Adult icu patient MARSI risk prediction method and system

Abstract

The invention discloses an adult icu patient MARSI risk prediction method and system, and relates to the technical field of MARSI risk assessment, and the method comprises the steps: collecting adult icu patient MARSI data; performing single-factor analysis on the data to obtain a significance result of each variable in the single-factor analysis; a Logistic multi-factor regression model is constructed; according to the constructed Logistic multi-factor regression model, obtaining an independent risk factor; and constructing a column graph risk prediction model based on the independent risk factors, and performing real-time risk prediction through the column graph risk prediction model. According to the method, MARSI risk factors of the adult ICU patient are analyzed, the risk prediction model is constructed, and external verification is performed, so that a screening tool is provided for early identification of the risk of medical viscose-related skin injury of the ICU patient, occurrence of medical viscose-related adverse events of the ICU patient is reduced, the economic burden of the patient is relieved, and the satisfaction degree of the patient is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of MARSI risk assessment, and more specifically to a method and system for predicting MARSI risk in adult ICU patients. Background Art

[0002] ICU is the main place for treating critically ill patients. Various drainage tubes, monitoring instruments, etc. need to be fixed with adhesive products. In addition, the patients are in critical condition and have low skin resistance. Wound exudate, sweat, disinfectants, etc. will irritate the patient's skin, making critically ill patients a high-risk group for medical adhesive-related skin injury (MARSI).

[0003] MARSI refers to the appearance of erythema with or without blisters, erosions or tears and other abnormal skin symptoms that last for 30 minutes or longer after the medical adhesive is removed. MARSI is common in the medical care process and has various types, which can be summarized into the following four types: mechanical injury (including epidermal exfoliation, tension injury and skin tear), dermatitis (contact dermatitis and allergic dermatitis), maceration, and folliculitis. Domestic and foreign studies have shown that the incidence of MARSI is 3.4%-38%, which varies depending on the research subjects. The incidence of MARSI in hospitalized infants and young children is 8%-25.6%, the incidence of MARSI in critically ill children is 38%, the incidence of MARSI in elderly patients is 3.9%-15.5%, the incidence of MARSI in elderly patients in ICU is 10.3%, and the incidence of MARSI in adult ICU patients is 10.96%-11.86%.

[0004] The occurrence of MARSI not only destroys the integrity of the skin, but also causes pain and discomfort to the patient, leading to local infection, increased nursing costs, reduced quality of life of the patient, and even unplanned extubation of the patient.

[0005] At present, the research and application of MARSI risk assessment are still in the initial stage and scattered. The multiple MARSI risk assessment scales formed in China are suitable for specific groups. Domestic research on MARSI assessment tools mainly targets newborns and PICC patients. There is no reliable risk prediction assessment tool for MARSI in adult ICU patients.

[0006] Therefore, how to propose a MARSI risk prediction method and system for adult ICU patients, analyze the risk factors of medical adhesive-related skin injuries in adult ICU patients, construct a risk prediction model and conduct external verification, provide a screening tool for early identification of the risk of medical adhesive-related skin injuries in ICU patients, reduce the occurrence of medical adhesive-related adverse events in ICU patients, reduce the economic burden of patients, and improve patient satisfaction, is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0007] In view of this, the present invention provides a method and system for predicting the risk of MARSI in adult ICU patients. By analyzing the risk factors of medical adhesive-related skin injuries in adult ICU patients, a risk prediction model is constructed and externally verified, providing a screening tool for early identification of the risk of medical adhesive-related skin injuries in ICU patients, reducing the occurrence of medical adhesive-related adverse events in ICU patients, reducing the economic burden of patients, and improving patient satisfaction. In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0008] A method for predicting the risk of MARSI in adult ICU patients, comprising:

[0009] Collect MARSI data of adult ICU patients;

[0010] Perform univariate analysis on MARSI data, calculate the correlation index between each variable and MARSI occurrence, and obtain the significance results of each variable in the univariate analysis;

[0011] The significant univariate analysis variables were used as predictor variables, and the occurrence or not of MARSI was used as the predicted variable to construct a Logistic multivariate regression model.

