The use of predictive model building genetic
algorithms and patient safety alert system - A Case Study of total parenteral nutrition
Abstract
Total budget in the health insurance more and more limited health care resources, how to use the hospital's medical database and information from the huge sums up the results can be accepted and the establishment of an effective model, not only to have more information of the data to help doctors determine and provide
appropriate nutrition care ability of patients to improve resistance to disease and reduce infection and thus enhance the efficacy of
treatment, but also a more efficient allocation of resources to maintain the quality of medical care.
Some can not get enough nutrition from the gut of patients by intravenous injection only way to obtain needed calories and all nutrients, total parenteral nutrition (TPN) is the five nutrients the body needs: carbohydrates, protein , fat, vitamins, minerals enter the body through the central venous catheter to maintain the body's metabolism. However, the nutritional status of patients during their treatment may have complications, hospital stay and survival rates are closely related. In fact, when the patient was caused by
infusion of total parenteral nutrition-related metabolic complications, treatment of disease will be more difficult to understand and prevent the early, regular clinical assessment is necessary.
In this study, since the conditions of the patients required the use of total parenteral nutrition, the last possibility of complications, with a retrospective chart review, an attempt to generalize the rules to construct for clinicians to use the prediction model and warning system, and look forward to the future of this area can improve the quality of care and can take into account the use of medical
resources.
