Please use this identifier to cite or link to this item: http://hdl.handle.net/11718/14120
Full metadata record
DC FieldValueLanguage
dc.contributor.authorJhajharia, Smita-
dc.contributor.authorPal, S.K.-
dc.contributor.authorVerma, Seema-
dc.contributor.authorKumar, Manish-
dc.date.accessioned2015-07-14T12:05:17Z-
dc.date.available2015-07-14T12:05:17Z-
dc.date.issued2015-
dc.identifier.citationJhajharia, S., Pal, S.K.,Verma, S., & Kumar, M. (2015). Predictive Analytics For Better Health And Disease Reduction. 1st IIMA International Conference on Advances in Healthcare Management Services. Indian Institute of Management, Ahmedabaden_US
dc.identifier.urihttp://hdl.handle.net/11718/14120-
dc.description.abstractPredictive analytics can be used effectively to evaluate enormous data generated by health care industry to extract useful information and establish relationships amongst the variables. In our country, health care providers have just began to hear of predictive analytics but are rapidly becoming aware that they have to make changes as the health care industry demands are changing. Unlike traditional statistical methods for data evaluation, Predictive Analytics uses artificial intelligence like statistical methods to reveal surprising associations which doctors would never even suspect. Hospitals, pharmaceutical companies and insurance providers will see changes from past treatment outcomes, latest medical research and databases like fewer complications, shorter hospital stay, fewer readmissions. We have chosen a cardiac surgery Centre in New Delhi, where about 650-700 children with cardiac defects are operated every year, out of which about 1/3rd have tetralogy of fallot cardiac defect. Tetralogy of fallot is most common cyanotic congenital heart disease which comprise of VSD, aortic override, pulmonary stenosis, RVH. Firstly, we have selected important clinical features, i.e., Age, sex, Prematurity, Nutritional Status, Hemoglobin and Aristotle score which can affect post operative ICU stay. Secondly, we are using Data Mining techniques to evaluate the particulars of each patient. Results: 450 patients underwent cardiac surgery for tetralogy of fallot from 2011 to 2013. Multiple linear Regression model identified age, male sex (p<0.042), malnutrition (p<0.020), prematurity (p<0.028) and higher hemoglobin (>21g/dl) (p<0.035) as independent factors predictive of increased ICU length of stay. When these five factors were analyzed in a regression model, the age (p<0.001) and Aristotle score (p<0.001) variable emerged as the strongest predictor of length of stay. Conclusions: Although patient factors were influential, the age was the most important determinant of ICU length of stay after cardiac surgery. It may be possible to reduce length of ICU stay by identifying ideal age of patients to undergo cardiac surgery and encouraging surgeons to take sex, history of prematurity, hemoglobin levels in consideration before planning surgery for best outcomes.en_US
dc.language.isoenen_US
dc.publisherIndian Institute of Management, Ahmedabaden_US
dc.relation.ispartofseriesIC 15;040-
dc.subjectCardiac Surgeryen_US
dc.subjectPredictive Analyticsen_US
dc.subjectICU Stayen_US
dc.subjectMultiple Linear Regression Modelen_US
dc.subjectAssociative Classificationen_US
dc.titlePredictive Analytics For Better Health And Disease Reductionen_US
dc.typeArticleen_US
Appears in Collections:1st IIMA International Conference on Advances in Healthcare Management Services

Files in This Item:
File Description SizeFormat 
CMHS-IC-15-040.pdf658.23 kBAdobe PDFView/Open


Items in IIMA Institutional Repository are protected by copyright, with all rights reserved, unless otherwise indicated.