Whole hospital intelligent nutrition risk screening method and system

Abstract

The invention relates to the technical field of intelligent medical treatment, in particular to a hospital-wide intelligent nutrition risk screening method and system. The method comprises the following steps: S1, acquiring patient information; s2, converting text information in the patient information into structured data to form first data, and extracting numerical information in the patient information to form second data; s3, calculating the nutrition condition of the patient according to the first data to obtain a first score; s4, calculating the nutrition condition of the patient according to the second data to obtain a second score; and S5, obtaining a first total score according to the first score and the second score, and determining the nutritional status of the patient according to the first total score N0. According to the invention, the patient nutrition risk screening coverage rate is more than 90%; the nutrition risk screening time of the patient is greatly shortened, and the screening efficiency is improved; in addition, the lag problem of nutrition risk screening of the patient is avoided, and real-time monitoring of nutrition risk screening of the patient is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent healthcare, and particularly to a method and system for intelligent nutritional risk screening for the whole hospital. Background Art

[0002] Malnutrition not only increases the risk of complications but may also lead to an extended hospital stay and increased medical costs. Therefore, timely nutritional screening and intervention are crucial for improving the clinical outcomes of patients. The purpose of nutritional screening is to identify those patients who may be affected by nutritional problems and thus impact clinical outcomes, providing technical support for subsequent issuance of nutritional prescriptions. Research shows that through nutritional screening and treatment, the probability of postoperative complications can be reduced, the hospital stay can be shortened, and the average hospitalization cost can be decreased, which has a significant impact on improving the treatment effect and the quality of life of patients.

[0003] In hospitals, in the face of numerous inpatients, although the professional team of nutritional screening nurses is growing, challenges still exist. Each screening takes a lot of time, and it is not possible to conduct nutritional risk screening for every patient in the whole hospital in a timely manner. The screening coverage rate is not high, resulting in the nutritional problems of patients not being screened out in a timely manner. In order to ensure that each patient can obtain timely and accurate nutritional assessment, an intelligent nutritional risk screening method and system for the whole hospital are invented, aiming to optimize the screening operation to improve efficiency and reduce the possibility of omission.

[0004] In addition, when current medical institutions conduct nutritional risk screening, it is mainly for patients within 24 hours of admission, lacking continuous monitoring. During the hospitalization of patients, due to the progression of the disease and treatments such as surgery and medication, the nutritional status of patients may decline significantly. Summary of the Invention

[0005] In order to solve the above technical problems existing in the prior art, the present invention provides a method and system for intelligent nutritional risk screening for the whole hospital.

[0006] To achieve the above object, the technical solution of the present invention is as follows:

[0007] In the first aspect, the present invention provides an intelligent nutritional risk screening method for the whole hospital, including:

[0008] S1. Obtain patient information;

[0009] S2. Convert the text information in the patient information into structured data to form the first data, and extract the numerical information in the patient information to form the second data;

[0010] S3. Calculate the nutritional status of the patient according to the first data to obtain the first score;

[0011] S4. Calculate the nutritional status of the patient according to the second data to obtain the second score;

[0012] S5. Obtain a first total score based on the first score value and the second score value, and determine the nutritional status of the patient according to the first total score N0.

[0013] Furthermore, convert the text information in the patient information into structured data, specifically including:

[0014] Extract the text information in the patient information according to the keyword database to form structured data; the keyword database includes disease names, symptom keywords, weight keywords, eating keywords, and negative keywords.

[0015] Furthermore, calculate the nutritional status of the patient based on the first data to obtain a first score value; specifically including:

[0016] Divide the keyword database into first-class data, second-class data, third-class data, and fourth-class data, compare the structured data with the corresponding first-class data, second-class data, third-class data, and fourth-class data, and obtain the number of fields F with the highest similarity to the first-class data in the structured data 1 , the number of fields F with the highest similarity to the second-class data 2 , the number of fields F with the highest similarity to the third-class data 3 , the number of fields F with the highest similarity to the fourth-class data 4 ;

[0017] Judge the magnitudes of the number of fields F 1 , F 2 , F 3 and F 4 , and obtain the maximum value among them,

[0018] If the maximum value is F 1 , then the first score value is 3;

[0019] If the maximum value is F 2 , then the first score value is 2;

[0020] If the maximum value is F 3 , then the first score value is 1;

[0021] If the maximum value is F 4 , then the first score value is 0.

[0022] Even further, the fourth-class data is the structured data extracted through negative keywords.

