Short-Term Blood Transfusion Outcomes in Preterm Infants admitted to Neonatal Intensive Care Unit (NICU): A Retrospective Analytical Study

Acknowledgements

We would like to thank all our neonates together with their parents and all the staff members (physicians and nurses) at Cairo University Children Hospitals.

Author's contributions

NG: Conceptualization; Investigation and writing the original draft. AA: Data curation; validation; software writing, review & editing. NG and AA:  Writing - review & editing. All authors have read and approved the final manuscript.

Conflict of interest

The author has no conflict of interests to declare.

Funding

This study received no special funding and was totally funded by the author.

Date received: 26^th March 2021, accepted 2^nd May 2021

Introduction

The improvement of neonatal intensive care has increased the survival rate among tiny premature infants [1], leading to a rise in the rate of RBC transfusions [2]. RBC transfusion increases tissue oxygenation through higher circulation of hemoglobin. The tinier the premature infants are, the more prone they are to cardiopulmonary compromise, frequent phlebotomies, subsequent anemia, and frequent RBC transfusions [3-6].

The guidelines for blood transfusion in the Neonatal Intensive Care Unit (NICU) are generalized and subjective [7,8]. Symptomatic anemia occurs and affects the clinical condition when the imbalance between oxygen delivery and consumption is observed; the hemoglobin is not at the standard level and varies with every preterm [9]. The decision for transfusion is made by the neonatologist, who expects the benefits to outweigh the risks. However, these decisions are not always evidence-based and highly dependent on experience and local guidelines [10].

In the previous two decades, the evidence showed the adverse effects of frequent RBC transfusions, including the risk of mortality, multi-organ system failure, and prolonged hospital stay. Preterm infants were more prone to tissue injury due to impaired tissue oxygenation. Neonatal studies compared RBC transfusion with the incidence of short-term outcomes such as BPD, NEC, ROP, and ICH [11-17].

This study was conducted to eradicate the inconsistency found so far and to elucidate the relationship between RBC transfusion and short-term morbidities and outcomes in preterm infants with gestation of less than or equal to 32 weeks and weighing less than or equal to 1500 grams. This study was conducted in line with the standard RBC transfusion protocol.

Methods

Inclusion Criteria: This research centered on infants having a gestational ages of less than or equal to 32 weeks and weighing less than or equal to1500 grams. They were admitted to Cairo University Children Hospital (Aburreesh El Mounira) from the 1^st of March 2018 to the 29^th of February 2020.

Neonatal data included gestational age, birth weight, sex, mode of delivery, and multiple gestations.

Maternal history included premature rupture of membranes, preeclampsia, gestational diabetes mellitus, and administration of antenatal medications like steroids.

Exclusion criteria: Infants who died during the first month and newborns with congenital anomalies.

RBC transfusion was implemented on a clinical basis and in line with the transfusion guidelines as stated in the Manual of Neonatal Care [18]. Packed RBCs (PRBC) were transfused at a dose of 15 mL/kg over two to three hours. The RBC units were filtered and irradiated. Feeding was skipped before and after RBC transfusion. During the study period, delayed cord clamping was rarely performed as many obstetricians did not follow the guidelines for delayed cord clamping.

Neonatal morbidities included respiratory distress syndrome (RDS), bronchopulmonary dysplasia (BPD), pneumothorax, patent ductus arteriosus (PDA), early-onset sepsis (EOS), late-onset sepsis (LOS), necrotizing enterocolitis (NEC) (stage≥2), intracranial hemorrhage (ICH) (grade ≥1), retinopathy of prematurity (ROP) and infectious episodes such as pneumonia, CNS infections, urinary tract infections, wound infections, infective endocarditis, myocarditis, septic arthritis and omphalitis. Hospital duration, oxygenation, total parenteral nutrition (TPN) duration, inotropes, and inhaled steroids were recorded.

RDS was diagnosed based on characteristic clinical symptoms and chest X‑rays. BPD diagnosis and severity were carried out depending on the severity-based definition of BPD by the National Institute of Child Health and Human Development (NICHD) [19]. At the time of assessment, infants with no oxygen need were considered as mild BPD. Moderate BPD was considered for cases requiring less than 30% oxygen while severe BPD was considered for cases needing positive pressure and/or oxygen support ≥30% [19].

Sepsis was diagnosed when a positive blood culture and systemic symptoms were present. Early-onset sepsis was defined as sepsis occurring at less than seven days and late-onset sepsis as sepsis occurring seven days or more after birth. NEC was diagnosed based on systemic symptoms and radiographic findings, with severity based on the staging criteria modified by Bell et al. [20]. Moreover, PDA was diagnosed by clinical symptoms and echocardiography. ROP was established by a neonatologist and confirmed by an ophthalmologist following a schedule of the American Academy of Ophthalmology [21]. Furthermore, ICH was identified via brain ultrasound [22–24]. An initial brain ultrasound was performed within 72 hours of birth and rescreening was performed at intervals of one to four weeks. In cases of abnormal brain ultrasound or neurological symptoms, brain MRI was performed prior to discharge.

