13 - Higher Blanket Temperature During Therapeutic Hypothermia is associated with worse outcomes in Neonatal Hypoxic-ischemic Encephalopathy.

Publication Number: 1011.13

Franklin Ducatez, Boston Children's Hospital, Brookline, MA, United States; Chuan-Heng Hsiao, Boston Children's Hospital, Boston, MA, United States; Rutvi Vyas, Boston Children's Hospital, Boston, MA, United States; Rina Bao, Boston Children's Hospital and Harvard Medical School, Boston, MA, United States; Ankush Kesri, Boston Children's Hospital, Cambridge, MA, United States; Seetha Shankaran, University of Texas at Austin Dell Medical School, Ann Arbor, MI, United States; Abbot R.. Laptook, Women & Infants Hospital of Rhode Island, North Kingstown, RI, United States; Michael Cotten, Duke University School of Medicine, Durham, NC, United States; P. Ellen grant, Boston Children's Hospital, Boston, MA, United States; Yangming Ou, Boston Children's Hospital, Boston, MA, United States

Background: Hypoxic-ischemic encephalopathy (HIE) affects 1-3‰ newborns in high-income countries (1). The Consortium Of MRI Biomarkers In Neonatal Encephalopathy (COMBINE) harmonized data from 2 therapeutic hypothermia (TH) trials (2); Late Hypothermia (LH) and Optimized Cooling (OC) (3, 4); led by NICHD Neonatal Research Network (2008-16). This comprehensive multisite database allows us to explore a potential predictor - the mean of blanket temperature, relative to the target temperature during TH - for not only general (5) but domain-specific HIE outcomes.

Objective: To associate blanket temperature and 5 neurodevelopmental outcomes in two phases using the harmonized LH+OC datasets.

Design/Methods: Patients w/o TH or followup, or wi/ ambiguous sex were excluded. All available blanket temperature data points were extracted. Measurements from the first 4h of TH were excluded. Relative blanket temperature (defined as blanket subtracting target temperatures) was computed from 4h to the end of TH (measured every hour in 5-12h, every 4 hours thereafter). Univariate associations between the relative blanket temperature mean (rbT-mean) and outcomes were tested with the Wilcoxon and Pearson tests. Predictive performance of rbT-mean for each outcome was assessed via AUC. Multivariate linear and logistic regressions models were used to evaluate the effect of rbT-mean, accounting for covariates: TH arm, sex, gestational age, birthweight, maternal education and use of antiepileptics (Table 1). In total, 5 outcomes were used at 18-22 months: moderate/severe disability or death, cerebral palsy, and Bayley-III cognition, motor, and language composite scores (Table 1).

Results: From 532 patients, 109 were excluded (85 w/o TH, 1 with sex ambiguity, and 23 w/o follow-up). In the resulting 427 neonates, 203 (49%) had normal outcomes, 86 (21%) had mild disability, 5 (1.2%) moderate, 66 (16%) severe disability, and 58 (14%) died (Table 1).
Univariate exploration showed that higher rbT-mean was associated with worse outcomes (Fig 1), with predictive performance (AUC 0.76 for disability; 0.72 for CP; Fig 2). Multivariate regressions with covariates showed associations between higher rbT-mean and adverse outcomes for disability level or death (OR 1.04 [95% CI 1.03-1.06]; adj p < 0.001), CP (OR 1.02 [1.005-1.035]; adj p = 0.008) and lower Bayley-III motor scores (β = -0.881; adj p = 0.02), but not for Bayley-III Language or Cognitive scores (Table 2).

Conclusion(s): Higher relative blanket temperature mean is associated with adverse neurodevelopmental outcomes (moderate/severe disability or death; cerebral palsy, and Bayley III Motor score).

Fig 1
Univariate association between the mean relative blanket temperature (rbT-mean) on 2 out of the 5 outcomes by 18-22 months. Panel A: outcome moderate/severe/death or not. Panel B: outcome cerebral palsy or not. Panels C-E: outcomes Bayley-III Cognitive, Language, and Motor scores.

Fig 2
Receiver Operation Characteristics curve, using the mean relative blanket temperature to predict moderate/severe disability or death (upper panel), or cerebral palsy (lower panel).

Tables
Table1_Table2.pdfTable 1. Characteristics in the harmonized LH+OC data. TH - therapeutic hypothermia.
Table 2. Effect estimation of the mean relative blanket temperature on 5 outcomes by 18-22 months; adjusted for the following cofounders: TH arm, sex, gestational age, birthweight, maternal education and use of antiepileptics.