Abstract
Background
Studies aimed at reducing neonatal anaemia or transfusing higher blood volumes did not find improvement in neurodevelopmental function at two years of age. This study investigated the relationship between the receipt, timing, and number of red blood cell (RBC) transfusions and neurodevelopmental outcomes among preterm infants.
Materials and methods
This is a retrospective review of preterm infants (gestational age <34 weeks) with a full neurodevelopmental assessment at 18–36 months corrected age from October 2008 to September 2020. Bayley Scales of Infant and Toddler Development, third edition and the Modified Checklist for Autism in Toddlers were collected. Multivariable regressions were used to evaluate neurodevelopmental outcomes.
Results
654 preterm infants were evaluated with a mean follow-up of 25 months. 295 infants (45%) received a total of 1,322 blood transfusions. After adjustment for gestational age, baseline morbidity, and socioeconomic status, receipt of RBC transfusion was associated with decreased two-year cognitive and motor function, but not language (p=0.047, 0.025, and 0.879, respectively). There was no significant difference in outcomes between receipt of transfusion in the first week of life compared to after. Number of transfusions was associated with decreased cognitive, language, and motor function (all p<0.001), and increased likelihood to develop severe neurodevelopmental impairment (adjusted-odds ratio, 1.09; 95% confidence interval, 1.03–1.15; p=0.004).
Discussion
Our study demonstrates an association between RBC transfusion and lower cognitive and motor outcomes at two-years after adjustment for prematurity and illness at birth. Increasing number of transfusions worsened neurodevelopmental outcomes.
Keywords: Bayley, premature, newborn, neonate, impairment
Introduction
Up to 90% of extremely low birth weight infants receive at least one red blood cell (RBC) transfusion during their neonatal intensive care unit (NICU) hospitalisation1–3. RBC transfusions are needed to offset blood loss with lab draws to prevent anaemia associated with poor feeding and breathing problems. Despite these benefits, there are risks of infection, extension of intraventricular haemorrhage (IVH), retinopathy of prematurity, bronchopulmonary dysplasia, and death4. Long-term outcomes of infants that receive RBC transfusion are limited.
Few studies have examined the neurodevelopmental outcomes of infants who received RBC transfusion in the NICU. Studies aimed at reducing neonatal anaemia5–7 or transfusing higher blood volumes8 did not find improvement in neurodevelopmental function at two years of age. Our primary aim was to examine the relationship between receipt of RBC transfusion in preterm infants and their subsequent neurodevelopment at two years of age. We also conducted secondary analyses of: 1) cumulative number of transfusions, 2) transfusions received in the first week of life, and 3) transfusions in a subcohort of extremely preterm infants with a gestational age <28 weeks.
Materials and methods
A retrospective review was conducted of preterm infants (gestational age <34 weeks) with a full neurodevelopmental assessment at 18–36 months corrected age from October 2008 to September 2020. Since lower gestational age infants are more likely to receive a blood transfusion and have increased neurodevelopmental impairment (NDI), all analyses were adjusted for gestational age. Infants with major congenital abnormalities or missing follow-up data were excluded. This study was approved by the institutional review board of Sharp Mary Birch Hospital for Women & Newborns.
Bayley Scales of Infant and Toddler Development, third edition and the Modified Checklist for Autism in Toddlers were collected. Assessments were performed in-person by certified neurodevelopmental specialists. Moderate-severe NDI was defined as: 1) any Bayley cognitive, language, or motor composite score below 85 (i.e. >1 standard deviation [SD] below the mean); 2) any vision or hearing impairment; and/or 3) an abnormal neurological exam, which is suggestive of cerebral palsy. Severe NDI was defined as: 1) any Bayley cognitive, language, or motor composite score below 70 (i.e. >2 SDs below the mean); 2) bilateral legal blindness or deafness; and/or 3) an abnormal neurological exam. Vision impairment was defined as any loss of visual acuity or control of eye muscles; legal blindness was defined as visual acuity of ≤20/200 with the best conventional correction. Hearing impairment was defined as any hearing loss as determined by a sound field hearing assessment or auditory brainstem responses; legal deafness was defined as hearing loss unresponsive to amplification. Abnormal neurological exam was defined as hyper- or hypotonicity in at least one arm or leg, associated with deficient control over movement or posture.