[0023] Furthermore, calculate the nutritional status of the patient based on the second data to obtain a second score value, specifically including:

[0024] Get the patient's age information. If the age is ≥18, the second score is 0. Then get the patient's height and weight, calculate the BMI value. If the BMI is ≤18.5 or BMI is ≥24, add 0.5 to the second score. If the second score is between 18.5-24, the second score remains unchanged.

[0025] If the age is less than 18, the second score is 1; then the patient's height and weight are obtained, and the BMI value is calculated. If BMI≤15.5 or BMI≥18.5, the second score is increased by 0.3; if, the second score is between 15.5-18.5, the second score remains unchanged.

[0026] Furthermore, a first total score is obtained according to the first score and the second score. Specifically, if the first score is 0, the first score is the first total score; otherwise, the first score and the second score are added together to obtain the first total score.

[0027] Further, according to the first total score N0, the nutritional status of the patient is determined, specifically including:

[0028] If N 0 =0, indicating no nutritional risk;

[0029] If 0<N 0 ≤1, indicating low risk of nutritional status;

[0030] If 1<N 0 ≤2.5, indicating moderate risk of nutritional status;

[0031] If 2.5<N 0 , indicating a high-risk nutritional status.

[0032] Furthermore, the patient information includes height, weight, age, chief complaint, current medical history, diagnosis record, admission time, doctor's advice, etc.

[0033] Furthermore, the hospital-wide intelligent nutrition risk screening method also includes:

[0034] S6. Update patient information regularly, and obtain the i-th total score N according to the updated patient information. i , according to the total score N of the ith i And the patient's historical nutritional status, calculate the current total score N 当 , according to the current total score N 当 , update the patient's nutritional status.

[0035] Furthermore, according to the total score N of the i i And the patient's historical nutritional status, calculate the current total score N 当 , including:

[0036] Get the patient's total historical score, recorded as N 0、N 1 ...N i-1 , calculate N 0 、N 1 ...N i-1 The average value N 平 and variance σ 2 , then the current total score N 当 for

[0037]

[0038] Furthermore, according to the current total score N 当 , update the patient's nutritional status, specifically:

[0039] If N 当 =0, indicating no nutritional risk;

[0040] If 0<N 当 ≤1, indicating low risk of nutritional status;

[0041] If 1<N 当 ≤2.5, indicating moderate risk of nutritional status;

[0042] If 2.5<N 当 , indicating a high-risk nutritional status.

[0043] Furthermore, the hospital-wide intelligent nutrition risk screening method also includes:

[0044] S7. Determine the total score N of the patient's history according to the patient's historical nutritional status. i The outliers in the data are obtained, and the number of outliers is obtained. If the number of outliers is greater than the set value, the structured data corresponding to the outliers is obtained, and the keywords corresponding to the outliers in the structured data are filtered out and updated.

[0045] In a second aspect, the present invention provides a hospital-wide intelligent nutrition risk screening device, including a nutrition screening module, the nutrition screening module including a processor, and the processor executes any of the above-mentioned hospital-wide intelligent nutrition risk screening methods.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The hospital-wide intelligent nutrition risk screening method provided by the present invention achieves a patient nutrition risk screening coverage rate of more than 90% by connecting the hospital HIS, LIS, and EMR systems; greatly shortens the patient nutrition risk screening time and improves the screening efficiency; in addition, the present invention also regularly updates patient information, updates the patient's nutrition risk status based on the patient's latest information, avoids the lag problem of patient nutrition risk screening, and realizes real-time monitoring of patient nutrition risk screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a logic diagram of the method of the present invention. DETAILED DESCRIPTION

[0049] The technical solution of the present invention will be clearly described below in conjunction with the accompanying drawings. Obviously, the described embodiments are not all embodiments of the present invention, and all other embodiments obtained by ordinary technicians in the field without making creative work are within the protection scope of the present invention.

[0050] It should be noted that, unless otherwise specifically stated, the relative arrangements of components and steps, and numerical expressions set forth in these embodiments should not be construed as limiting the scope of the present invention.

[0051] The following description of the exemplary embodiments is merely illustrative and is not intended to limit the present invention and its application or use in any sense. Techniques, methods and devices known to ordinary technicians in the relevant field may not be discussed in detail here, but where applicable, these techniques, methods and devices should be considered as part of this specification.

[0052] The present invention provides a hospital-wide intelligent nutritional risk screening method, such as Figure 1 As shown, including:

[0053] S1. Obtain patient information; patient information includes height, weight, age, chief complaint, current medical history, diagnosis record, admission time, doctor's orders, etc.

