Method and system for constructing a poor prognosis prediction model for infant bacterial meningitis

Abstract

The present invention discloses a method and system for constructing a poor prognosis prediction model for infant bacterial meningitis. The method comprises the following steps: determining independent risk factors for poor prognosis of bacterial meningitis; constructing a poor prognosis prediction model for bacterial meningitis using R software according to the weights of the variables of the determined independent risk factors for poor prognosis of bacterial meningitis, and drawing a model nomogram; obtaining the probability of recent complications of infant bacterial meningitis according to the sum of the scores of the variables; verifying the constructed poor prognosis prediction model for bacterial meningitis; and establishing a dynamic webpage version of the verified poor prognosis prediction model for bacterial meningitis using R language, so as to automatically calculate the probability of poor prognosis of a certain infant according to which independent risk factors the infant actually has. Effect: Screening clinical characteristics and constructing a prognosis prediction model for infant bacterial meningitis can provide early prognosis judgment for infant bacterial meningitis patients in clinic, and provide a reference basis for the formulation of clinical treatment plans in advance.

Description

Technical Field

[0001] The present invention relates to the field of biomedical technology, and in particular to a method and system for constructing a poor prognosis prediction model for infant bacterial meningitis. Background Art

[0002] Bacterial meningitis is a serious health condition affecting children, which can be rapidly fatal and survivors may have complications. For infants younger than 90 days with bacterial meningitis, they have similar clinical features, and the World Health Organization estimates that 402,414 infants younger than 90 days died from sepsis, meningitis, or both in 2015. Efforts to reduce this burden are particularly needed in Africa and Asia, where the incidence is highest. Even in developed countries such as Finland, France, Ireland, and the United Kingdom, infants have a higher risk of death from bacterial meningitis. However, there is still uncertainty about the risk factors for poor outcomes in infants with bacterial meningitis, and continued disease surveillance is considered a public health priority.

[0003] At present, there are very few literatures on the high-risk factors for predicting poor prognosis of bacterial meningitis in newborns, and the diagnostic criteria, definition criteria for poor prognosis, follow-up time and conclusions are not the same, and there is a lack of reports in the past 5 years. Liu Min, Lu Tiantian, Chen Xiuling, et al., Meta-analysis of poor prognostic factors for bacterial meningitis in newborns in my country, Journal of Practical Cardiovascular Diseases, 2020; 28(7): 95-102, conducted a meta-analysis of poor prognostic factors for bacterial meningitis in newborns in China. As of July 2019, 11 case-control studies were included, which only focused on newborns. There is still a lack of research on high-risk factors related to bacterial meningitis in infants less than 90 days old at home and abroad.

[0004] Therefore, the current clinical judgment on the prognosis of early bacterial meningitis in infants of this age group is still insufficient, which brings serious burden to the society and family economy. Therefore, developing a prediction model for the adverse prognosis of bacterial meningitis in infants less than 90 days old and making it into a small program will certainly be of great help in the early assessment of the prognosis of bacterial meningitis in infants and clinical management to improve the prognosis. Summary of the invention

[0005] The present invention aims to solve one of the technical problems in the related art at least to a certain extent. To this end, the purpose of the present invention is to provide a method and system for constructing a poor prognosis prediction model for infant bacterial meningitis.

[0006] To achieve one of the above purposes, a method for constructing a poor prognosis prediction model for infant bacterial meningitis according to an embodiment of the present invention comprises the following steps:

[0007] S1. Identify independent risk factors for poor prognosis in bacterial meningitis;

[0008] S2. According to the weights of the independent risk factors for poor prognosis of bacterial meningitis, a prediction model for poor prognosis of bacterial meningitis was constructed using R software, and a model nomogram was drawn;

[0009] S3. The prediction model obtains the probability of recent complications of infant bacterial meningitis based on the sum of the scores of each variable;

[0010] S4. Validate the constructed prediction model for poor prognosis of bacterial meningitis;

[0011] S5. Use R language to establish a dynamic webpage version of the verified bacterial meningitis adverse prognosis prediction model to automatically calculate the probability of adverse prognosis of a certain infant based on the independent risk factors that the infant actually has.

