IJCS | Volume 33, Nº3, May / June 2020

236 Lima et al. Unfractionated bovine heparin safety Int J Cardiovasc Sci. 2020; 33(3):235-242 Original Article anaphylactic reaction with facial edema, hypotension, tachycardia, nausea, urticaria, dyspnea, and death. 5,7,8 Unlike other countries, Brazil presented a lower risk of bovine spongiform encephalopathy and continued to use bovine heparin without significant concerns. Bovine heparin currently represents 40% of the heparin market in Brazil. 5 Bovine and porcine heparins differ in biological and pharmacological aspects and, consequently, their effects on blood coagulation. 3,5,9,10 Most available data concerning the safety of bovine heparin comes from lung samples, which fell by the wayside with in the 1980s. 3 Heparin from bovine lung has a higher risk of heparin- induced thrombocytopenia. 9 Bovine and porcine intestinal heparin have a similar molecular weight, but different anticoagulant and antithrombotic properties (porcine has higher activity). 3,10,11 A higher dose of bovine heparin is necessary in order to achieve similar effects, and higher protamine dose for neutralization. 3,12 A study in animals suggests that bovine heparin presents higher risk of bleeding. 3 The use of bovine UFH is on the rise, and more data concerning safety in humans is needed. This observational study reports the safety profile of bovine UFH in patients undergoing on-pump cardiac surgery. Methods This study was performed following Good Clinical Practices and in compliance with the Declaration of Helsinki of 1975, revised in 2008. The local Institutional Review Board approved the study protocol. We performed a single-center, descriptive, retrospective investigation using data collection from all patients who had on-pump cardiac surgery, using bovine UFH (supplied by Eurofarma Laboratories S.A.) between October 2008 and November 2012 and porcine UFH (supplied by Cristália and Blausiegel Laboratories) between June 2013 and December 2014 at the Heart Surgery Institute of Hospital Bom Jesus in Ponta Grossa, Paraná, Brazil. Patients were excluded from the study if they had received intravenous UFH two hours before the surgical procedure, and/or low molecular weight heparin or fondaparinux subcutaneously or oral anticoagulants within 12 hours before the surgical procedure, and/or fibrinolysis 48 hours before the surgical procedure. They were also not included in the absence of preoperative platelet count results, and/or cardiac surgery performed in the presence of severe sepsis, with high risk of widespread intravascular coagulation. Eligible patient data were collected from medical records by trained reviewers under the investigator’s supervision and were anonymized and stored in an electronic database instrument mainly designed for this purpose. Pre and postoperative data were collected, and the postoperative period was covered until hospital discharge or the seventh postoperative day, whichever occurred first. A questionnaire for assessing the quality of the records was also filled out for each medical history reviewed. Study outcomes were bleeding, thrombocytopenia defined as platelets counts lower than 150,000), postoperative vasoplegia, activated clotting time (ACT) shorter than 400 seconds and coagulation phenomena, such as blood clotting with bovine UFH. Data on porcine UFH were also collected only as exploratory means. Statistical analyses The present study is a preliminary study of the feasibility of a descriptive nature only. We did not calculate the sample size because it was an exploratory study. Therefore, we considered all patients who met the inclusion criteria. All descriptive analyses were stratified by period in which each type of heparin was used in our service (October 2008 to November 2012: bovine UFH, and June 2013 to December 2014: porcine UFH). Continuous variables were described by the number of participants evaluated. Mean and standard deviation and median and values range ​observed or quartiles Q1 and Q3, according to their distribution defined by the Shapiro-Wilk test, with alpha level of 0.05. Categorical variables were summarized by frequency distribution. The 95% confidence intervals (95% CI) were calculated for the one-time estimates. Results From 2008 to 2014, 790 patients underwent surgery at our medical service. Of these, we considered 428 (54.2%) for analysis (completedmedical record), and 269 (62.9%) were eligible for the study, meeting inclusion and exclusion criteria. Of the patients, 75.8% were treated with bovine UFH and 24.2% were treated with porcine UFH.