[0012] According to the constructed Logistic multivariate regression model, independent risk factors were obtained;

[0013] A nomogram risk prediction model was constructed based on independent risk factors, and the real-time risk prediction of MARSI in adult ICU patients was performed using the nomogram risk prediction model.

[0014] Optionally, the collecting of MARSI data of adult ICU patients includes determining the sample size:

[0015] Estimated overall rate based on sample survey:

[0016] Error δ = 0.04, P is the incidence of medical adhesive-related skin injuries in adult ICU patients, which is 0.1186, and α is the allowable error (i.e., the allowable difference between the specified sample incidence and the overall incidence) is 0.05. is the quantile of the standard normal distribution, where Z is a variable with unit standard deviation of the standard normal distribution, and α is the significance level (taken as 0.05). 1-α / 2 is because the significance level is divided into two parts, the α / 2 on the left is the first type of error (the error of rejecting the null hypothesis when it is actually true), and the 1-α / 2 on the right is the second type of error (the error of accepting the null hypothesis when it is actually false). When α=0.05, 1-α / 2=0.975, which means that it is necessary to look up the table to find the value of Z so that the area of ​​the normal distribution is greater than 0.975. This is to reject the null hypothesis at a 95% confidence level. The Z value at a 95% confidence level is 1.96. That is, the formula is n=1.96 2 ×0.1186(1-0.1186) / 0.042≈250.

[0017] Optionally, the MARSI data of adult ICU patients include observation indicators, which are: mechanical injury: skin peeling, skin lacerations, tension injury or blisters; dermatitis: contact dermatitis, allergic dermatitis; maceration; folliculitis.

[0018] Optionally, the collecting of MARSI data of adult ICU patients includes:

[0019] After the patient has been in the ICU for ≥24h, fill in the risk survey form;

[0020] After each change of adhesive product, observe whether there is skin damage. If there is skin damage, observe and evaluate again after X minutes. If the skin damage lasts for at least X minutes, it is recorded as MARSI.

[0021] Collect skin pictures, store them in codes and fill in the appendix of the risk survey form.

[0022] Optionally, it also includes preprocessing the collected data:

[0023] Check if the data is complete;

[0024] Handling missing values ​​and outliers;

[0025] Data transformations were performed, including encoding categorical variables into dummy variables.

[0026] Optionally, the risk questionnaire is obtained by searching the China National Knowledge Infrastructure, Wanfang Database, Pubmed and ScienceDirect databases to obtain risk factors for MARSI in adult ICU patients and formulate a risk questionnaire.

[0027] Optionally, performing univariate analysis on the MARSI data, calculating the correlation index between each variable and the occurrence of MARSI and obtaining the significance result of each variable in the univariate analysis includes:

[0028] SPSS26.0 and R3.5 software were used for statistical analysis, univariate analysis was performed for each variable, and the correlation index between each variable and the occurrence of MARSI was calculated;

[0029] Non-normally distributed quantitative data were expressed as medians and quartiles, and nonparametric tests or chi-square tests were used for inter-group comparisons. Count data were expressed as proportions (%). For discrete variables, the incidence of each level was calculated and compared. For continuous variables, the mean or median between different groups was calculated and compared.

[0030] According to the set significance level P, the statistical significance of the variables in each univariate analysis was judged, and the variables that met the significance level threshold were judged to be related to the occurrence of MARSI.

[0031] Optionally, it also includes the evaluation of the constructed Logistic multivariate regression model, including the evaluation of the model's goodness of fit and predictive ability, and the use of goodness of fit tests and model discrimination indicators to evaluate the model.

[0032] Optionally, validation of the nomogram risk prediction model is also included, and the discrimination and calibration of the nomogram risk prediction model are verified by the area under the ROC curve, drawing a calibration curve, and the Hosmer-Lemeshow chi-square test.

