Transfusion rates and disease spectrum in neonates treated with blood transfusion in China

Supplemental Digital Content is available in the text


Introduction
Neonatal transfusion therapy is used for the treatment of anemia, thrombocytopenia, and diseases associated with coagulation disorders of different etiologies. Since the "two-child policy" has come into effect in China, the number of women at advanced maternal age with high-risk pregnancy has increased rapidly, leading to an increase in the number of preterm neonates, lowbirth-weight(LBW) neonates, and neonates with postnatal defects. The two-child policy is a family planning policy implemented in China, which means that couples who meet the specified conditions are allowed to have "two children" since January 1, 2016. The couples are allowed to have only "one children" before that time. Blood transfusion therapy is particularly important for the treatment of LBW or very-lowbirth-weight(VLBW) neonates, in which the transfusion rate during hospitalization can reach more than 80%. [1] However, since there are no big data from neonatal transfusion research in China to aid in the formulation of national neonatal transfusion guidelines, the decisions for neonatal transfusion therapy in the clinical settings in China are mainly made on empirical basis or with reference to the transfusion guidelines from foreign countries. Some countries, such as the United Kingdom, the United States, Italy, and Australia, have developed guidelines or expert consensus on neonatal transfusion therapy that provide guidance on transfusions for neonatal thrombocytopenia and neonatal anemia. [2][3][4][5][6][7] However, the recommendations are inconsistent, and treatment decisions are often still based on clinical experience. [8] This study utilized a multicenter and a combined retrospective and prospective survey design to investigate the status of neonatal blood transfusion in China. The data on neonatal blood transfusions obtained can provide an evidence-based medical basis for the development of neonatal blood transfusion treatment guidelines in China.

Study population
(1) Study population. Between January 1, 2012 and December 31, 2016, all neonates hospitalized in the Department of Neonatology of 55 hospitals nationwide were included, and all studies that involve human subjects were approved by the ethics committee of Shaanxi Provincial People's Hospital.
The 55 selected hospitals should match any of the following conditions: (1) the number of neonates department beds are greater than 20 in the tertiary general hospital. (2) Children's hospital or maternal and child health hospital.
The hospitals were nationwide collected on the net, and the 55 participating hospitals were extracted at random from them according to the regions.

Disease classification
(1) Assignment of disease. Data of the top 10 diseases identified in the hospitalized neonates and the top 10 diseases identified in those treated with blood transfusion in each hospital were collected and sorted as follows: (1) The diseases were grouped according to the ICD.
(2) Assignment of disease: the top 10 assigned diseases in each hospital were assigned a value between 1 and 10 (10, 9, 8, . . . , 1 points) according to their rank. (3) To exclude the impact of hospital size and the number of discharged patients in the Department of Neonatology on the overall disease ranking, data weighting was conducted based on the disease assignments in each hospital, the number of discharged patients in the Department of Neonatology, and the total number of discharged neonates in all participating units.
The scores (r s ) of the top 10 diseases in this unit relative to the diseases in all participating units in the nation was calculated as follows: where r s represents the scores of the disease, p i as number of discharged neonatal patients in each participating unit during the survey reference period, p s as total number of discharged neonatal patients in all participating units during the survey reference period, and s r as disease type in discharged neonatal patients in each participating unit during the survey reference period. Each disease was assigned a value according to its rank (1st = 10 points, 2nd = 9 points, etc.).
(2) Rank order in the disease spectrum. The r s of the same disease from the 55 hospitals were classified and summed to obtain a total score for each disease. Finally, a disease spectrum was obtained by ranking different diseases according to their total scores.

Statistical analysis
All data were analyzed using the IBM SPSS 24.0 statistical software, and categorical data were expressed as rates and ratios.

Background information of the 55 hospitals
The 55 hospitals included in this study were located across 7 administrative regions, 21 provinces, 2 direct-controlled munici-  (Table 1).

Transfusion rates in neonates
Between January 1, 2012 and December 31, 2016, the total number of discharged neonates in the 55 hospitals nationwide was 541,128, and the number of neonates who received blood transfusion was 70,433, with an average transfusion rate of 13.02%. The blood transfusion rates in hospitals of different sizes varied between 1.21% and 27.38% (Table 1). In addition, there were considerable differences in the transfusion rates between different regions. The highest transfusion rate was in the Northeast region (17.99%), whereas the lowest transfusion rate was in the Northwest region (9.74%). The transfusion rates in other regions were between 10.60% and 16.22% ( Table 2). The data were grouped into quartiles based on the number of beds in the Department of Neonatology of the hospitals ( Table 3). The transfusion rate was the lowest (7.55%) in hospitals with less than 48 beds, and the highest (16.24%) in hospitals with more than 91 beds.
Statistical analysis by hospital type showed that the neonatal blood transfusion rate was 12.26% in general hospitals and 13.8% in women and children's hospitals. The overall blood transfusion rate in women and children's hospitals was greater than that in general hospitals (Table S1, http://links.lww.com/ MD/E128).

