Modelling the risk of emerging infections for blood transfusion

Abstract

There is extensive documentation of transfusion transmission of a number of emerging infectious diseases (EIDs). This suggests that current measures might be insufficient to eliminate the risk of EID transmission through blood products, and additional measures may be required in case an EID outbreak occurs. Implementing such measures, even if technically possible or readily available, is costly and supporting evidence for such measures usually only becomes available after an outbreak. In order to determine whether to implement such steps, and which ones, an assessment of the seriousness of this risk is needed at the beginning, or even before an actual outbreak. Mathematical models have been increasingly used to support decision-making. In blood transfusion safety however, models are tailored to a specific disease outbreaks and existing ones can be of limited use for transfusion regulators or public health officials who require a rapid assessment. A generic model that is applicable to any EID, can be used to assess the impact of an on-going outbreak for blood transfusion specifically, and is readily available via an easy-to-use platform to help transfusion regulators or public health officials is called for. This thesis describes the development and application of models that quantify the risk posed by emerging infections for blood safety. These models aim to support a rational and proportional response to threats. The first model allows the ranking of the risk of blood transmissible infections as regards their potential threat to transfusion safety. The second model, the European Up Front Risk Assessment Tool (EUFRAT), quantifies the transmission risk of an EID through blood transfusion in two different scenarios: (1) transmissions by donors exposed to a local outbreak, and (2) transmission by donors returning from travels to an outbreak affected area. This latter model also allows evaluation of potential safety measures during EID outbreaks. In this thesis the development, application, and (partial) validation of this second model is extensively described and discussed. We provide an illustration of how the model is generally applicable to a number of outbreaks of EID with short infectivity (dengue, chikungunya) and presumed long infectivity (Q fever). To show how well the model performs we compared model estimates with data from an independent source. We also discuss how the model can be further enhanced to support decision making in transfusion safety.