Out for blood: causal inference in clinical transfusion research

Abstract

Blood products, by many measures, have evolved to be an extremely safe and fundamental part of hospital care. Moreover, they are a valuable resource that should be respected and safeguarded. In this thesis, we studied the relation of donor and product characteristics with patient outcomes in detail. By using thorough epidemiological methods, we found that there are still causes for concern pertaining to donor characteristics and transfusion recipient outcomes. In all chapters, in addition to describing the results and the most relevant aspects for clinical transfusion practice, we extensively described study limitations, and in a number of chapters we acknowledge that methodological limitations preclude causal claims. It should be noted that the goal of the research included in this thesis is not to criticize the use of blood products as a whole. Rather, the continuous improvement of a therapy’s safety and effectivity for those in need of it is always justified, and is the ultimate target of the research described here.