Preoperative laboratory testing before pediatric neurosurgery: an NSQIP-Pediatrics analysis

Restricted access


The goal of this study was to evaluate clinical predictors of abnormal preoperative laboratory values in pediatric neurosurgical patients.


Data obtained in children who underwent a neurosurgical operation were extracted from the prospective National Surgical Quality Improvement Program–Pediatrics (NSQIP-P, 2012–2013) registry. Multivariable logistic regression evaluated predictors of preoperative laboratory values that might require further evaluation (white blood cell count < 2000/μl, hematocrit < 24%, platelet count < 100,000/μl, international normalized ratio > 1.4, or partial thromboplastin time > 45 seconds) or a preoperative transfusion (within 48 hours prior to surgery). Variables screened included patient demographics; American Society of Anesthesiologists (ASA) physical designation classification; comorbidities; recent steroid use, chemotherapy, or radiation therapy; and admission type. Predictive score validation was performed using the NSQIP-P 2014 data.


Of the 6556 patients aged greater than 2 years, 68.9% (n = 5089) underwent laboratory testing, but only 1.9% (n = 125) had a critical laboratory value. Predictors of a laboratory abnormality were ASA class III–V; diabetes mellitus; hematological, hypothrombotic, or oncological comorbidities; nutritional support; recent chemotherapy; systemic inflammatory response syndrome; and a nonelective hospital admission. These 9 variables were used to create a predictive score, with a single point assigned for each predictor. The prevalence of critical values in the validation population (NSQIP-P 2014) of patients greater than 2 years of age was 0.3% with a score of 0, 1.0% in those with a score of 1, 1.6% in those with a score of 2, and 6.2% in those with a score ≥ 3. Higher score was predictive of a critical value (OR 2.33, 95% CI 1.91–2.83, p < 0.001, C-statistic 0.76) and with the requirement of a perioperative transfusion (intraoperatively or within 72 hours postoperatively; OR 1.42, 95% CI 1.22–1.67, p < 0.001) in the validation population. Moreover, when the same score was applied to children aged 2 years or younger, a greater score was predictive of a critical value (OR 2.47, 95% CI 2.15–2.84, p < 0.001, C-statistic 0.76).


Critical laboratory values in pediatric neurosurgical patients are largely predicted by clinical characteristics, and abnormal preoperative laboratory results are rare in patients older than 2 years of age without comorbidities who are undergoing elective surgery. The NSQIP-P critical preoperative laboratory value scale is proposed to indicate patients with the highest odds of an abnormal value. The scale can assist with triaging preoperative testing based on the surgical risk, as determined by the treating surgeon and anesthesiologist.

ABBREVIATIONS ASA = American Society of Anesthesiologists; HL = Hosmer-Lemeshow; INR = international normalized value; NSQIP = National Surgical Quality Improvement Program; NSQIP-P = NSQIP–Pediatrics; PTT = partial thromboplastin time; RBC = red blood cell; SIRS = systemic inflammatory response syndrome; WBC = white blood cell.

Downloadable materials

  • Supplemental Table 1 (PDF 466 KB)

Article Information

Correspondence Shenandoah Robinson: Johns Hopkins Hospital, Baltimore, MD.

INCLUDE WHEN CITING Published online April 12, 2019; DOI: 10.3171/2018.12.PEDS18441.

Disclosures The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

© AANS, except where prohibited by US copyright law.



  • View in gallery

    Variations in the crude prevalence (and associated standard errors) of critical preoperative laboratory values in children aged at least 2 years by the NSQIP-P critical preoperative laboratory scale score.

  • View in gallery

    Receiver operating characteristic (ROC) curves evaluating the predictive capability of the NSQIP-P critical preoperative laboratory scale score in the validation populations of children aged 2–18 years (left) and less than 2 years (right).



All Time Past Year Past 30 Days
Abstract Views 53 53 53
Full Text Views 15 15 15
PDF Downloads 19 19 19
EPUB Downloads 0 0 0


Google Scholar