Abstract
Background:
Ethnic and racial disparities in neonatal outcomes have been well-documented, including higher risk of hypoxic-ischemic encephalopathy (HIE). Therapeutic hypothermia (TH) is the only approved treatment for infants with moderate to severe HIE and previous studies have shown mixed results regarding the impact of timing of initiation of TH on outcomes. These studies often include both inborn and outborn neonates and few minority patients.
Study design:
This retrospective cohort study of exclusively inborn neonates from a high-volume, urban, safety net hospital (SNH) serving primarily racial/ethnic minority patients assessed the impact of timing of initiation of TH on short-term outcomes.
Results:
There were 268 infants diagnosed with moderate or severe HIE from 2009 to 2023. After excluding patients for late cooling (n = 32), participation in a clinical trial (n = 41), and major comorbidities (n = 8), there were 187 patients for analysis. Similar to our neonatal population, this study cohort consisted of 94 % racial/ethnic minority patients. The average time to initiate TH was 4.4 ± 1.1 h of life (HOL) and 88 % of qualifying neonates received TH by 6 HOL. Those initiating TH at <4 HOL compared to 4–6 HOL were more likely to have severe HIE (p = 0.01). The adjusted OR for the primary outcome of in-hospital death was not associated with timing of initiation of TH [aOR = 0.75 (95 % CI 0.40–1.33); p = 0.33], nor were secondary outcomes of abnormal brain MRI or length of stay.
Conclusion:
In a vulnerable population from a high-volume SNH, timing of initiation of TH was not associated with short-term outcomes.
Keywords: Hypoxic ischemic encephalopathy, Neonatal asphyxia, Therapeutic hypothermia, Neonatal brain injury, Safety net hospital
1. Introduction
Hypoxic ischemic encephalopathy (HIE) remains a common birth complication from disrupted cerebral blood flow and oxygenation, with an estimated incidence of 2–3 per 1000 live births in the United States [1–3]. Despite increased awareness of racial and ethnic inequalities in healthcare, discrepancies continue to exist in the incidence, treatment, and outcomes from HIE [4]. Racial and ethnic minorities are more often treated at safety-net hospitals (SNH), which provide care to patients regardless of insurance status, ability to pay, or immigration status [5]. Overall quality of care in SNH neonatal intensive care units (NICU) has been shown to differ in significant ways, with one study showing higher rates of infection, necrotizing enterocolitis, and mortality compared to non-SNH NICUs [6]. There is limited information about the timeliness of treatment for neonates with HIE at SNHs.
Whole body therapeutic hypothermia (TH) is the standard of care for treating infants diagnosed with moderate to severe HIE [2,7–9]. TH decreases rates of mortality and neurodevelopmental disability (NDD) by protecting the brain from secondary energy failure if started within 6 h of life (HOL). Despite the significant impact TH has had on improving the lives of infants with HIE, it has not eliminated the risk of mortality or major morbidity, particularly for vulnerable populations like those cared for by SNHs [1,10–12]. Similar to other neuroprotective treatments, the timeliness of initiation of TH has been repeatedly shown to enhance its neuroprotective capacity in controlled pre-clinical experiments [13–18]. Timing of initiation of TH may also contribute to its degree of neuroprotection in neonates with HIE and ultimately influence their outcomes. The timing of TH initiation by 6 HOL in the major clinical trials (National Institute of Child Health and Human Development Neonatal Research Network [NICHD NRN] and Total Body Hypothermia for Neonatal Encephalopathy [TOBY] trials) was based on those animal studies showing decreased neuroprotection when started later [14–16,18,19]. A dedicated clinical trial of late TH (initiated at 6–24 HOL) showed a probability of decreased death or disability, but with uncertain effectiveness [20]. There has never been a prospective clinical trial to determine whether earlier initiation of TH may be more protective, and such a trial is no longer feasible. Analysis of the NICHD NRN and TOBY trials to determine the impact of time to start cooling is complicated by the rightward skew and lack of variability in time to initiate TH. The average time to initiate TH was 4–5 h after birth due to the time needed for appropriate diagnosis, consent, and randomization [2,21]. In the cooled group of the TOBY trial, over 70 % of neonates were randomized at 4–6 HOL [21].
Retrospective studies of outcomes based on time to initiate TH have produced mixed results [22,23]. A recent report of post-hoc analysis from the HEAL (High-dose Epo for asphyxia and encephalopathy) trial [24] did not show beneficial effects of earlier initiation of cooling on 2-year outcomes [10]. In a smaller retrospective study, earlier initiation of TH was associated with improved motor outcomes [22]. However, these studies are limited by inclusion of a mixture of inborn and outborn neonates, a majority of whom were White, and some of whom were enrolled in clinical trials of adjuvant treatments or altered cooling protocols. Given these limitations of prior studies, the objective of this study was to determine the association between time to initiate TH and short-term outcomes in an exclusively inborn population of neonates from a single center, high volume, SNH serving primarily Hispanic and non-Hispanic Black patients. To isolate the effects of timing of initiation of TH on short-term outcomes, we excluded infants who participated in clinical trials with an altered duration or depth of cooling therapy [25] as well as those who were enrolled in the HEAL trial and received adjuvant treatment for HIE [24].
2. Methods
2.1. Subjects: Inclusion and exclusion criteria
This retrospective cohort study included infants admitted to the neonatal intensive care unit (NICU) with moderate to severe HIE at Parkland Hospital from 2009 to 2023. This study was approved by the University of Texas Southwestern Institutional Review Board. All infants ≥35 weeks gestational age diagnosed with moderate or severe HIE as determined by the modified Sarnat exam [2,26] and treated with standard TH were included in the study. Our protocol for diagnosis of moderate to severe HIE closely follows that described in the NICHD NRN clinical trial. Briefly, neonates qualify for TH based on neurologic examination after meeting eligibility criteria. Infants were at least 35 weeks gestation and at least 1800 g at birth and met the physiologic criteria of pH ≤ 7.0 or base deficit ≥16 mEq/L on arterial cord blood sample or arterial blood sample within 1 HOL. For neonates with cord blood gases showing paO2 below the minimum value (<29), we considered those to likely be venous cord blood and instead used the infants first blood gas. Infants with pH between 7.01 and 7.15 or base deficit between 10 and 15.9 mEq/L qualified for a neurologic examination if there was clinical history of an acute perinatal event. All infants in the final cohort received comprehensive neurologic examination using the modified Sarnat scoring with numerical grading, as described previously [26]. All infants in the cohort were diagnosed with moderate or severe HIE based on modified Sarnat scoring and qualified for initiation of therapeutic hypothermia (presence of moderate or severe abnormality in at least 3 of 6 categories).
