Length of Storage of Red Blood Cells and Patient Survival After Blood Transfusion: A Binational Cohort Study

TY - JOUR

T1 - Length of Storage of Red Blood Cells and Patient Survival After Blood Transfusion

T2 - A Binational Cohort Study

AU - Halmin, Märit

AU - Rostgaard, Klaus

AU - Lee, Brian K

AU - Wikman, Agneta

AU - Norda, Rut

AU - Nielsen, Kaspar René

AU - Pedersen, Ole B

AU - Holmqvist, Jacob

AU - Hjalgrim, Henrik

AU - Edgren, Gustaf

PY - 2017/2/21

Y1 - 2017/2/21

N2 - Background: Possible negative effects, including increased mortality, among persons who receive stored red blood cells (RBCs) have recently garnered considerable attention. Despite many studies, including 4 randomized trials, no consensus exists. Objective: To study the association between the length of RBC storage and mortality in a large population-based cohort of patients who received transfusions, allowing detection of small yet clinically significant effects. Design: Binational cohort study. Setting: All transfusion recipients in Sweden and Denmark. Patients: 854 862 adult patients who received transfusions from 2003 to 2012. Measurements: Patients were followed from first blood transfusion. Relative and absolute risks for death in 30 days or 1 year in relation to length of RBC storage were assessed by using 3 independent analytic approaches. All analyses were conducted by using Cox proportional hazards regression. Results: Regardless of the analytic approach, no association was found between the length of RBC storage and mortality. The difference in 30-day cumulative mortality between patients receiving blood stored for 30 to 42 days and those receiving blood stored for 10 to 19 days was -0.2% (95% CI, -0.5% to 0.1%). Even among patients who received more than 6 units of RBCs stored for 30 days or longer, the hazard ratio of death was 1.00 (CI, 0.96 to 1.05) compared with those who received no such units. Limitation: Observational study; risk of confounding by indication. Conclusion: Consistent with previous randomized trials, this study found no association between the length of storage of transfused RBCs and patient mortality. Results were homogeneous, with differences in absolute mortality consistently less than 1% among the most extreme exposure categories. These findings suggest that the current practice of storing RBCs for up to 42 days does not need to be changed.

AB - Background: Possible negative effects, including increased mortality, among persons who receive stored red blood cells (RBCs) have recently garnered considerable attention. Despite many studies, including 4 randomized trials, no consensus exists. Objective: To study the association between the length of RBC storage and mortality in a large population-based cohort of patients who received transfusions, allowing detection of small yet clinically significant effects. Design: Binational cohort study. Setting: All transfusion recipients in Sweden and Denmark. Patients: 854 862 adult patients who received transfusions from 2003 to 2012. Measurements: Patients were followed from first blood transfusion. Relative and absolute risks for death in 30 days or 1 year in relation to length of RBC storage were assessed by using 3 independent analytic approaches. All analyses were conducted by using Cox proportional hazards regression. Results: Regardless of the analytic approach, no association was found between the length of RBC storage and mortality. The difference in 30-day cumulative mortality between patients receiving blood stored for 30 to 42 days and those receiving blood stored for 10 to 19 days was -0.2% (95% CI, -0.5% to 0.1%). Even among patients who received more than 6 units of RBCs stored for 30 days or longer, the hazard ratio of death was 1.00 (CI, 0.96 to 1.05) compared with those who received no such units. Limitation: Observational study; risk of confounding by indication. Conclusion: Consistent with previous randomized trials, this study found no association between the length of storage of transfused RBCs and patient mortality. Results were homogeneous, with differences in absolute mortality consistently less than 1% among the most extreme exposure categories. These findings suggest that the current practice of storing RBCs for up to 42 days does not need to be changed.

KW - Adolescent

KW - Adult

KW - Aged

KW - Aged, 80 and over

KW - Blood Preservation/adverse effects

KW - Databases, Factual

KW - Denmark

KW - Erythrocyte Transfusion/mortality

KW - Female

KW - Humans

KW - Male

KW - Middle Aged

KW - Proportional Hazards Models

KW - Rh-Hr Blood-Group System

KW - Sweden

KW - Time Factors

KW - Young Adult

U2 - 10.7326/M16-1415

DO - 10.7326/M16-1415

M3 - Journal article

C2 - 27992899

SN - 0003-4819

VL - 166

SP - 248

EP - 256

JO - Annals of Internal Medicine

JF - Annals of Internal Medicine

IS - 4

ER -