Article Text

Download PDFPDF

Red cell and platelet transfusions in neonates: a population-based study
  1. Jennifer R Bowen1,2,
  2. Jillian A Patterson3,
  3. Christine L Roberts3,
  4. James P Isbister2,
  5. David O Irving4,
  6. Jane B Ford3
  1. 1Department of Neonatology, Royal North Shore Hospital, St Leonards, New South Wales, Australia
  2. 2University of Sydney, Sydney, New South Wales, Australia
  3. 3Clinical and Population Perinatal Health Research, Kolling Institute, University of Sydney, Sydney, New South Wales, Australia
  4. 4Department of Research and Development, Australian Red Cross Blood Service, Sydney, New South Wales, Australia
  1. Correspondence to Jillian Patterson, Clinical & Population Perinatal Health Research, c/- University Dept of O&G, Building 52, Royal North Shore Hospital, St Leonards, NSW 2065, Australia; jillian.patterson{at}


Objectives This study aimed to describe the use of red cells, platelets and exchange transfusions among all neonates in a population cohort, to examine trends in transfusion over time and to determine transfusion rates in at-risk neonates.

Design Linked population-based birth and hospital data from New South Wales (NSW), Australia, were used to determine rates of blood product transfusion in the first 28 days of life. The study included all live births ≥23 weeks’ gestation in NSW between 2001 and 2011.

Results Between 2001 and 2011, 5326 of 989 491 live born neonates received a red cell, platelet or exchange transfusion (5.4/1000 births). Transfusion rates were 4.8 per 1000 for red cells, 1.3 per 1000 for platelets and 0.3 per 1000 for exchange transfusion. Overall transfusion rate remained constant from 2001 to 2011 (p=0.27). Among transfused neonates, 60% were <32 weeks' gestation (n=3210, 331/1000 births), 40% were ≥32 weeks' gestation (n= 2116, 2/1000 births) and 7% received transfusions in a hospital without a neonatal intensive care unit (NICU). Factors other than prematurity associated with higher transfusion rates were prior in utero transfusion (631/1000), congenital anomaly requiring surgery (440/1000) and haemolytic disorder (106/1000).

Conclusions In this population-based study, preterm neonates had a higher rate of transfusion than term neonates; however, 40% of those who received a transfusion were born ≥32 weeks' gestation and 7% were transfused in hospitals without an NICU. These findings need to be considered by transfusion services and personnel developing neonatal transfusion guidelines.

  • Epidemiology
  • Neonatology
  • Haematology
View Full Text

Statistics from

What is already known on this topic

  • Red cell and platelet transfusions are commonly used interventions in neonatal intensive care units, particularly among very low birthweight and preterm neonates.

  • Little is known about neonatal transfusion in term neonates or neonates outside neonatal intensive care units.

What this study adds

  • In this population-based study, 40% of neonates who received red blood cells or platelets were born ≥32 weeks' gestation.

  • High transfusion rates were seen in neonates with haemolytic disorders (132/1000) and congenital abnormalities requiring surgery (508/1000), as well as in those born preterm.


Blood product transfusions in neonates may be life-saving; however, blood products are a costly and limited resource and adverse effects are being increasingly recognised following neonatal transfusions.1 Transfusion rates differ over time and between populations2–5 due to variation in clinician practice and transfusion thresholds,6 ,7 and due to differences in case-mix within the populations being reported.

Previous reports of transfusion practices in neonates have predominantly studied patients admitted to a neonatal intensive care unit (NICU), with most studies focusing on transfusions in premature or very low birthweight neonates.2 ,4 ,5 Neonates born at term gestation or cared for in hospitals without an NICU may also require blood transfusions for reasons including antenatal or postnatal blood loss, hereditary blood disorders, haemolysis or need for major surgery; however, the rates of transfusion for neonates with these problems have not been well documented. Knowledge about use of blood transfusions in neonates is important for planning purposes for transfusion services and for personnel developing guidelines for neonatal transfusion. Population data provide a valuable source of data for all neonates, including those who are cared for outside the NICU environment and can be used to identify trends over time. Population data have been used to explore the age distribution of transfusion recipients,8 the distribution of blood use in medical and surgical specialties9 and transfusion rates in selected adult populations10 but have not previously been used to identify rates and trends in transfusion among neonates.

This study aimed to determine rates of red cell and platelet transfusion among neonates in a population cohort and to examine trends in transfusions over time. Secondary aims were to identify groups with pre-existing risk factors for transfusion and to provide information about blood product use in these groups.