Blood component transfusion rates in neonates
Of the 55 hospitals, 39 participated in the survey on neonatal blood component transfusion rates. The number of blood transfusions between January 1, 2014 and December 31, 2016 were calculated. The total number of hospitalized neonates in the 39 hospitals within the 3-year reference period was 259,886. About 24,536 neonatal patients received red blood cell suspension transfusion (9.44%), 1717 neonatal patients received platelet transfusion (0.66%), and 12,394 neonatal patients received plasma transfusion (4.77%) ( Table 4). The rates of red blood cell suspension (14.3%), plasma (6.75%), and platelet transfusions (1.41%) were highest in Northeast China. The lowest rate of red blood cell transfusion was in Southwest China (8.12%), the lowest rate of platelet transfusion was in Northwest China (0.41%), and the lowest rate of plasma transfusion was in East China (3.26%) ( Table 4). The rates of red blood cell suspension and platelet transfusions in neonates were lower in general hospitals than in women and children's hospitals, whereas the rate of plasma transfusion in both hospitals was the same (Table S2, http://links.lww.com/MD/E129).

Spectrum of diseases in hospitalized neonates and that in hospitalized neonates treated with blood transfusion
The disease spectrum in neonates was obtained by weighting the top 10 assigned diseases in each participating hospital each year and the total number of hospitalized neonatal patients in the 55 hospitals (Table 5). Between 2012 and 2016, the top 10 diseases in hospitalized neonates were, in rank order, as follows: pneumonia, prematurity, hyperbilirubinemia, respiratory distress syndrome, asphyxia, hemolytic disease, bacterial sepsis, hypoxicischemic encephalopathy, aspiration pneumonia, and dysphagia. The rank order in the disease spectrum of the hospitalized neonates varied in different regions. Among the hospitalized patients in Northeast China and North China, the number of preterm neonates was the highest and ranked 1st in the spectrum. In East China, South China, Northwest China, and Southwest China, the disease with the highest rank was hyperbilirubinemia, whereas that in Central China was pneumonia. Except for Southwest China, the top three leading causes of hospitalization in other regions were prematurity, hyperbilirubinemia, and pneumonia (Table S3, http://links.lww.com/MD/E130).
Data on the hospitalized neonates who received blood transfusions in the 55 hospitals were weighted, and a spectrum of diseases in neonates treated by blood transfusion was obtained ( Table 6). The top 10 diseases identified in neonates treated with blood transfusion were, in rank order, as follows: prematurity, pneumonia, hyperbilirubinemia, bacterial sepsis, respiratory distress syndrome, anemia, hemolytic disease, asphyxia, hemorrhage, and necrotizing enterocolitis. The rank order of diseases in neonates treated with blood transfusion varied in different regions. The disease that ranked first in Northeast China, North China, East China, South China, and Northwest China was prematurity, and those in Southwest China and Central China were hyperbilirubinemia and pneumonia, respectively (Table S4, http://links.lww.com/MD/E131).

Discussion
Recent clinical studies have found that the effect of the strongly advocated restrictive transfusion strategy on neurodevelopment remains uncertain. [8][9][10][11][12][13][14] In the era of evidence-based medicine, transfusion strategies require the support of strong evidence from clinical research. Generally, transfusions can be considered as long as the benefits outweigh the risks. However, the evolution of transfusion strategies in preterm neonates in recent decades lacks the support from clinical evidence. The criteria for determining blood transfusion need in critically ill neonates are imprecise, and opinions are divided among neonatologists. The use of the relatively objective parameters (such as hemoglobin or hematocrit) as guidance for blood transfusion also remains controversial. Current guidelines for neonatal blood transfusions are still in contention, and references for disease spectrum or the current status of neonatal blood transfusions are lacking. For these reasons, we conducted a multicenter study on neonatal blood transfusion in China. This study aimed to investigate the changes in the disease spectra in neonates in recent years, the results of which will allow us to understand and further explore the current status of neonatal blood transfusion in China.
This study analyzed the data on hospitalized neonates and blood transfusions between January 1, 2012 and December 31, 2016 in 55 hospitals in China. The transfusion rate was 13.02%. Neonates with younger gestational ages, lower birth weights, and more severe conditions have a higher probability of receiving blood transfusion therapy. Some researchers reported that 20% to 25% of neonates in the neonatal intensive care unit (NICU) received one or more platelet transfusions, [15] and a previous study showed that 532 of 952 (55.9%) very-low-birth-weight infants received blood transfusion. [16] A single-center study by Zhou et al [17] reported that the transfusion rate in neonates was 25.24% and that in preterm neonates was 38.68%. The present study shows that the average blood transfusion rate in the 55 www.md-journal.com Table 1 The basic information of the 55 hospitals and analysis of transfusion rates in neonates and the total number of discharged neonates between 2012 and 2016.

