A Transfusion Audit In An Ugandan Hospital

Abstract 4410

Over the past 20–30 years the quality of hospital transfusion services and the optimal use of blood components have been major preoccupations of hospitals in developed countries and are now beginning to be addressed in hospitals in developing countries. In order to assess the quality of transfusion practices in an university-affiliated hospital in Uganda and to identity areas of strength and areas requiring improvement we undertook an audit of all RBC, either whole blood (WB) or pediatric packed cell (PC), transfusions administered over a 72 hour week-day period in March 2010, at Mulago Hospital Complex (MHC), Kampala, Uganda. MHC, the teaching hospital affiliated with the College of Health Sciences, Makerere University, provides medical, surgical and obstetrical services; there are over 140,000 in-patient admissions per year of which approximately 15,000 are pediatric (i.e. < 12 year old) medical patients; over 30,000 deliveries are performed per year. MHC receives all its blood units from the Uganda Blood Transfusion Service (UBTS). MHC has 4 blood banks – the main hospital blood bank (MBB), an affiliated blood bank serving the pediatric medical wards (PBB) and independent blood banks serving the MHC cardiac and cancer institutes. This audit was limited to the MBB and the PBB. During the audit period the unit numbers of RBC units received by the MBB from the UBTS were recorded and for all RBC units recorded as leaving the MBB or the PBB the following information was recorded: whether or not a cross-match (XM) was performed, the ABO/Rh group of patients and units, the time of issue and time of beginning and ending the transfusion, the indication for the transfusion and the pre-transfusion (PT) hemoglobin (Hb) level, if performed. General observations on transfusion practices related to patient identification and documentation of transfusion in patient charts were made but not systematically collected for all transfusions. In total 225 units (WB 196, PC 29) were either present in the MBB/PBB or received from the UBTS during the audit period. Of these, 143 units (104 from the MBB, 39 from the PBB) were issued to the wards; 49 remained in the MBB or PBB at the end of the audit period; 33 units could not be accounted for (i.e. not recorded as issued but no longer in the MBB or the PBB). Of the 143 units issued, 102 were transfused (36 in the pediatric medical wards and the other 66 in other hospital wards), 5 were not transfused, 9 were issued to the cancer institute (so not further tracked), 11 were inadvertently not tracked and 16 proved impossible to trace after issuing (13 from MBB, 3 from PBB). In only 1 case was a group O unit transfused to a non-group O patient. For units issued from the PBB, XMs were performed for 87% of units while XMs were only performed for 54% of units issued from the MBB. The time from unit issue to the completion of transfusion was the similar whether or not a XM was performed – median (range) 5.4 (2.2-13.9) and 4 (2.3-6) hours for units issued from the MBB and PBB respectively. PT Hb levels were performed for 82% of transfusions in the pediatric medical wards; median (range) PT Hb level was 3.9 g/dL (1.9-9.6). A PT Hb level was obtained in only 13% of cases in the adult/sx/ob wards; median (range) PT Hb level was 5.9 g/dL (3.4-12.6). Indications for transfusion in the pediatric medical wards were: malarial anemia 64%, sickle cell anemia 18%, other specific anemias 8%, unable to determine 10%. Indications on the adult/sx/ob wards were: medical 39% (hematology 11%, malarial anemia 8%, other 20%), trauma/sx 31%, ob/gynecology 25%, unable to determine 5%. Only 5 patients received more than 1 unit (4 × 2 units and 1 × 3 units) within a 12 hour period or less. Among the informal observations, it was noted that labelling of blood specimens used in PT testing was often inadequate, patient identification was difficult to ensure because of lack of unique patient identifiers (e.g. birth dates are not recorded and the hospital does not use patient identification bands) and details of the transfusion (including timing, unit number) are often not recorded in the patient chart. In summary, in this sub-Saharan setting, it appears that RBC transfusions are likely usually given for appropriate indications (with the caveat that PT Hb levels are often not performed in adult patients) but that stricter controls regarding PT testing, patient and sample identification, unit administration, and documentation are required to ensure patient safety.