Day of transfusion and number of packed RBC transfusions were collected from the newborn record. Packed RBCs were transfused at a volume of approximately 15 mL/kg. Neonatal outcomes during initial hospitalisation were collected. A comparison of early exposure (first week of life) vs late (after 7 days of life until discharge for first transfusion) exposure was also performed. Baseline and clinical data between RBC transfused and non-transfused newborns are shown in Table I. Placental transfusion was assessed as delayed cord clamping ≥30 seconds or umbilical cord milking (milking the intact cord). The receipt of inotropes included dopamine, dobutamine, and/or epinephrine. Bronchopulmonary dysplasia was defined as receipt of any supplemental oxygen or nasal cannula flow greater than 2 liters per minute at 36 weeks corrected age. Haemoglobin (Hb) was recorded for three separate occasions: 1) at NICU admission, 2) prior to the first RBC transfusion (within 72 hours), and 3) nadir Hb during NICU stay. Completion of prenatal care, highest level of maternal education completed, and infants’ insurance type at 2-year follow-up were used to assess socioeconomic status9.
Table I.
Baseline and clinical characteristics
| Characteristics | Total (n=654) | No transfusion (n=359) | Transfusion (n=295) | p-value |
|---|---|---|---|---|
| Pubic Health Insurance, n. | 45% (282) | 43% (146) | 50% (136) | 0.086 |
| Maternal highest education completed: bachelor's degree | 82% (345) | 79% (182) | 86% (163) | 0.052 |
| Maternal age, yrs, mean ± SD | 31 ± 6 | 31 ± 6 | 30 ± 6 | 0.001 |
| No prenatal care | 3.0% (19) | 2.0% (7) | 4.1% (12) | 0.115 |
| Hispanic ethnicity | 47% (276) | 46% (150) | 48% (126) | 0.748 |
| Received ≥ 1 dose of antenatal steroids | 91% (583) | 91% (321) | 91% (262) | 0.876 |
| Rupture of membranes >18 hours | 23% (141) | 24% (81) | 22% (60) | 0.561 |
| Cesarean delivery | 78% (510) | 77% (277) | 79% (233) | 0.575 |
| Chorioamnionitis, histological | 29% (165) | 22% (71) | 37% (94) | <0.001 |
| Placental abruption | 7.8% (51) | 5.3% (19) | 10.8% (32) | 0.008 |
| Pre-eclampsia | 24% (156) | 26% (93) | 22% (63) | 0.178 |
| Maternal diabetes | 18% (117) | 18% (65) | 18% (52) | 0.874 |
| Female, n. | 47% (307) | 46% (166) | 48% (141) | 0.691 |
| Gestational age, wks, mean ± SD | 29.2 ± 2.7 | 30.8 ± 1.8 | 27.2 ± 2.2 | <0.001 |
| Gestational age <28 weeks | 32% (215) | 7.1% (27) | 63% (188) | <0.001 |
| Birth weight, g, mean ± SD | 1279 ± 468 | 1549 ± 413 | 1010 ± 446 | <0.001 |
| Small for gestational age * | 12% (78) | 9% (32) | 16% (46) | 0.009 |
| Singleton birth | 71% (481) | 69% (246) | 74% (218) | 0.132 |
| 5-minute Apgar score ≥7 | 89% (575) | 95% (337) | 82% (238) | <0.001 |
| DCC or UCM | 83% (254) | 85% (142) | 81% (112) | 0.247 |
| Length of stay in NICU, d, median (IQR) | 53 (35–83) | 37 (26–49) | 87 (66–111) | <0.001 |
| Hb at NICU admission, g/dL, mean ± SD | 15.7 ± 2.3 | 16.6 ± 2.1 | 14.7 ± 2.1 | <0.001 |
| Nadir Hb in the NICU, g/dL, mean ± SD | 10.4 ± 2.5 | 11.7 ± 2.6 | 8.8 ± 0.8 | <0.001 |
| Culture positive sepsis | 7.6% (50) | 2.5% (9) | 13.9% (41) | <0.001 |
| Peak serum bilirubin, mg/dL, mean ± SD | 8.5 ± 2.6 | 9.4 ± 2.4 | 7.4 ± 2.4 | <0.001 |
| Days of phototherapy, median (IQR) | 3 (0–4) | 3 (1–4) | 3 (0–4) | 0.701 |
| PDA requiring treatment | 24.8% (161) | 4.8% (17) | 49.1% (144) | <0.001 |
| ROP requiring avastin or surgery | 4.9% (31) | 0% (0) | 10.7% (31) | <0.001 |
