Percutaneous cannulated screw placement (PCSP) is a common method of fixation. In pelvic trauma neurovascular structures are in close proximity to the screw path. Pre-operative planning is needed to prevent injury. This study aims to the safety margin and accuracy of screw placement with computer navigation (CAS).

A control had no pathology in the pelvis but CT scans were performed for suspected trauma. The treated group had pelvic and acetabular fractures and were treated with CAS PCSP at our institution. Using a new technique involving CT 3D modelling of the whole (3D) safe corridor, the dimensions of the Posterior elements (PE) of the pelvic ring and the anterior column of the acetabulum (AC) were measured in the control group.

The accuracy of screw placement (deviation between the actual screw and planned screw) was measured in treated patient using a screenshot method and post-operative CTs. There were 22 control patients and 30 treated patients (40 screws).

The mean ± (standard deviation, SD) minimum measurement of the safe corridor at the PE was 15.6 ± 2.3 mm (range 11.6 mm to 20.2 mm) and at the AC was 5.9 ±1.6 mm (range 3.0 mm to 10.0 mm). The mean ± (SD) accuracy of screw placement was 6.1 ± 5.3 mm and ranged from a displacement of 1.3 mm to 16.1 mm. There was a notable correlation between Body Mass Index, duration of surgery and inaccuracy of screw placement in some patients. The largest inaccuracy of screw placement was due to reduction of the fracture during screw insertion, causing movement of the bone fragments relative to the array and therefore also the computerised screw plan.

There were no screw breakages, non-unions, neurological or vascular complications.

CAS PCSP is a safe and accurate technique. However, the safe corridor is variable and often very narrow. We recommend that the dimensions of the safe corridor be assessed pre-operatively in every patient using 3D modelling to determine the number and size of screw that can be safely placed.