Although our practice has always been to initiate TH as soon as possible after diagnosis of moderate or severe HIE in keeping with the abundant pre-clinical evidence, there are times when diagnosis may be delayed and/or an infant’s exam evolves to meet the criteria. In this study, neonates were excluded from analysis if TH was initiated after 6 HOL, in addition to these other exclusion criteria: TH was discontinued early due to transfer for extracorporeal membrane oxygenation (ECMO), infant was delivered via an ex-utero intrapartum treatment (EXIT) procedure, and/or the infant was found to have significant congenital comorbidities, as listed in Fig. 1. During the study period, our center participated in clinical trials investigating adjuvant treatment and alternative cooling strategies. Neonates involved in those trials who received therapies other than 33.5 °C for a total of 72 h were also excluded from our analysis. For eligible neonates at our center, TH began by 6 HOL with placement on servo-controlled cooling equipment (Blanketrol 2009–2023 or Arctic Sun 2023). In the time from birth to placement on the cooling blanket, passive cooling was avoided. Procedures such as central line placement and/or surfactant administration most commonly occurred prior to initiation of TH during the study period. Hypothermia to 33.5 °C was used for a total of 72 h and rewarming occurred at a rate of 0.5 °C per hour. Full neonatal montage electroencephalography (EEG) was used throughout TH and rewarming with initial treatment of seizures with phenobarbital.
Fig. 1. Flowchart of study population.
Study population showing 268 infants with HIE during the study period and 187 included in analysis after excluding those with late initiation of cooling (>6 HOL), participation in clinical trial that altered cooling treatment or included adjuvant therapy, early cessation of TH, and congenital anomalies.
2.2. Outcome measures
We examined time to initiation of cooling with in-hospital outcomes. The primary short-term outcome of interest was in-hospital death. We also assessed the association between time to initiation of cooling with seizures and background EEG, brain abnormalities attributable to HIE on MRI/MR spectroscopy (defined as injury to basal ganglia, watershed regions, restricted diffusion on diffusion-weighted imaging, hemorrhages not associated with delivery and elevated lactate peaks from basal ganglia/thalamus regions and parietal white matter), and hospital length of stay (LOS) [27]. As HIE commonly affects oromotor skills, we examined the short-term outcome of independence from tube feeding at discharge. In exploratory analyses of the subset of our cohort for whom neurodevelopmental follow up care was provided at our center, we determined the association of timing of initiation of TH with scores on the Bayley Scale of Infant Development completed between 18 and 30 months of age.
2.3. Data analysis
Maternal and infant characteristics were summarized with descriptive statistics. Continuous variables were presented as means with standard deviations or medians with interquartile ranges (IQRs), and categorical variables were presented as counts with percentages, grouped by those who were cooled within 4 HOL (TH <4 HOL group) and those who were cooled at 4–6 HOL (TH 4–6 HOL group). The 4 HOL cut-off point was chosen for consistency with the recent report of post-hoc analysis of timing of initiation of TH from the HEAL cohort [10]. For the primary outcome of in-hospital mortality, analyses were performed by TH group and with timing of TH treated as a continuous variable. We used logistic regression to adjust for covariates relevant to the outcome of death. Other short-term outcomes were compared between groups in similarly adjusted analyses including linear regression for the outcome of LOS and logistic regression for the outcome of abnormality on brain MRI. For adjusted models, missing data was handled by multiple imputation. A p-value of <0.05 was considered the threshold for statistical significance. All statistical analyses were conducted using R version 4.4.1 (The Foundation for Statistical Computing, Vienna, Austria, https://www.R-project.org).
3. Results
3.1. Study population
A total of 268 newborns were diagnosed with moderate to severe HIE within the study period using the modified Sarnat exam. Eighty infants were excluded from the study including 32 infants diagnosed with moderate to severe HIE after 6 HOL, resulting in 88 % of our HIE patients who initiated TH before the recommended time frame of 6 HOL during the 15-year time frame of the study. For those who initiated TH beyond 6 HOL, careful review of their medical records indicate that they were initially diagnosed with mild HIE or triaged to newborn nursery with progression of their neurologic examination that led to their diagnosis of HIE and qualification for TH. There were 116 diagnosed with moderate HIE in the cohort and 52 with severe HIE. Among the 187 newborns included in our analysis, there were 64 who received TH within 4 HOL (TH < 4 HOL group), and 123 who initiated TH between four and six HOL (TH 4–6 HOL group) [Fig. 1]. Maternal age, gravidity, parity, number of cesarean sections, and prenatal care did not differ between groups. Pregnancy and labor complications such as diabetes, hypertension, and chorioamnionitis were also similar, as were infant sex and gestational age. Markers of HIE severity including initial blood gas, Apgar scores, chest compressions, Sarnat score, and seizures were more prevalent in the TH < 4 HOL group (Table 1). There were 21 patients for whom cord blood gases showed paO2 greater than the minimum value (<29) and were considered likely venous cord blood. For these patients, the infant’s initial blood gas was reported and used in subsequent analyses.
Table 1.
