Background: Selective dorsal rhizotomy (SDR) reduces spasticity in children with cerebral palsy (CP). We analyzed potential preoperative predictors of complications after SDR via single-level laminectomy at the conus medullaris.
Methods: One hundred and forty SDRs performed in children (2-17 years) with CP were included in this retrospective study (March 2016 to July 2019). Of these children, 69% were ambulatory (Gross Motor Functional Classification System (GMFCS) II and III). Variables associated with wound dehiscence and infections, cerebrospinal fluid (CSF) leaks, and prolonged epidural pain management were analyzed statistically.
Results: Five children (3.6%) showed prolonged wound healing, which was associated with obesity (BMI z-score ≥ 1.64; odds ratio (OR) 24.4; 95% confidence interval (CI) 3-199; p = 0.003). Two cases (1.4%) had superficial surgical site infections (SSIs), which was associated with obesity (p = 0.004) and thrombocytopenia (< 180,000 G/l; p = 0.028). The area under the curve at ≥ 1.55 BMI z-score for SSI was 0.97 (95% CI 0.93-0.99, p = 0.024), with a sensitivity and specificity for SSI of 100 and 94.9%, respectively. CSF leaks occurred in four (2.9%) children, associated with age ≤ 5 years (p = 0.029). Fifteen (10.7%) children required prolonged (4-5 days) epidural pain treatment, which was associated with non-ambulatory GMFCS levels (IV and V) (OR 3.6; 95% CI 1.2-10.8; p = 0.008).
Conclusions: SDR is safe for all GMFCS levels. Obesity predicts prolonged wound healing and SSI. Prolonged pain management via epidural pain catheter is safe, but care should be taken with non-ambulatory children.
Keywords: Body mass index; Cerebral palsy; Infection; Obesity; Selective dorsal rhizotomy.