Long-term outcomes of very low birth weight infants with intraventricular hemorrhage: a nationwide population study from 2011 to 2019

Intraventricular hemorrhage (IVH) in preterm infants occurs when a germinal matrix hemorrhage ruptures into the lateral ventricle through the ependyma [1]. Grades 3 and 4 (G3 and G4) IVH increase the risk of long-term neurologic and neurodevelopmental disability, including seizures, cognitive and executive function impairment, and cerebral palsy (CP) [2‐4]. Advancements in neonatal intensive care, which have improved the survival rates of extremely preterm infants, have also resulted in a relative increase in the number of infants at high risk of developing IVH [5, 6]. The reported incidence of IVH in very low birth weight (VLBW) infants decreased from 50% in the early 1980s to around 20% in 2005 [7, 8]. However, more recent studies up to 2012 have reported a stable or increasing trend [3, 9]. According to data from the Korean Neonatal Network, the overall survival rate of the enrolled VLBW infants was 87% between 2013 and 2020. Although not statistically significant, the overall survival rate of preterm infants born in 2020 improved compared with that in the previous era [10]. The prevalence of overall IVH and IVH [≥ grade 2 (G2)] among VLBW infants was 42.5% and 17%, respectively. Among infants with a gestational age of less than 24 weeks, the prevalence of IVH (≥ G2) was 52% [11].

Posthemorrhagic ventricular dilatation (PHVD) is characterized by progressive ventricular enlargement and elevated intracranial pressure following hemorrhagic events. The incidence of PHVD with surgical intervention has remained stable nationwide and is reported to be 25%–50% in patients with IVH [12]. Although there is controversy regarding the impact and timing of neurosurgical interventions, direct removal or diversion of cerebrospinal fluid remains the primary approach to PHVD therapy [13]. Infants with severe IVH complicated by PHVD (40%–60%) are at significantly higher risk of poor neurodevelopmental outcomes, especially those who ultimately require a shunt (75%–88%) [14, 15]. Significant cognitive and motor impairments were found among infants with PHVD at 18–24 months’ corrected age [16‐19]. However, there are limited nationwide studies on PHVD.

A deeper understanding of the changing epidemiology of IVH and PHVD among VLBW infants may enhance the measurement of disease burden, allocation of public health resources, and development of clinical research objectives. Therefore, we undertook this study to determine the prevalence of IVH in VLBW infants and investigate the long-term neurodevelopmental outcomes associated with severe IVH over a 10-year period in Korea.

Over the past 10 years in Korea, there were 3805 VLBW patients with G2–4 IVH and 2038 ELBW infants. The prevalence rate gradually increased from 2010 to 2019, as illustrated in Fig. 1. Among these patients, 321 (10.5%) received treatment with a VP shunt or EVD. Specifically, 154 patients (4.7%) received only a VP shunt, 31 patients (0.9%) underwent only EVD, and 136 patients (4.0%) were treated with both a VP shunt and EVD. It is worth noting that a significant increase in VP shunt placements was observed in the year 2019.

The co-morbidities observed in these patients included HMD in approximately 3320 (87%), PDA in 796 (21%), NEC in 121 (3%), and PVL in 561 (15%). BPD and ROP exhibited significant decreases according to the results of the linear trend test. When dividing the 10-year period into two halves of 5 years each, the results indicated a significant decrease in BPD from 79 to 57%, sepsis from 47 to 34%, and ROP from 72 to 57% in the later time period. Additionally, an analysis of the long-term prognosis for children born between 2010 and 2017 revealed a declining trend in the prevalence of CP, autism, and seizure disorder over time (Table 1).

Among infants with IVH, an average of 92 patients (24%) were diagnosed with disabilities each year. The most prevalent disability diagnosis was cerebral lesion disorder, with an average of 66 patients (17%) registered annually, followed by intellectual disability in 15 patients (3%) and language disorder in eight patients (2%).

Over the 10-year period, it was confirmed that the prevalence of DD (P = 0.023), CP (P < 0.0001), epilepsy (P < 0.001), and visual impairment (P = 0.016) significantly decreased in patients requiring surgical treatment during the past decade (Fig. 2). Nevertheless, among complicated IVH patients who underwent surgical treatment (n = 404), a significantly higher incidence of HMD, BPD, PDA ligation, sepsis, NEC, PVL, and ROP was observed when compared with uncomplicated IVH patients who did not receive surgical treatment. Furthermore, patients who underwent surgery exhibited worse long-term neurological outcomes, including DD, CP, and seizure disorder, compared with the non-surgical group (P < 0.001) (Supplementary Table 1).