[0012] In addition, the poor prognosis prediction model method for infant bacterial meningitis according to the above embodiment of the present invention may also have the following additional technical features:

[0013] According to an embodiment of the present invention, step S1 specifically includes:

[0014] S11. Screen infants with bacterial meningitis who meet the exclusion criteria;

[0015] S12. Clinical data collection for infants with bacterial meningitis who meet the inclusion criteria;

[0016] S13. Multivariate Logistic regression analysis was performed on the collected clinical data to screen and identify independent risk factors for poor prognosis (complications or death at discharge). The independent risk factors identified included lethargy, convulsions, and cerebrospinal fluid white blood cell count>332×10 6 / L, cerebrospinal fluid protein>2.5g / L and cerebrospinal fluid glucose<1.6mmol / L.

[0017] According to one embodiment of the present invention, in step S11,

[0018] Inclusion criteria for infants with bacterial meningitis were: age of onset within 90 days and meeting the diagnostic criteria for bacterial meningitis;

[0019] The exclusion criteria for infants with bacterial meningitis were as follows: (1) infants with congenital neurodevelopmental malformations, such as meningocele and spina bifida; (2) infants with genetic metabolic diseases and multiple malformations; (3) infants with severe asphyxia, hypoxic-ischemic encephalopathy, bilirubin encephalopathy, grade IV intraventricular hemorrhage, trauma, and gestational age < 28 weeks or birth weight < 1000 g; (4) infants with incomplete clinical data due to various reasons, including demographic data, clinical data, and auxiliary examination data.

[0020] According to one embodiment of the present invention, in step S12, the collected clinical data include: demographic information, clinical symptoms, signs, complications, disease outcome information evaluated at discharge, and laboratory test data of infants with an age of onset within 90 days.

[0021] According to one embodiment of the present invention, in step S12,

[0022] The demographic data of infants with an onset age of less than 90 days at least include the infant's age, gestational age, gender and perinatal information;

[0023] The clinical symptoms include at least fever, vomiting and convulsions;

[0024] The physical signs include at least a full anterior fontanelle, neck resistance, and abnormal muscle tone;

[0025] The complications include at least ventriculitis, subdural effusion, hydrocephalus and intracranial hemorrhage;

[0026] The disease outcome information assessed at discharge included cure, death / abandonment, and the presence of neurological abnormalities;

[0027] The laboratory test data at least include routine blood tests, blood inflammatory indicators, and blood / cerebrospinal fluid bacterial culture, smears and other etiological tests.

[0028] According to an embodiment of the present invention, for step S2, in the drawn model nomogram, each variable corresponds to a point on the variable axis where the corresponding variable is located in the nomogram, and the point is used as a vertical line on the variable axis corresponding to the upper scoring scale to obtain the assigned value of the variable;

[0029] Among them, the symptom of drowsiness was assigned 75 points, and no drowsiness was assigned 0 points; convulsions were assigned 65 points, and no convulsions were assigned 0 points; cerebrospinal fluid leukocytes> 332×106 / L were assigned 50 points, and cerebrospinal fluid leukocytes ≤ 332×10 6 / L was assigned 0 points; CSF protein>2.5g / L was assigned 70 points, and CSF protein≤2.5g / L was assigned 0 points; CSF glucose<1.6mmol / L was assigned 100 points, and CSF glucose≥1.6mmol / L was assigned 0 points.

[0030] According to one embodiment of the present invention, for step S3, the sum of the scores of each variable is set to T, the probability of recent complications of infant bacterial meningitis is set to S, and S=T / sum of the highest scores of each variable.

[0031] According to an embodiment of the present invention, step S4 specifically includes:

[0032] S41. Analyze the discrimination of the model using the ROC curve, perform internal validation of the bacterial meningitis prediction model in the modeling cohort of the bacterial meningitis prediction model, and perform external validation of the bacterial meningitis prediction model in the validation cohort of the bacterial meningitis prediction model;

[0033] S42. Using the Bootstrap repeated sampling method (1000 times), the Hosmer-Lemeshow goodness of fit test was performed on the bacterial meningitis prediction model in the modeling cohort and the validation cohort;

[0034] S43. The decision curve analysis (DCA) method was used to evaluate the clinical applicability of the bacterial meningitis prediction model in the modeling cohort and the validation cohort.

[0035] To achieve the second of the above objectives, a system for constructing a poor prognosis prediction model for infant bacterial meningitis according to an embodiment of the present invention is constructed by the method for constructing a poor prognosis prediction model for infant bacterial meningitis described above.