Regions
Participating medical centers *

Types †
The number of total hospital beds (n) The number of beds in neonates department (n) The number of neonates who received blood transfusion (n) The total number of discharged neonate (n)  [18] This may be due to the difference in hospital size. Among the 55 hospitals included in the present study, 8 hospitals have a transfusion rate of more than 20%. The average transfusion rate in hospitals with more than 91 neonatal beds is 16.24%, which is much higher than the transfusion rate reported in hospitals with less than 48 neonatal beds (7.55%). These results show that hospitals with more neonatal beds have higher treatment capacities, more critically ill neonatal patients, and higher blood transfusion rates. This study also revealed that neonatal transfusion rates vary in different regions of China. For example, the highest transfusion rate is in the Northeast region (17.99%), and the lowest transfusion rate is in the Northwest region (9.74%). Therefore, regional factors should be considered when formulating expert consensus on neonatal transfusion. The results of the present survey show that among neonatal transfusions, the rate of red blood cell transfusion is 9.44%, which is significantly higher than the transfusion rates of plasma and platelets. However, there are few domestic and foreign reports on blood component transfusion rates in neonates. Platelet transfusion is often used to treat thrombocytopenia of different etiologies. It has been reported that thrombocytopenia occurs in 22% of neonates in the NICU, and the risk of intracranial hemorrhage is higher in preterm neonates with thrombocytopenia. However, the platelet transfusion rate   The data were grouped into quartiles based on the number of beds. R = transfusion rates (%), T/C = the number of neonates who received blood transfusion (n)/the total number of discharged neonates (n). Table 4 Blood component transfusion rates in neonates in different regions.

Regions
The total number of hospitals (n) The total number of discharged neonates (n) The number of patients received red blood cell suspension transfusion n (%) observed in our study was only 0.66%, and the underlying reason of such a low percentage warrants further investigation. In this study, the neonatal plasma transfusion rate was 4.77%. An epidemiological survey conducted by a British group showed that the plasma transfusion rate was 3.3% in hospitalized patients aged 0 to 19 years, which was lower than the transfusion rates of red blood cells and platelets. [9] Neonatal transfusion of plasma and its products are mainly used to supplement coagulation factors. As the levels of coagulation factors and plasminogen in neonates are low, they are at risk for both hemorrhage and thrombosis. This survey also shows that the types of blood components transfused in neonates indicate regional differences. The rates of red blood cell suspension (14.3%), plasma (6.75%), and platelet transfusions (1.41%) were highest in the Northeast China. So when we develop the neonatal blood transfusion treatment guidelines in China in the future, we should consider the geography of China.
The results of this survey show that the top 10 diseases identified in hospitalized neonates in China were summarized at Table 5. Compared with the disease types identified in discharged patients from the Department of Neonatology in the general hospitals described in the 2016 report on the statistical analysis of medical data [19] (pneumonia, prematurity, hyperbilirubinemia, asphyxia, respiratory distress syndrome, low birth weight, bacterial sepsis [septicemia], diarrhea, and transient tachypnea of the newborn). The results were basically consistent with the present study. The present study shows that respiratory diseases account for the largest number of diseases, with 4 (33.40%) of the top 10 diseases being respiratory diseases (pneumonia, respiratory distress syndrome, asphyxia, and aspiration pneumonia). This study also showed that preterm neonates account for 12.09% of all hospitalized neonates, which is slightly higher than the 11.10% incidence rate of preterm neonates reported in the "Global Action Report on Preterm Birth" in 2012, [20] and is significantly higher than the 7.8% preterm birth rate reported in the large-scale epidemiological survey on preterm neonates conducted in China in 2005. [21] Between 2012 and 2016, the top 10 diseases identified in hospitalized neonates treated with blood transfusion in 55 hospitals in China were summarized at Table 6. The understanding of the disease spectrum in neonates who received blood transfusions can assist the formulation of expert consensus on neonatal transfusions, but there are few domestic and foreign reports on the spectrum of diseases in neonates treated with blood transfusion. Our study has also shown that the top 10 diseases in hospitalized neonates only partially overlap with the top 10 diseases in those treated with blood transfusion. In other words, not all of the top 10 diseases identified in hospitalized neonates require transfusion therapy during hospitalization. Limitation of the study: This is a pilot study for a national multicenter cohort study on neonatal blood transfusion. The data collected do not include detailed data on the neonatal blood transfusion cases and transfusion data from neonates in the NICU of each hospital.

Conclusion
The study can provide supporting data for the next phase of study conducted on Chinese neonatal cohort and provide technical data and theoretical evidence for the development of neonatal blood transfusion strategies in clinical settings.