]. However, mean time to union following intramedullary nailing complicated by infection is significantly increased and occasionally revision surgery is required prior to fracture healing. Final and complete eradication of the infection will require removal of the implant and also frequently intramedullary reaming to help eradicate the infected material.

As soon as a fracture is diagnosed as a non-union, consideration should be given to removal of the intramedullary nail before on-going repetitive loading of the nail leads to implant breakage. Unfortunately predicting how long an intramedullary nail will endure in an un-united fracture before it breaks is dependent on too many variables to be predictable with any accuracy. However, removal of broken implants is often so much more complicated than removal of intact implants, that it is important to consider early surgery to avoid this scenario.

Limited studies have not shown these factors to be important for intramedullary nails [17, 18]. The concerns regarding neoplasia from arthroplasty practice do not seem to be applicable to intramedullary nails [19, 20]. Therefore patient and surgeons concerns about these factors are not likely to be indications for routine removal of intramedullary nails.

Removal of intramedullary nails many years later can be extremely difficult, additional surgery around the joint at the time of arthroplasty can increase the risk of infection, and complications from nail removal can lead to more complex arthroplasty procedures. These factors have led many surgeons to undertake routine removal of all intramedullary nails that might interfere with future arthroplasty. Unfortunately there is no published evidence to indicate whether this practice is of overall benefit or not.

Diaphyseal fracture and intramedullary nailing are not independent risk factors for future arthroplasty, but traumatic joint damage and malunion are. Therefore it would seem sensible at least to remove routinely intramedullary nails in patients with these risk factors for future arthroplasty. Similarly patients with inflammatory arthropathies should also have intramedullary nails removed routinely because of the increased risk of future arthroplasty in these conditions.

Unlike plate fixation, stress shielding has not been shown to be a problem after intramedullary nail fixation and bone is able to remodel well despite the presence of an intramedullary implant [21].

While litigation is not an indication for intramedullary nail removal, it has been shown to be associated with a higher incidence of nail removal [22, 23].

In addition to many of the factors considered for adults patients, growing children are at risk of problems from bone overgrowth around intramedullary implants. The risk of bone overgrowth has not been quantified and the optimal timing for nail removal has not been determined but it is generally accepted that the longer the period of growth with the nail in situ, the greater the risk of bone overgrowth. In addition to overgrowth at the insertion site, incarceration of the length of the nail in thick cortical bone has been described (Fig. 9.2a, b).

Reported complications associated with removal of intramedullary nails have included delayed wound healing, wound infection, further soft tissue damage, re-fracture, new pain syndromes, failure to relieve existing pain, and failure to remove the implant.

Complications vary according to anatomical site and it is useful to consider each bone separately.

Removal of an intramedullary nail is usually a routine procedure when the manufacturer and design of the implant is known, the manufacturer’s specific extraction equipment is available, and the implant is not broken. Unfortunately it is fairly common for the implant to be unknown or for the specific extraction equipment to be unavailable. Broken intramedullary nails occur occasionally and broken locking bolts are a common problem during nail removal. When any one of these scenarios occurs, special equipment and techniques may be required.

Universal nail extraction sets, broken nail removal instruments, and broken screw removal sets are commercially available to help overcome these problems. If the nail is unknown the specific extraction kit not available, or there is broken metalwork it is strongly advised not to proceed without universal specialist equipment being available.

The sequence of implant removal is important. The end of the nail for attachment of the extraction jig must be identified first. The presence of an end cap can be difficult to identify on radiographs and for this reason some surgeons do not recommend the use of end caps unless they are being used to extend the working length of the nail [27]. To maintain rotational stability while the extraction jig is attached to the nail, the final locking bolts should only be removed once the jig has been screwed into place. It must be remembered to remove all locking bolts prior to attempted nail extraction with the jig. Therefore anteroposterior and lateral radiographic views of the whole healed bone are indispensable before nail extraction. Attempts to remove nails with locking bolts in situ have resulted in serious iatrogenic fractures.

When the make of nail is unknown, adequate exposure to visualize the end of the nail is required. This enables an extraction jig with matching size and thread to be selected from the range of options on the universal extraction set. When a perfect match is not found, selecting a conical extraction device can be of benefit to achieve an interference fit. If bone ingrowth onto the nail or into the locking holes is suspected, it is recommended to use a mallet to drive the nail in further by a few millimetres to break the ingrowth before attempting an extraction manoeuvre.

Experienced authors have summarized a multitude of techniques to address broken implants in intact or healed long bones [28]. Removing the broken nail fragment using an open technique via the fracture site may be considered if the fracture site is to be opened for debridement and direct reduction during revision fixation. The following paragraphs describe techniques that do not expose the fracture site.