[0033] Optionally, a MARSI risk prediction system for adult ICU patients, comprising:

[0034] Acquisition module: used to collect MARSI data of adult ICU patients;

[0035] Analysis module: used to perform univariate analysis on MARSI data, calculate the correlation index between each variable and MARSI occurrence, and obtain the significance results of each variable in the univariate analysis;

[0036] Model building module: used to construct a Logistic multi-factor regression model by taking the significant univariate analysis variables as predictor variables and the occurrence or non-occurrence of MARSI as the predicted variable;

[0037] Judgment module: used to obtain independent risk factors based on the constructed Logistic multi-factor regression model;

[0038] Prediction module: used to construct a nomogram risk prediction model based on independent risk factors, and to perform real-time risk prediction of MARSI in adult ICU patients through the nomogram risk prediction model.

[0039] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method and system for predicting the risk of MARSI in adult ICU patients, which has the following beneficial effects:

[0040] The present invention proposes a MARSI risk prediction method for adult ICU patients, collects MARSI data of adult ICU patients; performs univariate analysis on MARSI data, calculates the relevant index of each variable and MARSI occurrence and obtains the significance result of each variable in the univariate analysis; uses the significant univariate analysis variable as the predictor variable, and uses whether MARSI occurs or not as the predicted variable to construct a Logistic multifactor regression model; obtains independent risk factors according to the constructed Logistic multifactor regression model; constructs a nomogram risk prediction model based on the independent risk factors, and performs real-time risk prediction of MARSI in adult ICU patients through the nomogram risk prediction model. The present invention analyzes the risk factors of medical viscose-related skin damage in adult ICU patients, constructs a risk prediction model and performs external verification, provides a screening tool for early identification of the risk of medical viscose-related skin damage in ICU patients, reduces the occurrence of medical viscose-related adverse events in ICU patients, reduces the economic burden of patients, and improves patient satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0042] Figure 1 This is a flow chart of a method for predicting MARSI risk in adult ICU patients provided by the present invention.

[0043] Figure 2 This is a structural framework diagram of a MARSI risk prediction system for adult ICU patients provided by the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] The embodiment of the present invention discloses a method for predicting the risk of MARSI in adult ICU patients. Figure 1 As shown, including:

[0046] Collect MARSI data of adult ICU patients;

[0047] Perform univariate analysis on MARSI data, calculate the correlation index between each variable and MARSI occurrence, and obtain the significance results of each variable in the univariate analysis;

[0048] The significant univariate analysis variables were used as predictor variables, and the occurrence or not of MARSI was used as the predicted variable to construct a Logistic multivariate regression model.

[0049] According to the constructed Logistic multivariate regression model, independent risk factors were obtained;

[0050] A nomogram risk prediction model was constructed based on independent risk factors, and the real-time risk prediction of MARSI in adult ICU patients was performed using the nomogram risk prediction model.

[0051] Furthermore, the collection of MARSI data of adult ICU patients includes determining the sample size:

[0052] Estimated overall rate based on sample survey:

[0053] Error δ = 0.04, P is the incidence of medical adhesive-related skin injuries in adult ICU patients, which is 0.1186, and α is the allowable error (i.e., the allowable difference between the specified sample incidence and the overall incidence) is 0.05. is the quantile of the standard normal distribution, where Z is a variable with unit standard deviation of the standard normal distribution, and α is the significance level (taken as 0.05). 1-α / 2 is because the significance level is divided into two parts, the α / 2 on the left is the first type of error (the error of rejecting the null hypothesis when it is actually true), and the 1-α / 2 on the right is the second type of error (the error of accepting the null hypothesis when it is actually false). When α=0.05, 1-α / 2=0.975, which means that it is necessary to look up the table to find the value of Z so that the area of ​​the normal distribution is greater than 0.975, in order to reject the null hypothesis at a 95% confidence level. The Z value at a 95% confidence level is 1.96. That is, the formula is n=1.962×0.1186(1-0.1186) / 0.042≈250.

[0054] Furthermore, the MARSI data of adult ICU patients include observation indicators, which are: mechanical injury: skin peeling, skin lacerations, tension injuries or blisters; dermatitis: contact dermatitis, allergic dermatitis; maceration; folliculitis.

