The current trend in kyphosis correction is for “every level” instrumentation to achieve intraoperative stability, correction, fusion and implant longevity.

We evaluate the medium term follow up of a low implant density (LID) construct.

All patients with adolescent kyphosis (idiopathic or Scheurmann's) on our deformity database were identified. Radiographs and records were analysed for neurological complications, correction and revision.

The constructs included were all pedicle screw anchors with multiple apical chevron osteotomies and a proximal and distal “box” of 6 to 8 screws. A four rod cantilever reduction manoeuvre with side to side connectors completed the construct. Kyphosis for any other cause was excluded. Follow up less than 12 months was excluded.

23 patients were identified with an average follow up 27 months (72 to 12 months) and a mean implant density of 1.1 (53.5% of “available” pedicles instrumented).

There was 1 false positive neurophysiological event without sequelae (4%).

There were no proximal junctional failures (0%).

There were no pseudarthroses or rod breakages (0%).

There was 1 loss of distal rod capture (early set screw failure) (4%). This was revised uneventfully.

There were 4 infections requiring debridement (early series).

Average initial correction was 44% (77.7 degrees to 43.5 degrees) with a 1% loss of correction at final follow up (43.5 to 44.0 degrees). The fulcrum bending correction index was 107% (based on fulcrum extension radiographs). 85% of curves had a fulcrum flexibility of less than 50%.

The average cost saving compared to “every level “instrumentation was £5700 per case.

This paper shows that a LID construct for kyphosis has technical outcomes as good as high density constructs. The obvious limitation of the study is the small number of patients in the cohort.

The infection rates have improved with changes to perioperative process in the later series of patients. We do not believe these are a consequence of the construct itself.