| 11 th Annual Pediatric Critical Care Colloquium |
| Session/Time | Technology/Monitoring/Sat, 10:00 - 11:00 AM |
| Title | Survey of Transfusion Practices in Pediatric Intensive Care Units |
| Author | C Laverdiere, F Gauvin, PC Hdbert, H Hume, BJ Toledano, C Infante-Rivard, L Maillard, and J Lacroix |
| Affiliation | Division of Pediatric Critical Care Sainte-Justine Hospital, Montreal, Quebec, Canada |
| Introduction | Physicians have traditionally used a hemoglobin (Hb) threshold of 10.0 g/dl as a threshold to transfuse real blood cells,, yet no scienttic rational for this threshold has been advanced in the past and even less'.data is present in pediatrics. We therefore wanted to describe the red blood cell transfusion practices of pediatric intensive care physicians and to characterize what clinical factors were detern-flnant in these practices. |
| Method | Scenario-based survey was distributed. and completed among Canadian and French-speaking European critical care physicians working in tertiay-care pediatric intensive care units. The 4 scenarios included a case of bronchiolitis, septic shock, a polytrauma a nd postoperative care of a corrective surgery for cyanotic congenital. cardiopathy. The transfusion threshold and how rnany cc/kg, would be transfused under the clivln conditions were evaluated. Physician and PICU characteristics were appraised such as are, gender, language spoken, academic rank, number, number of years in ICU practice, number of PICU beds, number of admissions per year and averaoe PRISM score. The questionnaire evaluated 15 determinants of red blood cell transfusions including; Hb concentration, age, gender, PaO2, lactic acidosis, PRISM 11 score, active bleeding, thrornbocytopenia, disseminated inlravscular coagulation and impending surgery. |
| Result | A response rate of 66'7o was obtained (155/325). The survey docun)eiited an association between year of graduation and Lransfusion practice; no other respondent characteristics was sicnif-icakit. No PICU characteristics was statistically significant. The univariate analysis demonstrated that diagnosis, age, PaO2, lactaterrda, PRISM score, active bieedini,, Lhrotiibt)cytor,,enia, DIC and surgery were considered determinant of transfusion threshold while the patient's gender and if the patient was the physician's son/daughter did not alter the transfusion threshold. We also found that the volume of red blood cells transfused was not related to the thresold level which suggests that red blood cells are not optimally utilized |
| Conclusion | The striking variation in practice patterns among critical care practitioners that we documented strongly supports the need for further clinical studies, these studies must at least take into account the determinants of transfusion reported in this survey. |
Use your browser's back button to return to the appropriate index of abstracts...
Back to PCCC 98 Abstract Introduction | Back to PCCC 1998