[Evaluation of erythrocyte concentrate prescription in a university hospital center].

This prospective study, based on declaratory data, evaluates the appropriateness of red blood cell transfusion prescriptions in a university hospital. Local recommendations written after data collection and the analysis of prescriptions using a blinded method limited the bias related to the declaratory data. The results show that the rate of unjustified prescriptions is 4.2% (95% CI: 2.2%; 6.2%). This rate is statistically (P = 0.032) lower in the department of surgery (1.3%) than in the department of medicine (5.7%). This rate tends to decrease according to the experience of the prescriber (P = 0.06) and varies significantly according to the hemoglobin levels (P = 0.03). The logistic regression, integrating these three parameters, confirms that only the hemoglobin level is significantly related (P < 0.003) to the appropriateness of RBC transfusions. This study also highlights problems not linked to prescriptions, and the hospital created a quality assurance program as a result.