[0036] According to one embodiment of the present invention, a system for constructing a poor prognosis prediction model for infant bacterial meningitis according to an embodiment of the present invention is applied to perform early poor prognosis prediction for infant bacterial meningitis to obtain the risk of poor prognosis of bacterial meningitis patients when they are discharged from hospital.

[0037] The beneficial effects of the present invention are:

[0038] The method for constructing a poor prognosis prediction model for infant bacterial meningitis provided in the present application, in specific implementation, is carried out through the following steps: determining independent risk factors for poor prognosis of bacterial meningitis; constructing a poor prognosis prediction model for bacterial meningitis using R software according to the weights of each variable of the determined independent risk factors for poor prognosis of bacterial meningitis, and drawing a model nomogram; the prediction model obtains the probability of recent complications of infant bacterial meningitis according to the sum of the scores of each variable; verifies the constructed poor prognosis prediction model for bacterial meningitis; and uses R language to establish a dynamic webpage version of the verified poor prognosis prediction model for bacterial meningitis, so as to automatically calculate the probability of poor prognosis of a certain infant according to which independent risk factors the infant actually has. The poor prognosis prediction model for infant bacterial meningitis can be easily constructed, and a prediction model for the prognosis of infant bacterial meningitis can be constructed by screening clinical characteristics, which can provide early prognosis judgment for infant bacterial meningitis patients in clinical practice, and provide a reference basis for the formulation of clinical treatment plans in advance.

[0039] In other words, the system software constructed through this prediction model can be used to conduct early and rapid assessment of prognosis screening and guide treatment for infants with bacterial meningitis, which will help improve the prognosis, making this application very valuable in clinical practice.

[0040] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. 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 some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0042] Figure 1 It is a flowchart of the steps of the method for constructing a poor prognosis prediction model for infant bacterial meningitis of the present invention;

[0043] Figure 2 is a step flow chart of step S1;

[0044] Figure 3 is a step flow chart of step S4;

[0045] Figure 4 This is a screening display diagram for screening out infants with bacterial meningitis that meet the exclusion criteria in step S11.

[0046] Figure 5 It is a list chart of multi-factor Logistic regression analysis of clinical data collected in step S13;

[0047] Figure 6 is a display diagram of the model nomogram drawn in step S2;

[0048] Figure 7 It is a display diagram of analyzing the discrimination degree of the model using the ROC curve in the step S41;

[0049] Figure 8 The figure is a calibration curve display diagram of performing Hosmer-Lemeshow goodness of fit test on the bacterial meningitis prediction model in the modeling cohort and the validation cohort respectively by using the Bootstrap repeated sampling method (1000 times) in step S42;

[0050] Fig. 9In step S43, a decision curve display diagram is used to evaluate the clinical applicability of the bacterial meningitis prediction model in the modeling cohort and the validation cohort respectively by using the decision curve analysis method (DCA);

[0051] Fig.10 In step S5, the verified poor prognosis prediction model for bacterial meningitis is used to establish a dynamic webpage version using R language to automatically calculate the probability of a poor prognosis of a certain infant based on which independent risk factors the infant actually has.

[0052] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0053] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the drawings of the specification, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the drawings are exemplary and are intended to be used to explain the present invention, but cannot be understood as limiting the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present invention.

[0054] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “circumferential”, “radial”, etc., indicating orientations or positional relationships are based on the orientations or positional relationships shown in the drawings of the specification, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0055] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0056] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0057] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0058] The following describes in detail the method for constructing a poor prognosis prediction model for infant bacterial meningitis according to an embodiment of the present invention with reference to the accompanying drawings.

[0059] Embodiment 1

[0060] Reference Figures 1 to 10 As shown, according to the method for constructing a poor prognosis prediction model for infant bacterial meningitis provided by an embodiment of the present invention, the control Figure 1 As shown, the following steps are included:

[0061] S1. Identify independent risk factors for poor prognosis in bacterial meningitis;

[0062] S2. According to the weights of the independent risk factors for poor prognosis of bacterial meningitis, a prediction model for poor prognosis of bacterial meningitis was constructed using R software, and a model nomogram was drawn;

[0063] S3. The prediction model obtains the probability of recent complications of infant bacterial meningitis based on the sum of the scores of each variable;

[0064] S4. Validate the constructed prediction model for poor prognosis of bacterial meningitis;

[0065] S5. Use R language to establish a dynamic webpage version of the verified bacterial meningitis adverse prognosis prediction model to automatically calculate the probability of adverse prognosis of a certain infant based on the independent risk factors that the infant actually has.

