Font Size: a A A

Building Intelligent Model Of Demand Of Allogenic Red Blood Cells In Total Hip Arthroplasty

Posted on:2018-07-12Degree:MasterType:Thesis
Country:ChinaCandidate:X L SunFull Text:PDF
GTID:2334330515961917Subject:Clinical Laboratory Science
Abstract/Summary:PDF Full Text Request
ObjectiveThe aim of study is to explore an effective, fast method to clean the Transfusion intelligent management and evaluation system by computer technology. It is possible to take advantage of effective data to do data mining, including to describe the transfusion situation of the patients undergoing THA and establish the predicted transfusion formula respectively, and then to explore the blood conservation strategies, aiming to promote the establishing and development of the refining, individualized transfusion scheme.Methods1?Maybe the scientific methods could be found by learning the theory of the data cleaning, and then process of data cleaning could be designed aiming at three stages.Combining the computer technology and manual work to deal with the problematic data,by "screen" "diagnose" "compile", and acquire the data type and format ultimately to make data analyzed and rinsed.2?Transfusion records are gained from the Transfusion intelligent management and evaluation system, retrospective analysis is made among 4250 cases from Year 1999 to 2015, including patients' entire information. Compare the differences of the allogenic red blood among years by SPSS software, and analyze the clinical transfusion situation for the past few years.3?After selecting 1064 patients' complete records who undergone THA, clean and summary the data of patients ,including the basic information, operative information, vital signs, testing indexes, and transfusion information. It can be got that the influencing factors associated with perioperative transfusion by consulting articles, and formula which predict the transfusion dosage by multivariate regression analysis and Artificial Intelligence preoperatively.4?The study was retrospective which contained the patients undergoing total hip replacement during the period January 2013 to June 2015 in the PLA hospital, and 181 patients were divided into two groups according to the usage of TXA, the TXA group contained 88, the control group contained 93. Compare the indexes between two groups,such as pre-or post-operative Hb, fall in Hb, mean number of allogenic packed cells transfused, intraoperative blood loss, postoperative drainage.Results1?In accordance with the data cleaning diagram, the data from the Transfusion intelligent management and evaluation system are dealt with "screen" "diagnose""compile",the data of 99 messages including basic information,transfusion information,vital signs, pre/post-operative testing information, which cover character, figure and so on.After getting all data, the necessary data of cases could be gathered on the basis of data research requirement.2?The rate of transfusion following hip joint operation was 55.72% of all patients.The mean transfusion dosage was 3.17±1.27 units. During 1999-2015 period,the mean length of stay decreased gradually. The transfusion rate was between 40%-70%approximately. Year 2006 has the highest mean transfusion dosage -4 units, and from year 2009, the mean application dosage was less than 3 units. Year 2009 and 2010 had the most RBC transfusion dosages of unilateral and bilateral hip arthroplasty respectively.3?We get total clean datum of 1064 cases undergoing THA from year 2000 to 2015,allogenic transfusion is associated with gender, age, operative procedure, BMI and preoperative Hb. According to multivariate regression analysis, NN, GBDT, it can be observed GBDT is optimum model.4?In comparison, there was no significant difference in preoperative Hb(P=0.850),intraoperative blood loss(P=0.072), postoperative drainage(P=0.214), and mean number of allogenic packed cells transfused(P=0.091),while there were significant differences in postoperative Hb, and fall in Hb.ConclusionIt is effective to utilize computer technology and data learning to clean the Transfusion intelligent management and evaluation system. It is effective that GBDT is used to predict transfusion volumn. Using of low dose tranexamic acid is effective in reducing the fall of postoperative Hb, while it couldn't reduce the requirement of allogenic blood transfusion in total hip replacement surgery. Maybe high dose tranexamic acid can be effective to reduce the transfusion requirement.
Keywords/Search Tags:Data cleaning, Transfusion intelligent management and evaluation system, AI, Total hip arthroplasty(THA), Transfusion, Tranexamic acid
PDF Full Text Request
Related items