Quantitative prediction and intervention system method for long-term prognosis of patient

Abstract

The invention discloses a quantitative prediction and intervention method for long-term prognosis of a patient. The method comprises: newly added patient information is inputted to update a database;a risk factor is selected by using a training database; COX risk grading and LCA risk grading are combined to obtain a comprehensive risk grade of the patient and a risk score calculation method of the patient is determined; a comprehensive risk grade and a risk score of the patient are calculated; and when the patient is discharged from hospital, the doctor carries out intervention based on the comprehensive risk grade and the risk score. In addition, the invention also provides a quantitative prediction and intervention system for long-term prognosis of a patient. The method and system havethe following beneficial effects: variable selection is carried out and estimation randomness is considered fully; the model is reliable; correction and supplementing are carried out by using the LCArisk grading and COX risk grading, so that grading focuses on the own characteristics of the patient; the grading method is reliable; the influence of low reliability of occurrence of adverse events of the patient for long time is eliminated; discharge education and doctor intervention are carried out on the patient effectively; and thus the risk of the patient after discharge from the hospital isreduced.

Description

technical field [0001] The invention relates to a long-term prognosis quantitative prediction and intervention system and method for patients. Background technique [0002] Patients remain at risk after discharge from the hospital. Studies have pointed out that in the United States, about 20% of patients will be re-admitted within 30 days after discharge, and 34.3% of Chinese patients with heart failure will be re-admitted or die within one year. Lack of effective interventions at discharge and lack of continuum of care after discharge is a major source of adverse events. Therefore, it is necessary to study the risk factors of patients after discharge and take systematic intervention measures. [0003] Existing studies on the risk of patients after discharge are limited to the establishment of predictive models, and mainly focus on short-term predictive models of death. The disadvantages of using this method are: (1) Generally, the traditional logistic regression model st...

AI Technical Summary

Problems solved by technology

The disadvantages of using this method are: (1) Generally, the traditional logistic regression model step-by-step method is used to select risk factors. Once a variable is eliminated at a certain step, it may not have the opportunity to enter the model again. The criteria for variables are related to the criteria for allowing variables to enter, and important variables may be missed, resulting in an unstable selection model for this method; (2) Use the actual observations of whether adverse events occur after the patient is discharged to establish a model, and the premise is that all The observed adverse events are reasonable, but in the long-term model, this assumption itself has great limitations. Compared with the short-term, whether patients will experience adverse events in the long-term has greater uncertainty, it is very likely Adverse events occurred in patients who were discharged in good physical condition, whereas no adverse events occurred in patients who were discharged in poor physical condition

The model is built only based on the observed outcome events, without taking into account the uncertainty of the long-term occurrence of adverse events, which may result in the model being unable to capture the essential characteristics of the data

Images