The most common mechanism of injury in reported cases is a fall onto the back or buttocks, although more indolent precipitating injuries are described. Direct trauma to the pelvis or hyperextension of the hip may result in muscular tears leading to iliacus haematoma. Two separate syndromes of lumboscacral plexus compression are recognised.²¹ The entire plexus can be compressed within the psoas muscle resulting in weakness of those muscles supplied by the obturator and femoral nerves. Alternatively, the femoral nerve is compressed as it passes through the closed fibrous compartment formed by the iliac fascia and the ileum. As haematoma expands within this compartment, the femoral nerve becomes compromised.²⁰ Goodfellow et al²⁰ injected fluid into this compartment, successfully demonstrating subsequent compression of the femoral nerve. They also showed that maximal iliac compartment volume is associated with flexion and external rotation of the hip. These deformities have been reported in cases in the literature but are frequently absent even in patients with marked neuropathy.

Although infrequent in children and young adults, the possibility of intra-tumoural haemorrhage should be considered in patients presenting with spontaneous haematomas. The most frequent underlying diagnosis is malignant fibrous histiocytoma with cases being described in the iliopsoas muscle.^{22 23} Patients with spontaneous haematomas that do not follow the expected course of resolution should be treated with particular suspicion. It is good practice to monitor patients with spontaneous haematomas to ensure complete resolution to confirm the absence of the underlying tumour.

In the present case, the cause of the haematoma is unclear. The patient recalled a fall onto her buttocks 5 days prior to the onset of any neurological symptoms. However, the progression of neurological symptoms was rapid occurring 1 h after onset of groin pain precipitated by running and 5 days following the initial fall. It is feasible that running may have enlarged a haematoma initially caused by a fall. The use of an aspirin-based tablet may have exacerbated the problem. We plan to monitor the patient until radiographical resolution of the haematoma to exclude an underlying tumour, although no other clinical features suggestive of malignancy were evident.

Haematomas have been successfully managed with both non-operative treatment and surgical evacuation of the clot. The majority of reported cases of iliacus haematoma producing femoral nerve palsy have been managed operatively. Non-operative treatment has been recommended in patients on anticoagulation or with bleeding disorders with surgical decompression of the haematoma advocated in cases of iliacus haematoma resulting from trauma.^{10 12 20} Kumar et al¹² recommended evacuation of the iliacus haematoma before the onset of femoral nerve palsy to ensure early recovery and stop the potential pressure effect of the haematoma on the lumbar nerve roots. Takami et al¹⁰ stated that operative decompression is indicated in cases of large iliacus haematomas, although parameters defining haematoma size have not been reported. Surgical exploration and haematoma evacuation carry a risk of immediate neurovascular injury, further bleeding once the clot is removed and infection. The major surgery required to expose these deep structures has a significant impact on rehabilitation and return to sport. Decompression should be considered in the presence of progressing signs and symptoms. If non-operative management is chosen, serial neurological examinations and haematological parameters reflecting ongoing blood loss must remain stable.¹⁷

In the case presented, non-operative treatment was considered along with radiologically guided minimally invasive drainage and open surgical drainage. Non-operative treatment was elected in our case as the patient demonstrated early signs of improvement. Percutaneous drainage was not undertaken because, although the majority of the haematoma was deep to iliacus, there was evidence of inferior extension of haematoma/free-fluid in contact with the neurovascular bundle that may have not been amenable to percutaneous draining. In addition, due to the presumed viscosity of the clot, it was questionable whether adequate decompression could be achieved percutaneously. In the presence of progressing neurological signs, we would have proceeded with open surgical decompression. There was a dramatic reduction in the size of the haematoma at 6 weeks with complete resolution of the femoral nerve paralysis, although variable lengths of recovery have been reported in the literature.