This article is more than 5 years old. Information may no longer be current.
Transfusion timing, volume do not affect mortality among African children with severe anemia
Click Here to Manage Email Alerts
Administering 20 mL/kg or 30 mL/kg blood transfusions immediately produced similar clinical outcomes over 6 months among African children with severe anemia compared with a triggered strategy, according to results of two randomized trials published in The New England Journal of Medicine.
“[Shortages] of blood are a huge problem [in sub-Saharan Africa], with many children dying while awaiting a blood transfusion,” Kathryn Maitland, MD, PhD, professor of tropical pediatric infectious diseases and director of Centre of African Research and Engagement at Institute for Global Health Innovation of Imperial College in London, told HemOnc Today. “The bottom line is that many countries there have less than five units per 1,000 population donated. WHO estimates this should be above 20 units per 1,000 population.”
WHO recommends not performing transfusions for African children with uncomplicated severe anemia, defined as a hemoglobin level of 4 g/dl to 6 g/dl and no signs of clinical severity. Once hemoglobin drops below 4 g/dl, WHO recommends a 20 mL/kg transfusion, regardless of hemoglobin level.
Because of high in-hospital mortality rates, Maitland and colleagues conducted four randomized trials — two of which they are currently reporting — to assess immediate vs. triggered transfusion, transfusion volume, postdischarge supplementation with micronutrients, and postdischarge prophylaxis with trimethoprim-sulfamethoxazole.
Immediate vs. triggered transfusion
In a randomized, controlled trial, Maitland and colleagues randomly assigned 1,565 Ugandan and Malawian children aged 2 months to 12 years (median age, 26 months) with uncomplicated severe anemia to immediate transfusion (n = 778) with 20 ml/kg or 30 ml/kg of whole-blood equivalent or to no immediate transfusion (control group; n = 787).
Patients in the control group received a 20-mg/kg transfusion when their hemoglobin dropped to below 4 g/dl or they demonstrated new signs of clinical severity.
More than half (n = 984; 62.9%) of the children had malaria.
Mortality at 28 days served as the study’s primary endpoint.
During primary hospitalization, all children in the immediate group underwent transfusion (median time to transfusion, 1.3 hours) compared with 49% (n = 386) of children in the control group (median time to transfusion, 24.9 hours).
Mean total blood transfused per child was 314 ± 228 ml in the immediate group and 142 ± 224 ml in the control group.
After 28 days, seven patients in the immediate group and 13 patients in the control group had died (HR = 0.54; 95% CI, 0.22-1.36). After 180 days, 35 patients in the immediate group and 47 patients in the control group had died (HR = 0.75; 95% CI, 0.48-1.15).
Mean length of hospital stay was 0.9 days longer in the control group compared with the immediate group.
Mean hemoglobin levels were 2.42 g/dL higher after 48 hours in the immediate group than in the control group.
No differences in readmission rates or serious adverse events were observed.
“Overall, there was no evidence of differences in clinical outcomes between the children who received immediate transfusion and those who did not,” Maitland and colleagues wrote. “Although we cannot rule out the possibility of a small mortality benefit with immediate transfusion, mortality was very low with both strategies, and immediate transfusion failed to lower the rate of readmission during the following 6 months, which remained high in both groups.”
Transfusion volume
In a factorial, open-label trial, Maitland and colleagues randomly assigned 3,196 Ugandan and Malawian children aged 2 months to 12 years (median age, 37 months) to receive an immediate transfusion of 30 mL/kg (high-volume group; n = 1,598) or 20 mL/kg (low-volume group; n = 1,588). All children had a hemoglobin level of less than 6 g/dL and severity features, and 2,050 (64.1%) had malaria.
Mean volume of total blood transfused was 475 ± 385 mL in the higher-volume group and 353 ± 348 mL in the lower-volume group.
Mortality at 28 days served as the study’s primary endpoint.
Results showed that 55 children (3.4%) in the high-volume group and 72 (4.5%) in the low-volume group died before 28 days (HR = 0.76; 95% CI, 0.54-1.08), which did not reach statistical significance.
Mortality rates remained comparable at 90 days (high-volume group, n = 93; low-volume group, n = 114; HR = 0.81; 95% CI, 0.61-1.06) and 180 days (n = 134 vs. 154; HR = 0.86; 95% CI, 0.68-1.08).
Time until discharge from the hospital was shorter in the high-volume group than in the low-volume group (HR = 1.12; 95% CI, 1.04-1.2).
Researchers observed significant differences in outcomes according to the presence or absence of fever — greater than 37.5°C — at screening (P = .001). Among 1,253 children who had a fever at the time of transfusion, mortality was higher in the high-volume group (HR = 1.91; 95% CI, 1.04-3.49). Among 1,943 children who did not have a fever, mortality was lower in the high-volume group (HR = 0.43; 95% CI, 0.27-0.69).
“In most children who had a fever at screening ... it persisted over the 8 to 10 hours of admission despite receiving antibiotics, antimalarials and antipyretics,” Maitland told HemOnc Today. “It is not OK to wait until the fever has abated. These children are very critically sick will likely die before fever defervescence.
“We have no explanation for this, but what we do have is very strong scientific evidence with statistical confidence that the transfusion volume is working differently in the two groups,” Maitland added. “Higher [volume] is better for those without a fever but is harmful in the group with a fever.”
Overall, researchers observed no differences in readmission rates, serious adverse events or hemoglobin recovery at 180 days.