Using linked population-based birth and hospital data from New South Wales (NSW), Australia, a cohort of all live births of at least 23 weeks' gestation occurring between 2001 and 2011 was established. NSW is the most populous state in Australia, with approximately 7.5 million residents and 90 000 births per year. Neonates were followed from birth until 28 days or first discharge (whichever was longer).

Data were available from two sources: ‘birth data’ and ‘hospital data’. The Perinatal Data Collection is a statutory population-based collection of all births in NSW of at least 20 weeks' gestation or 400 g birth weight and includes information on pregnancy, birth and maternal characteristics (‘birth data’). The Admitted Patients Data Collection (‘hospital data’) is a census of public and private hospital separations in NSW and contains diagnoses and procedures for each separation coded according to the 10th revision of the International Classification of Diseases, Australian Modification11 and the Australian Classification of Health Interventions.12 Hospital separation records are available for all neonates (‘birth admission’) and additional records are created for neonates readmitted to hospital. The hospital data were linked to the birth data to provide information on the birth, transfers and subsequent admissions until 28 days after birth for all neonates. The NSW Centre for Health Record Linkage performed probabilistic data linkage between the two data sets.13 Linkage proportions were >98%.


The first 20 procedure codes in hospital data were used to identify admissions involving transfusion of blood products, defined as red cells, platelets or exchange transfusions. Coding criteria for fresh frozen plasma and coagulation factors changed during the study period and precluded inclusion in our study. Procedure codes for blood product administration have been shown to have good ascertainment (sensitivity/specificity red blood cell transfusion: 83.1%/99.9%, platelets: 73.1%/100.0%).14 Transfusion may have occurred in one or more hospital admissions; however, each neonate was only counted once. Some transfusions occurring after 28 days (in neonates not discharged until later) have been included. If any of the neonatal transfusions occurred in a hospital without an NICU, these were counted as non-NICU transfusions. A hospital was considered to have an NICU if it routinely provided assisted ventilation for neonates <32 weeks' gestation.

Maternal age, pregnancy and neonatal characteristics were identified in birth data. Maternal transfusions and pregnancy hypertension were identified in mother's hospital data (birth admissions). Neonatal conditions were identified using diagnoses and procedures coded in hospital data (see online supplementary appendix 2). Variables in these data sets are known to be reliably reported with few missing values.15

Identification of neonates at risk of transfusion

Neonates with risk factors for transfusion were identified in two ways. First, neonates were categorised based on gestational age and rates of transfusion for each gestational age group were determined. Second, neonates with other specific antenatal, congenital or early neonatal conditions associated with need for transfusion were identified in the data set and a hierarchy was developed based on transfusion rates associated with each condition in order to group neonates according to the most likely indication for transfusion. The hospital diagnoses codes were checked progressively for these conditions, including (in order): severe fetal anaemia requiring in utero transfusion, fetal blood loss, birth trauma with haemorrhage, hereditary red blood cell disorders (sickle cell disorders, thalasaemia, other hereditary anaemia); hereditary coagulation disorders (Factor VIII and Factor XI deficiency, other coagulation defects);16 major congenital cardiac and non-cardiac abnormalities requiring surgery in the neonatal period involving opening of a body cavity;16 and haemolytic disorders (immune haemolytic disease of fetus and newborn, Rh incompatibility, kernicterus, G6PD deficiency, jaundice due to excessive haemolysis), preterm birth and assisted ventilation (for neonates ≥32 weeks' gestation). Neonates with multiple diagnoses were only counted in one of these groups. Rare hereditary blood disorders were identified before more common conditions in order to capture the frequency of rare disorders. The use of red blood cell and platelet transfusions was then determined for all neonates and for neonates with gestational age <32 weeks and ≥32 weeks in each of these high-risk groups.

Statistical methods

Rates are calculated per 1000 live births and proportions are proportion of live births, unless specified otherwise. Trends were assessed using the Cochran Armitage test for trend. All analyses were performed in SAS V.9.3.


There were 989 491 neonates born ≥23 weeks' gestation in NSW between January 2001 and December 2011, including 9688 (1%) neonates born <32 weeks and 979 803 (99%) born ≥32 weeks. Neonates with missing gestational age (151) were excluded. There were 5326 neonates given at least one blood product transfusion, giving a transfusion rate of 5.4 per 1000 live births. Among all neonates, transfusion rates were 4.8 per 1000 for red cells, 1.3 per 1000 for platelets and 0.3 per 1000 for exchange transfusions. Of the 5326 neonates who received a blood product, 3956 (74.3%) received red cells only. Most neonates received a transfusion of blood products during only one admission (99.2%), with the majority of transfusions being during the birth admission (93.2% of transfusions). Transfusions were predominantly given to neonates who were admitted to a hospital with an NICU; however, 7% (n=385) of transfused neonates received at least one transfusion in a hospital without an NICU, and 12.5% of exchange transfusions were given in non-intensive care environments.

