Image guidance is associated with lower rates of pseudoarthrosis diagnosis after long-segment posterior spinal instrumentation: a propensity-matched cohort study

The purpose of this study was to assess the association between intra-operative navigation and short- and long-term outcomes following long-segment posterior spinal instrumentation.

This was a retrospective propensity-matched cohort study of adults undergoing long-segment posterior spinal instrumentation utilizing a multi-institutional electronic health record database. Inclusion criteria were patients undergoing thoracolumbar posterior spinal instrumentation greater than seven levels with and without intra-operative navigation. Exclusion criteria were < 18 years old, spinal neoplasms, trauma, inflammatory, or infectious bone conditions. The index surgery date range was restricted a priori to 2015 through 2020 to reduce era-related confounding in navigation exposure and outcome ascertainment and to allow adequate observation time for each outcome window. Cohorts were matched 1:1 on 40 covariates, with balance assessed by standardized mean differences. Categorical outcomes were compared with Chi-squared tests and time-to-event outcomes with Kaplan-Meier estimation, log-rank testing, and Cox proportional-hazards models. Because multiple outcomes were tested within each time window, the Holm-Bonferroni procedure was applied to control the family-wise error rate.

Propensity score matching resulted in 1,646 patients in each cohort. At thirty days, the navigated cohort had a higher rate of blood transfusion (12.4% versus 9.2%; OR 1.40, 95% CI 1.12-1.75; P = 0.003), which remained significant after Holm-Bonferroni correction. Higher observed rates of acute kidney injury and myocardial infarction in the navigated cohort were nominal only and did not survive correction, and there was no significant difference in deep venous thrombosis or sepsis. At one, two, and five years, patients in the non-navigated cohort demonstrated a significantly higher risk of pseudoarthrosis diagnosis than patients in the navigated cohort (one year OR 1.42, 95% CI 1.10-1.82, P = 0.006; two years OR 1.44, 95% CI 1.15-1.81, P = 0.001; five years OR 1.38, 95% CI 1.13-1.69, P = 0.002), each surviving Holm-Bonferroni correction. However, there were similar rates of reoperation at five-years between the two cohorts (OR 0.94, 95% CI 0.75-1.18; P = 0.465) with similar rates of adjacent segment disease (OR 0.98, 95% CI 0.77-1.24; P = 0.820).

Image guidance was associated with a lower rate of pseudoarthrosis diagnosis through five years following long-segment posterior spinal instrumentation. This was not accompanied by a reduction in reoperation, and short-term medical outcomes were largely similar with mixed differences in selected complications.