Intraventricular hemorrhage (IVH) is a common and severe complication of prematurity due to the fragility of the germinal matrix and the immature cerebrovascular system[1] and remains a significant cause of morbidity and mortality, particularly in infants born before 32 weeks of gestation or weighing <1500 g [2]. Despite substantial advances in neonatal care, IVH continues to affect up to 50% of very low birth weight infants in many settings, with severe grades of IVH strongly associated with poor neurodevelopmental outcomes, such as cerebral palsy, cognitive impairment, and sensory deficits [3,4]. IVH is a multifactorial condition influenced by antenatal, perinatal, and postnatal factors, including antenatal exposure to steroids, mode of delivery, tracheal intubation, and inotropic use [5,6].

Most epidemiological data on IVH originate from high-income countries with well-established neonatal networks with consistent quality of care and distinct population characteristics. However, in low- and middle-income countries (LMICs), the incidence and outcomes of IVH may differ due to resource constraints, limited access to specialized neonatal care, and disparities in the implementation of evidence-based practices [7].

Brazil, as an LMIC with heterogeneity in perinatal care, offers a unique setting to understand the burden of IVH in preterm neonates. The Brazilian Neonatal Research Network has documented that 30.4% of preterm infants in Brazil experience IVH, with significant variations in severity and outcomes across neonatal intensive care units (NICUs). Moreover, trends from 2013 to 2018 indicate an increasing incidence of IVH, reflecting the increased rate of survival of lower GA infants, the complexity of neonatal care, and possible gaps in resource allocation and clinical protocols [8].

To address this knowledge gap, the authors conducted a multicenter cohort study in four Brazilian NICUs, including both private and public centers, to identify the rates and severity of IVH among infants born at < 32 weeks’ gestation or < 1500 g, and secondarily, to identify the demographic, maternal, prenatal, delivery, and neonatal risk factors for IVH in these Brazilian NICUs.

This observational, prospective cohort study was conducted in four NICUs in Brazil from September 2023 to September 2024. The study was approved by the Research Ethics Committees of all participating hospitals and followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guideline.

One center was based in the state of Goiás (Central-West region) and three in São Paulo (Southeast region). Brazil’s healthcare system is characterized by a dual structure with both public and private sectors. Of the four participating NICUs, three were public hospitals integrated into the Brazilian Unified Health System (Sistema Único de Saúde – SUS), and one was a private facility. All centers are referral institutions for high-risk pregnancies and provide level III neonatal care. The inclusion of centers from different geographic regions and with distinct organizational structures and resource availability was intentional, aiming to capture the heterogeneity of neonatal care in Brazil and enhance the external validity of the findings. Prior to the study initiation, training sessions were held with the center’s investigators to standardize the data collection process. A shared protocol and centralized database were used to ensure data consistency across centers.

The study population consisted of preterm infants born at < 32 weeks’ gestation or with birth weight < 1500 g, admitted to the participating NICUs during the one-year study period. Infants with congenital malformations or genetic syndromes were excluded.

Antenatal, in-hospital, and outcome data were collected from the patients' medical records, including information on invasive interventions in the NICU, such as the use of sedatives or analgesics, inotropes, fluid bolus, and mechanical ventilation. Antenatal steroid exposure was categorized as complete when two doses were administered, irrespective of the time interval between doses. Exposure was considered incomplete when fewer than two doses were administered, regardless of timing. The most severe IVH grade identified on cranial ultrasound (cUS) during NICU stay was included in the analysis. The diagnosis and classification of IVH were based on Papile’s classification system [9]. Grades I and II IVH were considered to be mild IVH, and grades III and IV were considered to be severe IVH [9]. Post-hemorrhagic ventricular dilatation was also classified as a severe finding.

Descriptive analyses were conducted using frequencies for categorical variables and means, medians, standard deviations, and interquartile ranges for continuous variables, according to data distribution. The results of cUS were divided into normal (no evidence of IVH), mild, and severe. To avoid exclusion of high-risk infants, those who died before cUS assessment were included in a composite outcome of severe IVH or death prior to cUS. Infants with leukomalacia and no IVH were excluded from this comparison.