Maternal and neonatal characteristics of the cohort.
| Overall Cohort | TH < 4 HOL (n = 64) | TH 4–6 HOL (n = 123) | p-value | |
|---|---|---|---|---|
|
| ||||
| Maternal | ||||
| Age, yr mean ± SD | 28 ± 10 | 27 ± 7 | 28 ± 7 | 0.49 |
| Race/ ethnicity, n (%) | 0.03 | |||
| Hispanic | 130 (69) | 36 (56) | 94 (76) | |
| Non-Hispanic Black | 41 (22) | 22 (34) | 20 (16) | |
| Non-Hispanic White | 12 (6) | 5 (8) | 7 (6) | |
| Asian | 4 (2) | 1 (2) | 3 (2) | |
| Gravidity, median (IQR) | 2 (1,3) | 2 (1, 4) | 2 (1, 3) | 0.82 |
| Parity, median (IQR) | 1 (0,2) | 0 (0, 2) | 1 (0, 2) | 0.31 |
| Cesarean delivery, n (%) | 135 (72) | 49 (77) | 86 (69) | 0.30 |
| Prenatal care, n (%) | 177 (94) | 61 (95) | 116 (94) | 0.68 |
| Pregnancy & Labor | ||||
| Complications, n (%) | ||||
| Diabetes | 22 (12) | 7 (11) | 15 (12) | 0.82 |
| Hypertension | 67 (36) | 24 (38) | 43 (35) | 0.70 |
| Choriamnionitis | 50 (27) | 19 (30) | 31 (25) | 0.51 |
| Infant | ||||
| Male sex, n (%) | 100 (53) | 32 (50) | 68 (55) | 0.53 |
| Gestational age, median (IQR) | 39 (37, 40) | 39 (38, 40) | 39 (37, 40) | 0.20 |
| Birth weight, median (IQR) | 3240 (2830, 3588) | 3370 (3048, 3685) | 3190 (2764, 3570) | 0.03 |
| Cord gas, median (IQR) | ||||
| pH | 7.0 (6.93, 7.13) | 7.0 (6.9, 7.21) | 7.0 (6.93, 7.12) | 0.17 |
| Base deficit | −17.2 (−21.8, −12.4) | −17.1 (−24.5, −11.0) | −17.1 (−21.3, −12.3) | 0.87 |
| First postnatal gas, median | ||||
| pH | 7.16 (7.08, 7.25) | 7.08 (6.88, 7.22) | 7.18 (7.11, 7.26) | 0.002 |
| Base deficit | −15.0 (−19.0, −12.0) | −19.0 (−22.3, −14.0) | −15.0 (−18.0, −12.0) | ≤0.001 |
| Worst base deficit, median (IQR) | −19.0 (−22.8, −15.0) | − 22.0 (−25.7, −17.5) | −19.0 (−22.0, −15.3) | 0.02 |
| Apgar score, median (IQR) | ||||
| 1 min. | 2 (1, 3) | 1 (1,2) | 2 (1, 4) | <0.001 |
| 5 min. | 5 (3, 7) | 3 (2, 6) | 6 (3, 7) | <0.001 |
| 10 min. | 6 (4, 7) | 4 (3, 6) | 6 (4,7) | 0.003 |
| Chest compressions in delivery room, n (%) | 30 (16) | 16 (26) | 14 (12) | 0.01 |
| Degree of encephalopathy on admission, n (%) | ||||
| Moderate | 116/168 (69) | 35/61 (57) | 81/107 (76) | 0.01 |
| Severe | 52/168 (31) | 26/61 (43) | 26/107 (24) | |
| Sarnat score, median (IQR) | 11 (10,13) | 13 (11, 16) | 11 (10, 13) | 0.001 |
| Seizures, n (%) | 65 (41) | 32 (53) | 33 (33) | 0.02 |
SD: standard deviation; IQR: interquartile range.
Race/ethnicity varied between groups with the TH 4–6 HOL group having a higher proportion of Hispanic mothers compared to the TH < 4 HOL group. The median time to initiate cooling in the two TH groups differed significantly [3.31 (2.58, 3.67) vs. 5.12 (4.55, 5.6) HOL; p < 0.0001] and the time to reach target temperature showed a similar difference [4.60 (4.41, 5.03) vs. 6.35 (5.58, 7.02) HOL; p < 0.0001]. Average admission temperature was at the lower end of normothermia in the cohort, 36.8 ± 0.8 °C. During initiation of TH, temperatures in half of patients fell below the target range (mean lowest temperature was 32.7 ± 1.0 °C), but temperatures remained within the goal range throughout the remainder of cooling therapy.
3.2. Exploratory analyses of racial/ethnic differences
Further exploration of racial/ethnic differences demonstrated a small difference in median time to initiate TH between groups [3.9 (3.2, 4.6) HOL in Non-Hispanic Black patients (n = 41) vs. 4.2 (3.8, 4.6) HOL in Non-Hispanic White patients (n = 12) vs. 4.6 (3.7, 5.5) HOL for Hispanic patients (n = 130) vs. 5.0 (4.4, 5.3) HOL in Asian neonates (n = 4); p = 0.04]. In pairwise comparisons, the difference between Non-Hispanic Black and Hispanic infants was statistically significant (p = 0.007), but other comparisons did not differ. Clinical variables associated with increased severity of HIE, including chest compressions in the delivery room, Apgar scores at 1 or 5 min, blood gas values, modified Sarnat score, EEG background, and degree of injury on MRI did not differ significantly by race/ethnicity (Table 2).
Table 2.
