Epidemiology and risk factors for thrombosis in children and newborns: systematic evaluation and meta-analysis

In newborns and children, thrombosis is often underdiagnosed [1]. Neonatal and pediatric arterial or venous thrombosis ranges between 2.6 and 6.4 cases per 100,000 per year [2, 3]. The incidence of thrombosis in the pediatric population is highest in newborns. Thromboses tend to occur in very sick neonates, particularly preterm neonates [4]. pediatric thrombosis has occurred more often in intensive care units(ICU) [5]. After neonatal thrombosis has received attention, case–control studies and cohort studies have gradually increased [6], but the results of different studies on the factors affecting thrombosis are different [7].

In recent years, significant progress has been in research on the influencing factors of thrombosis in children [8]. Studies have shown that several factors contribute to the incidence of thrombosis in this population, including genetic predisposition, underlying medical conditions, and environmental factors [9]. However, the level of the studies varied, and the accuracy of the individual results needs to be further determined [10, 11]. For example, the latest large cohort study found differences in risk factors for venous versus arterial thrombosis in neonates and central access device characteristics with neonatal thrombosis in ICU [8]. Whether male and female, ethnicity, antenatal maternal history, central venous catheters, catheters of different sites, or catheter length are all factors contributing to thrombosis in children and newborns is controversial [12]. And as the incidence increases, thrombosis in children is no longer rare, and its incidence in ICU needs to be further determined [13].

Currently, there are no meta-analyses of the overall factors associated with thrombosis in the ICU. The available meta-analyses of thrombosis versus catheter and thrombophilia only explored a single factor of ductal [14] or thrombophilia [15]. However, the risk factors for thrombosis have not been conclusively determined. This study aimed to identify the risk factors for thrombosis in children and neonates, allowing clinicians to develop more effective preventative and therapeutic strategies for reducing its incidence.

This systematic review and meta-analysis were conducted by the Preferred Reporting Program for Systematic Reviews and Meta-Analyses (PRISMA) statement. The study protocol was prospectively registered (PROSPERO ID: CRD 42022333449).

We use meta-analysis method for the first time to explore the ranking of factors that influence the incidence of thrombosis in ICU. This meta-analysis summarized the incidence of thrombosis in ICU children to be approximately 2%. We found that the most significant influencing factor for thrombosis in children and newborns was CVC (OR = 3.66), followed by Surgery (OR = 2.25), mechanical ventilation (OR = 2.1), Infection/sepsis (OR = 1.95), gestational age (OR = 1.5), Respiratory distress (OR = 1.39), and different ethnicities (OR = 0.88). Next, we will discuss based on the ranking of influencing factors and analyze the incidence of thrombosis in ICU children.

