The patient was taken up for an arthroscopic retrieval of the bullet with a preparedness of an arthrotomy in case of failure. The patient was positioned supine on the fracture table with a standard mid-line perineal post. In spite of the thin built of the patient, distraction at the hip joint was difficult and required heavier traction. The standard antero-lateral portal and the lateral (supra trochanteric and anterior partrochanteric) portals were used. A 30° scope was used. In the absence of the arthroscopic pump, the effect of gravity was used in distending the hip joint by raising the height of the irrigation fluid. The visualization of the bullet was initially not easily possible due to chronicity having resulted in the growth of organized scar tissue around the projectile. Debridement of the hyper-trophic synovium was done with the shaver, resulting in some bleeding inside the joint, making the arthroscopic picture blurred. The scope was placed adjacent to the bullet with the help of the image intensifier [Figure 2]. Visualization was possible by lowering the blood pressure of the patient, and raising the fluid pressure inside the hip joint by enhancing the height of the fluid bottles and lowering the operation table. The surrounding area was cleared with the shaver blade and thermal ablator leading to visualization of the bullet, which was lying firmly impacted in the supero-lateral acetabular wall.

It was not possible to move the bullet with the probe or the canula of the arthroscopic trocar. A long thin osteotome from the revision hip arthroplasty set was used to disimpact the bullet [Figure 3]. The disimpaction of the bullet left a well circumscribed crater in the acetabular wall [Figure 4]. The bullet was manipulated with the probe to a convenient location in the hip [Figure 5], where grasping was possible with a loose body grasper.

Arthroscopic examination of periphery of the joint was made with 30° and 70° scopes to rule out any other loose fragments. No major defect in the femoral head was noticed except for a small abrasion in the supero-posterior area. The total duration of the procedure was 70 minutes. The three portals used in the procedure were sutured [Figure 6]. The removed bullet was roughly spherical shaped with approximate one cm diameter [Figure 7].

During the procedure, the distraction of the hip joint required more traction than we normally used in hip arthroscopy. This may be due to the formation of scar tissue around the projectile resulting from a delay of four months. However, the patient being of a thin built with less muscles, we were able to overcome this, although it resulted in temporary neuropraxia in the perineal region due to pressure of the perineal post.

We used a thin long curved osteotome for disimpacting the projectile from the acetabulum. Before introducing the osteotome, we had already unsuccessfully tried moving the bullet with a probe and a blunt arthroscopic trocar repeatedly. Before abandoning the procedure and changing over to open arthrotomy, the attempt of using the osteotome proved fertile. However, we were very cautious during the introduction of the osteotome and used fluoroscopic navigation to avoid slipping of the osteotome. After the disimpaction, we used a probe hook to manipulate the disimpacted bullet towards the convenient location in the joint where the grasper could be easily introduced. The thin built of the patient further facilitated the reach of our instruments into the hip joint.

The compelling reason of surgical extraction of an intraarticular bullet in the present case was excruciating pain suffered by the patient and his inability to lead a normal life. The location of the bullet in the hip joint resulting in catching of the bullet in certain positions of the thigh had made the patient's life miserable. Even otherwise, the extraction of the bullet from the hip joint has been indicated to avoid the long term risk of mechanical arthritis, and systemic lead toxicity caused due to dissolution and absorption of the lead through the synovial fluid.^4,8–10 The risk of being a potential nidus of infection, although mentioned in literature,^4,11 was probably negligible in our case at four months post-injury.

We have not used a pre-operative CT scan in the present case. However, the two standard antero-posterior and lateral views of the preoperative X rays and the visualization of the joint through a complete arch of rotation of C arm, per operatively, helped us to rule out the other intraarticular pathologies even in the medial articular space where the arthroscopic visualization is difficult.³ Although we agree that the availability of the facility of CT scan pre-operatively is an additional advantage, for more accurate planning and better pre-operative documentation, absence of the facility should not be a limiting factor for undertaking the surgery.

With successful arthroscopic retrieval, we could avoid several disadvantages of an open procedure including significantly diminished blood loss, cosmetic incisions, decreased risk of osteonecrosis of the femoral head and shortened recovery time.^6,7 All the potential complications of the hip arthroscopy were discussed with the patient before undertaking the procedure. There is a report of cardiac arrest resulting from extravasation of the irrigation fluid through the fracture line into the abdomen during the arthroscopic removal of a loose body from the hip joint of a patient with an acetabular fracture.¹² A theoretic risk of developing the same complication by introduction of the fluid through the bullet tract has also been expressed.⁴ In the present case, however, having no fracture line, and being undertaken at a delay of four months, resulting in complete healing of the bullet tract, this risk was negligible. Other risks to be kept in mind include perineal numbness/necrosis related to the traction post, radiation exposure and injury to the femoral neurovascular bundle anteriorly and sciatic nerve posteriorly.