| Bronchopulmonary dysplasia | 19.9% (130) | 3.3% (12) | 40.0% (118) | <0.001 |
| IVH ≥ grade 3 | 2.6% (17) | 1.1% (4) | 4.4% (13) | 0.009 |
| Received inotropes | 17.0% (111) | 3.1% (11) | 33.9% (101) | <0.001 |
| Highest level of care in delivery room | ||||
| Supplemental oxygen | 77% (501) | 75% (269) | 79% (232) | 0.264 |
| CPAP | 76% (497) | 85% (305) | 65% (192) | <0.001 |
| Positive pressure ventilation | 58% (381) | 43% (154) | 77% (227) | <0.001 |
| Intubation | 42% (276) | 25% (88) | 64% (188) | <0.001 |
| Respiratory care in the NICU | ||||
| Received surfactant in DR or NICU | 60% (394) | 41% (146) | 84% (248) | <0.001 |
| Supplemental oxygen at 36 weeks | 18.3% (117) | 2.6% (9) | 37.4% (108) | <0.001 |
| High-flow nasal cannula at 36 weeks | 6.7% (42) | 0.9% (3) | 13.7% (39) | <0.001 |
| CPAP at 36 weeks | 6.1% (39) | 0.6% (2) | 12.8% (37) | <0.001 |
| Mechanical ventilation at 36 weeks | 2.0% (13) | 0.3% (1) | 4.2% (12) | 0.001 |
| Days on mechanical ventilation, median (IQR) | 1 (0–4) | 0 (0–1) | 5 (2–24) | <0.001 |
Birth weight of less than the 10th percentile for gestational age using intrauterine growth curves by Olsen et al.20
CPAP: continuous positive airway pressure; DCC: delayed cord clamping; DR: delivery room; Hb: haemoglobin; IQR: interquartile range; IVH: intraventricular haemorrhage; NICU: neonatal intensive care unit; PDA: patent ductus arteriosus, ROP: retinopathy of prematurity; SD: standard deviation.
Means ± SDs and medians with interquartile ranges were used for parametric and nonparametric data, respectively. Student’s t-test, Mann-Whitney U, Kruskal-Wallis, or Chi-squared tests were used as appropriate. Bivariable and backwards-stepwise multivariable regression models were performed to adjust for socioeconomic status as well as the greater prematurity and higher morbidity of infants in the transfusion group. Logistic and linear regression models were used to predict dichotomous and continuous outcomes, respectively. Odds ratios (OR), β-coefficients, and 95% confidence intervals (CI) were calculated. Statistical significance was defined as a two-sided p-value <0.05.
Results
Over the 12-year study period, 654 high-risk preterm infants were evaluated. For this overall cohort, 307 (47%) were female, 78 (12%) were small for gestational age, and 575 (89%) had a 5-minute Apgar score ≥7 (Table I). 295 infants (45%) received a total of 1,322 blood transfusions. Fifty percent (146/295) received a transfusion within the first week of life, the median number of transfusions given was 3 (1–7) units, and the median cumulative transfusion volume per birthweight was 86.0 (16.3–130.9) mL/kg. Infants who received ≥1 transfusion had a lower mean gestational age (27.2 vs 30.8 weeks) and Hb at NICU admission (14.7 vs 16.6 g/dL) compared to those who did not (Table I). The mean Hb prior to their first transfusion was 10.2 (±1.6) g/dL. After adjustment for gestational age, transfused infants were still more frequently diagnosed with bronchopulmonary dysplasia, retinopathy of prematurity, IVH ≥ grade 3, culture positive sepsis, and requiring inotropes during hospitalisation. Maternal characteristics were similar between transfused and non-transfused infants, except for chorioamnionitis and placental abruption, which were more likely with the transfusion group. Proxies of socioeconomic status –no prenatal care, highest maternal education completed, and public health insurance– were also similar between groups.
