Transfusion demand in COVID-19 patients from the Korean population: a nationwide study in South Korea.

As patients infected with severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) often present coagulopathy depending on disease severity, appropriate prevention and treatment for haemodynamics control are most essential for coronavirus disease 2019 (COVID‐19) management. Although several papers warned that, as expected, there was a decrease in the number of blood donations in the COVID‐19 pandemic, , few studies investigated the actual prevalence of blood transfusions requested in hospitalized COVID‐19 patients, , and no publication has reported nationwide results yet. Here, we examined the transfusion demand in COVID‐19 patients using the Korean nationwide COVID‐19 patient database generated by the government.

The Ministry of Health and Welfare of Korea and the Health Insurance Review and Assessment (HIRA) Service of Korea have jointly released the de‐identified nationwide COVID‐19 patient data using the Korean National Health Insurance System. This database contains all claim records (including a five‐year period before hospitalization) for 7 512 individuals who were confirmed for COVID‐19 and for whom hospitals issued claims to the HIRA until 15 May 2020. Confirmed COVID‐19 cases were identified following the Korean Classification of Diseases seventh revision (KCD‐7) code, which is a modified version of the tenth revision of the International Classification of Diseases (ICD‐10). We identified all the transfusion procedures performed for COVID‐19 patients before and during the treatment periods. The procedure codes for transfusion practices included all types of transfusion products based on KCD‐7 codes (from X1001 to X2516). Additionally, we divided transfusion events into four subgroups according to transfusion products; red blood cell (RBC), platelet concentrate (PC), fresh‐frozen plasma (FFP), and cryoprecipitate. As the database includes detailed information regarding demographic characteristics, diagnoses, prescriptions, procedures, and patient outcomes, we extracted data for associated risk factors and clinical outcome parameters for COVID‐19 patients. For associated risk factors, age, gender, insurance type, comorbidities, and anticoagulant medication history were considered. Comorbidities were defined on the basis of claim codes within one year before the index date and evaluated on the basis of the Charlson Comorbidity Index. Four clinical outcome parameters during the COVID‐19 treatment periods assessed in this study were intensive care unit (ICU) admission, ventilator usage, oxygen therapy, and all‐cause mortality. The study protocol for analysis of de‐identified patient data was exempted from review by the Institutional Review Board of National Health Insurance Service Ilsan Hospital (NHIMC 2020‐04‐029).

The proportions of associated risk factors and clinical outcome parameters in the COVID‐19 patient group who required blood transfusion therapy were compared with those not receiving any transfusion using Student’s t‐test or the chi‐squared test. Multivariate logistic regression models were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for all‐cause mortality according to transfusion therapy after adjustment of other factors. All tests were two‐tailed and differences were considered statistically significant at P values of <0·05. All statistical analyses were performed using SAS software (version 9.4; SAS Institute, Cary, NC, USA).

Interestingly, 1·2% of all patients (n = 93) required blood transfusion during COVID‐19 treatment periods (Table I). While the proportion of RBC transfusion was predominant (n = 88), PC and FFP transfusion were also observed (n = 28 and n = 18, respectively) with 30 patients receiving multiple types of transfusion product. The median amounts of RBC, PC, FFP, and cryoprecipitate received in the transfusion group during COVID‐19 treatment periods were 3, 11, 4, and 27 units, respectively.

Table I

Demographic and clinical characteristics of Korean COVID‐19 patients.

