TW201430754A - Using the simple parameter scoring system to predict nosocomial infection model - Google Patents

Abstract

The invention is to provide a high accuracy rate of nosocomial infections prediction system which is suitable for the prediction of nosocomial infection in hospitalized patients, to assess effectively the patient's condition and to assess their infection status, in order to facilitate the physician to pre-disposal to reduce the risk of infection and improve patient safety, further reduce medical waste. Nosocomial infections are closed related with increased morbidity, mortality, prolonged hospitalization, and medical costs. A simply, reliable prediction of nosocomial infections prediction model has very important clinical significance. Therefore, this invention is to develop a Logistic Regression to derive the scoring system to predict nosocomial infections, This method creates a simple scoring system which compared with a complex artificial neural network model it result no statistical differences, hence verified its usefulness. This simple scoring system consists of seven input variables: nasogastric tubes, urethral catheters, arterial catheters, central venous catheters, use of systemic corticosteroids, stress ulcer prevention medications and hemodialysis. The required information of this seven variables can be provided via the hospital information system database, and produce a set of ratings data via the simple scoring system, used to predict the possibility of nosocomial infection, the higher the score, the higher the possibility of infection, and the lower the score the lower possibility of infection. The criteria for this score may be defined according to different hospitals, and to let the medical staffs to utilize this simple system actually in clinical care.

Description

A model for predicting nosocomial infections using a simple parameter scoring system

The present invention provides a predictive degree of risk of nosocomial infection in hospitalized patients, and can provide preventive care measures at an early stage to avoid and reduce disease progression.

The hospital's in-hospital infection status is predicted to effectively assess the patient's condition and assess the status of the infection, so that the physician can pre-dispose to reduce the risk of infection and improve patient safety, further reducing medical waste. Nosocomial infections are associated with increased morbidity, mortality, prolonged hospital stay, and medical costs. At present, the prevention and treatment of nosocomial infections is often judged based on the past experience of the hospital medical team. There is no quantitative evaluation concept of the application of information systems to the risk level of infection among different patients.

The present invention provides a predictive degree of risk of nosocomial infection in hospitalized patients, and can provide preventive care measures at an early stage to avoid and reduce disease progression.

Therefore, the object of the present invention is to quantify the level of infection risk by a simple parameter scoring system based on relevant clinical data such as treatment or treatment received by the patient, and to provide a warning function for preventive intervention measures for nosocomial infections of medical staff. Thus, the present invention uses a simple parameter scoring system to predict a simple model of nosocomial infection risk, and is suitable for assisting medical personnel in finding a patient population that may be infected in a hospital. The model includes seven variable input modules and a weighted specific gravity calculation module. An output module for risk assessment scores. The variable input module can provide a user interface to input by itself or automatically imported from the hospital information system data. The input parameter variables include nasogastric tubes, urinary catheters, arterial catheters, central venous catheters, use of systemic corticosteroids, pressure ulcer preventives, hemodialysis. The weighted specific gravity calculation module, a total of seven imported The parameters are: stomach weight 3, catheter weight 2, arterial catheter weight 1, central venous catheter weight 1, total systemic corticosteroid weight 3, pressure ulcer preventive drug weight 2, hemodialysis The weight is 2. The seven parameters of the preceding paragraph are verified by the artificial neural network and the logistic regression system. The weight ratio is proved to be more than 90% of the predicted parameters through complex calculations.

The foregoing and technical description of the present invention will be apparent from the

Referring to Figures 1 and 2, the simple parameter scoring system of the present invention predicts a model of nosocomial infection, including a medical information system, a standard parameter setting, a simple scoring system, and a warning notification system.

First, from the medical information system, provide medical information, including nasogastric tube, catheter, arterial catheter, central venous catheter, use of systemic corticosteroids, pressure ulcer preventive drugs, hemodialysis and other seven variables.

Standard parameter setting, the default value is the weight of the stomach tube is 3, the weight of the catheter is 2, the weight of the arterial catheter is 1. The weight of the central venous catheter is 1. The weight of the systemic corticosteroid is 3. The weight of the preventive drug for the pressure ulcer is 2. The hemodialysis weight is 2.