[0055] Furthermore, the collection of MARSI data of adult ICU patients includes:

[0056] After the patient has been in the ICU for ≥24h, fill in the risk survey form;

[0057] After each change of adhesive product, observe whether there is skin damage. If there is skin damage, observe and evaluate again after X minutes. If the skin damage lasts for at least X minutes, it is recorded as MARSI.

[0058] Collect skin pictures, store them in coding and fill in the risk survey form.

[0059] Specifically, in a specific embodiment, the factors related to the occurrence of MARSI include: endogenous factors and exogenous factors;

[0060] Endogenous factors include: (1) Age: The high-risk groups for MARSI are the elderly and infants. Infants and young children have incomplete skin development, sparse stratum corneum, and incomplete defense function. Allergens, chemical irritants, and microorganisms can easily enter the body through the skin, causing skin allergies or infections. Elderly patients have thinner epithelial layers, weakened cohesion at the dermal-epidermal junction, reduced collagen and elastin, water loss, reduced skin elasticity and repair ability, and other factors that make them more susceptible to MARSI. (2) Gender: Many studies have shown that women are more likely to develop MARSI because female patients have delicate skin and thinner stratum corneum than men. (3) History of allergies: Studies have shown that patients with a history of skin allergies are 12.33 times more likely to develop local adhesive-related contact dermatitis than those without a history of skin allergies. (4) Skin dryness: Patients with dry skin have a 2.691-fold higher risk of developing MARSI than those with normal skin dryness, and patients with moist skin have a 5.957-fold higher risk of developing MARSI than those with normal skin dryness. (5) Skin edema: When the skin is edematous, the water content between the cells in the dermis increases, and the adhesion between the cells decreases, which is less than the adhesion between the skin and the adhesive, increasing the risk of MARSI. (6) Underlying diseases: When patients have diabetes, chronic kidney disease, chronic heart disease, tumors, malnutrition, critical illness, etc., the body's nervous, immune, metabolic and other functions are affected, and the skin immune barrier will also be damaged, and the probability of MARSI is higher. (7) Nutritional status: Studies have found that nutritional indicators such as nutritional risk, low body weight and hypoproteinemia are associated with the risk of MARSI in elderly ICU patients. In patients with poor nutritional status, the junction between the epidermis and the dermis becomes thinner, and collagen and elastic fibers are reduced, which makes the skin more sensitive to external friction and the use of adhesives and more prone to mechanical skin damage.

[0061] Exogenous factors: Several studies have shown that ICU nurses have a low level of awareness of MARSI, and most nurses lack relevant knowledge in selecting, sticking, and removing medical adhesives. In addition, improper sticking and removing techniques of medical staff are high-risk factors for MARSI. There are many types of adhesive products on the market (such as polyurethane, acrylate, silicone, etc.). Different types of viscose products have different characteristics and their effects on the skin are also different. Some tapes have poor air permeability, and the water inside the viscose is not easy to evaporate. The skin is prone to immersion and irritation symptoms, and mechanical damage is prone to occur. When adhesive tapes with a harder backing texture are used on edematous skin or joint activities, it is easy to cause skin damage. Therefore, it is important to reasonably select viscose products and choose the right sticking site for patient skin protection. Disinfectants can keep the skin clean and reduce bacterial growth to a certain extent, and play a protective role, but excessive cleaning can reduce the natural oil on the skin surface, thereby causing dry skin. Studies have shown that disinfectants containing quaternary ammonium and aldehyde components can easily damage the skin's epidermal protective layer and cause irritant dermatitis.

[0062] Furthermore, it also includes preprocessing the collected data:

[0063] Check if the data is complete;

[0064] Handling missing values ​​and outliers;

[0065] Data transformations were performed, including encoding categorical variables into dummy variables.

[0066] Furthermore, the risk questionnaire is developed by searching domestic and foreign databases such as CNKI, Wanfang Database, Pubmed, ScienceDirect, etc. to find out the risk factors for MARSI in adult ICU patients.

