Premium
A consensus‐based tool for ranking the risk of blood‐transmissible infections
Author(s) -
Oei Welling,
Neslo Rabin,
Janssen Mart P.
Publication year - 2016
Publication title -
transfusion
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.045
H-Index - 132
eISSN - 1537-2995
pISSN - 0041-1132
DOI - 10.1111/trf.13656
Subject(s) - medicine , disease , risk assessment , asymptomatic , infectious disease (medical specialty) , intensive care medicine , blood transfusion , transmissibility (structural dynamics) , ranking (information retrieval) , risk analysis (engineering) , immunology , computer science , machine learning , computer security , physics , vibration isolation , quantum mechanics , vibration
BACKGROUND Emerging infectious diseases (EIDs) pose a threat to blood transfusion safety. Despite a lack of evidence, safety interventions may be required. However, what should decision makers base their decisions on? A model was developed that allows valuing the perceived risk of an EID for blood safety as derived from a group of experts. The model requires estimates of four disease characteristics and the accuracy of these estimates. STUDY DESIGN AND METHODS Sixteen selected experts ranked 24 hypothetical diseases, each comprising a quantitative estimate of four characteristics: transfusion transmissibility, proportion of asymptomatic infectious phase, prevalence of infection, and disease impact. Each of the characteristics was expressed at one of six predefined levels with varying ranges of uncertainty. The model was derived using probabilistic inversion and was applied to value the perceived risk of most currently known EIDs relevant to blood transfusion. RESULTS The model demonstrated that transmissibility and prevalence are the most important risk drivers. However, disease impact and likelihood of transmission during the asymptomatic phase of infection are more important when the disease characteristics are unknown. In the ranking of currently known EIDs, diseases that have been identified previously as posing a serious risk to blood transfusion appear at the top of the list. CONCLUSION With the current model, the perceived risk of EIDs for transfusion safety can be determined for both known and unknown diseases, even when little information is available. Extension of the expert base, further model development and validation, and continuous updating of the model are recommended.