[0066] Based on the above, it can be clearly understood that, in its specific implementation, the present application provides a method for constructing a poor prognosis prediction model for infant bacterial meningitis. After the above steps S1, S2, S3, S4 and S5, a poor prognosis prediction model for infant bacterial meningitis can be easily constructed. By screening clinical characteristics and constructing a prediction model for the prognosis of infant bacterial meningitis, early prognosis judgment can be provided for infants with bacterial meningitis in clinical practice, and a reference basis can be provided for the formulation of clinical treatment plans in advance. In other words, the software constructed by the prediction model can be used to conduct early and rapid prognosis assessment and screening and guide treatment of infants with bacterial meningitis, which helps to improve the prognosis, making the present application very valuable in clinical practice.

[0067] Furthermore, in the specific implementation, Figure 2 As shown, according to one embodiment of the present invention, step S1 specifically includes:

[0068] S11. Screen infants with bacterial meningitis who meet the exclusion criteria;

[0069] S12. Clinical data collection for infants with bacterial meningitis who meet the inclusion criteria;

[0070] S13. Multivariate Logistic regression analysis was performed on the collected clinical data to screen and identify independent risk factors for poor prognosis (complications or death at discharge). The independent risk factors identified included lethargy, convulsions, and cerebrospinal fluid white blood cell count>332×10 6 / L, cerebrospinal fluid protein>2.5g / L and cerebrospinal fluid glucose<1.6mmol / L.

[0071] Specifically, according to an embodiment of the present invention, in step S11,

[0072] Inclusion criteria for infants with bacterial meningitis were: age of onset within 90 days and meeting the diagnostic criteria for bacterial meningitis;

[0073] The exclusion criteria for infants with bacterial meningitis were as follows: (1) infants with congenital neurodevelopmental malformations, such as meningocele and spina bifida; (2) infants with genetic metabolic diseases and multiple malformations; (3) infants with severe asphyxia, hypoxic-ischemic encephalopathy, bilirubin encephalopathy, grade IV intraventricular hemorrhage, trauma, and gestational age < 28 weeks or birth weight < 1000 g; (4) infants with incomplete clinical data due to various reasons, including demographic data, clinical data, and auxiliary examination data.

[0074] For the above-mentioned step S11, the screening display diagram for screening infants with bacterial meningitis that meet the exclusion criteria is shown in the attached diagram of the present application specification. Figure 4 shown.

[0075] The screening subjects were infants with bacterial meningitis (BM) within 90 days of age who were admitted to Peking University First Hospital, Kunming Children's Hospital (January 1, 2019 to December 31, 2020), Peking University Shenzhen Hospital (January 1, 2015 to December 31, 2021), Shenzhen Children's Hospital (January 1, 2017 to December 31, 2021) and Shenzhen Maternal and Child Health Hospital (January 1, 2019 to December 31, 2021) from January 1, 2007 to December 31, 2021 as the modeling group.

[0076] The validation subjects were infants with bacterial meningitis (BM) who met the same inclusion and exclusion criteria and were admitted to Kunming Children's Hospital and Shenzhen Children's Hospital from January 1 to October 31, 2022 as the validation group.

[0077] For the multi-factor Logistic regression analysis list diagram of the collected clinical data in step S13, please refer to the attached manual of this application. Figure 5 As shown;

[0078] Since the variation range of measurement data is quite wide, the measurement data with statistical differences in univariate analysis were converted into corresponding binary variables using the optimal cutoff value of the ROC curve as the boundary value.

[0079] Multivariate logistic stepwise regression analysis showed that lethargy [OR: 2.826 (1.584, 5.043), p = 0.000], convulsion [OR: 2.609 (1.378, 4.942), p = 0.003], increased WBC count in cerebrospinal fluid (> 332 × 106 / L) [OR: 1.895 (1.085, 3.310), p = 0.025], increased cerebrospinal fluid protein (> 2.5 g / L) [OR: 2.822 (1.530, 5.202), p = 0.001] and decreased cerebrospinal fluid glucose (< 1.6 mmol / L) [OR: 3.941 (2.211, 7.024), p = 0.000] were independent risk factors for poor prognosis in infants with bacterial meningitis (BM).