[0067] Furthermore, the univariate analysis of the MARSI data, calculating the correlation index between each variable and the occurrence of MARSI and obtaining the significance results of each variable in the univariate analysis include:

[0068] SPSS26.0 and R3.5 software were used for statistical analysis, univariate analysis was performed for each variable, and the correlation index between each variable and the occurrence of MARSI was calculated;

[0069] Non-normally distributed quantitative data were expressed as medians and quartiles, and nonparametric tests or chi-square tests were used for comparison between groups. Count data were expressed as cases (%). For discrete variables, the incidence of each level was calculated and compared. For continuous variables, the mean or median between different groups was calculated and compared.

[0070] According to the set significance level P, the statistical significance of the variables in each univariate analysis was judged, and the variables that met the significance level threshold were judged to be related to the occurrence of MARSI.

[0071] Furthermore, it also includes the evaluation of the constructed Logistic multivariate regression model, including the evaluation of the model's goodness of fit and predictive ability, and the use of goodness of fit tests and model discrimination indicators to evaluate the model.

[0072] Furthermore, it also includes the validation of the nomogram risk prediction model, and the discrimination and calibration of the nomogram risk prediction model are verified by the area under the ROC curve, drawing the calibration curve, and the Hosmer-Lemeshow chi-square test. As a commonly used clinical prediction model, the nomogram predicts disease risk based on the patient's relevant clinical data through multifactor analysis and cumulative points, and the final result is presented in the form of a graph. To a certain extent, it replaces complex and cumbersome calculation formulas. The nomogram is applied to clinical external evaluation with high sensitivity and specificity.

[0073] In a specific embodiment, a method for predicting the risk of MARSI in adult ICU patients comprises the following steps:

[0074] Step 1: Research Subject

[0075] Inclusion criteria: ① Age >18 years old; ② ICU stay ≥24 hours; ③ Patients using medical adhesive. Exclusion criteria: ① Incomplete data; ② Previous skin disease; ③ Skin infection or skin injury at admission or transfer

[0076] Step 2: Sample size calculation

[0077] The calculation formula for estimating the overall rate based on a sample survey is: The allowable error was set to δ = 0.04. According to the incidence of medical adhesive-related skin injuries in adult ICU patients ranging from 10.96% to 11.86%, α = 0.05, and Z 1-0.05 / 2 =1.96, P=0.1186, and the calculated sample size n=250. Considering the exclusion criteria, the loss rate was set at 15%, and the sample size of this part was determined to be 288 cases.

[0078] Step 3: Observe the indicators

[0079] The evaluation criteria are based on the MARSI standards issued by the International Skin Care and Ostomy Care Association.

[0080] Mechanical damage:

[0081] ① Skin peeling: When removing adhesive tape or sticky dressings, one or more layers of the skin's stratum corneum are also removed. The damage is generally superficial and irregular in shape. The skin may be shiny, and erythema or blisters may form.

[0082] ② Skin laceration: partial or full separation of skin layers due to shear force, friction and / or blunt force.

[0083] ③ Tension injury or blisters: When the adhesive tape or dressing used is hard and improperly adhered, shear force is generated when the skin swells or the joints move, resulting in separation of the epidermis and dermis and the appearance of tension blisters.

[0084] dermatitis:

[0085] ① Contact dermatitis: non-allergic dermatitis caused by chemical substances, which occurs in the contact area with the adhesive tape or dressing, has clear boundaries, and may cause redness, swelling, blisters, etc., and lasts for a short time.

[0086] ② Allergic dermatitis: A cell-mediated immune response caused by stimulation from certain components in the adhesive tape or the adhesive tape backing. Typical manifestations are erythema, blisters, and itching, which occur in the contact area with the adhesive tape or dressing, or over a wider area, and last for about a week.

[0087] Maceration: Skin wrinkles, white or gray due to prolonged exposure to moist, airtight environments such as adhesive bandages or dressings. As the skin softens, its permeability increases, making it more susceptible to damage from shear forces and irritants.

[0088] Folliculitis: An inflammatory reaction of the hair follicles due to shaving or bacterial retention, manifested as redness and swelling around the hair follicles, which may be non-purulent papules or pustules.