Moderate-severe NDI was higher in the transfusion group compared to the no transfusion group (58 vs 47%, p=0.003), as was severe NDI (21 vs 15%, p=0.028; Table II). In bivariable analyses, both receipt of transfusion and number of transfusions (per 1-unit increase) were associated with nearly all neurodevelopmental impairments, as well as the following: gestational age, female, haemoglobin at NICU admission, IVH ≥ grade 3, receipt of inotropes, retinopathy of prematurity requiring avastin or surgery, patent ductus arteriosus requiring treatment, and receipt of prenatal care (Table III, only NDI shown as outcomes). These factors were used in the multivariable analyses to control for socioeconomic status and baseline severity of illness. In multivariable analysis, receipt of RBC transfusion was not associated with moderate-severe NDI, severe NDI, or autism (p=0.897, 0.130, 0.108, respectively) after adjusting for socioeconomic status and baseline severity of illness (Table IV). For specific Bayley domains, receipt of transfusion was associated with modestly decreased composite scores for both two-year cognitive (adjusted-β [a-β], −2.34 [95% CI, −4.64 to −0.04]; p=0.047) and motor function (a-β, −2.87 [95% CI, −5.38 to −0.36]; p=0.025), but not language (a-β, −0.29 [95% CI, −3.99 to 3.41]; p=0.879). The total number of units of blood transfused was associated with all neurodevelopmental outcomes after adjusting for gestational age, baseline morbidity, and socioeconomic status. For every 1-unit of transfused RBCs, the adjusted odds ratio (a-OR) for moderate-severe NDI increased by 18% (a-OR, 1.18 [95% CI, 1.10–1.27]; p<0.001) and severe NDI by 9% (a-OR, 1.09 [95% CI, 1.03–1.15]; p=0.004; Table IV). For every 1-unit transfused there was a decrease in the cognitive score by 0.65 points (a-β, −0.65 [95%CI, −0.95 to −0.35]; p<0.001), language by 0.63 points (a-β, −0.63 [95% CI, −0.98 to −0.29]; p<0.001), and motor by 0.76 points (a-β, −0.76 [95% CI, −1.08 to −0.45]; p<0.001). In the final step of multivariable analysis, autism was associated exclusively with two risk factors: RBC transfusion per 1-unit increase (a-OR, 1.13 [95% CI, 1.04–1.23]; p=0.005) and maternal age (a-OR, 0.94 [95% CI, 0.89–0.99]; p=0.026). In a separate sub analysis, receipt of blood transfusion in the first week of life was not associated with cognitive, language, or motor scores after multivariable adjustments (p=0.410, 0.545, 0.579, respectively).
Table II.
Neurodevelopmental outcomes at 2-year follow-up
| Outcomes | Total (n=654) | No Transfusion (n=359) | Transfusion (n=295) | p-value |
|---|---|---|---|---|
| Corrected age at follow-up, mo, mean ± SD | 25 ± 3 | 25 ± 3 | 25 ± 3 | 0.522 |
| Hospitalised since discharge, n. | 17% (112) | 9.5% (34) | 27% (78) | <0.001 |
| Underwent surgery since discharge | 10% (66) | 7.8% (28) | 13% (38) | 0.030 |
| Cognitive composite score, mean ± SD | 95 ± 14 | 96 ± 14 | 92 ± 13 | <0.001 |
| Language composite score, mean ± SD | 90 ± 17 | 91 ± 17 | 87 ± 15 | 0.003 |
| Motor composite score, mean ± SD | 95 ± 14 | 98 ± 13 | 92 ± 13 | <0.001 |
| Abnormal neurological exam | 12% (77) | 8.5% (30) | 16% (47) | 0.003 |
| Neurosensory impairment | 18% (116) | 14% (50) | 22% (66) | 0.005 |
| Moderate-Severe NDI (comprehensive) | 52% (340) | 47% (168) | 58% (172) | 0.003 |
| Severe NDI (comprehensive) | 18% (116) | 15% (53) | 21% (63) | 0.028 |
| Positive or Suspected Autism Screen (M-CHAT) | 10% (42) | 7.9% (19) | 13% (23) | 0.106 |
M-CHAT: modified checklist for autism in Toddlers; NDI: neurodevelopmental impairment; SD: standard deviation.
Table III.