[0080] In this way, the screening process of this application for screening out infants with bacterial meningitis who meet the exclusion criteria will be very detailed and clear, taking all aspects into consideration, so that infants with bacterial meningitis who meet the exclusion criteria can be screened out very accurately, laying a solid foundation for the construction of a poor prognosis prediction model for infant bacterial meningitis.

[0081] Furthermore, in a specific implementation, according to an embodiment of the present invention, in step S12, the collected clinical data include: demographic information of infants with an age of onset within 90 days, clinical symptoms, signs, complications, disease outcome information evaluated at discharge, and laboratory test data.

[0082] Preferably, according to an embodiment of the present invention, in step S12,

[0083] The demographic data of infants with an onset age of less than 90 days at least include the infant's age, gestational age, gender and perinatal information;

[0084] The clinical symptoms include at least fever, vomiting and convulsions;

[0085] The physical signs include at least a full anterior fontanelle, neck resistance, and abnormal muscle tone;

[0086] The complications include at least ventriculitis, subdural effusion, hydrocephalus and intracranial hemorrhage;

[0087] The disease outcome information assessed at discharge included cure, death / abandonment, and the presence of neurological abnormalities;

[0088] The laboratory test data at least include routine blood tests, blood inflammatory indicators, and blood / cerebrospinal fluid bacterial culture, smears and other etiological tests.

[0089] In this way, when the present application collects clinical data on infants with bacterial meningitis that meet the inclusion criteria, the collected clinical data is also very detailed and clear, and all aspects are taken into consideration, so that the clinical data of infants with bacterial meningitis that meet the inclusion criteria can be collected very accurately, so as to further lay a solid foundation for the construction of a poor prognosis prediction model for infant bacterial meningitis.

[0090] Furthermore, in this technical solution, according to one embodiment of the present invention, for step S2, Figure 6 As shown in the model nomogram, each variable corresponds to the point on the variable axis where the corresponding variable is located in the nomogram, and the point is used as a vertical line on the variable axis corresponding to the upper scoring scale to obtain the assigned value of the variable;

[0091] Among them, the symptom of drowsiness was assigned 75 points, and no drowsiness was assigned 0 points; convulsions were assigned 65 points, and no convulsions were assigned 0 points; cerebrospinal fluid leukocytes> 332×106 / L were assigned 50 points, and cerebrospinal fluid leukocytes ≤ 332×10 6 / L was assigned 0 points; CSF protein>2.5g / L was assigned 70 points, and CSF protein≤2.5g / L was assigned 0 points; CSF glucose<1.6mmol / L was assigned 100 points, and CSF glucose≥1.6mmol / L was assigned 0 points.

[0092] On this basis, according to one embodiment of the present invention, for step S3, the sum of the scores of each variable is set to T, the probability of recent complications of infant bacterial meningitis is set to S, and S=T / sum of the highest scores of each variable.

[0093] The sum of the highest scores of each variable = 75 points (sleepiness) + 65 points (convulsions) + 50 points (cerebrospinal fluid white blood cells> 332 × 10 6 / L) + 70 points (CSF protein>2.5g / L) + 100 points (CSF glucose<1.6mmol / L) = 360 points.

[0094] That is, for step S3, the sum of the scores of the variables is set to T, the probability of the occurrence of recent complications of infant bacterial meningitis is set to S, and S=T / 360.

[0095] In addition, in the specific implementation, the control Figure 3 As shown, according to one embodiment of the present invention, the step S4 specifically includes S41, S42 and S43;

[0096] S41. Analyze the discrimination of the model using the ROC curve, perform internal validation of the bacterial meningitis prediction model in the modeling cohort of the bacterial meningitis prediction model, and perform external validation of the bacterial meningitis prediction model in the validation cohort of the bacterial meningitis prediction model;

[0097] Among them, the control Figure 7 As shown, it is a display diagram of analyzing the discrimination of the model using the ROC curve in step S41.

[0098] In the Figure 7 A total of 568 cases were included in the modeling cohort, and the ROC curve was used to analyze the discrimination of the model, and the complication rate was found to be 43.5% (247 / 568); a total of 96 cases were included in the validation cohort, and the ROC curve was used to analyze the discrimination of the model, and the complication rate was found to be 30.2% (29 / 96), which was significantly lower than that of the validation cohort (p<0.05). The baseline data of the two cohorts were basically balanced.