Figure 3 illustrates the percentage of children with IVH who experienced growth below the 10th percentile and abnormal developmental outcomes as assessed by the K-DST. A total of 24% of infants exhibited height (HT) below the 10th percentile, 23% had weight (WT) below the 10th percentile, and 23% had a head circumference (HC) below the 10th percentile. Additionally, 13% of infants required further work-up, while 31% required close observation. When dividing study periods into two groups based on a period of years, 2010–2013 (358 infants) and 2014–2017 (329 infants), a significant decrease in the number of infants with weight below the 10th percentile (28% to 15%, P < 0.001) and requiring further work-up (13% to 11%, P < 0.05) was observed (Fig. 3a). Among the 404 infants with IVH who underwent surgical treatment, a significantly higher incidence of growth below the 10th percentile was observed, including HT (31% vs 24%, P < 0.001), WT (36% vs 23%, P < 0.001), and HC (36% vs 21%, P < 0.001). These infants also exhibited a higher occurrence of abnormal developmental outcomes (13% vs 11%, P < 0.05) (Fig. 3b). When dividing the years into two halves and by surgical treatment, IVH patients who underwent surgical therapy during the earlier 4 years exhibited poorer growth and more, requiring further work-up (P < 0.001) (Fig. 3c).

We observed a consistent rise in IVH incidence over the past decade, despite a decrease in mortality among VLBW infants. Notably, both short-term outcomes (such as BPD and ROP) and long-term outcomes (including CP and seizure disorder) have shown improvement over this period in VLBW infants with IVH. However, among infants with IVH who underwent surgical treatment (10%), there was a tendency towards more severe co-morbidities and long-term sequelae, such as CP, DD, and seizure disorder. Additionally, these infants exhibited a higher prevalence of growth below the 10th percentile in HT, WT, and HC, as well as poorer developmental test results, compared with those without treatment. It is noteworthy that significant improvements were observed when dividing the study period into two halves of 4 years each. This study, encompassing almost all infants, as per information from the Korean Statistical Information Service, provides valuable insights into the nationwide epidemiology of IVH over the past decade.

Advancements in neonatal intensive care have led to improved survival rates for extremely preterm infants [23]. Stoll et al. argued that the effort to save more high-risk infants has resulted in an increased proportion of IVH. They reported that the decreasing trend of IVH rates, which reached 14% in 2002, either stabilized or showed a relative increase, reaching 16% in 2007 [5, 6]. The global incidence range of IVH G3 and G4 was reported to be 6%–22%, with variations by region (Europe: 6%–17%; North America: 11%–22%; Asia: 10%–14%; Oceania: 12%–13%) [14]. IVH G2 had an incidence range of 5%–19%. Among infants with a gestational age of less than 24 weeks, the incidence of IVH was close to 37%. Generally, the incidence of IVH was inversely related to gestational age. In our cohort, the overall prevalence of IVH (≥ G2) in VLBW infants was 12.1%. It is worth noting that Stoll et al. observed increasing trends in both BPD and IVH, whereas our cohort exhibited a decreasing trend in BPD [5]. This disparity may be attributed to the fact that their measurements were taken from a slightly earlier era, during which ventilator strategies for BPD and post-natal corticosteroid administration have evolved over the past decade [24, 25].

IVH is closely linked to increased mortality and neurodevelopmental abnormalities. Notably, neonates with IVH G3 and G4 are at high risk of experiencing long-term neurologic and neurodevelopmental disabilities, including seizures, cognitive and executive function impairment, and CP [2‐4]. In Korea, the mortality rate among VLBW infants stood at approximately 20%, with IVH-related mortality accounting for 7% of this total. Subsequent follow-up studies, conducted until the age of 22, unveiled those individuals with IVH faced a 4.2-fold increased risk of developing CP [11]. Furthermore, the risk of cardiovascular, respiratory, digestive, eye and ear issues, as well as malignant tumors, continued to rise, leading to a lifelong burden for patients [26‐28]. IVH patients exhibited a 2.68-fold increased risk of having an intelligence quotient (IQ) below 2 SD, a 4.45-fold increased risk of motor dysfunction, and a 2.91-fold increased risk of encountering difficulties in any academic skills (below 2 SD) [29‐31]. In this study, CP was diagnosed in 837 patients with IVH (28%), and DD was observed in 1095 patients (36%). While it is about known that CP occurs in about 20–30% of cases after IVH, long-term outcomes following ventricular drainage have not been extensively studied [32]. The data from a single-center cohort revealed that the mortality rate of IVH was 36.9%. Among the 77 surviving patients, the rate of surgical treatment after PHVD was 18%, and the incidence of CP was 55.9% (49.4%) [33].

