Methodological challenges in observational transfusion research: lessons learned from the Scandinavian Donations and Transfusions (SCANDAT) database. (14th December 2016)

• Record Type: Journal Article
• Authors: Edgren, G.; Rostgaard, K.; Hjalgrim, H.
• Is Part Of: ISBT science series. Volume 12:Number 1(2017)
• Page Start: 191
• Page End: 195
• Publication Date: 2016-12-14
• Subjects: epidemiology -- transfusion ‐ transmissible infections -- transfusion medicine (in general); Blood -- Periodicals; Blood -- Transfusion -- Periodicals; Immunohematology -- Periodicals; Immunopathology -- Periodicals; 615.39
• DOI: 10.1111/voxs.12342
• Languages: English
• ISSNs: 1751-2816

Abstract

Abstract : Background and Objectives: An increasing number of investigations into different aspects of transfusion therapy are being conducted using data extracted from a wide range of electronic data sources. Results of such observational studies, however, have in some instances been conflicting, possibly as an effect of variable attention to methodological and statistical detail. In this review, we will introduce and discuss some recent examples of important methodological problems. Materials and Methods: We reviewed the recent literature of observational studies of possible deleterious effects of aspects of transfusion therapy, including donor factors, donation factors and component factors. From this literature, we identified a number of potential methodological problems and assessed these using data from the Scandinavian donations and transfusions (SCANDAT2) database. Results: The main finding of our review relates to the importance of maintaining sufficient attention to epidemiological and statistical methods. The most important factor to consider is perhaps the number of transfusions which is often found to be an important confounding factor in this field. We also review recently applied methods for studying disease transmission through blood transfusion. Conclusion: While observational research in the transfusion medicine field has great potential for improving transfusion safety, pitfalls are plentiful and easy to overlook. In this review, we have gone over some methodological issues that we feel are especially important to consider for researchers who are involved in this type of research.