[0089] If it is difficult to distinguish between contact dermatitis and allergic dermatitis, invite a dermatologist to help differentiate, or perform a patch test. If the same adhesive can cause MARSI in different parts of the same patient, it can be considered allergic dermatitis.

[0090] Step 4: Information Collection

[0091] ① The researcher and an intravenous therapy specialist nurse served as trainers to train the department's responsible nurses on MARSI-related knowledge and the precautions for filling out the "Medical Adhesive-Related Skin Injury Risk Survey Form for Adult ICU Patients", and provided supervision and guidance.

[0092] ② After the patient has been admitted to the ICU for ≥24 hours, the responsible nurse shall fill in the "Medical Adhesive-Related Skin Injury Risk Survey Form for Adult ICU Patients" item by item according to unified standards and methods. After each change of the adhesive product, observe whether there is skin damage. If there is skin damage, two members of the research team will observe and evaluate again after 30 minutes. Skin damage lasting 30 minutes or longer will be recorded as MARSI. The research team members will also collect skin pictures, encode and store them, and fill in the "Appendix to the Medical Adhesive-Related Skin Injury Risk Survey Form for Adult ICU Patients".

[0093] Step 5: Quality Control

[0094] ① Personnel control: Before the investigation and research began, the researchers conducted MARSI-related theoretical training for all nurses through multimedia and pictures; using the method of operational training, ICU nursing staff all used a unified adhesive sticking method and removal technique, that is, using a single-handed film tension-free sticking method, a high-lift platform method to fix the catheter, and 0° or 180° to remove the adhesive product. The degree of mastery of MARSI theoretical knowledge and operational methods by all ICU nursing staff was evaluated through theoretical and operational assessments.

[0095] ②Material control: All viscose materials are viscose products purchased uniformly by the hospital, including: 3M disposable ECG electrodes, Xinpigu (type C1, type L1, type A3), Kanglidi sterile dressing (6cm*7cm), Jinbang medical self-adhesive dressing, Yijiliang hydrocolloid dressing, Yijile foam dressing, CVC disposable dressing pack B5 type.

[0096] Step 6: Data processing

[0097] (1) Collect relevant data: Collect relevant data on MARSI in adult ICU patients, including risk factors (such as age, gender, disease severity score, etc.) and whether MARSI occurs or not.

[0098] (2) Data cleaning and processing: Check whether the data is complete, handle missing values ​​and outliers, and perform appropriate data transformation, such as encoding categorical variables into dummy variables.

[0099] (3) Univariate analysis: SPSS26.0 and R3.5 software were used for statistical analysis. Non-normally distributed quantitative data were expressed as medians and quartiles, and non-parametric tests or chi-square tests were used for inter-group comparisons. Count data were expressed as cases (%). Univariate analysis was performed on the risk factors for MARSI in adult ICU patients. Univariate analysis was performed on each variable, and the correlation indicators between each variable and the occurrence of MARSI were calculated, such as chi-square test, t-test or Mann-Whitney U test. For discrete variables, the incidence rate of each level was calculated and compared; for continuous variables, the mean or median between different groups was calculated and compared.

[0100] (4) Determine significance: Determine the statistical significance of each variable in the univariate analysis based on the set significance level (usually 0.05). For variables with (P < 0.05), they are judged to be related to the occurrence of MARSI, and the variables with (P < 0.05) in the univariate analysis are included in the logistic multivariate regression analysis to determine the risk factors for adult ICU patients.

[0101] (5) Constructing a logistic multivariate regression model: Taking the significant univariate analysis variables as predictor variables and the occurrence or non-occurrence of MARSI as the predicted variable, a logistic multivariate regression model was constructed. The model was constructed using stepwise regression or full model regression methods.

[0102] (6) Model evaluation: Evaluate the constructed logistic multivariate regression model, including the model's goodness of fit and predictive ability. Use goodness of fit tests, such as the Pearson chi-square test and the Likelihood Ratio chi-square test, and model discrimination indicators, such as AUC value, sensitivity, and specificity, to evaluate the quality of the model.

[0103] (7) Interpretation of results: Based on the results of the constructed logistic multivariate regression model, explain the impact of each variable on the occurrence of MARSI, including the coefficient of the variable or the adjusted odds ratio, confidence interval, and p-value.

