Prediction method for dystrophy risk factors after anorectal plasty

Abstract

The invention discloses a method for predicting dystrophy risk factors after anorectal plasty, and belongs to the technical field of dystrophy risk prediction. The prediction method comprises the following steps: collecting clinical data of patients suffering from anorectal deformity, dividing the patients into a malnutrition group and a non-malnutrition group, and screening out prediction factors possibly influencing the nutrition state; candidate factors are screened out through univariate regression analysis, and then independent influence factors of malnutrition after anorectal plasty are identified through multivariate logistic regression analysis; and establishing a Nomo model for predicting the postoperative malnutrition risk through the identified independent influence factors to obtain the risk probability of the anorectal plasty postoperative malnutrition, and then evaluating the efficiency of the Nomo model. The method can effectively predict risk factors of malnutrition after anorectal plasty.

Description

Technical Field

[0001] This application relates to the technical field of malnutrition risk prediction, and more specifically, to a method for predicting risk factors of malnutrition after anorectoplasty. Background Art

[0002] Congenital anorectal malformation (ARM) is a congenital disease involving the anus and rectum, with a prevalence in the population of approximately 3.26 per 10,000 newborns. Complex ARMs include anal atresia without fistula, rectovesical bulbous fistula, rectoprostatic fistula, rectovesical neck fistula, and cloacal malformation, etc. These types have posed great challenges to pediatric surgeons in terms of preoperative diagnosis, surgical techniques, and postoperative management. In the past few decades, the development of nursing and surgical techniques has significantly reduced the mortality rate of complex ARMs, but a series of problems that occur after anorectoplasty have gradually attracted attention, among which the most concerned include defecation function and quality of life. However, postoperative malnutrition can easily cause more dangerous situations for patients, and there is less research on the postoperative nutritional status assessment and influencing factors of patients with complex ARMs in the current existing technologies.

[0003] Based on this, this application provides a method for predicting risk factors of malnutrition after anorectoplasty. Summary of the Invention

[0004] This application provides a method for predicting risk factors of malnutrition after anorectoplasty, which can effectively predict the risk factors of malnutrition after anorectoplasty.

[0005] This application is implemented as follows: This application provides a method for predicting risk factors of malnutrition after anorectoplasty, including the following steps: Collect the clinical data of patients with anorectal malformation, and divide the patients into a malnutrition group and a non-malnutrition group, and screen out the predictive factors that may affect the nutritional status; Screen out candidate factors through univariate regression analysis, and then use multivariate logistic regression analysis to identify the independent influencing factors of malnutrition after anorectoplasty; establish a Nomo model for predicting the risk of postoperative malnutrition through the identified independent influencing factors, obtain the risk probability of malnutrition after anorectoplasty, and then evaluate the efficacy of the Nomo model.

[0006] In a possible implementation, the predictive factors include: the type of anorectal malformation patients, gestational age, birth weight, feeding method, place of residence, caregiver, caregiver's education level, congenital heart disease, multiple congenital malformations, history of systemic malformation surgery, defecation function at 3 months after stoma, and malnutrition at different ages.

[0007] In a possible implementation, the types of patients with anorectal malformations include rectovesical bulbous fistula or rectoprostatic fistula; the feeding methods include breast milk, artificial feeding, or mixed feeding; malnutrition at different ages includes malnutrition at the first age, the second age, and the third age; the first age is the age corresponding to the time of stoma creation, the second age is the age corresponding to the time of PSARP surgery, and the third age is the age corresponding to the time of stoma closure.

[0008] In a possible implementation, the steps of screening candidate factors through univariate regression analysis include: using RStudio software to perform statistical analysis on the predictive factors; presenting categorical data as n (%) and using the chi-square test for comparison; evaluating the normality of numerical data through the Shapiro-Wilk test: if it conforms to a normal distribution, it is presented as the mean ± standard deviation (SD) and the Student's t-test is used; if it does not conform to a normal distribution, it is presented as the interquartile range and the Mann-Whitney test is used.