Risk factors for moderate-severe NDI and severe NDI using bivariable logistic regression model
| Risk factors | Moderate-severe NDI | Severe NDI | ||
|---|---|---|---|---|
| p-value | OR (95% CI) | p-value | OR (95% CI) | |
| Received RBC transfusion | 0.003 | 1.590 (1.165–2.169) | 0.029 | 1.568 (1.048–2.346) |
| Number of RBC transfusions, per 1-unit increase | <0.001 | 1.159 (1.096–1.225) | <0.001 | 1.105 (1.051–1.162) |
| First RBC transfusion received within week 1 of life | 0.073 | 0.651 (0.408–1.040) | 0.137 | 0.651 (0.369–1.147) |
| Maternal age, per 1-year increase | 0.002 | 0.960 (0.936–0.985) | 0.086 | 0.972 (0.941–1.004) |
| Female | 0.002 | 0.616 (0.452–0.840) | 0.011 | 0.585 (0.386–0.885) |
| Gestational age, per 1-week increase | <0.001 | 0.879 (0.829–0.933) | <0.001 | 0.867 (0.805–0.933) |
| Birth weight, per 100-g increase | <0.001 | 0.937 (0.905–0.969) | 0.004 | 0.933 (0.890–0.978) |
| Nadir Hb in the NICU | 0.018 | 0.925 (0.868–0.987) | 0.066 | 0.917 (0.837–1.006) |
| Hb at NICU admission | 0.033 | 0.929 (0.868–0.994) | <0.001 | 0.847 (0.775–0.926) |
| Hb prior to first RBC transfusion | 0.079 | 1.146 (0.984–1.334) | 0.074 | 1.180 (0.984–1.416) |
| IVH ≥ grade 3 | 0.020 | 4.436 (1.262–15.585) | <0.001 | 9.272 (3.355–25.630) |
| Received inotropes | 0.001 | 2.083 (1.356–3.198) | 0.001 | 2.213 (1.381–3.548) |
| ROP requiring avastin or surgery | 0.001 | 5.235 (1.984–13.815) | 0.002 | 3.386 (1.591–7.208) |
| Days on ventilation | <0.001 | 1.032 (1.016–1.048) | <0.001 | 1.027 (1.012–1.042) |
| Received placental transfusion | 0.343 | 0.747 (0.409–1.365) | 0.715 | 1.175 (0.494–2.793) |
| 5-minute Apgar ≥7 | 0.757 | 0.924 (0.560–1.524) | 0.793 | 0.918 (0.484–1.742) |
| PDA requiring treatment | 0.026 | 1.507 (1.050–2.163) | 0.001 | 2.145 (1.395–3.299) |
| Culture positive sepsis | 0.999 | 1.001 (0.562–1.782) | 0.051 | 1.914 (0.996–3.677) |
| Received surfactant | 0.729 | 1.057 (0.773–1.446) | 0.658 | 1.098 (0.726–1.659) |
| Choriomnionitis, histological | 0.520 | 1.126 (0.784–1.618) | 0.394 | 1.228 (0.766–1.969) |
| Placental abruption | 0.887 | 1.042 (0.588–1.848) | 0.986 | 0.993 (0.469–2.102) |
| Highest maternal education completed: bachelor's degree | 0.741 | 0.920 (0.560–1.510) | 0.250 | 0.517 (0.290–0.921) |
| Public health insurance | <0.001 | 2.119 (1.532–2.929) | 0.054 | 1.500 (0.993–2.266) |
| No prenatal care | 0.062 | 2.677 (0.953–7.522) | 0.001 | 4.552 (1.805–11.481) |
Hb: haemoglobin; NDI: neurodevelopmental impairment; NICU: neonatal intensive care unit; PDA: patent ductus arteriosus; RBC: red blood cell; IVH: intraventricular haemorrhage; ROP: retinopathy of prematurity; SD: standard deviation; UCM: umbilical cord milking.
Table IV.
