Blood transfusion utilization and recipient survival at Hospital das Clinicas in São Paulo, Brazil

Abstract

Background The characteristics of blood recipients including diagnoses associated with transfusion and post-transfusion survival are unreported in Brazil. The goals of this analysis were: 1) to describe blood utilization according to clinical diagnoses and patient characteristics at a large public hospital, Hospital das Clinicas (HC), a tertiary teaching hospital and trauma center in the city of Sao Paulo; 2) to determine the factors associated with survival of blood recipients. Methods A retrospective cross-sectional analysis was conducted on all inpatients in 2004. Data came from three sources were merged: HC electronic admission files, blood issue files, and the national death registry. The first two files consist of data about patient characteristics, clinical diagnosis, and transfusion information. Analyses comparing transfused and non-transfused patients were conducted. The third file was used to determine survival status of recipients up to three years after last transfusion. Logistic regression was conducted among transfused patients to examine survival curves and characteristics associated with follow up patient survival. Results In 2004, 30,779 patients were admitted to HC, with 3,835 (12.4%) transfused. These patients had 10,479 transfusions episodes, consisting of 39,561 transfused components; 16,748 (42%) red cells, 15,828 (40%) platelets and 6,190 (16%) plasma. The median number of components transfused was 3 (range 1 – 656) per patient admission. Mortality during hospitalization was dramatically different for patients whose admissions included transfusion or not (24% vs. 4%). After 1 year, 56% of transfusion recipients were alive. The multivariable model of factors associated with mortality following transfusion showed that the most significant factors in descending order were hospital ward, increasing age, increasing number of components transfused, and type of components received. Conclusion Ward and transfusion are markers of underlying medical conditions and are associated with the probability of survival. Platelet transfusions are common and likely reflect the types of patients treated at HC. This comprehensive blood utilization study, first of its kind in Brazil can help in developing transfusion policy analyses in South America.

Assessment of transfusion utilization and patient outcomes at the largest referral and university hospital in Addis Ababa, Ethiopia

“…In developed countries, on the other hand, plasma use was high, in Sao Paulo, Brazil (16%), Denmark (12Á6-19%) and Spain (6Á9%). Likewise, only 4 (1Á1%) of the transfusions in this study were platelets, whereas in Sao Paulo, Denmark and Sweden platelet transfusions as a per cent of all transfusions were 40%, 5Á6% and 7Á1%, respectively [12][13][14].…”

“…The results of this study in Ethiopia and noting the values above seen in other countries indicate a need for change in blood component usage in Ethiopia to maintain such therapeutic standards in blood component utilization. Similarly, the use of packed red blood cells in our study was only 5·5%; however, in Brazil, 42% of patients received red blood cells, Sweden (74%), Denmark (56–96%) and France (90%) .…”

Blood Component Use in a Sub-Saharan African Country: Results of a 4-Year Evaluation of Diagnoses Associated With Transfusion Orders in Namibia

“…Similar studies describing blood use have been conducted in developing countries but are generally limited in sub-Saharan Africa [19][20][21][22]. The few available studies from the region are restricted to specific wards or regional or tertiary hospitals, and only cover short periods [23,24].…”

Accuracy of a probabilistic record-linkage methodology used to track blood donors in the Mortality Information System database

“…Some of our studies have made exclusive use of the Fundação Pró-Sangue Hemocentro de São Paulo (FPS) blood bank donors’ and recipients’ routine secondary data, while others have made use of such data in conjunction with primary data collected within the context of the international collaboration Retrovirus Epidemiology Donor Study (REDS II) 9,10 and The Recipient Epidemiology and Donor Evaluation Study-III (REDS-III). We are now particularly interested in performing longitudinal studies in which we need to ascertain some long term outcomes (e.g.…”