Multinomial logistic regression was performed to evaluate associations between clinical variables and IVH severity. Crude associations were estimated using univariate models. Subsequently, domain-specific multivariable models were constructed a priori based on clinical plausibility. A baseline demographic model included gestational age (GA), sex, mode of delivery, and hospital sector. Separate models evaluated prenatal exposures, perinatal variables, and early postnatal interventions within the first 72 h of life, all adjusted for baseline variables. GA and birth weight were not included simultaneously due to their biological collinearity. Multicollinearity was assessed using the variance inflation factor (VIF), and variables with VIF > 5 were excluded. Results are presented as relative risk ratios (RRR) with 95% confidence intervals (CI), and statistical significance was defined as p < 0.05.

This prospective multicenter study conducted in four Brazilian NICUs revealed a 36.4% incidence of IVH among preterm infants born at < 32 weeks' gestation or weighing < 1500 g, with the combined outcome of severe IVH or death prior to cUS occurring in 20%. Infants with this combined outcome were characterized by lower birth weight and GA, lower Apgar scores, and lower rates of cesarean delivery, and were less likely to be exposed to a complete course of antenatal corticosteroids and more likely to have undergone advanced resuscitation in the delivery room and more invasive interventions in the NICU, such as use of sedatives or analgesics, inotropes, fluid bolus, and mechanical ventilation. Hypothermia was prevalent across all groups, but particularly elevated among those with severe IVH or death. In multinomial regression analyses, lower GA, advanced resuscitation in the delivery room, and early use of inotropes were independently associated with severe IVH or death.

The IVH rates seen in this study align with global and local reports. A previous meta-analysis of 64 studies including data from 9633 preterm infants ≤ 32 weeks GA born after 2007, reported a global IVH occurrence of 31% (95% CI 25- 36%) and 11% (95% CI 8–14%) of severe IVH [2]. A multicenter cohort study including very low birth weight newborns < 1500 g, published by the Brazilian Network on Neonatal Research, reported a similar overall rate of IVH in 30.4%, with 9.8% classified as severe [8]. These findings highlight that a substantial proportion of preterm infants remain at high risk for severe brain injury.

In this analysis, infants with severe IVH or death presented more frequently with lower birth weight, GA, and Apgar scores. In the adjusted multinomial analyses, GA emerged as the most consistent independent predictor of severe IVH or death. These findings are consistent with previously identified risk profiles for adverse neonatal outcomes [10,11]. The fragility of the vessels in the germinal matrix at early GA and the instability of cerebral blood flow place preterm newborns at a higher risk for brain injury [10].

In the studied cohort, cesarean delivery appeared less frequent among infants with severe IVH or death; however, this association was not maintained after adjustment for other perinatal factors. Previous studies have reported conflicting findings regarding the protective role of cesarean delivery in extremely preterm infants [12–14]. While some suggest that cesarean delivery may reduce hemodynamic instability during birth, as vaginal delivery may expose fragile cerebral vessels to significant fluctuations in cerebral blood flow and oxygen saturation, others have shown that the association disappears after adjustment for GA and perinatal condition [13–16]. These findings suggest that the relationship between delivery mode and IVH risk is likely confounded by the clinical circumstances leading to preterm birth rather than reflecting a direct causal effect of delivery method.

While some risk factors for IVH are inherent to the unpredictability of preterm birth, many are modifiable through clinical practice. Of particular concern in the studied cohort were the low rate of antenatal corticosteroid administration (47%), high rates of inotrope use (25%), delivery room intubation (43%), and postnatal hypothermia (up to 88% within 72 h), all of which are strongly associated with increased risk for IVH and adverse neonatal outcomes [12–17].

Previous studies have demonstrated that the use of antenatal corticosteroids significantly reduces the incidence of IVH and improves survival in preterm infants by enhancing pulmonary and cerebrovascular stability [15–18]. As shown in a large population-based cohort from the California Perinatal Quality Care Collaborative (CPQCC), including 44,028 preterm infants, antenatal corticosteroid exposure was the only factor independently associated with a reduced risk of severe IVH across all GA groups [15]. Although infants exposed to a complete course of antenatal corticosteroids in the present cohort showed lower crude rates of severe IVH or death, this association was not maintained in adjusted analyses. This may reflect the relatively small number of severe IVH cases or residual confounding related to the timing of steroid administration and perinatal condition. Nonetheless, the relatively low rate of complete antenatal corticosteroid exposure observed in this cohort (47%) highlights an important opportunity for improvement in perinatal care.