The simple scoring system can directly retrieve the above seven variables of the hospital information system, and use the conversion module of the simple scoring system to weight the score data of each patient and compare the set of hospital infection standard values (3 points). The system automatically determines that the risk of hospital infection is divided into high-risk groups and low-risk groups. Provide medical staff with early intervention and prevention measures and warnings. The standard value hospital of the preceding paragraph may also modify the threshold based on the actual patient's disease combination or empirical experience. Value, as the standard value of the in-hospital assessment score.

The warning informs the system that the patient with high-risk nosocomial infection in the simple scoring system will be sent back to the medical information system for medical staff to inquire. It can even be designed to use the notification method such as newsletter and E-mail to inform the medical staff early. Interventions to prevent infections.

11‧‧‧Standard parameter setting

12‧‧‧ Medical Information System

13‧‧‧Simple scoring system

14‧‧‧Warning notification system

21‧‧‧ Nasogastric tube variables: Whether the patient receives nasogastric tube treatment, if there is one, give one point, if not, then zero

22‧‧‧ catheter changes: whether the patient receives catheter treatment, if there is one, give one point, if not, then zero

23‧‧‧Arterial catheter variable: whether the patient receives arterial catheter treatment, if there is one, then one point is zero

24‧‧‧ central venous catheter variable: whether the patient is treated with central venous catheter, if there is one, then one point, if not, zero

25‧‧‧ Use systemic corticosteroid variables: whether the patient is treated with systemic corticosteroids, if any, one point, if none, zero

26‧‧‧Pressure ulcer preventive drug variables: whether the patient receives treatment for pressure ulcer preventive drugs, if any, give one point, if not, then zero

27‧‧‧ Hemodialysis variables: whether the patient receives hemodialysis treatment, if there is one, give one point, if not, then zero

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a flow chart illustrating an exemplary embodiment of a simple parameter scoring system of the present invention for predicting nosocomial infections.

Figure 2 is a block diagram of a simple scoring system illustrating the seven variable parameter setting principles.

11‧‧‧Standard parameter setting

12‧‧‧ Medical Information System

13‧‧‧Simple scoring system

14‧‧‧Warning notification system

21‧‧‧ Nasogastric tube variables

22‧‧‧ catheter changes

23‧‧‧Arterial catheter variables

24‧‧‧ central venous catheter variables

25‧‧‧Use of systemic corticosteroid variables

26‧‧‧Pressure ulcer preventive drug variables

27‧‧‧hemodialysis variables

Claims (5)

A predictive system for in-hospital infection with high accuracy, which is suitable for predicting the in-hospital infection status of patients in hospitals, so as to effectively assess the condition of patients and assess their infection status, so that doctors can pre-dispose to reduce the risk of infection and improve patient safety, and further reduce medical treatment. waste. Includes: a medical information system, a standard parameter setting, a simple scoring system, and a warning notification system. Provide medical data according to the medical information system described in item 1 of the patent application, including nasogastric tube, catheter, arterial catheter, central venous catheter, use of systemic corticosteroids, preventative pressure ulcers, hemodialysis and other seven variables . According to the standard parameter setting mentioned in item 1 of the patent application scope, the preset value is the weight of the nasogastric tube, the weight of the catheter is 2, the weight of the arterial catheter is 1. The weight of the central venous catheter is 1. The weight of the whole body corticosteroid is used. 3. The weight of the pressure ulcer prevention drug is 2, and the weight of the hemodialysis is 2. According to the simple scoring system described in item 1 of the patent application scope, according to the seven variables of the hospital information system, and using the conversion module of the simple scoring system to weight the score data of each patient, the system automatically determines the patient. High and low risk groups of nosocomial infections. The standard value hospital of the preceding paragraph may also modify the threshold according to the actual patient's disease combination or empirical experience as the standard value of the in-hospital evaluation score. According to the warning notification system described in item 1 of the patent application scope, patients with high-risk nosocomial infections in the simple scoring system will be transmitted back to the medical information system for medical personnel to inquire, and may even be designed as a newsletter or E-mail. Such as the active notification method, the warning informs the medical staff to early intervention in the prevention of infection.