[0104] (8) Determination of risk factors:

[0105] (a) The results of the logistic regression multivariate analysis based on the constructed logistic multivariate regression model showed that the independent risk factors for MARSI included: skin edema, history of MARSI, history of skin allergies, summer, moist skin, and patient self-care ability. Based on these six risk factors, the constructed logistic regression model: its equation is P = e (-1.857+1.079*皮肤水肿+1.437*MARSI史+1.188*皮肤过敏史+0.940*夏季+0.943*皮肤潮湿-1.137*患者自我护理能力良好) / 1+e (-2.261+1.079*皮肤水肿+1.437*MARSI史+1.188*皮肤过敏史+0.940*夏季+0.943*皮肤潮湿-1.137*患者自我护理能力良好), e is a natural number, approximately equal to 2.71828182. The corresponding values ​​of skin edema are shown in Table 1, and the scores and corresponding probabilities are shown in Table 2.

[0106] Table 1

[0107]

[0108] Table 2

[0109] Total score Probability 55 0.1 110 0.2 150 0.3 180 0.4 210 0.5 235 0.6 265 0.7 305 0.8 360 0.9

[0110] Based on statistically significant variables, risk factors affecting the occurrence of MARSI in adult ICU patients were determined, including skin edema, history of MARSI, history of skin allergy, summer, moist skin, and patient self-care ability. These variables were considered to be independent predictors of the occurrence of MARSI.

[0111] (b) The “rms” package in R language (R3.6.2) was used to draw a nomogram for risk prediction of MARSI in adult ICU patients, and the “pROC” package was used to draw the receiver operating characteristic curve (ROC curve) of the nomogram model to evaluate the predictive efficiency and clinical applicability of the nomogram model.

[0112] (c) The discrimination and calibration of the prediction model were verified by the area under the ROC curve, calibration curve, and Hosmer-Lemeshow chi-square test:

[0113] 1) Calculate the area under the ROC curve (AUC): Use the pROC package or ROCR package in R language to calculate the area under the ROC curve (AUC). The area under the ROC curve measures the ability of the model to distinguish between observations with events and observations without events.

[0114] 2) Draw a calibration curve: Use the calibrate package or rms package in R to draw a calibration curve. The calibration curve represents the comparison between the actual incidence of the observed results and the incidence predicted by the model.

[0115] 3) Perform Hosmer-Lemeshow chi-square test: The Hosmer-Lemeshow chi-square test was performed using the logistf package in R language to assess the calibration of the model.

[0116] d) The collected risk factors are introduced into the established risk prediction nomogram to calculate the probability of MARSI.

[0117] In a specific embodiment, a MARSI risk prediction system for adult ICU patients, such as Figure 2 As shown, including:

[0118] Acquisition module: used to collect MARSI data of adult ICU patients;

[0119] Analysis module: used to perform univariate analysis on MARSI data, calculate the correlation index between each variable and MARSI occurrence, and obtain the significance results of each variable in the univariate analysis;

[0120] Model building module: used to construct a Logistic multi-factor regression model by taking the significant univariate analysis variables as predictor variables and the occurrence or non-occurrence of MARSI as the predicted variable;

[0121] Judgment module: used to obtain independent risk factors based on the constructed Logistic multi-factor regression model;

[0122] Prediction module: used to construct a nomogram risk prediction model based on independent risk factors, and to perform real-time risk prediction of MARSI in adult ICU patients through the nomogram risk prediction model.

[0123] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0124] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the risk of MARSI in adult ICU patients. It is characterized in that include: Collect MARSI data of adult ICU patients; Perform univariate analysis on MARSI data, calculate the correlation index between each variable and MARSI occurrence, and obtain the significance results of each variable in the univariate analysis; The significant univariate analysis variables were used as predictor variables, and the occurrence or not of MARSI was used as the predicted variable to construct a Logistic multivariate regression model. According to the constructed Logistic multivariate regression model, independent risk factors were obtained; A nomogram risk prediction model was constructed based on independent risk factors, and the real-time risk prediction of MARSI in adult ICU patients was performed using the nomogram risk prediction model.