In the RBC transfusion group, the number of transfusions, age, and Hb (Hct) level at time of the first transfusion were further reviewed. The number of PRBC transfusions within seven days (Group1) and 30 days of life (Group2) were recorded for each patient.

Ethical consideration

This study’s protocol has been approved by the Ethics Committee of the Faculty of Medicine, Cairo University, and complies with the provision of the Declaration of Helsinki in 1964 and its later amendments or comparable ethical standards. Informed written consent has been obtained from the parentsof each child before their enrollment in this study.

Statistical analysis

Data analysis was performed using the IBM SPSS program version 21. Quantitative variables were described as mean, standard deviation (SD), median, and interquartile range (IQR) while qualitative variables were defined as number and percentage. A Chi-square test was employed to compare qualitative variables; the Pearson correlation was used to test linear relations between variables. A P‑value less than or equal to 0.05 was considered significant, and less than or equal to 0.01 was deemed highly significant.

Results

During the two-year study period, 161 premature infants with less than or equal to 32 weeks of gestation and less than or equal to 1500 gm were admitted to our NICU. There were 91 males and 70 females. Table 1(a, b, and c) displays that the (a) average gestational age was 30.8±1.2 (mean± standard deviation); (b) average birth weight in grams was 1287±144.1; (c) average admission period in days was 44.4±16.4; (d) average duration of oxygen in days was 25.3±14.3; (e) average duration of caffeine citrate was 32.4±16.5; and (f) average duration of mechanical ventilation, CPAP, oxygen blender, head box, and incubator oxygen was 16.6±16.6, 9.8±7.5, 5.7±3.2, 2.7±1.4, 2.8±1.9, respectively.

Table 1 also shows that the average duration of TPN in days was 17.7±11.7; the average onset of trophic feeding in days was 2.3±1.3; the average intake of inotropes duration in days was 12.4±7.6; and average inhaled steroids duration in days was 17±12.6. Furthermore, the average Hgb and Hct on admission were 15± 2.7 and 43.9±8.1, respectively; while at the time of the first transfusion, the average Hgb and Hct were 9.5±1.6 and 27.3±4.3. Additionally, the average blood volume transfused was 54.9±30.2; the average infectious episodes were 1.9±1.2; and the average maternal age in years was 26.8±4.

There were 139 RDS cases (86.3%), 155 VLBW infants (96.3%), 6 ELBW cases (3.7%), 29 PROM cases (18%), and 41 multiple gestation cases (25.5%). Antenatal steroid use was found in 45 cases (28%), and oxygen was administered for 140 neonates in the form of CPAP, blender, mechanical ventilation, head box, and incubator oxygen (124 (77%), 127 (78.9%), 58 (36%), 113 (70.2), 126 (78.3%), respectively). TPN was administered in 141 cases (87.6%) while inotropes were given in 83 cases (51.6%). PDA was diagnosed in 57 cases (35.4%) and was treated in 55 cases (34.2%).

Early onset sepsis (EOS) developed in 65 cases (40.4%) while late onset sepsis (LOS) appeared in 126 cases, accounting for 78.3%. There was also NEC in 34 cases (21.1%), ICH in 45 cases (34.2%), jaundice in 92 cases (57.1%), ROP in 102 cases (63.4%), BPD in 60 cases (37.3%), and pneumothorax in 36 cases (22.5%). Forty‑five cases (28%) received surfactant. Forty-three cases received blood in the first week accounting for 26.7%, and 86 cases in the first month (53.4%). Thirty-five cases (21.7%) were delivered by vaginal delivery and 126 cases (78.3%) by caesarian section. Maternal illnesses documented in the records were 11 DM (6.8%) and 19 preeclampsia (11.8%) as shown in Table 1.

Blood transfusion in the first week was predominantly associated with the incidence of NEC, ICH, ROP and its degrees of severity as well as with BPD and its degree of severity, regardless of the number of transfusions. However, the outcome was significant in addition to the number of transfusions, as shown in table 2 and table 3.

In the first month, blood transfusion was significantly correlated with the incidence of NEC, ICH, and ROP and its degree of severity. Blood transfusion was also substantially related to the incidence of BPD although no statistical significance with the severity of BPD was found. The number of transfusions in the first month was considerably linked to the incidence and severity of BPD and the outcome as shown in table 4 and table 5.