Indicators of Severity of Hypoxic Ischemic Encephalopathy by Race/ Ethnicity.
| Hispanic (n = 130) | Non-Hispanic Black (n = 41) | Non-Hispanic White (n = 12) | Asian (n = 4) | p-value | |
|---|---|---|---|---|---|
|
| |||||
| Chest compressions in the delivery room, n (%) | 15 (12) | 4 (11) | 1 (10) | 0 (0) | 0.93 |
| Apgar score, median (IQR) | |||||
| 1 min. | 2 (1, 4) | 1 (1, 2) | 1 (1, 2) | 1 (1, 2) | 0.57 |
| 5 min. | 5 (3, 7) | 4 (2, 6) | 3 (2, 5) | 3 (3, 5) | 0.38 |
| Cord blood gas, median (IQR) | |||||
| pH | 7.02 (6.93, 7.14) | 7.00 (6.93, 7.15) | 7.14 (7.03, 7.22) | 7.00 (6.97, 7.06) | 0.19 |
| Base deficit | −17.5 (−21.7, −11.5) | −16.15 (−22.4, −12.5) | −14.2 (−17.1, −18.6) | −17.1 (−18.2, −14.8) | 0.49 |
| First postnatal blood gas, median (IQR) | |||||
| pH | 7.20 (7.10, 7.28) | 7.18 (7.07, 7.23) | 7.14 (7.06, 7.19) | 7.09 (7.05, 7.12) | 0.50 |
| Base deficit | −16.0 (−21.5, −11.0) | −16.0 (−24.0, −14.0) | −18.0 (−28.0, −16.2) | −17.0 (−18.0, −15.5) | 0.46 |
| Modified Sarnat score, median (IQR) | 11 (10, 13) | 12 (10, 13) | 11 (10,11) | 11 (10,12) | 0.46 |
| Degree of Encephalopathy, n (%) | |||||
| Moderate | 70 (56) | 25 (68) | 7 (70) | 3 (100) | 0.28 |
| Severe | 25 (20) | 6 (16) | 0 (0) | 0 (0) | 0.28 |
| Abnormal background EEG, n (%) | 51 (40) | 15 (41) | 5 (50) | 0 (0) | 0.44 |
| Seizure, n (%) | 43 (33) | 14 (34) | 7 (58) | 1 (25) | 0.42 |
| Abnormal brain MRI, n (%) | 68 (52) | 21 (57) | 7 (70) | 2 (67) | 0.98 |
| Death | 12 (10) | 6 (16) | 2 (20) | 0 (0) | 0.49 |
IQR: interquartile range; EEG: electroencephalogram; MRI: magnetic resonance imaging.
3.3. Primary outcome
In unadjusted analysis, the primary short-term outcome of in-hospital mortality was significantly higher in those cooled at <4 HOL compared to those cooled at 4–6 HOL (18 % vs. 7 %; p = 0.03). Treating timing of initiation of TH as a continuous variable demonstrated a persistent negative association between time to initiate TH and death before discharge (Fig. 2). In logistic regression modeling, however, the adjusted OR (aOR) for the primary outcome of in-hospital death was not associated with timing of initiation of TH (aOR = 0.99 (95 % CI 0.99–1.00); p = 0.41]. Higher risk of mortality was associated with severity of HIE and race/ethnicity, with non-Hispanic White neonates having increased aOR of death compared to minority populations (Fig. 3A). Of note, mortality in the cohort was limited to 20 patients [2 of 12 (17 %) non-Hispanic White, 6 of 41 (15 %) non-Hispanic Black, 12 of 130 (9 %) Hispanic neonates].
Fig. 2. Survival to Hospital Discharge and Timing of Initiation of Therapeutic Hypothermia.
Violin plot of timing of initiation of TH as a continuous variable in neonates with in-hospital mortality (orange) versus those who survived to discharge (blue).
Fig. 3. Forest Plots of Short-Term Outcomes.
Logistic regression adjusting for maternal pregnancy complications (teen pregnancy or advanced maternal age, diabetes in pregnancy, hypertensive disorders of pregnancy), race/ ethnicity (minority vs. non-Hispanic White patients), female sex, base deficit of first blood gas, severity of HIE by modified Sarnat score, timing of initiation of TH, and presence of seizures, for (A) in-hospital mortality and (B) abnormal brain MRI. (C) Linear regression of hospital LOS adjusted for the same variables.
3.4. Secondary outcomes
For our whole cohort, median LOS was 15 days with a range of 6 to 75 days. With regards to the secondary short-term outcomes we assessed, in our whole cohort, there were 22 % with electrographic seizures, 61 % with an abnormal background EEG and over half (52 %) with abnormal brain MRI consistent with HIE. There was no difference in the need for tube feeding at discharge between the two groups (14 % vs. 9 %; p = 0.36). In adjusted models accounting for relevant clinical variables, the outcome of abnormality on brain MRI was not associated with timing of initiation of TH, but instead was only significantly associated with severity of HIE (Fig. 3B). Similarly, linear regression modeling of the LOS outcome only found a significant association with severity of HIE (Fig. 3C).
3.5. Neurodevelopmental outcomes in limited subgroup
As clinical trial patients were excluded from this analysis and the birth hospital differed from the clinic hospital system, less than half of the medical records for neurodevelopmental testing results of our patient cohort were available. A total of 57 of the 187 patients in the cohort had Bayley Scales of Infant Development III (BSID III) testing at 18–30 months. In those for whom BSID III scores were available, there were 19 %, 46 %, and 11 % with moderate or severe neurodevelopmental delays in cognitive, language, and motor domains, respectively. Mild cognitive delays were diagnosed in 46 %, mild language delays in 26 %, and mild motor delays in 22 % of the subgroup for whom neurodevelopmental testing was available. These neurodevelopmental outcomes did not differ between patients in the TH < 4 HOL or TH 4–6 HOL groups (Table 3).
Table 3.