Risk factors for moderate-severe NDI and severe NDI in last step of backwards-stepwise multivariable logistic regression model
| Risk factors | Moderate-severe NDI | Risk factors | Severe NDI | ||
|---|---|---|---|---|---|
| p-value | OR (95%CI) | p-value | OR (95%CI) | ||
| RBC transfusion | removed at step 1 | RBC transfusion | removed at step 4 | ||
| Female | 0.001 | 0.565 (0.406–0.787) | Female | 0.005 | 0.500 (0.310–0.807) |
| Gestational age, per 1-week increase | 0.002 | 0.900 (0.843–0.962) | Gestational age, per 1-week increase | 0.009 | 0.885 (0.807–0.970) |
| IVH ≥ grade 3 | 0.052 | 3.614 (0.990–13.190) | Hb at NICU admission | 0.012 | 0.872 (0.784–0.970) |
| ROP requiring avastin or surgery | 0.025 | 3.211 (1.154–8.936) | IVH ≥ grade 3 | <0.001 | 8.515 (2.938–24.683) |
| No prenatal care | 0.003 | 4.830 (1.686–13.839) | |||
| Number of RBC transfusions, per 1-unit increase | <0.001 | 1.181 (1.095–1.274) | Number of RBC transfusions, per 1-unit increase | 0.004 | 1.087 (1.026–1.151) |
| Female | <0.001 | 0.540 (0.386–0.755) | Female | 0.002 | 0.471 (0.290–0.765) |
| IVH ≥ grade 3 | 0.107 | 2.953 (0.790–11.039) | Hb at NICU admission | 0.008 | 0.868 (0.781–0.964) |
| ROP requiring avastin or surgery | 0.108 | 2.391 (0.825–6.929) | IVH ≥ grade 3 | <0.001 | 8.433 (2.882–24.673) |
| PDA requiring treatment | 0.069 | 0.635 (0.389–1.037) | No prenatal care | 0.002 | 5.414 (1.896–15.461) |
Risk factors included in step 1 of all multivariable models: gestational age, female, Hb at NICU admission, IVH ≥ grade 3, received inotropes, ROP requiring avastin or surgery, PDA requiring treatment, no prenatal care. IVH: intraventricular haemorrhage; NDI: neurodevelopmental impairment; NICU: neonatal intensive care unit; PDA: patent ductus arteriosus; ROP: retinopathy of prematurity.
In the <28-week subgroup, receipt of blood transfusion was not associated with moderate-severe NDI or severe NDI (p=0.652 and 0.769, respectively). There was, however, an increased risk of NDI and autism with increased number of transfusions. Each 1-unit RBC transfusion increased the risk of moderate-severe NDI (a-OR, 1.15 [95% CI, 1.05–1.27]; p=0.004), severe NDI (a-OR, 1.09 [95% CI, 1.01–1.18]; p=0.022), and autism (a-OR, 1.24 [95% CI, 1.06–1.44]; p=0.008).
Discussion
The present study found an inverse relationship between receipt of RBC transfusion and neurodevelopmental outcomes after adjustments for gestational age, socioeconomic, demographic, and clinical confounders. Total number of transfusions also demonstrated an inverse association.
Our findings are in contrast to an observational study of extremely premature infants in the Netherlands, which found no relationship between differing transfusion volumes and neurodevelopment8. However, the study used parental surveys rather than in-person standard assessments to determine outcomes. Our study may have been able to detect smaller, subtle differences by using standardised Bayley Scales assessments done in-person. Another study found that preterm infants transfused using liberal haemoglobin thresholds presented with smaller brain volumes 12 years later10. These studies did not control for known confounders of neurodevelopmental delays, such as IVH and/or socioeconomic status9,11. They analysed differences in transfusion volume and threshold, but not the receipt of transfusion itself, nor the timing or accumulation of transfusion exposures.
If performing fewer blood transfusions improves outcomes, it would seem logical that a threshold definition for when to provide a transfusion would be beneficial. Studies on transfusion and neurodevelopmental outcomes are unclear on the risk of Hb thresholds for transfusion. A randomised clinical trial of 56 preterm infants found that children assigned to a liberal Hb threshold performed worse for verbal and memory function than those assigned to a restrictive threshold12. A 2009 observational study of extremely low birth weight infants, however, determined that transfusion Hb thresholds were not associated with severe NDI, language function, or motor function, but was related to a 4.3 point decrease in cognitive function (p=0.03)5. Most recently, in the two largest randomised trials to date, the Transfusion of Prematures trial and the ETTNO trial did not demonstrate any differences in 2-year neurodevelopmental outcomes between Hb threshold groups6,7. This is also consistent with the failure of erythropoietin to improve cognitive or other neurodevelopmental outcomes despite increasing red-cell mass13. In context with the present study, transfusion receipt and/or cumulative number, rather than Hb threshold, may be more indicative of subsequent neurodevelopmental function.
