Hemoglobin Levels Pre- and Posttreatment as a Surrogate for Disease Severity in Early-Onset Scoliosis.

Study DesignProspective cohort.ObjectiveTo compare preoperative hemoglobin levels to postoperative hemoglobin levels in patients with early-onset scoliosis (EOS).Summary of Background DataElevated hemoglobin (Hgb) may be a marker for preoperative hypoxia in patients with EOS and thoracic insufficiency syndrome (TIS). The changes in Hgb level after treatment may be a surrogate marker for improved oxygenation.MethodsBecause normal levels of Hgb vary with patient age, Hgb z scores were calculated by dividing age-adjusted mean Hgb levels by the age-adjusted standard deviation. Elevated Hgb was defined by a hemoglobin z score >1. Patients with a baseline Hgb value measured before initial implantation with at least one follow-up measurement, at 6, 12, or 18 months, were included in longitudinal analysis. Change in Hgb z score as well as change in curve magnitude over time was assessed using piecewise linear mixed modeling for patients with elevated Hgb and those without.ResultsTwo hundred sixty-seven patients with EOS were treated surgically over the study period. Average age at initial implantation was 6.8 years. Forty-eight (18%; 95% confidence interval = 13.7%, 23.2%) subjects had an elevated Hgb (z score > 1) level before implantation procedure. Hgb levels decreased in subjects with elevated Hgb from implantation to 6 months (p < .001) with no change in Hgb from 6 to 12 months (p = .46) or from 12 to 18 months (p = .59), but an overall decrease from preoperative to 18 months (p < .001). There was no change in Hgb levels for subjects without elevated Hgb from implantation to 6 months (p = .94), from 6 to 12 months (p = .61), or from 12 to 18 months (p = .78).ConclusionsIn some patients with EOS and TIS, there appears to be significant positive impact on oxygenation from distraction instrumentation as evidenced by a meaningful proxy measurement: improvement in abnormal preoperative Hgb levels after surgery.Level of EvidenceIII.