Timing of Initiation of Cooling and Neurodevelopmental Outcomes in Cohort Subgroup.
| Bayley Scales of Infant Development III | Whole Cohort (n = 57) | TH < 4 HOL (n = 14) | TH ≥ 4 HOL (n = 43) | p-value |
|---|---|---|---|---|
|
| ||||
| Cognitive, n (%) | 0.15 | |||
| ≤70 | 11 (19) | 5 (36) | 6 (14) | |
| 71–85 | 26 (46) | 4 (28) | 22 (51) | |
| >85 | 20 (35) | 5 (36) | 15 (35) | |
| Language, n (%) | 0.46 | |||
| ≤70 | 26 (46) | 8 (57) | 18 (42) | |
| 71–85 | 15 (26) | 2 (14) | 13 (30) | |
| >85 | 16 (28) | 4 (29) | 12 (28) | |
| Motor, n (%) | ||||
| ≤70 | 6 (11) | 1 (7) | 5 (12) | 0.86 |
| 71–85 | 12 (22) | 3 (21) | 9 (22) | |
| >85 | 37 (67) | 10 (71) | 27 (66) | |
4. Discussion
The primary finding of our analysis was that timing of initiation of TH was not associated with the primary outcome of in-hospital mortality in our exclusively inborn population at a SNH treating an underserved population. This result, however, seems to be influenced by the earlier initiation of TH in those with severe HIE, despite statistical attempts to correct for this bias. Conversely, our results may arise from inherit differences in the patients who succumbed to their illness, not accounted for in this dataset. Similarly, timing of initiation of TH was not associated with higher risk of abnormal brain MRI or longer hospital LOS. Another key result of our study was the high proportion of infants at our large urban SNH who were correctly identified with moderate or severe HIE and started TH by 6 HOL. This is reassuring given the disparities in healthcare accessibility and outcomes in minority patient populations and is an important confirmation of the findings of prior studies in our unique NICU population [23,28,29]. Importantly, a total of 88 % of neonates who qualified for TH during the 15-year study period received treatment with TH within 6 HOL. All neonates initiating TH beyond 6 HOL were either triaged to newborn nursery after delivery or diagnosed with mild HIE initially, and then their neurologic status declined. A recent report described factors associated with delayed TH including birth outside a cooling center, less severe clinical presentation, and birth at a low-volume hospital center [30].
A lack of effect of the timing of initiation of TH on the in-hospital outcomes of mortality, abnormal brain MRI, and hospital LOS is likely explained by the unknown timing of perinatal insult in most neonates in our cohort and has similarly complicated the analyses from other studies [23,29]. This uncertainty precludes our ability to accurately measure the effects of time to initiate TH and potentially explains the mixed results in similar studies [10,22,28,29,31]. Specifically, Thoresen et al. demonstrated improved motor outcomes at 18–20 months of age among 65 surviving newborns who received TH as early as 3 HOL compared to those treated at 3–6 HOL [22]. Conversely, a retrospective study including 68 newborns treated with TH did not show a significant effect of early TH initiation on improved short-term outcomes including death or MRI injury [28]. Most recently, Rao et al. demonstrated from the large cohort of the HEAL trial that there was no significant difference in risk of death or neurodevelopment impairment between those infants who reached target temperature before vs. after 4 HOL [10]. Only a population with acute onset and precise timing of perinatal injury, such as those with placental abruption, may recapitulate animal studies sufficiently to answer the question on effects of timing of initiation of TH. However, data from a group of neonates with a specific inciting event such as placental abruption, may not produce results that are generalizable to other etiologies of HIE.
Our adjusted analysis showed that severity of HIE was the only variable associated with increased risk of death, abnormal brain MRI, and increased hospital LOS. In modified Sarnat scoring, severe HIE typically corresponds to a score between 15 and 18 where at least three categories of the exam are scored as severe. However, there were a total of 52 infants with severe HIE in our cohort, meaning that 64 % of infants with severe HIE did survive to discharge. This demonstrated that even complicated statistical models are not able to accurately predict outcomes for these patients, likely due to the crudeness of our clinical exams and complexity of neonatal brain injury and recovery. Clearly, additional studies and more refined tools are needed to assess this and other forms of neonatal brain injury.
There were two unexpected outcomes in our analyses, 1) Hispanic neonates in our cohort initiated TH later than other groups, and 2) as a whole, our minority patients had a significantly lower risk of in-hospital mortality compared to our non-Hispanic White patients. We hesitate to draw any conclusions from these secondary exploratory analyses for which our study was not powered. The difference in timing of initiation of TH by racial/ethnic category was not explained by differences in severity of HIE. This association was explored because those with increased severity of HIE typically begin TH earlier owing to their more acute presentation at birth [22,23]. We can best explain this finding by the imbalanced numbers in each population with a significantly smaller non-Hispanic White population in our cohort. Similarly, the finding of reduced risk of mortality for minorities in our adjusted models is likely explained by the low number of overall deaths during the 15-year timeframe of our analysis, and, again by the very few non-Hispanic White patients which comprised only 6 % of the cohort. Further research is needed to more fully explore this finding.
The limitations of this study center on its retrospective design and our attempt to collect data spanning two hospital systems. In an effort to include the maximum years of data from patients with HIE treated with TH, we encountered missing data due to incomplete electronic medical records, particularly prevalent from the early years of our study time frame. Additionally, a significant number of infants were excluded from this analysis due to participation in prospective trials [24,25,32], termination of cooling early due to clinical deterioration, or for presence of congenital anomalies. These excluded patients represented a total of 30 % of the study population, and therefore, may have had significant impacts on our results. Furthermore, our neurodevelopmental outcomes data was reduced relative to our typical follow up rates, which is attributable to patients changing names between birth hospital and the separate clinic hospital system making recovery of those results not possible. Our missing data also resulted from reduced adherence to recommended follow-up during the COVID-19 pandemic. Due to these limitations, our primary and secondary outcomes focused on in-hospital data, which was readily available. We feel this data adds a valuable perspective to the HIE field owing to the high volume and predominantly minority population served at our urban SNH.
