Construction method and system for patient in-hospital bad prognosis prediction model and application

Abstract

The invention belongs to the technical field of model construction, and discloses a method and a system for constructing a hospital adverse prognosis prediction model of a patient and application of the method. Plasma cell factors of a diagnosed patient of CSA-AKI are measured; the method comprises the following steps: identifying a key blood predictive factor of a complex end point event in a bad outcome in a hospital by adopting LASSO regression and a random forest quantile classifier RFQ, internally verifying a model containing the selected predictive factor through a bootstrapping re-sampling method, and externally verifying in an independent queue. The model provided by the invention still has a good discrimination effect when being applied to a verification queue, and a column line diagram is made to be convenient for clinical use. Decision curve analysis shows that the new model increases more net benefits in the aspect of evaluating the adverse hospital outcome risk of the patient compared with the new model which purely adopts the change value of serum creatinine.

Description

technical field [0001] The invention belongs to the technical field of model building, and in particular relates to a method, system and application for building a predictive model for poor prognosis of patients in a hospital. Background technique [0002] Currently, cardiac surgery-associated acute kidney injury (CSA-AKI), defined as sudden renal dysfunction after cardiac surgery, is a common adverse postoperative complication, occurring in approximately 5-42% of patients undergoing cardiac surgery . Despite remarkable progress in the early recognition of CSA-AKI, a large number of patients with CSA-AKI have progressively worsening renal function until renal replacement therapy is required, and the length of stay in the intensive care unit (ICU) of these patients and high mortality. Therefore, identifying the risk of such adverse in-hospital outcomes in patients with CSA-AKI can aid in patient management and the development of new therapies. . [0003] However, changes ...

AI Technical Summary

Problems solved by technology

[0006] (1) Inflammatory markers such as IL-6, IL-4, IL-8, IL-18, IL-10, TNF-α, IFN-γ, and MCP-1 show potential for early identification of disease, but their prognosis Research on predictive power is limited

[0007] (2) There is almost no comprehensive research in the prior art, evaluating a variety of blood immune and inflammatory cytokines and identifying key cytokines combined with common clinical indicators for predicting the prognosis of adult patients with CSA-AKI

Images