The mechanism(s) behind the relationship between RBC transfusions and abnormal neurodevelopmental outcomes are unclear. Transfusion has been implicated with producing inflammatory cytokines from endothelial damage and reactive oxygen species in preterm infants14. Moreover, the lifespan of transfused erythrocytes is nearly halved in neonatal recipients compared to adults (70 vs 120 days, respectively)15, thus potentially overloading the ability of premature infants to clear the toxic RBC breakdown products. Among transfused infants who experienced IVH ≥ grade 3, impaired neuronal migration from the germinal matrix may be a contributing factor16. Finally, as demonstrated in several studies17–19, volume expansion from transfusion may lead to the development of IVH ≥ grade 3 or the transfusion itself may have been in response to a drop in Hb due to severe IVH.
To our knowledge, this is the largest study examining the relationship between the receipt, timing, and number of RBC transfusions and subsequent neurodevelopmental outcomes in preterm infants. Another distinction of the present study includes adjustments for haemoglobin and socioeconomic status in addition to baseline morbidity. However, there are several limitations. We were unable to collect data on confounders such as erythropoietin use and age of administered RBC transfusions. Infants who received a blood transfusion were sicker and more immature; these morbidities could have reflected their resultant long term neuromorbidity. Given our robust sample size we adjusted for these confounders. The study institution did not employ standardised transfusion practices, which permitted varying tolerances of neonatal anaemia, but all within U.S. standards of care. Finally, our study is limited to those infants who returned for neurodevelopmental evaluations.
Conclusions
Our study demonstrates an association between RBC transfusion and lower cognitive and motor outcomes at two-years after adjustment for prematurity and illness at birth. Increasing number of transfusions worsened neurodevelopmental outcomes. Prospective studies utilizing blood-saving strategies (placental transfusion or minimal blood draws) to avoid blood transfusion and improve long term outcomes are needed to confirm this association.
Footnotes
AUTHORSHIP CONTRIBUTIONS
TGL conceived and designed the study, collected the data, performed the analysis, drafted the initial manuscript; JS designed the study, collected the data, edited the manuscript; RY designed the study; collected the data; edited the manuscript; SF designed the study, collected the data, edited the manuscript; AM curated the data, collected the data, edited the manuscript; DP designed the study; performed the analysis, edited the manuscript, provided supervision/oversight; AK conceived and designed the study, edited the manuscript, provided supervision/oversight.
The Authors declare no conflicts of interest.
References
- 1.Keir AK, Yang J, Harrison A, et al. Canadian Neonatal Network. Temporal changes in blood product usage in preterm neonates born at less than 30 weeks’ gestation in Canada. Transfusion. 2015;55:1340–6. doi: 10.1111/trf.12998. [DOI] [PubMed] [Google Scholar]
- 2.Maier RF, Sonntag J, Walka MM, et al. Changing practices of red blood cell transfusions in infants with birth weights less than 1000 g. J Pediatr. 2000;136:220–4. doi: 10.1016/s0022-3476(00)70105-3. [DOI] [PubMed] [Google Scholar]
- 3.Strauss RG. Practical issues in neonatal transfusion practice. Am J Clin Pathol. 1997;107:S57–63. [PubMed] [Google Scholar]
- 4.Howarth C, Banerjee J, Aladangady N. Red Blood Cell Transfusion in Preterm Infants: Current Evidence and Controversies. Neonatology. 2018;114:7–16. doi: 10.1159/000486584. [DOI] [PubMed] [Google Scholar]
- 5.Whyte RK, Kirpalani H, Asztalos EV, et al. Neurodevelopmental outcome of extremely low birth weight infants randomly assigned to restrictive or liberal hemoglobin thresholds for blood transfusion. Pediatrics. 2009;123:207–13. doi: 10.1542/peds.2008-0338. [DOI] [PubMed] [Google Scholar]
