Risk factors for gastrointestinal complications after spinal fusion in children with cerebral palsy.

DESIGN Prospective cerebral palsy (CP) registry review. OBJECTIVES (1) Evaluate the incidence/risk factors of gastrointestinal (GI) complications in CP patients after spinal fusion (SF); and (2) investigate the validity of the modified Clavien-Dindo-Sink classification. BACKGROUND Perioperative GI complications result in increased length of stay (LOS) and patient morbidity/mortality. However, none have analyzed the outcomes of GI complications using an objective classification system. METHODS A prospective/multicenter CP database identified 425 children (mean, 14.4 ± 2.9 years; range, 7.9-21 years) who underwent SF. GI complications were categorized using the modified Clavien-Dindo-Sink classification. Grades I-II were minor complications and grades III-V major. Patients with and without GI complications were compared. RESULTS 87 GI complications developed in 69 patients (16.2%): 39 minor (57%) and 30 major (43%). Most common were pancreatitis (n = 45) and ileus (n = 22). Patients with preoperative G-tubes had 2.2 × odds of developing a GI complication compared to oral-only feeders (OR 2.2; 95% CI 0.98-4.78; p = 0.006). Similarly, combined G-tube/oral feeders had 6.7 × odds compared to oral-only (OR 6.7; 95% CI 3.10-14.66; p < 0.001). The likelihood of developing a GI complication was 3.4 × with normalized estimated blood loss (nEBL) ≥ 3 ml/kg/level fused (OR 3.41; 95% CI 1.95-5.95; p < 0.001). Patients with GI complications had more fundoplications (29% vs. 17%; p = 0.03) and longer G-tube fasting periods (3 days vs. 2 days; p < 0.001), oral fasting periods (5 days vs. 2 days; p < 0.001), ICU admissions (6 days vs. 3 days; p = 0.002), and LOS (15 days vs. 8 days; p < 0.001). LOS correlated with the Clavien-Dino-Sink classification. CONCLUSION Gastrointestinal complications such as pancreatitis and ileus are not uncommon after SF in children with CP. This is the first study to investigate the validity of the modified Clavien-Dindo-Sink classification in GI complications after SF. Our results suggest a correlation between complication severity grade and LOS. The complexity of perioperative enteral nutritional supplementation requires prospective studies dedicated to enteral feeding protocols. LEVEL OF EVIDENCE Therapeutic-level III.