Despite the lack of association between timing of initiation of TH and outcomes in this study and others, we caution against the misinterpretation of these studies. Following an accurate diagnosis of HIE, we continue to recommend the prompt initiation of TH for those who qualify, as it is the singular therapy proven to improve outcomes in these patients. Withholding or delaying this effective treatment is never desirable. We acknowledge that limitations of clinical research differ from the controlled environment of preclinical studies and encourage additional research in this field to more precisely identify patient populations with variable response to TH. Pre-clinical data clearly showed that early initiation of TH was beneficial – in neonatal rats and fetal sheep, the effectiveness of neuronal sparing decreased significantly with delay of TH [15,18] and TH was ineffective after 3 HOL in neonatal piglets [17,19]. As much of our knowledge regarding the pathophysiology of neuronal rescue by TH arises from these animal models, an earlier initiation of TH may also improve neuroprotection in neonates, which may or may not be apparent in the short-term outcomes that have been assessed to date. Even small differences in neuronal survival may have longer-reaching impacts than current studies have measured. Lessons taken from neonates with congenital heart disease, including those with singular operations in the neonatal period, show that periods of inadequate cerebral oxygenation increase the risk of adverse outcomes in later childhood, adolescence, and may even persist into adulthood [33–36]. Thus, the study of treatments and timing of initiation of them for neonates with HIE cannot stop at the toddlerhood assessments; rather, important information can be gained from follow up into later childhood, adolescence, and beyond.
5. Conclusion
In-hospital outcomes were not associated with the timing of initiation of TH in our cohort of exclusively inborn neonates with HIE from a high-volume delivery SNH that treats primarily an underserved and vulnerable population. Unlike preclinical research, the timing of perinatal insult varies in clinical practice, which potentially explains why some infants benefit more than others from TH. The impacts of timing of initiation of TH may become apparent in studies limited to neonates with a known onset of perinatal injury. Ultimately, our study highlights the need for further research including effects of timing of TH initiation in specific subsets of neonates and should include longer term outcomes.
Footnotes
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Lina F. Chalak reports financial support was provided by Crystal Charity Ball. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
CRediT authorship contribution statement
Christine C. Pazandak: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Maria Grajeda Martinez: Writing – review & editing, Methodology, Data curation. Megan E. Whiting: Writing – review & editing, Formal analysis, Data curation. Srinivas Kota: Writing – review & editing, Writing – original draft, Visualization, Data curation. L. Steven Brown: Validation, Methodology, Formal analysis. Lina F. Chalak: Writing – review & editing, Supervision, Methodology. Rachel L. Leon: Writing – review & editing, Writing – original draft, Visualization, Supervision, Methodology, Formal analysis, Data curation.
Data availability
The dataset generated and analyzed during the current study is not publicly available due to privacy and confidentiality considerations. However, it may be obtained from the corresponding author upon reasonable request, subject to appropriate data protection and confidentiality agreements.
References
- [1].Park J, Park SH, Kim C, Yoon SJ, Lim JH, Han JH, Shin JE, Eun HS, Park MS, Lee SM, Growth and developmental outcomes of infants with hypoxic ischemic encephalopathy, Sci. Rep. 13 (2023) 23100, 10.1038/s41598-023-50187-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE, McDonald SA, Donovan EF, Fanaroff AA, Poole WK, Wright LL, Higgins RD, Finer NN, Carlo WA, Duara S, Oh W, Cotten CM, Stevenson DK, Stoll BJ, Lemons JA, Guillet R, Jobe AH, Whole-Body Hypothermia for Neonates with Hypoxic–Ischemic Encephalopathy, N. Engl. J. Med. 353 (2005) 1574–1584, 10.1056/NEJMcps050929. [DOI] [PubMed] [Google Scholar]
- [3].Chalak LF, Bitar L, Kota S, Perinatal hypoxic-ischemic encephalopathy among a large public hospital population, JAMA Netw. Open 7 (2024) e2444448, 10.1001/jamanetworkopen.2024.44448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Parchem JG, Rice MM, Grobman WA, Bailit JL, Wapner RJ, Debbink MP, Thorp JM, Caritis SN, Prasad M, Tita ATN, Saade GR, Sorokin Y, Rouse DJ, Tolosa JE, Eunice Kennedy Shriver National Institute of child health human development maternal-fetal medicine units (MFMU) network, racial and ethnic disparities in adverse perinatal outcomes at term, Am. J. Perinatol. 40 (2023) 557–566, 10.1055/s-0041-1730348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Hefner JL, Hogan TH, Opoku-Agyeman W, Menachemi N, Defining safety net hospitals in the health services research literature: a systematic review and critical appraisal, BMC Health Serv. Res. 21 (2021) 278, 10.1186/s12913-021-06292-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Liu J, Pang EM, Iacob A, Simonian A, Phibbs CS, Profit J, Evaluating Care in Safety net Hospitals: clinical outcomes and neonatal intensive care unit quality of Care in California, J. Pediatr. 243 (2022) 99–106.e3, 10.1016/j.jpeds.2021.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Azzopardi D, Brocklehurst P, Edwards D, Halliday H, Levene M, Thoresen M, Whitelaw A, TOBY study group, the TOBY study, Whole body hypothermia for the treatment of perinatal asphyxial encephalopathy: a randomised controlled trial, BMC Pediatr 8 (2008) 17, 10.1186/1471-2431-8-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Jacobs SE, Whole-body hypothermia for term and near-term newborns with hypoxic-ischemic encephalopathy: a randomized controlled trial, Arch. Pediatr. Adolesc. Med. 165 (2011) 692, 10.1001/archpediatrics.2011.43. [DOI] [PubMed] [Google Scholar]
