Purpose: To investigate the value of various clinical signs to identify occult hip fracture.

Methodology: MRI register was searched to identify all patients who had MRI scan between July 2000–June 2006 for suspected occult hip fracture. 64 patients were identified. 33 patients had occult proximal femoral fractures. 27 patients had no fracture.

Results: 19 patients with fracture were not independently mobile compared to 6 patients without fracture. 7 patients with fracture and 2 patients without fracture were able to weight bear. 13 patients with fracture and 10 patients without fracture had unrestricted straight leg raise ability. 7 patients with fracture and 16 patients without fracture had no pain on axial loading. Fisher exact and chi square analysis was performed; with Bonferroni correction for multiple comparisons (10 tests) p< 0.005 was deemed significant. Pain on axial loading of limb and pre-fracture patient mobility were associated with fracture (p< 0.005). Both factors had positive predictive value = 0.76 and post-test probability of disease given a negative test = .30. Predictive values remained the same when both factors were considered together.

Discussion: These data indicate that although patients who were independently mobile before the fall and who do not have pain on axial compression of the limb are less likely to have a fracture, these signs alone or in combination will not exclude a fracture. It is essential to perform MRI scanning of patients with severe hip pain but normal x rays after fall as it does not seem possible to clinically exclude fracture.