[0099] In the Figure 7 The area under the ROC curve of the modeling cohort was 0.824 (95% CI: 0.7882-0.8594). The clinical data of the validation cohort were included in the prediction model for external validation, and the area under the ROC curve was 0.839 (95% CI: 0.751-0.9273), indicating that the model had good discrimination ability.

[0100] S42. Using the Bootstrap repeated sampling method (1000 times), the Hosmer-Lemeshow goodness of fit test was performed on the bacterial meningitis prediction model in the modeling cohort and the validation cohort;

[0101] Among them, the control Figure 8 As shown, it is a calibration curve display diagram of the Hosmer-Lemeshow goodness of fit test of the bacterial meningitis prediction model in the modeling cohort and the verification cohort respectively using the Bootstrap repeated sampling method (1000 times) in the step S42.

[0102] Internal validation of the modeling cohort: The calibration curve of the modeling cohort was close to the reference line, and the goodness of fit test showed a good fit (χ 2 =3.3095, df=6, p=0.7691).

[0103] External validation of the validation cohort: The calibration curve goodness-of-fit test of the validation cohort also showed a good fit (χ 2 =3.7163, df=5, p=0.5909).

[0104] S43. The decision curve analysis (DCA) method was used to evaluate the clinical applicability of the bacterial meningitis prediction model in the modeling cohort and the validation cohort.

[0105] Among them, the control Fig. 9 As shown, it is a decision curve display diagram of evaluating the clinical applicability of the bacterial meningitis prediction model in the modeling cohort and the verification cohort respectively using the decision curve analysis (DCA) in step S43.

[0106] In the Fig. 9 In the study, decision curve analysis was used. In both the modeling cohort and the validation cohort, the decision curves were basically higher than the two extreme curves. The results showed that clinical intervention had a higher net benefit when the threshold probabilities of the modeling cohort and the validation cohort were 10% to approximately 100%, respectively.

[0107] It can be seen that after the above steps S41, S42 and S43, the constructed bacterial meningitis poor prognosis prediction model has been well verified, making the bacterial meningitis poor prognosis prediction model constructed in this application consistent with reality.

[0108] It should be added that, for the step S5, Fig.10In step S5, the verified poor prognosis prediction model for bacterial meningitis is used to establish a dynamic webpage version using R language to automatically calculate the probability of a poor prognosis of a certain infant based on which independent risk factors the infant actually has.

[0109] Embodiment 2

[0110] To achieve the second objective above, a system for constructing a poor prognosis prediction model for infant bacterial meningitis according to an embodiment of the present invention is constructed by the method for constructing a poor prognosis prediction model for infant bacterial meningitis described in the first embodiment above.

[0111] According to one embodiment of the present invention, a system for constructing a poor prognosis prediction model for infant bacterial meningitis according to an embodiment of the present invention is applied to perform early poor prognosis prediction for infant bacterial meningitis to obtain the risk of poor prognosis of bacterial meningitis patients when they are discharged from hospital.

[0112] Comparison Fig.10 As shown, when the present application is subsequently applied, the probability of a poor prognosis of a certain infant can be automatically calculated based on which independent risk factors the infant actually has.

[0113] For example, in Fig.10 In the dynamic web version of the bacterial meningitis adverse prognosis prediction model system software shown, if an infant has the following three independent risk factors for adverse prognosis, including symptoms of drowsiness, convulsions, and cerebrospinal fluid protein>2.5g / L, they will be checked. Fig.10 The dynamic web-based bacterial meningitis poor prognosis prediction model system shown will automatically generate the probability of poor prognosis of bacterial meningitis S=(75+65+70) / 360=58.3%.

[0114] For example, in Fig.10 In the dynamic web version of the bacterial meningitis adverse prognosis prediction model system software shown, if an infant has three independent risk factors for adverse prognosis, including convulsions, cerebrospinal fluid protein>2.5g / L and cerebrospinal fluid glucose<1.6mmol / L, then they will be checked. Fig.10 The dynamic web-based bacterial meningitis poor prognosis prediction model system shown will automatically generate the probability of poor prognosis of bacterial meningitis S=(65+70+100) / 360=65.3%.

[0115] In this way, the verified bacterial meningitis adverse prognosis prediction model is used to establish a dynamic web page version using R language, so that the probability of adverse prognosis of a certain infant can be automatically and quickly calculated based on the independent risk factors that the infant actually has, making the application very intuitive and clear to use, and very quick and time-saving.