- [9].Simbruner G, Mittal RA, Rohlmann F, Muche R, neo.nEURO.Network trial participants, systemic hypothermia after neonatal encephalopathy: outcomes of neo.nEURO.Network RCT, Pediatrics 126 (2010) e771–e778, 10.1542/peds.2009-2441. [DOI] [PubMed] [Google Scholar]
- [10].Rao R, Comstock BA, Wu T-W, Mietzsch U, Mayock DE, Gonzalez FF, Wood TR, Heagerty PJ, Juul SE, Wu YW, Time to reaching target cooling temperature and 2-year outcomes in infants with hypoxic-ischemic encephalopathy, J. Pediatr. 266 (2024) 113853, 10.1016/j.jpeds.2023.113853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Youn Y-A, Kim JH, Yum S-K, Moon C-J, Lee I-G, Sung IK, The hospital outcomes compared between the early and late hypothermia-treated groups in neonates, J. Matern. Fetal Neonatal Med. 29 (2016) 2288–2292, 10.3109/14767058.2015.1083548. [DOI] [PubMed] [Google Scholar]
- [12].Natarajan G, Pappas A, Shankaran S, Laptook AR, Walsh M, McDonald SA, Ehrenkranz RA, Tyson JE, Goldberg RN, Bara R, Higgins RD, Das A, Munoz B, Effect of inborn vs. outborn delivery on neurodevelopmental outcomes in infants with hypoxic-ischemic encephalopathy: secondary analyses of the NICHD whole-body cooling trial, Pediatr. Res. 72 (2012) 414–419, 10.1038/pr.2012.103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Gunn AJ, Gunn TR, Gunning MI, Williams CE, Gluckman PD, Neuroprotection with prolonged head cooling started before postischemic seizures in fetal sheep, Pediatrics 102 (1998) 1098–1106, 10.1542/peds.102.5.1098. [DOI] [PubMed] [Google Scholar]
- [14].Nedelcu J, Klein MA, Aguzzi A, Martin E, Resuscitative hypothermia protects the neonatal rat brain from hypoxic-ischemic injury, Brain Pathol. 10 (2000) 61–71, 10.1111/j.1750-3639.2000.tb00243.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Gunn AJ, Gunn TR, De Haan HH, Williams CE, Gluckman PD, Dramatic neuronal rescue with prolonged selective head cooling after ischemia in fetal lambs, J. Clin. Invest. 99 (1997) 248–256, 10.1172/JCI119153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Sabir H, Scull-Brown E, Liu X, Thoresen M, Immediate hypothermia is not neuroprotective after severe hypoxia-ischemia and is deleterious when delayed by 12 hours in neonatal rats, Stroke 43 (2012) 3364–3370, 10.1161/STROKEAHA.112.674481. [DOI] [PubMed] [Google Scholar]
- [17].Karlsson M, Tooley JR, Satas S, Hobbs CE, Chakkarapani E, Stone J, Porter H, Thoresen M, Delayed hypothermia as selective head cooling or whole body cooling does not protect brain or body in newborn pig subjected to hypoxia-ischemia, Pediatr. Res. 64 (2008) 74–78, 10.1203/PDR.0b013e318174efdd. [DOI] [PubMed] [Google Scholar]
- [18].Thoresen M, Bagenholm R, Loberg EM, Apricena F, Kjellmer I, Posthypoxic cooling of neonatal rats provides protection against brain injury, Arch. Dis. Child. Fetal Neonatal Ed. 74 (1996) F3–F9, 10.1136/fn.74.1.F3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Haaland K, Løberg EM, Steen PA, Thoresen M, Posthypoxic hypothermia in newborn piglets, Pediatr. Res. 41 (1997) 505–512, 10.1203/00006450-199704000-00009. [DOI] [PubMed] [Google Scholar]
- [20].Laptook AR, Shankaran S, Tyson JE, Munoz B, Bell EF, Goldberg RN, Parikh NA, Ambalavanan N, Pedroza C, Pappas A, Das A, Chaudhary AS, Ehrenkranz RA, Hensman AM, Van Meurs KP, Chalak LF, Khan AM, Hamrick SEG, Sokol GM, Walsh MC, Poindexter BB, Faix RG, Watterberg KL, Frantz ID 3rd, Guillet R, Devaskar U, Truog WE, Chock VY, Wyckoff MH, McGowan EC, Carlton DP, Harmon HM, Brumbaugh JE, Cotten CM, Sanchez PJ, Hibbs AM, Higgins RD, Eunice H Kennedy Shriver National Institute of Child, N. Human Development Neonatal Research, Effect of Therapeutic Hypothermia Initiated After 6 Hours of Age on Death or Disability Among Newborns With Hypoxic-Ischemic Encephalopathy: A Randomized Clinical Trial, JAMA 318 (2017) 1550–1560. doi: 10.1001/jama.2017.14972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Azzopardi DV, Strohm B, Edwards AD, Dyet L, Halliday HL, Juszczak E, Kapellou O, Levene M, Marlow N, Porter E, Thoresen M, Whitelaw A, Brocklehurst P, Group TS, Moderate hypothermia to treat perinatal asphyxial encephalopathy, N. Engl. J. Med. 361 (2009) 1349–1358, 10.1056/NEJMoa0900854. [DOI] [PubMed] [Google Scholar]
- [22].Thoresen M, Tooley J, Liu X, Jary S, Fleming P, Luyt K, Jain A, Cairns P, Harding D, Sabir H, Time is brain: starting therapeutic hypothermia within three hours after birth improves motor outcome in asphyxiated newborns, Neonatology 104 (2013) 228–233, 10.1159/000353948. [DOI] [PubMed] [Google Scholar]