[0054] S2. Convert text information in the patient information into structured data to form first data, and extract numerical information in the patient information to form second data;

[0055] Convert text information in patient information into structured data, including:

[0056] According to the keyword database, text information in the patient information is extracted to form structured data; the keyword database includes disease names, symptom keywords, weight keywords, eating keywords and negative keywords. Disease names include cerebral infarction, acute appendicitis, amebiasis, mediastinal tuberculoma, etc.; symptom keywords include nausea, vomiting, constipation, etc.; weight keywords include significant weight loss, weight loss, etc.; eating keywords include poor appetite, eating difficulties, etc.; negative keywords include no weight loss, no diarrhea, etc.

[0057] S3. Calculate the nutritional status of the patient according to the first data to obtain a first score; specifically including:

[0058] Divide the keyword database into the first type of data, the second type of data, the third type of data, and the fourth type of data. Compare the structured data with the corresponding first type of data, second type of data, third type of data, and fourth type of data, and obtain the number of fields F with the highest similarity to the first type of data in the structured data. 1 , the number of fields F with the highest similarity to the second type of data 2 , the number of fields F with the highest similarity to the third type of data 3 , the number of fields F with the highest similarity to the fourth type of data 4 ;

[0059] Judge the sizes of the number of fields F 1 , F 2 , F 3 and F 4 , and obtain the maximum value among them.

[0060] If the maximum value is F 1 , then the first score is 3;

[0061] If the maximum value is F 2 , then the first score is 2;

[0062] If the maximum value is F 3 , then the first score is 1;

[0063] If the maximum value is F 4 , then the first score is 0.

[0064] If there are cases with the same value, obtain the first score in the order of priority from high to low F 1 ≥F 2 ≥F 3 ≥F 4 .

[0065] Among them, the first type of data includes disease names, symptom keywords, weight keywords, and eating keywords, and the disease names, symptom keywords, weight keywords, and eating keywords in the first type of data are all keywords that have a greater impact on nutritional evaluation. For example, cerebral infarction, acute appendicitis, liver metastasis after rectal cancer surgery, typhoid bacillus septicemia, inability to walk, vomiting after eating, significant weight loss, emaciation, difficulty in eating, and almost unable to eat.

[0066] The second type of data includes disease names, symptom keywords, weight keywords, and eating keywords, and the disease names, symptom keywords, weight keywords, and eating keywords in the second type of data are all keywords that have a medium impact on nutritional evaluation. For example, secondary cancer of the liver lymph nodes, secondary malignant tumor of the liver, secondary cancer of the liver, etc., dysphagia, progressive weight loss, gradual weight loss, persistent lack of improvement in appetite, and small food intake.

[0067] The third type of data includes disease names, symptom keywords, weight keywords, and eating keywords, and the disease names, symptom keywords, weight keywords, and eating keywords in the third type of data are all keywords that have less impact on nutritional evaluation. For example, ameboma, tuberculoma of the lung, tuberculoma of the pleura, tuberculoma of the mediastinum, acute pyelonephritis, diabetic ketoacidosis, etc., constipation, nausea, slight weight loss, relatively light weight loss, poor appetite, slightly poor diet, poor appetite, etc.

[0068] The fourth type of data is structured data extracted through negative keywords. For example, no emaciation, no weight loss, no nausea and vomiting, no diarrhea, normal diet, good appetite.

[0069] S4. Calculate the nutritional status of the patient based on the second data to obtain the second score; specifically including:

[0070] Obtain the patient's age information. If the age ≥ 18, the second score is 0; then obtain the patient's height and weight, calculate the BMI value. If the BMI ≤ 18.5 or the BMI ≥ 24, the second score is increased by 0.5; if the BMI is between 18.5 - 24, the second score remains unchanged;

[0071] If the age is less than 18, the second score is 1; then obtain the patient's height and weight, calculate the BMI value. If the BMI ≤ 15.5 or the BMI ≥ 18.5, the second score is increased by 0.3; if the BMI is between 15.5 - 18.5, the second score remains unchanged.

[0072] S5. Obtain the first total score based on the first score and the second score, and determine the nutritional status of the patient according to the first total score N0.

[0073] The specific method is: if the first score is 0, the first score is the first total score; otherwise, add the first score and the second score to obtain the first total score.

[0074] If N 0 = 0, it indicates no nutritional risk;

[0075] If 0 < N 0 ≤ 1, it indicates a low nutritional risk;

[0076] If 1 < N 0 ≤ 2.5, it indicates a medium nutritional risk;

[0077] If 2.5 < N 0 , it indicates a high nutritional risk.