[0009] In a possible implementation, the screened candidate factors include: feeding method, caregiver, malnutrition at the third age, congenital heart disease, systemic malformation surgery, and defecation function.

[0010] In a possible implementation, the independent influencing factors screened out include: caregiver, malnutrition at the third age, systemic malformation surgery, and defecation function.

[0011] In a possible implementation, the steps of evaluating the efficacy of the Nomo model include: plotting the receiver operating characteristic curve and calculating the Harrell concordance index based on the characteristic curve to evaluate the discrimination of the Nomo model; using the Hosmer-Lemeshow goodness-of-fit test and the calibration curve to evaluate the calibration of the Nomo model, and plotting the decision curve to evaluate its clinical effectiveness.

[0012] In a possible implementation, the Harrell concordance index of the Nomo model is 0.960.

[0013] In a possible implementation, the corrected Harrell concordance index of the Nomo model is 0.957.

[0014] In a possible implementation, the result of the Hosmer-Lemeshow goodness-of-fit test is χ2 = 1.827, P = 0.873.

[0015] The beneficial effects of this application are at least as follows: The prediction method for risk factors of malnutrition after anorectal plasty in this application has good prediction efficacy, can early identify high-risk malnourished patients and risk factors, help doctors take preventive and intervention measures, adopt preventive and treatment strategies as early as possible, and provide reference for parental guidance. BRIEF DESCRIPTION OF THE DRAWINGS To more clearly illustrate the technical solutions of the embodiments of this application, the following will briefly introduce the drawings required in the embodiments. It should be understood that the following drawings only show some embodiments of this application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other relevant drawings can also be obtained based on these drawings.

[0016] Figure 1 It is the flow chart of screening the research population for clinical data in the embodiment of this application; Figure 2 It is the specific cumulative distribution diagram of the defecation function of the research population in the embodiment of this application; Figure 3 It is the ROC curve analysis for predicting influencing factors of malnourished patients Figure 4 It is the variable allocation diagram of the Nomo model for predicting malnutrition; Figure 5 It is the prediction performance diagram of the Nomo model. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The following will describe the implementation solutions of this application in detail in combination with the embodiments. However, those skilled in the art will understand that the following embodiments are only used to illustrate this application and should not be regarded as limiting the scope of this application. For those not specified in the embodiments, they are carried out according to conventional conditions or conditions recommended by the manufacturer. For reagents or instruments not specified by the manufacturer, they are all conventional products that can be obtained through commercial purchase.

[0018] The following specifically describes the prediction method for risk factors of malnutrition after anorectal plasty in the embodiments of this application: This application provides a prediction method for risk factors of malnutrition after anorectal plasty, including the following steps: (1) Collect the clinical data of patients with anorectal malformations, and divide the patients into a malnutrition group and a non-malnutrition group, and screen out the predictive factors that may affect the nutritional status.

[0019] This study adopted a retrospective analysis of the cases of male patients with anorectal atresia associated with rectovesical bulbous fistula (RB) / rectoprostatic fistula (RP) who underwent posterior sagittal anorectoplasty (PSARP) in the neonatal surgery department of the hospital. It should be noted that only those patients with complete clinical data and willing to cooperate with follow-up were included. Among them, the exclusion criteria were: patients who underwent unplanned surgeries due to postoperative complications (such as rectal prolapse and secondary megarectum); patients with severe deformities affecting the management of anorectal malformations (such as solitary kidney with renal failure and hermaphroditism); patients with missing information. Specifically, Figure 1 is the flow chart for patient inclusion and exclusion in this study.

[0020] Among them, the predictive factors that may affect nutritional status were screened out, including: the type of anorectal malformation in patients, gestational age, birth weight, feeding method, place of residence, caregiver, the educational level of the caregiver, congenital heart disease (CHD), multiple congenital malformations (MCD), history of surgery for systemic malformations, bowel function at 3 months after stoma formation, and malnutrition at different ages.

