A high rate of failure of the internal fixation of unstable spinal fracture
s in complete cord injured patients was noted in patients referred to the S
alisbury Spinal Centre who had been stabilised with a Hartshill rectangle.
This prompted a review of the operative notes, radiographs and clinical out
comes of all patients referred to the centre with a Hartshill rectangle in
situ. All patients identified with a complete spinal cord injury and Hartsh
ill rectangle were identified. Forty-three such patients referred from 13 d
ifferent centres were found. Pre- and postoperative radiographs were assess
ed for Fracture pattern and for spinal correction. Operative outcome in ter
ms of pain and complications relating to surgery were identified. The most
recent radiographs were assessed for signs of loss of reduction or stabilis
ation. Follow-up averaged 84 months (range 36-132 months). Or the 43 identi
fied patients, 19 were found to have unsatisfactory stabilisation. Persiste
nt pain, broken implants and worsening kyphosis were the main complications
. The failure to use bone graft at the time of stabilisation was significan
tly (P < 0.001) related to risk of failure. The application and use of the
Hartshill is not a technically challenging procedure: however, if the syste
m is to be used, it must be used correctly. Failure to correctly apply the
rectangle and to use bone graft will lead to an unacceptably high rate of f
ailure.