Objectives: In 1995, a fatal transfusion complication led to the developmen
t of a quality-assurance programme (QAP). Initial analysis showed important
deficiencies in practice patterns for both nurses and physicians. No writt
en guidelines had been previously produced.
Study design: Evaluation of practice patterns and quality-assurance program
me.
Methods: A QAP was prepared and implemented, a booklet describing completel
y the transfusion process and institutional policies and providing practica
l information was edited and the medical and non-medical personnel was info
rmed and trained on a regular basis. Implementation was evaluated during a
three-year period by measuring previously defined surrogate endpoints of th
e quality of the transfusion process.
Results: Implementation became rapidly close to 100% either for well-accept
ed requirements (i.e. patient's identity reporting on transfusion forms) or
when physicians were highly motivated (i.e autologous blood transfusion).
For other indicators, implementation progressed more slowly (i.e. traceabil
ity of labile blood products) although significant efforts were made to imp
rove the computerised system of data collection and although hospital admin
istration showed a very positive behaviour. Finally, other indicators showe
d only partial implementation. Several practical barriers (perceived lack o
f time, excessive increase in writing requirements) were described for tran
sfusion checklist or transfusion prescription form but more in-depth insuff
iciencies (patients' information) were also observed.
Conclusion: The success of a change of practice patterns relies on hospital
administration positive role, education and feedback, written and immediat
ely available guidelines, employment of specially trained personnel, long-s
tanding actions. Future accreditation of hospitals based on well-defined an
d well-implemented procedures will also be a major help to increase the qua
lity of the transfusion process. (C) 2001 Editions scientifiques et medical
es Elsevier SAS.