[0078] S6. Regularly update the patient information, and obtain the i-th total score N i , the calculation method of the i-th total score N i is the same as that of the first total score.

[0079] Then according to the total score N i And the patient's historical nutritional status, calculate the current total score N 当 , according to the current total score N 当 , update the patient's nutritional status. Specifically include:

[0080] Get the patient's total historical score, recorded as N 0 、N 1 ...N i-1 , calculate N 0 、N 1 ...N i-1 The average value N 平 and variance σ 2 , then the current total score N 当 for

[0081]

[0082] If N 当 =0, indicating no nutritional risk;

[0083] If 0<N 当 ≤1, indicating low risk of nutritional status;

[0084] If 1<N 当 ≤2.5, indicating moderate risk of nutritional status;

[0085] If 2.5<N 当 , indicating a high-risk nutritional status.

[0086] S7. Determine the total score N of the patient's history according to the patient's historical nutritional status. i The outliers in the data are obtained, and the number of outliers is obtained. If the number of outliers is greater than the set value, the structured data corresponding to the outliers is obtained, and the keywords corresponding to the outliers in the structured data are filtered out and updated.

[0087] In a second aspect, the present invention provides a hospital-wide intelligent nutrition risk screening device, including a nutrition screening module, the nutrition screening module including a processor, and the processor executes any of the above-mentioned hospital-wide intelligent nutrition risk screening methods.

[0088] Embodiment 1

[0089] Patient information: height 171cm, weight 74kg, male, age 55. The patient suddenly developed chills and fever without obvious cause, accompanied by distension and discomfort in the left waist and abdomen, nausea and vomiting, without frequent urination, urgency, or gross hematuria. He was admitted to the outpatient clinic with acute pyelonephritis and left ureteral stones. Since the onset of the disease, he has had poor appetite, poor mental and sleep, and no significant weight loss.

[0090] Convert the text part of the patient information into structured data to form the first data, and the first data is:

[0091] {"disease": ["acute pyelonephritis", "left ureteral calculus"],

[0092] "symptom": ["nausea", "vomiting"],

[0093] "eating": ["poor appetite"],

[0094] "negative": ["no significant weight loss"]}

[0095] Obtain the number of fields F 1 、F 2 、F 3 and F 4 ,F 1 is 0, F 2 is 0, F 3 is 3, F 4 is 1; then the first score is 1.

[0096] The patient's age is 55, so the second score is 0; then calculate the patient's BMI value, and the BMI is 25.3, so the second score is increased by 0.5. Therefore, the final second score is 0.5.

[0097] Finally, the first total score of this patient is 1.5, which is within the range of 1 - 2.5. Therefore, this patient is at medium risk of nutritional status.

[0098] Example 2

[0099] Patient information: height 165 cm, weight 46 kg, male, age 16. The patient has dry mouth and polydipsia without obvious inducement, with a daily water intake of more than 3000 ml, accompanied by polyuria, urine volume of more than 2500 ml, nocturia 2 times / night, weight loss of 5 kg, accompanied by easy hunger, fatigue, poor spirit, dizziness, easy to get sleepy, accompanied by multiple oral ulcers, no obvious polyphagia, no visual acuity decline, blurred vision, no headache, visual field defect, no numbness, tingling in the extremities, no nausea, vomiting, no abdominal pain, diarrhea, no cough, expectoration, chills, fever, no palpitations, chest tightness, chest pain. The patient was admitted to the hospital with diabetic ketoacidosis. Since the onset of the disease, the diet has been slightly poor, the consciousness is clear, the spirit is poor, the sleep is poor, the stool is normal, the urine is as described above, and the weight has decreased by 5 kg.

[0100] Convert the text part of the patient information into structured data to form the first data, and the first data is:

[0101] {"disease": ["diabetic ketoacidosis"],

[0102] "symptom": [" "],

[0103] "Eating": ["Slightly poor diet"],

[0104] "Negative": ["No nausea or vomiting, no abdominal pain or diarrhea"]}

[0105] Obtain the number of fields F 1 , F 2 , F 3 and F 4 , F 1 is 0, F 2 is 1, F 3 is 2, F 4 is 2; according to the priority, the first score is 1.

[0106] If the patient's age is 16, the second score is 1; then calculate the patient's BMI value. If the BMI is 16.9, the second score remains unchanged. Therefore, the final second score is 1.

[0107] Finally, the first total score of this patient is 2, which is within the range of 1 - 2.5. Therefore, this patient has a medium risk of nutritional status.