In some cases, infants who are unable to tolerate a permanent shunt due to specific conditions may undergo a temporizing surgical procedure to divert cerebrospinal fluid until these issues are resolved. Published rates of conversion from a temporary device to a permanent shunt in this population vary, ranging from just over 50% to as high as 85% [16, 17, 34‐36]. In 2019, the Early Versus Late Intervention Study (ELVIS) Trial group in the Netherlands demonstrated the highest recorded conversion rate, at 26% [37, 38]. In this study, temporary devices were used in 1.2% of IVH VLBW infants, while permanent shunts were required in 4.8% of cases. The conversion rate from temporary to permanent devices was 42.3%.

Notably, the risk of a poor neurodevelopmental outcome is significantly higher when severe IVH is complicated by PHVD, with a prevalence ranging from 40 to 60%. Further risk is observed in infants who eventually require a shunt, with prevalence as high as 75% to 88% [14, 15]. In the pre-surfactant era, up to 82% of infants with PHVD who survived developed significant neurologic impairments, including CP [39]. In our study, CP was reported in 245 patients (76%) in the surgical intervention group, which was higher than the 77 patients (24%) in the non-surgical intervention group.

In our study, the K-DST identified 16 patients (12.8%) categorized as requiring “further work-up”. When dividing the study period into the first half and the more recent half, the percentage was 12.7% from 2010 to 2013 and 11.3% from 2014 to 2017. These findings suggest that improved neonatal care and high-risk infant follow-up care have led to a significant reduction in abnormal neurodevelopmental results. However, among patients who underwent surgical treatment, the percentage of those categorized as “further work-up” was 13%, compared with 11% of infants without treatment. This highlights the need for careful consideration and early intervention in such cases.

This study boasts several strengths, notably its nationwide scope, enabling a comprehensive epidemiological analysis of IVH and PHVD in VLBW infants across all live births. Additionally, the analysis included long-term growth and developmental screening data until the age of six years.

However, it is crucial to acknowledge certain limitations in this study. Due to the nature of national health claim data, access to individual patient data was not possible. The study design was observational, and patient inclusion relied solely on ICD-10 codes, which may have introduced labeling errors or missed diagnoses from various hospitals that could not be rectified. The surgically treated group can be considered as having PHVD; however, it is important to note that treatment approaches vary among different institutes, introducing an additional challenge in drawing conclusive findings. We can only make assumptions that patients who receive surgical treatment might represent a more complex spectrum of the disease. Controlling for the heterogeneity and subjectivity in measuring scales across various units and clinicians was not feasible in this retrospective observational study. Furthermore, recent studies have proposed distinct outcomes associated with periventricular hemorrhagic infarction, suggesting the implication of different etiologies, as described by Volpe et al., rather than the traditional description by Papile et al. [4, 39, 40]. The diagnosis was applied according to the ICD-10 code inputted by the hospital. It was not possible to distinguish which diagnostic criteria were used between the Papile and Volpe criteria. To further enhance our understanding, future data collection efforts should categorize cases according to the provided diagnostic classification. While the K-DST was used as a developmental screening tool, the Bayley Scales of Infant Development were not employed as a diagnostic tool. Additionally, the study did not assess the association between outborn birth status and mortality or morbidity. Lastly, the duration of follow-up data was relatively short, potentially leading to an underestimation of developmental disorders that typically manifest after the age of three.

In conclusion, despite advancements in current neonatal care, the neurodevelopmental outcomes of infants with IVH, especially those who undergo surgical treatment, continue to be a matter of concern. It is imperative to prioritize specialized care for patients receiving surgical treatment and closely monitor their growth and development after discharge to improve developmental prognosis.