Shifts of transfusion demand in cardiac surgery after implementation of rotational thromboelastometry-guided transfusion protocols: analysis of the HEROES-CS observational, prospective open cohort database
In this study, Kuiper et al. compared outcome after cardiac surgery in terms of bleeding, rethoracotomies, transfusion needs and economic costs during two separate periods: the intervention period with implementation of ROTEM-guided bleeding management and the control period with conventional laboratory tests to guide haemostatic interventions. The study gives a relevant description of an everyday implementation of new monitoring device/method, in contrast to data obtained from randomised clinical trials.
Observations confirm previously reported results after ROTEM implementation, namely less bleeding and decreased transfusion needs, in particular red blood cells. There was no difference in terms of rethoracotomies and mortality. Hospital length of stay and economic costs were considerably reduced.
There are some inborn limitations in the study design and translation of results. The proportion of patients receiving tranexamic acid was lower in the historic patient cohort, and the decision to administer tranexamic during the intervention period was independent of ROTEM results. Tranexamic acid has previously been shown to significantly reduce both bleeding and transfusion needs. Furthermore, the use of FFP was replaced to a large extent by the administration of fibrinogen concentrate. This tendency can be related to ROTEM measurements, but it also reflects a general change in practice in many cardiac centres to reduce the volume load and the need for blood products. It is therefore not unequivocal whether the observed decrease in bleeding complications and transfusion needs is related to ROTEM monitoring as it may also be caused to a certain degree by a change in practice, namely the increased use of antifibrinolytics and fibrinogen concentrate.
The study clearly demonstrates the difficulties associated with comparing two time periods during which potential confounders, such as a change in clinical practice, have occurred independently of the investigation of interest. However, the authors should be congratulated for evaluating the impact of ROTEM-guided interventions on everyday management. It would be interesting to see if data are reproducible in a setting where baseline management is comparable in the two study cohorts.
– Hanne Berg Ravn