The need for advanced resuscitation in the delivery room was strongly associated with severe IVH or death in the adjusted analysis. This association likely reflects the severity of cardiorespiratory compromise immediately after birth, which may lead to abrupt fluctuations in cerebral blood flow during the transitional period. Previous observational studies have similarly demonstrated that extensive resuscitation and multiple intubation attempts in the delivery room are associated with increased risk of severe IVH in very preterm infants [15,19]. These findings underscore the importance of gentle stabilization strategies and minimizing hemodynamic instability during the immediate postnatal transition.

Similarly, early hemodynamic instability also appears to play a central role in IVH pathogenesis [17]. In the adjusted models, the use of inotropes within the first 72 h of life remained independently associated with severe IVH or death. Fluctuations in cerebral blood flow and systemic blood pressure during the early neonatal period have been strongly implicated in the development of germinal matrix hemorrhage [19–21]. A previous prospective study of 497 extremely preterm infants (≤ 29 weeks) found that early inotrope use was associated with significantly increased odds of severe brain injury, including IVH of any grade [21].

Hypothermia was highly prevalent in the studied cohort, occurring in nearly half of the infants in the delivery room and in the majority during the first 72 h after admission. Maintaining normothermia in the immediate postnatal period is critical, as hypothermia is linked to impaired cardiorespiratory transition and increased morbidity in preterm neonates [22,23]. A systematic review including over 300,000 very preterm infants reported that admission hypothermia was associated with increased mortality and higher risk of IVH, bronchopulmonary dysplasia, retinopathy of prematurity, and sepsis [23]. Although hypothermia was not independently associated with IVH severity in these models, its high prevalence underscores the importance of improving thermal management practices in the delivery room and early NICU care.

Taken together, the present findings highlight urgent opportunities for improvement in perinatal and neonatal practices in Brazilian NICUs. Multiple studies have consistently demonstrated that evidence-based interventions significantly reduce the incidence of IVH in preterm infants [24,25]. A recognized strategy to mitigate these multifactorial risks is to implement care bundles [17,26–28]. Rather than focusing on single interventions, bundles combine multiple evidence-based practices and training across the perinatal continuum to maximize neuroprotection. A previously published quality improvement study in Canada has shown a substantial reduction in the use of inotropes, fluid boluses, and opioids. These changes were associated with a 69% decrease in IVH rates [17]. Increasing the uptake of these evidence-based practices has the potential to yield substantial improvements in survival without severe morbidity among the most vulnerable preterm newborns [4,10,14].

A major strength of this study lies in its prospective design and inclusion of both public and private hospitals in an LMIC. This approach allows for a more comprehensive and realistic representation of neonatal care across different healthcare settings in Brazil, capturing systemic challenges and institutional variability. The multicenter nature and standardized methodology enhance the external validity of the findings and support their application in guiding local and national strategies for IVH prevention in similar contexts. However, the study sample may not fully represent all regions of the country, as differences in NICU resources and clinical practices could affect generalizability.

Nonetheless, certain limitations must be acknowledged. First, as an observational study, causal inferences cannot be firmly established. Second, a significant proportion of infants (8.2%) did not undergo cUS, possibly due to early mortality or limited access to imaging in the first DOL, potentially leading to underestimation of IVH incidence. Third, although several variables traditionally considered risk factors for severe IVH did not reach statistical significance in multinomial analyses, the sample size and number of outcome events may have limited the statistical power to detect modest associations; therefore, these findings should be interpreted with caution. These limitations underscore the need for systemic improvements in education, clinical practice, and access to early neuroimaging in Brazilian NICUs to improve morbidity and mortality due to IVH.

In conclusion, this multicenter cohort study highlights the persistent burden of IVH among preterm infants in Brazilian NICUs and identifies key factors associated with severe IVH or death, including lower GA, advanced delivery room resuscitation, and early hemodynamic instability. These findings support the development and implementation of standardized IVH prevention bundles, focused on optimizing antenatal corticosteroid administration, thermoregulation, gentle resuscitation, and judicious hemodynamic management for preterm infants in Brazil. Future studies should prospectively assess the impact of such interventions on IVH incidence and neonatal outcomes in Brazilian NICUs.

This work was supported by the Stanford Center for Innovation in Global Health.

Data availability

The data that support the findings of this study are available from the corresponding author.