RESUMO
Background: Different user groups regard systematic reviews as reliable and valuable sources for answering research questions. For systematic reviews to fulfill their purpose, methodological quality in all stages are of importance. The studies identified in a systematic search form the basis of the review, thus the search process methodology is important for both performing and reporting the search. The purpose of the present study was to evaluate the quality of non-Cochrane systematic reviews by analyzing how they perform and report the search. This is exemplified by systematic reviews on eye movement desensitization and reprocessing (EMDR), a trauma-focused therapy commonly used for post-traumatic stress disorder (PTSD). Methods and Results: We examined the method chapters of 20 systematic reviews on the subject, and rated their searches and reporting using relevant elements from the Cochrane Handbook and PRISMA. We found inadequacies in the methods employed for searching and reporting the search strategy, which could have been avoided by greater adherence to guiding documents for performing systematic reviews. Conclusion: Our findings raise important questions for future debate on the risk of omitting studies, thus impairing the conclusions in a systematic review. For clinical purposes, researchers should investigate if, and how, the search strategy in a systematic review affects the body of knowledge and the results.
RESUMO
We have conducted a randomized controlled study where 164 patients were randomized to receive autologous salvaged blood collected by Sangvia™ Blood Salvage System or allogeneic red cell concentrates if transfusion was indicated by clinical judgement. The study was powered to detect if transfusion of autologous blood reduced the occurrence of postoperative infections. We found no statistical significant difference in postoperative infection rate between the groups, but this may be due to the fact that postoperative infections were diagnosed in only five patients. Increased C-reactive protein concentrations slightly above level of significance indicate that autologous blood transfusions stimulate the patient's immune system. However, there was no indication of increased transfusion reaction rate, including febrile reactions, in the autologous group. Transfusion of autologous blood did not reduce the use of allogeneic red cell concentrates. The mean use of allogeneic red cell concentrates was 0.93 units (both groups combined), indicating that the transfusion policy may have been too liberal. There was a highly significant inverse correlation between pre-operative haemoglobin concentration and transfusion of allogeneic blood. In a patient population with a low frequency of postoperative infection, a larger study is needed to clarify if autologous salvaged blood protects against postoperative infections.