“It is important that neither the use of whole blood nor longer storage age of donor blood adversely affected 28-day or 180-day mortality,” Maitland and colleagues wrote. “Thus, component preparation, recently introduced to blood-transfusion services in sub-Saharan Africa at a substantial cost, does not appear to be essential for safe transfusion practice.”
No change in practice
Because the results of both studies were negative, there does not appear to be a reason to change transfusion practices.
Whether these results are relevant to children outside of sub-Saharan Africa remains to be seen. However, it is possible that these strategies could help reduce costs while maintaining good care for children living in sub-Saharan Africa, Julie R. Ingelfinger, MD, pediatrician and senior consultant in pediatric nephrology at Massachusetts General Hospital, wrote in an accompanying editorial.
“The primary outcomes in the first and second randomizations in the trial were negative, suggesting that, given the randomization groups, there is no credible reason to transfuse immediately or to transfuse a higher volume of blood, at least in pediatric populations in regions such as these two sub-Saharan countries,” Ingelfinger wrote. “Perhaps such information might conserve funds and still support good care.” – by John DeRosier
For more information:
Kathryn Maitland, MD, PhD can be reached by email at kathryn.maitland@gmail.com.
Disclosures: Maitland reports no relevant financial disclosures. Please see the studies for all other authors’ relevant financial disclosures. Ingelfinger reports no relevant financial disclosures.
Perspective
Back to Top
Jed Gorlin, MD
I had the honor of working in Rwanda to help rewrite its transfusion guidelines as part of the President's Emergency Plan for AIDS Relief (PEPFAR) program. Rwanda was ahead of other sub-Saharan African countries because it had satisfactory collection and testing, but the blood supply was still nowhere near where it needed to be. Knowing how precious blood was, we came up with the transfusion trigger of 6 g/dL of hemoglobin, which is a whole gram lower than in the United States. If you don’t have enough blood, you have to use it wisely.
That gives important background to these studies. We know very few people who have anemia die because of hemoglobin above 5 g/dL. You might die by loss of blood volume before loss of hemoglobin concentration. A lot of people die because their hemoglobin is 3 g/dL or below and they can’t compensate anymore; they are already pumping blood as fast as they can.
A study conducted in 2015 tested children in Uganda with severe anemia with hemoglobin below 5 g/dL and lactates above 5 mmol/L. Transfusion helped regardless of whether the blood was old or fresh. That proved older blood could work in places where there are absolute emergencies.
In the study by Maitland and colleagues of immediate vs. triggered transfusion, immediate transfusion was done in the control group, in which half of children underwent transfusion anyway. Children who had lower hemoglobin by far were transfused more to start.
Had they done this study the exact same way with hemoglobin of 3 g/dL to 5 g/dL, no doubt the children would do better. The problem is a study like that is not ethical because some children would die and, by the time you find children that low, they may be even lower.
Many transfusions in Africa happen by looking at how pale a child is, making an intelligent guess and transfusing. If you are in an ideal situation with a rapid turnaround of tests, then blood will be used more wisely. The turnaround time matters for tests. My guess is that this study was funded to a point where they had significantly better-than-average turnaround times. But in real-life settings in Africa, immediate access to labs that make that decision is not a guarantee.
The other thing I was confused about was that the researchers did not appear to analyze those who got transfused vs. those who didn’t. They compared the two groups, but 50% of the delayed transfusion group got transfused reasonably rapidly, so I think there should be a subgroup analysis.
Regarding the other paper on volume, in the United States, the vast majority of red cell transfusions are packed red cells. In parts of Africa where you don’t have centrifuges or electricity, you leave the blood whole, which further complicates things when you don’t have enough blood on the shelf.
The problem with 20 mL/kg vs 30 mL/kg is that if you have a big boy who is 100 kg, 30 kg would be 3 L. No one would do that, particularly with whole blood.
When you think about the fluid volume, the way you compensate for severe anemia is by increasing the blood volume to pump it around as fast as possible. But if you do it too fast, you put the patient at risk for cardiac overload and congestive heart failure. Transfusions also are about rate. If you transfuse slowly enough and the kidneys are working well enough, then the extra volume may be well-tolerated.
The results show the children with fever did better with lower volume, whereas children without fever did better with higher volume. Children with fever already have a higher heart rate, so the heart is going to have less ability to pump faster.
Overall, 20 ml/kg vs. 30 ml/kg isn’t that much of a difference; all you have to do is get the hemoglobin above 5 g/dL and the children will do well. I wouldn’t have predicted it would have shown any difference.
Blood collection absolutely has increased in countries where PEPFAR or WHO programs have been implemented. I’m sad to say that PEPFAR — for all practical purposes — has ended. The problem with “America First” is that everyone else is last, so — if anything — things have begun to slip over the last 2 years. WHO says if you are not collecting 10 units per 1,000 population, you don’t have enough blood. In most of these Central African countries, this number is three or four units per 1,000, which is not even close to what they need. The problem is that they don’t have the infrastructure for it.
Rwanda, which has enough blood, has begun to deliver it by drones. There are people doing heroic things, but resources are lacking.
I’d like to congratulate the authors for doing these studies. It’s hard enough to do a study at a hospital here in America, but it’s particularly hard in places where resources are limited. It shows what good science and good clinical care can do when they are given the resources.
Dhabangi A, et al. JAMA. 2015;doi:10.1001/jama.2015.13977.
Read more about
Click Here to Manage Email Alerts