[0108] The above specific implementation manners are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the examples, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A hospital-wide intelligent nutritional risk screening method, characterized in that: Including: S1. Obtain patient information; S2. Convert the text information in the patient information into structured data to form the first data, and extract the numerical information in the patient information to form the second data; S3. Calculate the nutritional status of the patient based on the first data to obtain the first score; S4. Calculate the nutritional status of the patient based on the second data to obtain the second score; S5. Obtain the first total score based on the first score and the second score, and determine the nutritional status of the patient according to the first total score N0.

2. The hospital-wide intelligent nutritional risk screening method according to claim 1, characterized in that: Converting the text information in the patient information into structured data specifically includes: Extract the text information in the patient information according to the keyword database to form structured data; wherein the keyword database includes disease names, symptom keywords, weight keywords, eating keywords, and negative keywords.

3. The hospital-wide intelligent nutritional risk screening method according to claim 2, characterized in that: Calculating the nutritional status of the patient based on the first data to obtain the first score; specifically including: Divide the keyword database into the first type of data, the second type of data, the third type of data, and the fourth type of data, compare the structured data with the corresponding first type of data, the second type of data, the third type of data, and the fourth type of data, and obtain the number of fields F1 with the highest similarity to the first type of data, the number of fields F2 with the highest similarity to the second type of data, the number of fields F3 with the highest similarity to the third type of data, and the number of fields F4 with the highest similarity to the fourth type of data in the structured data; Judge the sizes of the number of fields F1, F2, F3, and F4, and obtain the maximum value among them. If the maximum value is F1, the first score is 3; If the maximum value is F2, the first score is 2; If the maximum value is F3, the first score is 1; If the maximum value is F4, the first score is 0.

4. The hospital-wide intelligent nutritional risk screening method according to claim 1, characterized in that: Calculating the nutritional status of the patient based on the second data to obtain the second score, specifically including: Obtain the patient's age information. If the age ≥ 18, the second score is 0; then obtain the patient's height and weight, calculate the BMI value. If the BMI ≤ 18.5 or the BMI ≥ 24, the second score is increased by 0.5; if the second score is between 18.5 - 24, the second score remains unchanged; If the age is less than 18, the second score is 1; then obtain the patient's height and weight, calculate the BMI value. If the BMI ≤ 15.5 or the BMI ≥ 18.5, the second score is increased by 0.3; if the second score is between 15.5 - 18.5, the second score remains unchanged.

5. The hospital-wide intelligent nutritional risk screening method according to claim 1, characterized in that: Obtain the first total score based on the first score and the second score. The specific method is: if the first score is 0, the first score is the first total score; otherwise, add the first score and the second score to obtain the first total score.

6. The hospital-wide intelligent nutritional risk screening method according to claim 1, characterized in that: Determine the nutritional status of the patient according to the first total score N0, specifically including: If N0 = 0, it indicates no nutritional risk; If 0 < N0 ≤ 1, it indicates a low nutritional risk; If 1 < N0 ≤ 2.5, it indicates a medium nutritional risk; If 2.5 < N0, it indicates a high nutritional risk.

7. The hospital-wide intelligent nutritional risk screening method according to claim 1, characterized in that: It also includes: S6. Update patient information regularly, and obtain the i-th total score N according to the updated patient information. i , according to the total score N of the ith i As well as the patient's historical nutritional status, calculate the current total score N, and update the patient's nutritional status based on the current total score N.

8. The hospital-wide intelligent nutritional risk screening method according to claim 7, characterized in that: According to the total score N i And the patient's historical nutritional status, calculate the current total score N, including: Get the patient's total historical score, recorded as N0, N1...N i-1 , calculate N0, N1, ... N i-1 The mean value N and the variance σ 2 , then the current total score N is 9. The hospital-wide intelligent nutritional risk screening method according to claim 1, characterized in that: It also includes: S7. Determine the total score N of the patient's history according to the patient's historical nutritional status. i The outliers in the data are obtained, and the number of outliers is obtained. If the number of outliers is greater than the set value, the structured data corresponding to the outliers is obtained, and the keywords corresponding to the outliers in the structured data are filtered out and updated.

10. A hospital-wide intelligent nutrition risk screening device, comprising a nutrition screening module, characterized in that: The nutrition screening module includes a processor, and the processor executes the全院智能营养风险筛查方法 described in any one of claims 1 - 9. It should be noted that there may be an inaccuracy in the last line where "全院智能营养风险筛查方法" is directly used without a proper English translation. It might need to be accurately translated according to the actual method name in the relevant context. Here it is left as is for the purpose of following the translation rules strictly based on the provided text.

Images