Characteristics	Total No.	Transfusion during COVID‐19 treatment period		P value
		Yes No. (%)	No No. (%)
Total number	7 512	93 (1·2)	7 419 (98·8)
Transfusion product*
Red blood cell		88 (94·6)
Platelet concentrate		28 (30·1)
Fresh frozen plasma		18 (19·4)
Cryoprecipitate		4 (4·3)
Age, mean (standard deviation), years	46	68·2 (13·2)	45·7 (19·7)
Sex
Men	3 055	49 (52·7)	3 006 (40·5)	0·019
Women	4 457	44 (47·3)	4 413 (59·5)
Insurance type
National Health Insurance	6 888	79 (84·9)	6 809 (91·8)	0·029
Medical aids	624	14 (15·1)	610 (8·2)
Charlson Comorbidity Index
0	3 709	13 (14·0)	3 696 (49·8)	<·001
1	1 591	12 (12·9)	1 579 (21·3)
2	635	13 (14·0)	622 (8·4)
3+	1 577	55 (59·1)	1 522 (20·5)
Outcome parameters
ICU admission	914	73 (78·5)	841 (11·3)	<·001
Ventilator usage	121	55 (59·1)	66 (0·9)	<·001
Oxygen therapy	905	76 (81·7)	829 (11·2)	<·001
All‐cause mortality	227	45 (48·4)	182 (2·5)	<·001
Associated medical history within five years
Anticoagulant medication	278	18 (19·4)	260 (3·5)	<·001
Transfusion history	167	20 (21·5)	147 (2·0)	<·001

COVID‐19, coronavirus disease 2019; ICU, intenstive care unit.

Patients may have received more than one kind of transfusion product.

When anticoagulant medication and transfusion history before the COVID‐19 diagnosis were investigated for the transfusion group, 70% of patients who required transfusion (n = 65) did not have any history of transfusion or anticoagulant therapy in the preceding five years, suggesting that in the majority of patients who required transfusion during treatment periods unstable haemodynamics was independently caused by COVID‐19 itself.

Furthermore, multivariable logistic regression analysis revealed that the all‐cause mortality rate was statistically significantly increased in the patients who received transfusion therapy during COVID‐19 management after adjusting for age, sex, insurance type, Charlson Comorbidity Index, ICU admission, ventilator usage and oxygen therapy (adjusted OR 1·94; 95% CI 1·02–3·70; P value <0·05).

Taken together, the actual proportion of patients requiring transfusion therapy is 1·2%, which is reassuring in that no extreme additional blood requirement for transfusion is expected due to a dramatic increase in COVID‐19 patients who might present unstable haemodynamics. Our results support the previous findings of a recent study by Doyle et al. reporting that blood component usage due to COVID‐19 remains low. As this study represents nationwide results rather than an analysis of hospitalized patients, blood demand in the entire COVID‐19 patient population, ranging from asymptomatic individuals to patients with severe symptoms, presented lower requirements of transfusion practice. A recent study by Berzuini et al. revealed greater transfusion requirements in COVID‐19 patients with direct antiglobulin test (DAT) positivity compared to those with DAT‐negative results. Although the authors suggested autoimmune haemolytic anaemia as the major underlying pathomechanism for RBC transfusion, our results indicate that the transfusion demand for RBC is higher than those for other blood components possibly due to anaemia during the COVID‐19 treatment period.

Coagulopathy is another haematological problem frequently reported in COVID‐19 patients. While high rates of venous thrombosis and disseminated intravascular coagulation are considered as risk factors for COVID‐19 severity, a recent study by Capecchi et al. reported acquired thrombotic thrombocytopenic purpura associated with COVID‐19. In line with the previous study, our results revealed that PC and cryoprecipitate demand was less than 35% among COVID‐19 patients who received transfusion. Anticoagulant medication prescribed before COVID‐19 diagnosis also did not result in either less mortality or lower transfusion demand in our study.

Furthermore, there is enough evidence supporting that a poor prognosis is associated with blood transfusion during COVID‐19 treatment independent of previously known risk factors. Although COVID‐19 infection status itself might naturally cause haemodynamic changes, severe imbalances requiring transfusion therapy are not observed remarkably more prevalently when compared with other infections. , As our Korean nationwide results might represent real‐world transfusion demand in countries with similar strict strategies for early detection of SARS‐CoV‐2 infection, more studies from countries with higher mortality and infection rates are urgently required.

Author contributions

JHR and JY designed the study. SAL performed statistical research. JHR performed data analysis. JHR, SAL, CHH, and JY wrote the manuscript. JY supervised the study.

Conflicts of interest

The authors declare to have no potential conflicts of interest regarding the present work.