Blood product transfusion was more common among earlier gestations, with a crude transfusion rate in neonates born at 23–27 weeks' gestation of 582.8 per 1000 live births compared with 1.4 per 1000 at term (table 1). Among neonates with an Apgar score <7 at 5 min, the crude transfusion rate was 77.2 per 1000. Singletons had lower transfusion rates than multiple births (4.5/1000 vs 34.3/1000). Higher crude transfusion rates were also seen in neonates whose mother had a blood product transfusion during the birth admission (18.0/1000), neonates who were small for gestational age (9.9/1000) and whose mother had hypertension (9.3/1000).

Table 1

Characteristics of neonates receiving red cells, platelets or exchange transfusion, 2001–2011

The overall rate of transfusions of blood products in neonates stayed constant between 2001 and 2011 (2001: 5.0/1000; 2011: 5.4/1000; p for trend 0.27). A significant increase was seen in transfusion of platelets (p=0.03), and this primarily occurred among neonates born ≥32 weeks (2001: 0.6/1000; 2011: 0.7/1000; p for trend 0.01). There was no overall change in red cell transfusions (p=0.07) over the period, although use increased among neonates born before 32 weeks (2001: 271.9/1000; 2011: 338.0/1000; p for trend 0.02). Altogether 264 neonates received an exchange transfusion, with the rate of exchange transfusion decreasing from 0.42/1000 in 2001 to 0.11/1000 in 2011 (p<0.001) (figure 1).

Figure 1

Trend in red cell, plasma and exchange transfusion in (A) neonates <32 weeks’ gestation and (B) ≥32 weeks’ gestation, New South Wales (NSW), 2001–2011. Note that figures are on different scales.

Among neonates with risk factors for transfusion, red cell or platelet transfusions were given to 63.1% of neonates who had received an in utero transfusion prior to birth, 44.0% of those with a major congenital anomaly requiring surgery and 10.6% of those with a haemolytic disorder (table 2). Transfusion rates were higher for neonates undergoing cardiac surgery (55.0%) than for those requiring surgery for other major congenital abnormalities (17.3%).

Table 2

Blood product use among neonates at risk of transfusion*, 2001–2011

Twenty-seven per cent of very preterm neonates (<32 weeks) who received a transfusion had an identified risk factor for transfusion other than prematurity (see online supplementary appendix table 1). The majority of very preterm neonates who required a transfusion received red blood cells (n=3104, 97%), with a smaller proportion receiving platelets (16%) or exchange transfusion (1%).

Of neonates born at term or moderate preterm gestation (≥32 weeks) who required a transfusion, 64% had an identified risk factor for transfusion other than prematurity (see online supplementary appendix table 1). The most common conditions associated with transfusion in term or moderately preterm neonates were haemolytic disorders (31%) and surgery for major congenital abnormality (24%). Of the 36% (n=760) who received a transfusion in the absence of at-risk conditions, 472 (62%) required mechanical ventilation for severe respiratory illness. Three-quarters (77%) of term or moderately preterm neonates who required a transfusion received red cells and 36% received platelets.


This study provides the first reported population-level data on the use of red cell and platelet transfusions in the neonatal period. Transfusions occurred in 1 in every 185 neonates born in NSW between 2001 and 2011; with an overall transfusion rate of 5.4/1000 live births. This rate is higher than at any other time during childhood8 and similar to some reported rates in high-risk adult populations.10 ,17 Among the neonates transfused, 88.9% received a red cell transfusion and 23.9% received platelets, either alone or in combination with a red cell transfusion.

The majority (60%) of neonates who received a transfusion were born before 32 weeks' gestation with transfusion rates of 583/1000 in neonates <28 weeks' gestation and 213/1000 in neonates 28–31 weeks' gestation. These transfusion rates are consistent with other studies reporting red blood cell transfusions in 50–61% of neonates <32 weeks' gestation 4 ,18 with higher rates (>85%) reported among neonates of birth weight <1000 g, even with restrictive transfusion thresholds.5–7

Although transfusion rates were lower in more mature neonates, a substantial proportion (40%) of neonates receiving transfusions were born ≥32 weeks' gestation. Haemolytic disorders and major congenital anomalies requiring surgery were present in more than half of the term or moderately preterm neonates who received transfusions, with many of these children receiving both red cells and platelets.