[0116] In addition, the nomogram model constructed in this application based on the independent risk factors for predicting poor prognosis of infants with bacterial meningitis (BM) screened out in the modeling group includes five important clinical indicators: sleepiness, convulsions, cerebrospinal fluid white blood cell count, cerebrospinal fluid protein and cerebrospinal fluid glucose levels, which are all indicators that are easily obtained in the clinic, with strong feasibility and practicality, and can be intuitively used to predict the prognosis risk of individual diseases. From the nomogram score, from high to low, they are: cerebrospinal fluid glucose level, sleepiness, cerebrospinal fluid protein level, convulsions, cerebrospinal fluid white blood cell count, which also reflects the weight of each indicator in the prognosis evaluation, which is basically consistent with the results of previous studies summarized in the previous article.

[0117] This model was externally validated in infants with bacterial meningitis in two children's hospitals in the following year (2022). The incidence of complications in the validation cohort was significantly lower than that in the modeling group (30.2% vs 43.5%, p < 0.05), reflecting that with the development of medicine, the prognosis of the disease has improved, but about one-third of young infants with bacterial meningitis (BM) still have poor short-term prognosis, and the situation remains relatively serious, which is basically consistent with the current reality.

[0118] At the same time, since the validation cohort is a local children's hospital, it includes referral patients from primary hospitals below the second level, avoiding the limitations and type bias of patients admitted to general hospitals and maternal and child institutions, and has a certain degree of representativeness. In the validation cohort, the prediction model also showed a high prediction performance (AUC: 0.839, 95% CI: 0.751-0.9273), and the model has good fit and accuracy. It is used to predict the adverse prognosis risk of infant bacterial meningitis (BM), and has certain clinical applicability and promotion.

[0119] Other embodiments and the like are not described here by way of example.

[0120] In summary, the method and system for constructing a poor prognosis prediction model for infant bacterial meningitis provided by the present application, when implemented specifically, includes the following steps: S1, determining independent risk factors for poor prognosis of bacterial meningitis; S2, constructing a poor prognosis prediction model for bacterial meningitis using R software according to the weights of each variable of the determined independent risk factors for poor prognosis of bacterial meningitis, and drawing a model nomogram; S3, the prediction model obtains the probability of recent complications of infant bacterial meningitis according to the sum of the scores of each variable; S4, verifying the constructed poor prognosis prediction model for bacterial meningitis;

[0121] S5. The verified poor prognosis prediction model for bacterial meningitis is used to establish a dynamic webpage version using R language to automatically calculate the probability of a poor prognosis of a certain infant based on the independent risk factors that the infant actually has. This can easily build a poor prognosis prediction model for infant bacterial meningitis. By screening clinical characteristics and building a prediction model for the prognosis of infant bacterial meningitis, it can provide early prognosis judgment for infant bacterial meningitis patients in clinical practice and provide a reference for the formulation of clinical treatment plans in advance.

[0122] In other words, the system software constructed through this prediction model can be used to conduct early and rapid assessment of prognosis screening and guide treatment for infants with bacterial meningitis, which will help improve the prognosis, making this application very valuable in clinical practice.

[0123] Furthermore, the method and system for constructing a poor prognosis prediction model for infant bacterial meningitis provided by the present application are indeed extremely practical and have excellent use effects, which means that the present application will inevitably have great market promotion value, and the present application will inevitably be very popular and will surely be effectively popularized.

[0124] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0125] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the inventive concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A method for constructing a poor prognosis prediction model for infant bacterial meningitis, characterized in that: The following steps are involved: S1. Identify independent risk factors for poor prognosis in bacterial meningitis; S2. According to the weights of the independent risk factors for poor prognosis of bacterial meningitis, a prediction model for poor prognosis of bacterial meningitis was constructed using R software, and a model nomogram was drawn; S3. The prediction model obtains the probability of recent complications of infant bacterial meningitis based on the sum of the scores of each variable; S4. Validate the constructed prediction model for poor prognosis of bacterial meningitis; S5. Use R language to establish a dynamic webpage version of the verified bacterial meningitis adverse prognosis prediction model to automatically calculate the probability of adverse prognosis of a certain infant based on the independent risk factors that the infant actually has.

2. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 1, characterized in that: The step S1 specifically includes: S11. Screen infants with bacterial meningitis who meet the exclusion criteria; S12. Clinical data collection for infants with bacterial meningitis who meet the inclusion criteria; S13. Multivariate Logistic regression analysis was performed on the collected clinical data to screen and identify independent risk factors for poor prognosis (complications or death at discharge). The independent risk factors identified included lethargy, convulsions, and cerebrospinal fluid white blood cell count>332×10 6 / L, cerebrospinal fluid protein>2.5g / L and cerebrospinal fluid glucose<1.6mmol / L.

3. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 2, characterized in that: In step S11, Inclusion criteria for infants with bacterial meningitis were: age of onset within 90 days and meeting the diagnostic criteria for bacterial meningitis; The exclusion criteria for infants with bacterial meningitis were as follows: (1) infants with congenital neurodevelopmental malformations, such as meningocele and spina bifida; (2) infants with genetic metabolic diseases and multiple malformations; (3) infants with severe asphyxia, hypoxic-ischemic encephalopathy, bilirubin encephalopathy, grade IV intraventricular hemorrhage, trauma, and gestational age < 28 weeks or birth weight < 1000 g; (4) infants with incomplete clinical data due to various reasons, including demographic data, clinical data, and auxiliary examination data.

4. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 2, characterized in that: In step S12, the collected clinical data include: demographic information, clinical symptoms, signs, complications, disease outcome information evaluated at discharge, and laboratory test data of infants within 90 days of onset.

5. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 4, characterized in that: In step S12, The demographic data of infants with an onset age of less than 90 days at least include the infant's age, gestational age, gender and perinatal information; The clinical symptoms include at least fever, vomiting and convulsions; The physical signs include at least a full anterior fontanelle, neck resistance, and abnormal muscle tone; The complications include at least ventriculitis, subdural effusion, hydrocephalus and intracranial hemorrhage; The disease outcome information assessed at discharge included cure, death / abandonment, and the presence of neurological abnormalities; The laboratory test data at least include routine blood tests, blood inflammatory indicators, and blood / cerebrospinal fluid bacterial culture, smears and other etiological tests.

6. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 1, characterized in that: For step S2, in the drawn model nomogram, each variable corresponds to a point on the variable axis where the corresponding variable is located in the nomogram, and the point is used as a vertical line on the variable axis corresponding to the upper scoring scale to obtain the assigned value of the variable; Among them, the symptom of drowsiness is assigned 75 points, and no drowsiness is assigned 0 points; convulsions are assigned 65 points, and no convulsions are assigned 0 points; cerebrospinal fluid white blood cells> 332×10 6 / L is assigned 50 points, and cerebrospinal fluid white blood cells ≤ 332×10 6 / L was assigned 0 points; CSF protein>2.5g / L was assigned 70 points, and CSF protein≤2.5g / L was assigned 0 points; CSF glucose<1.6mmol / L was assigned 100 points, and CSF glucose≥1.6mmol / L was assigned 0 points.

7. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 6, characterized in that: For step S3, the sum of the scores of each variable is set to T, the probability of recent complications of infant bacterial meningitis is set to S, and S=T / the sum of the highest scores of each variable.

8. The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 7, characterized in that: The step S4 specifically includes: S41. Analyze the discrimination of the model using the ROC curve, perform internal validation of the bacterial meningitis prediction model in the modeling cohort of the bacterial meningitis prediction model, and perform external validation of the bacterial meningitis prediction model in the validation cohort of the bacterial meningitis prediction model; S42. Using the Bootstrap repeated sampling method (1000 times), the Hosmer-Lemeshow goodness of fit test was performed on the bacterial meningitis prediction model in the modeling cohort and the validation cohort; S43. The decision curve analysis (DCA) method was used to evaluate the clinical applicability of the bacterial meningitis prediction model in the modeling cohort and the validation cohort.

9. A system for constructing a poor prognosis prediction model for infant bacterial meningitis, characterized in that: The method for constructing a poor prognosis prediction model for infant bacterial meningitis according to any one of claims 1 to 8 is constructed.

10. The system for constructing a poor prognosis prediction model for infant bacterial meningitis according to claim 9, characterized in that: It is used to predict the early adverse prognosis of infant bacterial meningitis and to obtain the risk of adverse prognosis of patients with bacterial meningitis when they are discharged from hospital.

Images