Survivorship analysis of pedicle spinal instrumentation.

Between 1985 and 1989, the senior author performed 120 consecutive pedicle instrumentation cases--78 Steffee (VSP) procedures and 42 procedures using Cotrel-Dubousset instrumentation. Posterolateral or posterior fusions using autogenous iliac bone graft were performed across the instrumented vertebrae in all cases. Survivorship analysis was used to calculate a predicted cumulative rate of success for this series of patients over 10 years postoperative follow-up. The criteria of failure of pedicular instrumentation or "death" of an implant were defined as 1) screw bending, 2) screw breakage, 3) infection, 4) loosening of implants, 5) any rod or plate hardware problems, or 6) removal of hardware due to a neurologic complication. Out of 526 pedicle screws (175 Cotrel-Dubousset screws, 351 VSP screws) there were 22 problem screws (22/526 = 4.18%). Six screws had bent, none were infected, 16 screws had broken, and none were loose. The 22 problem screw events occurred in 12 patients. In seven patients, the instrumentation failure was an incidental radiographic finding, in that patients had a solid posterolateral fusion. The remaining five patients had screw breakage in association with a pseudarthrosis. Life table calculations predicted the survivorship of instrumentation without complications would be 80% at 10 years postoperative follow-up. Actuarial analysis predicted the survivorship of solid posterolateral fusion at 90% at 10 years follow-up. This survivorship rate is similar to those predicted at 10 years follow-up for other more widely used orthopedic surgical implants such as total hip arthroplasty components.