2. A method for predicting MARSI risk in adult ICU patients according to claim 1, It is characterized in that The collection of MARSI data of adult ICU patients includes determining the sample size: Estimated overall rate based on sample survey: Error δ = 0.04, P is the incidence of medical adhesive-related skin damage in adult ICU patients, α is the allowable error, is the quantile of the standard normal distribution, where Z is a variable with unit standard deviation of the standard normal distribution, α is the significance level, and 1-α / 2 divides the significance level into two parts. The α / 2 on the left is the first type error, and the 1-α / 2 on the right is the second type error.

3. A method for predicting MARSI risk in adult ICU patients according to claim 1, It is characterized in that The MARSI data of adult ICU patients include observation indicators, which are: mechanical injury: skin peeling, skin laceration, tension injury or blisters; dermatitis: contact dermatitis, allergic dermatitis; maceration; folliculitis.

4. A method for predicting MARSI risk in adult ICU patients according to claim 1, It is characterized in that The collection of MARSI data of adult ICU patients includes: After the patient has been in the ICU for ≥24 hours, fill in the risk survey form; After each change of adhesive product, observe whether there is skin damage. If there is skin damage, observe and evaluate again after X minutes. If the skin damage lasts for at least X minutes, it is recorded as MARSI. Collect skin pictures, encode and store them.

5. A method for predicting MARSI risk in adult ICU patients according to claim 4, It is characterized in that It also includes preprocessing of collected data: Check if the data is complete; Handling missing values ​​and outliers; Data transformation was performed, including coding categorical variables into dummy variables.

6. A method for predicting MARSI risk in adult ICU patients according to claim 4, It is characterized in that The risk questionnaire was developed by searching CNKI, Wanfang Database, Pubmed and ScienceDirect databases to obtain risk factors for MARSI in adult ICU patients.

7. A method for predicting MARSI risk in adult ICU patients according to claim 1, It is characterized in that The univariate analysis of the MARSI data, calculation of the correlation index between each variable and the occurrence of MARSI and obtaining the significance results of each variable in the univariate analysis include: Statistical analysis was performed using software, univariate analysis was performed on each variable, and the correlation index between each variable and the occurrence of MARSI was calculated; Non-normally distributed quantitative data were expressed as medians or quartiles, nonparametric tests or chi-square tests were used for inter-group comparisons, and count data were expressed as proportions. For discrete variables, the incidence of each level was calculated and compared. For continuous variables, the mean or median between different groups was calculated and compared. According to the set significance level P, the statistical significance of the variables in each univariate analysis was judged, and the variables that met the significance level threshold were judged to be related to the occurrence of MARSI.

8. A method for predicting MARSI risk in adult ICU patients according to claim 1, It is characterized in that It also includes the evaluation of the constructed Logistic multivariate regression model, including the evaluation of the model's fit and predictive ability, and the use of goodness of fit tests and model discrimination indicators to evaluate the model.

9. A method for predicting MARSI risk in adult ICU patients according to claim 1, Features: It also includes validation of the nomogram risk prediction model, and verifies the discrimination and calibration of the nomogram risk prediction model through the area under the ROC curve, drawing the calibration curve, and the Hosmer-Lemeshow chi-square test.

10. A MARSI risk prediction system for adult ICU patients. It is characterized in that include: Acquisition module: used to collect MARSI data of adult ICU patients; Analysis module: used to perform univariate analysis on MARSI data, calculate the correlation index between each variable and MARSI occurrence, and obtain the significance results of each variable in the univariate analysis; Model building module: used to construct a Logistic multi-factor regression model by taking the significant univariate analysis variables as predictor variables and the occurrence or non-occurrence of MARSI as the predicted variable; Judgment module: used to obtain independent risk factors based on the constructed Logistic multi-factor regression model; Prediction module: used to construct a nomogram risk prediction model based on independent risk factors, and to perform real-time risk prediction of MARSI in adult ICU patients through the nomogram risk prediction model.

Images