The incidence rate of ICU thrombosis is 2%, which is consistent with previous studies, but higher than the incidence rate of ordinary ward thrombosis reported in other studies [34]. The incidence of neonatal thrombosis tends to increase annually and is strongly associated with the placement of catheters [35]. In the Pediatric Health Information System database, the incidence in neonates was 44 per 10,000 admissions between 2001 and 2007 [36]. It had increased to 75 in 2014 (70% increase) [36]. CVC is one of the most important influencing factors, which is consistent with previous findings of CVC on thrombosis. Previous studies found that one in five children with CVCs experienced CVC-related deep vein thrombosis (DVT) [14]. Nearly all thrombosis-related deaths are associated with CVCs and at least 85% of thrombosis are related to CVCs [37]. At present, a large-scale multi-Centre RCT experiment is needed to detect the effective prevention of CVC-induced thrombosis. Surgery is a very important influencing factor, and there is no meta-analysis to discuss the effect of surgery on pediatric thrombosis. In newborns, due to the immaturity of the coagulation and inflammatory response system, when damaged, due to its hypercoagulable state, as well as the use of adult plasma in the newborns, there is a predisposition to severe postoperative bleeding and thrombosis and systemic inflammatory response [38, 39]. Increased probability of thrombosis after organ transplantation and pediatric patients undergoing liver transplantation are inherently prone to thrombosis [40]. Mechanically ventilated patients may develop thrombosis, which is a consensus in previous studies [41]. Some studies have demonstrated a high prevalence of deep vein thrombosis in mechanically ventilated COVID-19 patients [42]. Mechanical ventilation alters the negative thoracic pressure state and decreases venous blood return and cardiac output, most likely leading to lower limb venous thrombosis [43]. In neonates, lower limb blood return is maintained mainly by negative pressure in the thorax, contraction of calf muscles and venous valves to prevent blood backflow [44]. It is advisable to strictly test coagulation and administer anticoagulation early in the postoperative period and in mechanically ventilated children or neonates to prevent thrombosis and reduce mortality [45]. For infection and sepsis, Jacques in a large prospective cohort of 2,305,380 adults undergoing surgical procedures in 374 hospitals of all types in the United States from 2005–2012 [46], found that in all surgical procedures, patients with preoperative systemic inflammatory response syndrome or any sepsis were three times more likely to have postoperative arterial or venous thrombosis. This is consistent with the findings of this paper. Respiratory distress is an important influencing factor. In child respiratory distress, research finding thrombomodulin increases, increasing the probability of thrombosis [47]. Respiratory distress syndrome is also often combined with mechanical ventilation, significantly improving the prognosis of children with respiratory disease by maintaining adequate gas exchange [42]. Gestational age is also an influence on thrombosis, and newborns born at term have a more developed coagulation system and are less likely to develop thrombosis than preterm infants [48, 49]. However, gestational age, as a factor affecting neonatal thrombosis, had less significant effects on thrombosis in children and adults [50]. Moreover, no clinical studies of gestational age as an influencing factor in thrombosis studies in children and adults. Prospectively investigated the incidence of cerebral sinus veins in infants. They found that for infants born at less than 29 weeks gestational age, systemic serial cranial US showed a very high incidence of CSVT at 4.4%. Prematurity remains the most significant risk factor for developing germinal matrix-intraventricular haemorrhage (GM-IVH) [49]. The development of the haemostatic system in neonates is an age-dependent process, with low levels of procoagulant and anticoagulant factors in preterm infants [51]. Different ethnicity also affects the incidence of neonatal thrombosis. Some studies suggest that this is because the number of NICUs varies between countries [52]. The vast differences in NICU and ICU ratios and intensive care services also lead to differences in outcomes after treatment of preterm infants, post-operative children, and other patients admitted to the ICU, resulting in different morbidity rates [2].

The study results present a baseline value for the incidence of thrombosis in children in ICU units. In addition, the included studies included a wide range of populations and different sites of thrombosis, giving incidence data for a wide range of thrombosis.

A strength of this study is the inclusion of conference articles and articles by the same authors studying different populations in different years, the uniformity of statistical methods and study methodology, the high quality of the included studies, the number of people included at 1,371,608, the separate analysis of thrombosis in children and neonates in the ICU rather than the inclusion of patients from all wards, and the relevance of the study. This paper is the first to examine all factors other than catheterization factors in thrombosis in children. Few studies have examined the factors influencing thrombosis in children and neonates in the ICU and the incidence. The literature analysis shows that although there are fewer published articles and fewer clinical trials related to thrombosis factors in the pediatric ICU, the sample size is large, the results of case studies and cohort studies are definitive, and the relevant influencing factors are uniform. At the same time, the conclusions of this paper are reliable as determined by sensitivity analysis and the cut-and-patch method and have some clinical guidance value.

There are some limitations to this study. Our search strategy may have missed some studies that were not published or not found. These cohorts may differ from the clinical characteristics of patients outside the clinical trial setting (ICU). Some important influencing factors were not analyzed because there was only one literature or no literature provided data. Therefore, we did not discuss the following influencing factors: inherited/congenital thrombotic disorders, premature neonates or not, and the influence of maternal pregnancy complications. This may ignore the role of these important factors. The Robinson [28] report, which analyzed 1,158,755 infants and was published in 2021, had a significant bias in the incidence study because it included many patients and took a retrospective cohort analysis with reliable outcomes. Finally, other possible limitations include potential interactions between thrombosis and other factors (e.g. year of publication, country), For limitations arising from the inclusion of both prospective and retrospective analyses and There is little evidence for significant effect estimate differences between Retrospective observational studies and Prospective observational studies, regardless of specific observational study design, heterogeneity, or inclusion of studies of pharmacological interventions [53].

In summary, the incidence of thromboembolism using current rates is approximately 1% to 2% in the ICU. Influencing factors include CVC, Surgery, mechanical ventilation, Infection/sepsis, gestational age, Respiratory distress, and different ethnicities. Reference to the values and incidence of influencing factors can be used as a basis for ICU management and treatment of thrombosis, as well as alerting the clinic to what influencing factors need to be noted for early prophylactic treatment of different patients.