[0021] Among them, the types of anorectal malformations in patients included rectovesical bulbous fistula or rectoprostatic fistula. The feeding methods included breast milk, artificial feeding, or mixed feeding. The caregivers were divided into parents or others, and the place of residence was divided into urban or rural. Malnutrition at different ages included malnutrition at the first age, the second age, and the third age; the first age corresponded to the age at stoma formation, the second age corresponded to the age at PSARP surgery, and the third age corresponded to the age at stoma closure. In addition, a final nutritional assessment was also conducted on the patients after stoma closure.

[0022] Among them, the height and weight of all children were measured and recorded according to standard methods and compared with the "Growth Standards for Chinese Children", expressed as standard deviation scores (Z-scores), adjusted for age and gender (including height-for-age ratio (HAZ) and weight-for-age ratio (WAZ)). According to the World Health Organization (WHO) standards, when the HAZ and WAZ scores were less than -2, malnutrition was considered to exist.

[0023] Bowel function was evaluated using the revised Rintala questionnaire, which was performed by two experienced surgeons. For children under 3 years old, the revised Rintala questionnaire included 6 questions: whether they could control defecation, whether they had the urge to defecate, defecation frequency, defecation contamination, fecal incontinence, and constipation. According to the scores, the patients were divided into four categories: normal (15 - 17), good (9 - 14), fair (4 - 8), and poor (0 - 3).

[0024] Congenital heart disease is diagnosed by echocardiogram results and is classified into three categories according to the abnormalities present and their severity: none (no abnormalities), mild (patent foramen ovale with significant shunt, secundum atrial septal defect, and / or small ventricular septal defect), and severe (other defects). It should be noted that patent ductus arteriosus is defined as the persistence of the ductus arteriosus 1 month after birth or during cardiac surgery within 1 month after birth. In addition, atrial septal defect is defined as a heart defect in which blood still flows between the left and right atria even after 6 months of age.

[0025] Multiple congenital malformations are defined as the presence of three or more congenital malformations, including congenital anorectal malformations, or the patient has a syndrome confirmed by a clinical geneticist.

[0026] (2) Candidate factors were screened out by univariate regression analysis, and then multivariate logistic regression analysis was used to identify the independent influencing factors of malnutrition after anorectoplasty; a Nomo model for predicting the risk of postoperative malnutrition was established based on the identified independent influencing factors, and the risk probability of malnutrition after anorectoplasty was obtained. Then, the efficacy of the Nomo model was evaluated.

[0027] Exemplarily, the steps of screening out candidate factors by univariate regression analysis include: using R Studio software to perform statistical analysis on the predictive factors; categorical data are expressed as n (%) and compared using the chi-square test; numerical data are evaluated for normality by the Shapiro-Wilk test: if it conforms to a normal distribution, it is expressed as the mean ± standard deviation (SD) and the Student's t-test is used; if it does not conform to a normal distribution, it is expressed as the interquartile range and the Mann-Whitney test is used.

[0028] The candidate factors screened out include: feeding method, caregiver, malnutrition at the third age, congenital heart disease, systemic malformation surgery, and bowel function.

[0029] The independent influencing factors screened out include: caregiver, malnutrition at the third age, systemic malformation surgery, and bowel function.

[0030] The steps for evaluating the efficacy of the Nomo model include: plotting the receiver operating characteristic curve and calculating the Harrell concordance index based on the characteristic curve to evaluate the discrimination of the Nomo model; using the H-L goodness-of-fit test and calibration curve to evaluate the calibration of the Nomo model, and plotting the decision curve to evaluate its clinical effectiveness.

[0031] Among them, the Harrell concordance index of the Nomo model is 0.960. The corrected Harrell concordance index (C-index) calculated by 1000 times of bootstrap resampling is 0.957.

[0032] Among them, the result of the H-L goodness-of-fit test is χ2 = 1.827 and P = 0.873.

[0033] The following further describes in detail the prediction method for risk factors of malnutrition after anorectal plasty according to this application in combination with embodiments.