As shown in table 6, a moderate positive correlation was identified between the duration of oxygen in days, the duration of mechanical ventilation, the duration of TPN, the use of inhaled steroids, and infectious episodes and the number of transfusions in the first month. Moreover, there was a weak positive correlation between the duration of inotropes intake and the number of transfusions in the first month.

Additionally, in the first week, there was a moderate positive correlation between the number of transfusions and the duration of mechanical ventilation and a moderate negative relationship between the number of transfusions and CPAP duration. The number of transfusions in the first month showed a moderate negative correlation to the birth weight and gestational age. Age at time of the first transfusion showed a moderate positive correlation to birth weight and a weak positive correlation to gestational age in weeks as shown in table 7.

Table (8) compared the incidence of morbities in relation to first week (Group 1) and month (Group 2) transfusions

Figures (1-4) described the incidence of individual morbidity in relation to first week and month transfusions

Discussion

The study revealed that RBC transfusion in the first week of life in preterm infants whose gestational age was less than or equal to 32 weeks and whose weight was less than or equal to 1500 grams was associated with an increased risk of death. This finding is consistent with the results provided by Dos Santos et al. [11] and Wang et al. [25]. This can be explained by pro-inflammatory mechanisms and is considered to be one of the associations of transfusion-related immunomodulation according to the study of Vamvakus and Blajchman [26].

Blood transfusion during the first week and month was significantly associated with the incidence of NEC, ICH, and ROP along with its severity, regardless of the number of transfusions; mortality was significant in addition to the number of transfusions. Moreover, blood transfusion in the first week was largely connected to the incidence and severity of BPD, and only to the incidence of BPD at one month, regardless of the number of transfusions. However, the number of transfusions only became significant at one month. Both the incidence and severity were significant for the number of transfusions.

Regarding NEC, this research agreed with those reported by mny other studies [12, 13 & 27], but contradicts the studies of Wang et al. [25] and Wallestein et al. [28]. Severe anemia causes a reduction in mesenteric blood flow and eventually hypoxia. The re-oxygenation induced by blood transfusion causes reperfusion injury [29-32].

Concerning ICH, this study concurs with the researches [3, 16] but disagrees with other reports [8, 27]. The main cause for development or progression of ICH in relation to transfusion is unclear and needs further future studies. Though, it might be related to fluctuations in blood pressure in relation to transfusion, coagulation defect or might be the reason or related to the initial reason for transfusion.

With BPD, this paper’s outcome coincided with the findings of many studies [7, 8, 27, 32 & 33] but opposes Wang et al.’s results [25]. Transfusion increases oxidative injury caused by, an increase in non-transferrin bound iron, and inflammatory mediators present in stored blood products.

In reference to ROP, this research was consistent with other reports [9, 14, 15 & 33] who reported a correlation between the grades of ROP (more than or equal to two) and transfusions in ELBW and VLBW infants (more than or equal to two); however, this study opposes the findings of others reports [4, 7 & 27].

Blood transfusions increase the risk of ROP by two mechanisms: (a) by increasing retinal oxygen supply and (b) by increasing oxygen free radicals through free iron overload. Despite these explanations, there are conflicting studies that compare ROP and anemia [10, 34 & 35].

Regarding mortality, RBC transfusions in the first week was essentially related to death as shown in the research of Dos Santos et al. [11] and Wang et al. [25].The reason for mortality in relation to first week transfusion is unclear and may be related to the entity of prematurity itself.

Vamvakus et al. [26] suggested that transfusion might be associated with multi-organ system failure, pro-inflammatory mechanisms, and transfusion-related immunomodulation. These explanations may not explain fully the mortalities in this study as neonates included, died long time after the first week transfusions and even more than one month of age. This finding necessitates further prospective studies to detect the reason behind first week transfusions and 100% mortality as is the case in our study.

From the results of this study we recommend the following: It is of the utmost importance to limit the number of samplings in the preterm infants as they are the most common cause of RBC transfusion. Emphasis should be made on the use of a delayed cord clamping policy and early enteral feeding together with iron supplementation. Moreover, transfusions in the first week should be limited to infants designated as extremely preterm. Further research is needed to establish comprehensive guidelines specifying the timing of transfusions in relation to development and severity and calling for close follow-up of the aforementioned morbidities.

Limitations: The limitations of this retrospective study are linked to the improper recording of data that has impeded the achievement of better results. Moreover, several factors, including all the complications of prematurity other than those studied such as hypothermia, neonatal sepsis and nosocomial infections demonstrated the same outcome and could not be properly evaluated, except through statistical analysis.

Conclusions

RBCs transfusion should be limited to the extremely indicated preterm infants especially in the first month of life with special emphasis on the first week with close follow up of transfusion associated morbidities like NEC, ROP, BPD and ICH.