- 6.Kirpalani H, Bell EF, Hintz SR, et al. Higher or Lower Hemoglobin Transfusion Thresholds for Preterm Infants. N Engl J Med. 2020;383:2639–51. doi: 10.1056/NEJMoa2020248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Franz AR, Engel C, Bassler D, et al. Effects of Liberal vs Restrictive Transfusion Thresholds on Survival and Neurocognitive Outcomes in Extremely Low-Birth-Weight Infants: The ETTNO Randomized Clinical Trial. JAMA. 2020;324:560–70. doi: 10.1001/jama.2020.10690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.von Lindern JS, Khodabux CM, Hack KEA, et al. Long-term outcome in relationship to neonatal transfusion volume in extremely premature infants: a comparative cohort study. BMC Pediatr. 2011;11:48. doi: 10.1186/1471-2431-11-48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McGowan EC, Laptook AR, Lowe J, et al. Developmental Outcomes of Extremely Preterm Infants with a Need for Child Protective Services Supervision. J Pediatr. 2019;215:41–9e4. doi: 10.1016/j.jpeds.2019.07.063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nopoulos PC, Conrad AL, Bell EF, et al. Long-term outcome of brain structure in premature infants: effects of liberal vs restricted red blood cell transfusions. Arch Pediatr Adolesc Med. 2011;165:443–50. doi: 10.1001/archpediatrics.2010.269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Behrman RE, Butler AS. Institute of Medicine (US) Committee on Understanding Premature Birth and Assuring Healthy Outcomes, editor. Neurodevelopmental, health, and family outcomes for infants born preterm. In: Behrman RE, Butler AS, editors. Preterm Birth: Causes, Consequences, and Prevention. Washington (DC): National Academies Press (US); 2007. [PubMed] [Google Scholar]
- 12.McCoy TE, Conrad AL, Richman LC, et al. Neurocognitive profiles of preterm infants randomly assigned to lower or higher hematocrit thresholds for transfusion. Child Neuropsychol. 2011;17:347–67. doi: 10.1080/09297049.2010.544647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Juul SE, Comstock BA, Wadhawan R, et al. A Randomized Trial of Erythropoietin for Neuroprotection in Preterm Infants. N Engl J Med. 2020;382:233–43. doi: 10.1056/NEJMoa1907423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Keir AK, McPhee AJ, Andersen CC, Stark MJ. Plasma cytokines and markers of endothelial activation increase after packed red blood cell transfusion in the preterm infant. Pediatr Res. 2013;73:75–9. doi: 10.1038/pr.2012.144. [DOI] [PubMed] [Google Scholar]
- 15.Pearson HA. Life-span of the fetal red blood cell. J Pediatr. 1967;70:166–71. doi: 10.1016/s0022-3476(67)80410-4. [DOI] [PubMed] [Google Scholar]
- 16.Ballabh P. Intraventricular hemorrhage in premature infants: mechanism of disease. Pediatr Res. 2010;67:1–8. doi: 10.1203/PDR.0b013e3181c1b176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Baer VL, Lambert DK, Henry E, et al. Among very-low-birth-weight neonates is red blood cell transfusion an independent risk factor for subsequently developing a severe intraventricular hemorrhage? Transfusion. 2011;51:1170–8. doi: 10.1111/j.1537-2995.2010.02980.x. [DOI] [PubMed] [Google Scholar]
- 18.Baer VL, Lambert DK, Henry E, et al. Red blood cell transfusion of preterm neonates with a Grade 1 intraventricular hemorrhage is associated with extension to a Grade 3 or 4 hemorrhage. Transfusion. 2011;51:1933–9. doi: 10.1111/j.1537-2995.2011.03081.x. [DOI] [PubMed] [Google Scholar]
- 19.dos Santos AMN, Guinsburg R, Procianoy RS, et al. Variability on red blood cell transfusion practices among Brazilian neonatal intensive care units. Transfusion. 2010;50:150–9. doi: 10.1111/j.1537-2995.2009.02373.x. [DOI] [PubMed] [Google Scholar]
- 20.Olsen IE, Groveman SA, Lawson ML, Clark RH, Zemel BS. New intrauterine growth curves based on United States data. Pediatrics. 2010;125:e214–24. doi: 10.1542/peds.2009-0913. [DOI] [PubMed] [Google Scholar]