- [23].Rao R, Comstock BA, Wu T-W, Mietzsch U, Mayock DE, Gonzalez FF, Wood TR, Heagerty PJ, Juul SE, Wu YW, Time to reaching target cooling temperature and 2-year outcomes in infants with hypoxic-ischemic encephalopathy, J. Pediatr. 266 (2024) 113853, 10.1016/j.jpeds.2023.113853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Wu YW, Comstock BA, Gonzalez FF, Mayock DE, Goodman AM, Maitre NL, Chang T, Van Meurs KP, Lampland AL, Bendel-Stenzel E, Mathur AM, Wu T-W, Riley D, Mietzsch U, Chalak L, Flibotte J, Weitkamp J-H, Ahmad KA, Yanowitz TD, Baserga M, Poindexter BB, Rogers EE, Lowe JR, Kuban KCK, O’Shea TM, Wisnowski JL, McKinstry RC, Bluml S, Bonifacio S, Benninger KL, Rao R, Smyser CD, Sokol GM, Merhar S, Schreiber MD, Glass HC, Heagerty PJ, Juul SE, Trial of Erythropoietin for Hypoxic–Ischemic Encephalopathy in Newborns, N. Engl. J. Med. 387 (2022) 148–159. doi: 10.1056/NEJMoa2119660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Shankaran S, Laptook AR, Pappas A, McDonald SA, Das A, Tyson JE, Poindexter BB, Schibler K, Bell EF, Heyne RJ, Pedroza C, Bara R, Van Meurs KP, Grisby C, Huitema CM, Garg M, Ehrenkranz RA, Shepherd EG, Chalak LF, Hamrick SE, Khan AM, Reynolds AM, Laughon MM, Truog WE, Dysart KC, Carlo WA, Walsh MC, Watterberg KL, Higgins RD, Eunice H Kennedy Shriver National institute of child, N. Human development neonatal research, effect of depth and duration of cooling on deaths in the NICU among neonates with hypoxic ischemic encephalopathy: a randomized clinical trial, JAMA 312 (2014) 2629–39. doi: 10.1001/jama.2014.16058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Chalak LF, Nguyen K-A, Prempunpong C, Heyne R, Thayyil S, Shankaran S, Laptook AR, Rollins N, Pappas A, Koclas L, Shah B, Montaldo P, Techasaensiri B, Sanchez PJ, Sant´ ‘Anna G, Prospective research in infants with mild encephalopathy identified in the first six hours of life: neurodevelopmental outcomes at 18–22 months, Pediatr. Res. 84 (2018) 861–868, 10.1038/s41390-018-0174-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Wisnowski JL, Bluml S, Panigrahy A, Mathur AM, Berman J, Chen P-SK, Dix J, Flynn T, Fricke S, Friedman SD, Head HW, Ho CY, Kline-Fath B, Oveson M, Patterson R, Pruthi S, Rollins N, Ramos YM, Rampton J, Rusin J, Shaw DW, Smith M, Tkach J, Vasanawala S, Vossough A, Whitehead MT, Xu D, Yeom K, Comstock B, Heagerty PJ, Juul SE, Wu YW, McKinstry RC, HEAL study group, integrating neuroimaging biomarkers into the multicentre, high-dose erythropoietin for asphyxia and encephalopathy (HEAL) trial: rationale, protocol and harmonisation, BMJ Open 11 (2021) e043852, 10.1136/bmjopen-2020-043852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Dresbach T, Rigoni V, Groteklaes A, Hoehn T, Stein A, Felderhoff-Mueser U, Mueller A, Sabir H, The impact of time to initiate therapeutic hypothermia on short-term neurological outcomes in neonates with hypoxic-ischemic encephalopathy, Children (Basel) 11 (2024) 686, 10.3390/children11060686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Guillot M, Philippe M, Miller E, Davila J, Barrowman NJ, Harrison M-A, Ben Fadel N, Redpath S, Lemyre B, Influence of timing of initiation of therapeutic hypothermia on brain MRI and neurodevelopment at 18 months in infants with HIE: a retrospective cohort study, Bmjpo 3 (2019) e000442, 10.1136/bmjpo-2019-000442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Beck J, Debillon T, Guellec I, Vilotitch A, Loron G, Bednarek N, Ancel P-Y, Pierrat V, Ego A, Healthcare organizational factors associated with delayed therapeutic hypothermia in neonatal hypoxic-ischemic encephalopathy: the LyTONEPAL cohort, Eur. J. Pediatr. 182 (2023) 181–190, 10.1007/s00431-022-04666-7. [DOI] [PubMed] [Google Scholar]
- [31].Leon RL, Krause KE, Sides RS, Koch MB, Trautman MS, Mietzsch U, Therapeutic hypothermia in transport permits earlier treatment regardless of transfer distance, Am. J. Perinatol. 39 (2022) 633–639, 10.1055/s-0040-1718372. [DOI] [PubMed] [Google Scholar]
- [32].Chalak LF, Kang S, Kota S, Liu H, Liu Y, Juul SE, Wu YW, Evaluation of neurovascular coupling during neuroprotective therapies: a single site HEAL ancillary study, Early Hum. Dev. 183 (2023) 105815, 10.1016/j.earlhumdev.2023.105815. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Pfitzer C, Buchdunger LA, Helm PC, Blickle MJ, Rosenthal L-M, Ferentzi H, Berger F, Bauer UMM, Schmitt KRL, Education of children with cyanotic congenital heart disease after neonatal cardiac surgery, Ann. Thorac. Surg. 112 (2021) 1546–1552, 10.1016/j.athoracsur.2020.07.072. [DOI] [PubMed] [Google Scholar]
- [34].Bellinger DC, Wypij D, duPlessis AJ, Rappaport LA, Jonas RA, Wernovsky G, Newburger JW, Neurodevelopmental status at eight years in children with dextro-transposition of the great arteries: the Boston circulatory arrest trial, J. Thorac. Cardiovasc. Surg. 126 (2003) 1385–1396, 10.1016/s0022-5223(03)00711-6. [DOI] [PubMed] [Google Scholar]
- [35].Chen H, Yan Y, Li C, Zheng X, Wang G, Jin Z, Shi G, He X, Tong X, Chen H, Zhu Z, Inattention and hyperactivity in children and adolescents with repaired D-transposition of the great arteries: prevalence, perioperative risk factors, and clinical outcomes, Front. Cardiovasc. Med. 9 (2022) 937311, 10.3389/fcvm.2022.937311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Mercer-Rosa L, Zhang X, Tanel RE, Rychik J, Fogel MA, Paridon SM, Goldmuntz E, Perioperative factors influence the long-term outcomes of children and adolescents with repaired tetralogy of Fallot, Pediatr. Cardiol. 39 (2018) 1433–1439, 10.1007/s00246-018-1913-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The dataset generated and analyzed during the current study is not publicly available due to privacy and confidentiality considerations. However, it may be obtained from the corresponding author upon reasonable request, subject to appropriate data protection and confidentiality agreements.