Gaps in neonatal blood transfusion data highlighted
MedWire News: A review of trials evaluating red blood cell (RBC) transfusions in neonates has identified several areas of uncertainty regarding optimal practice.
Many of the 27 trials examined by Vidheya Venkatesh (Cambridge University Hospitals, UK) and colleagues failed to report on primary neonatal clinical outcomes, including mortality, neurologic development, and respiratory morbidity (ie, chronic lung disease [CLD]), which would be considered of primary importance to clinicians.
"The design of future RCTs [randomized controlled trials] can be informed by the lessons from this review," they remark in the British Journal of Haematology.
The researchers systematically reviewed RCTs of RBC transfusion in neonates, as this is a common form of therapy for premature and extremely low-weight babies.
They identified three trials comparing transfusion with placebo or no transfusion, in a total of 97 patients, but none reported on the clinically important outcomes of mortality, neurologic development, or CLD.
Of four RCTs that compared different doses or administration schedules in a total of 136 patients, only one reported on mortality, with no significant differences observed. No other clinically important outcomes were reported.
There were six trials, enrolling 679 neonates, which compared transfusion thresholds. Meta-analyses of these trials showed that there were no significant differences between liberal and restrictive thresholds in mortality (relative risk [RR]=1.22) or CLD (RR=0.99). Two trials assessed neurodevelopment outcomes, but with conflicting results.
The researchers note that the confidence intervals in the meta-analyses were wide and encompassed clinically meaningful differences, which "combined with the uncertainties in neurological events and neurodevelopmental follow-up, reiterates the pressing need for further research to define optimal thresholds for safe RBC transfusion in neonates."
The largest group of RCTs evaluated different RBC products, including leukocyte-reduction (three trials with 70 neonates), or the use of different storage or dilution media (seven trials with 221 neonates). These trials were generally small (overall mean sample size = 30) and did not report on clinically relevant outcomes. Furthermore, the reporting of adverse events was inconsistent.
"In summary, this comprehensive review has identified many areas of uncertainty in optimal neonatal RBC transfusion practice," Venkatesh and co-authors remark.
They conclude that endpoints in future studies need to include long-term neonatal neurodevelopmental outcomes, as well as standardized neonatal definitions for all relevant adverse events.
By Laura Cowen