[0034] This application provides a prediction method for risk factors of malnutrition after anorectal plasty, including the following steps: (1) Collect the clinical data of patients with anorectal malformations, divide the patients into a malnutrition group and a non-malnutrition group, and screen out the predictive factors that may affect the nutritional status.

[0035] In this embodiment, a total of 87 patients met the inclusion and exclusion criteria (RB: 53 cases; RP: 34 cases), and the overall malnutrition rate was 27.6%. Among them, 33 cases (37.9%) had congenital heart disease, 14 cases (16.1%) of which had undergone cardiac surgery, and 14 cases (16.1%) were classified as severe heart disease. It should be noted that 27.6% of the patients underwent surgical treatment related to congenital malformations (Table 1), and the most common surgery was cardiac surgery (58.3%).

[0036]

[0037] Among them, the predictive factors that may affect the nutritional status include: the type of anorectal malformation patients, gestational age, birth weight, feeding method, place of residence, caregiver, caregiver's education level, congenital heart disease, multiple congenital malformations, history of systemic malformation surgery, bowel function at 3 months after stoma, and malnutrition at different ages.

[0038] Among them, the types of anorectal malformation patients include rectovesical bulbous fistula or rectoprostatic fistula. The feeding methods include breast milk, artificial feeding, or mixed feeding. The caregivers are divided into parents or others, and the places of residence are divided into urban or rural. Malnutrition at different ages includes malnutrition at the first age, the second age, and the third age; the first age is the age corresponding to the time of stoma, the second age is the age corresponding to the PSARP surgery, and the third age is the age corresponding to the closure of the stoma. In addition, a final nutritional assessment should be performed on the patient after the stoma is closed.

[0039]

[0040] By analyzing the defecation function, we found that at 1 year old, the defecation function of the patients was poor in 5.75%, fair in 13.8%, good in 55.2% and normal in 25.3%. Most (80.5%) cases had good defecation function, while patients with poor early postoperative defecation function seemed to be more prone to malnutrition ( Figure 2 ).

[0041] (2) Candidate factors were screened out by univariate regression analysis, and then multivariate logistic regression analysis was used to identify the independent influencing factors of malnutrition after anorectoplasty.

[0042] The steps of screening candidate factors by univariate regression analysis included: using R Studio software to perform statistical analysis on the predictive factors; categorical data were expressed as n (%) and compared by chi-square test; numerical data were evaluated for normality by Shapiro-Wilk test: if it conformed to normal distribution, it was expressed as mean ± standard deviation (SD) and Student's t-test was used; if it did not conform to normal distribution, it was expressed as interquartile range and Mann-Whitney test was used.

[0043] The candidate factors screened out included: feeding method, caregiver, malnutrition at the third age, congenital heart disease, systemic malformation surgery and defecation function (P < 0.1).

[0044] The above candidate factors were included in the logistic regression analysis, and the independent influencing factors of malnutrition after anorectoplasty identified by multivariate logistic regression analysis (refer to Table 3) included: caregiver, malnutrition at the third age, systemic malformation surgery and defecation function (P < 0.05). There were no significant differences in ARM type, birth weight, gestational age, place of residence, feeding method, educational level, malnutrition (second age) and MCD between the non-malnutrition group and the malnutrition group (all P > 0.5).

[0045] Then collinearity diagnosis was performed on the above four variables (caregiver, malnutrition at the third age, systemic malformation surgery and defecation function), and the variance inflation factors (VIF) were 1.168, 1.334, 1.252 and 1.139 respectively (all VIF < 5), indicating that there was no multicollinearity relationship.

[0046]

[0047] Among them, CHD is congenital heart disease; β, regression coefficient; SE, standard error; OR, odds ratio; 95% CI, 95% confidence interval.