There was no significant change in the overall transfusion rate over the study period; however, there was a trend to an increase in red blood cell transfusion in neonates <32 weeks' gestation and an increase in platelet transfusion rate in neonates ≥32 weeks' gestation. In contrast, other studies have reported a decrease in red blood cell transfusion rates over a similar time period following the implementation of standard guidelines for transfusion.19 Although the risk of haemolytic transfusion reactions and transfusion-associated infections are low in neonates with modern transfusion practices,20 studies have identified other potential risks with neonatal red blood cell and platelet transfusions, 1 ,21 making it important that blood products are given according to appropriate guidelines to ensure that benefits of transfusions outweigh risks.

The exchange transfusion rate decreased over the course of the study. In the past, exchange transfusion was common in neonates with haemolysis due to Rh incompatibility; however, the incidence of haemolytic disease has fallen significantly with the widespread use of anti-D immunoglobulin22 ,23 and the need for exchange transfusion has decreased further in those with haemolysis with the introduction of in utero transfusions for severe fetal anaemia, more effective phototherapy devices and the use of intravenous immune globulin in neonates approaching exchange transfusion levels.

Despite a significant reduction in the need for exchange transfusions, neonates with haemolytic disorders remain at risk of transfusion. Among neonates who received an in utero transfusion, 63% (n=41) required a postnatal transfusion and 10.6% (n=1227) of neonates with an identified haemolytic disorder received blood products in the neonatal period. Although the risks of a blood product transfusion are less than those associated with an exchange transfusion, these children remain at risk of adverse transfusion outcomes.1

A high proportion (44.0%) of neonates undergoing surgery also received transfusions. In neonates undergoing surgery, even a relatively small loss of blood may have a major impact on tissue perfusion and a need for blood transfusion due to their low circulating blood volume. For children undergoing surgery for congenital cardiac disorders, additional blood is needed to ‘prime’ the extracorporeal circuits required for cardiopulmonary bypass and treatment with plasma and platelets may be required to treat coagulation disorders that can be induced by the bypass procedures.24 Processes that aim to reduce the use of blood products have been instituted in many centres; however, these children remain a group with high transfusion requirements.25

As transfusions in neonates may have long-term consequences and blood products are a limited resource, it is important that we understand which neonates are receiving blood product transfusions. Strengths of this study include the use of population-based rather than NICU-based data to identify all neonates receiving transfusions and the ascertainment of use of both red blood cells and platelet transfusions over the first 28 days of life. Administration of blood and blood products is accurately reported in adults,14 and therefore, likely to be accurately reported in these data; however, there is potential for some underascertainment. Limitations of this study are the lack information about the indication for transfusion, volume of blood product transfused and transfusion of coagulation factors. Some of the reported characteristics of transfused infants (table 1) are likely to also reflect prematurity (eg, multiple births). Furthermore, without additional clinical detail it was not possible to ascertain whether neonates were receiving transfusions based on generally accepted guidelines.

In summary, in this population-based study, 5.4/1000 neonates received a red cell, platelet or exchange transfusion, with no change in overall transfusion rates over the period of the study. Transfusion rates were higher in very preterm neonates (331/1000); however, 40% of those who received a transfusion were born ≥32 weeks' gestation and 7% were transfused in hospitals without an NICU. These neonates are not always identified in studies of neonatal transfusion rates in NICUs; however, they form a substantial subgroup of neonates who need to be further investigated and considered by transfusion services providing blood products and by personnel developing neonatal transfusion guidelines.


We thank the NSW Ministry of Health for access to the population health data and the NSW Centre for Health Record Linkage for linking the data sets.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors JRB: assisted in study design, provided expert input, drafted the initial manuscript and approved the final manuscript as submitted. JAP: assisted in study design, performed the analysis, revised the manuscript and approved the final manuscript for submission. CLR: involved in study design, revised the manuscript and approved the final manuscript for submission. JPI and DOI: provided important content knowledge, reviewed the manuscript and approve the final manuscript for submission. JBF: conceptualised the study, revised the manuscript and approved the final manuscript for submission.

  • Funding This work was supported by a Partnership Grant from the Australian National Health and Medical Research Council NHMRC (#1027262), the Australian Red Cross and the NSW Clinical Excellence Commission. Christine Roberts is supported by a NHMRC Senior Research Fellowship (#1021025). JF is supported by an ARC Future Fellowship (#120100069).

  • Competing interests None declared.

  • Ethics approval NSW Population and Health Services Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.