[0048] Based on the results of multivariate logistic regression analysis, the predictive value of these factors for malnutrition was explored. The results of ROC curve analysis for caregivers, malnutrition (third age), other systemic deformity surgeries, and defecation function showed that the area under the curve (AUC) was 0.687 (95% CI: 0.578 - 0.782), 0.661 (95% CI: 0.551 - 0.759), 0.799 (95% CI: 0.699 - 0.877), and 0.826 (95% CI: 0.746 - 0.905) ( Figure 3 ), and the specific predictive values for each factor are shown in Table 4. Defecation function had significantly better predictive performance in identifying patients with postoperative malnutrition, with a sensitivity of 54.2%, a specificity of 93.7%, a positive predictive value of 76.5%, and a negative predictive value of 84.3%. This indicates that patients with poor / moderate defecation function around 1 year of age are more likely to develop malnutrition.

[0049]

[0050] (3) Through the above analysis, we found that the predictive value (sensitivity and specificity) of single factors was not ideal. Therefore, in this application, a Nomo model for predicting the risk of postoperative malnutrition was established based on the identified independent influencing factors (refer to Figure 4 ), and the risk probability of malnutrition after anorectoplasty was obtained. Among them, the assignment of each variable in the Nomo model: caregiver: other = 1, parent = 0; malnutrition 3: yes = 1, no = 0; other surgeries: yes = 1, no = 0; defecation function: poor = 1, moderate = 2, good = 3, normal = 4.

[0051] (4) Evaluate the efficacy of the Nomo model.

[0052] The steps for evaluating the efficacy of the Nomo model include: plotting the receiver operating characteristic curve and calculating the Harrell concordance index based on the characteristic curve to evaluate the discrimination of the Nomo model; using the H-L goodness-of-fit test and calibration curve to evaluate the calibration of the Nomo model, and plotting the decision curve to evaluate its clinical effectiveness.

[0053] Among them, the Harrell concordance index (C-index) of the Nomo model was 0.960 (95% CI: 0.902 - 0.991, Figure 5 A). To verify the accuracy of this Nomo model, the corrected C-index calculated through 1000 bootstrap resamplings was 0.957, which had good value. The results of the H-L goodness-of-fit test showed that there was no significant difference between the predicted incidence of malnutrition and the actual incidence (χ2 = 1.827, P = 0.873). At the same time, the calibration curve showed that the predicted results were consistent with the actual situation (Figure 5 B). In addition, decision curve analysis showed that when the probability threshold of malnutrition predicted by the Nomo model > 0.02, using this model would bring net benefit compared with the "all treated" strategy or the "no treatment" strategy ( Figure 5 C).

[0054] That is to say, the relationship of the caregiver, malnutrition at the time of stoma closure, surgical history of other systemic malformations, and defecation function are independent influencing factors for postoperative malnutrition in patients with complex anorectal malformations (ARMs). The Nomo model based on these factors has good accuracy and discrimination in evaluating individualized postoperative malnutrition prediction. It shows that the prediction method of risk factors for postoperative malnutrition of the present application has good prediction efficacy, can early identify high-risk malnourished patients and risk factors, help doctors take preventive and intervention measures, take preventive and treatment strategies as early as possible, and provide reference for parental guidance.

[0055] The above are only specific embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for predicting risk factors of malnutrition after anorectal plasty, characterized in that: The following steps are involved: The clinical data of patients with anorectal malformation were collected, and the patients were divided into malnutrition group and non-malnutrition group to screen out the predictive factors that may affect the nutritional status; Candidate factors were screened out through univariate regression analysis, and then multivariate logistic regression analysis was used to identify the independent influencing factors of malnutrition after anorectal plasty; a Nomo model for predicting the risk of postoperative malnutrition was established based on the identified independent influencing factors, the risk probability of malnutrition after anorectal plasty was obtained, and then the effectiveness of the Nomo model was evaluated.

2. The method for predicting risk factors for malnutrition after anorectoplasty according to claim 1, characterized in that: The predictive factors included the type of anorectal malformation patient, gestational age, birth weight, feeding method, place of residence, caregiver, caregiver's education level, congenital heart disease, multiple congenital malformations, history of systemic malformation surgery, defecation function at 3 months of stoma, and malnutrition at different ages.

3. The method for predicting risk factors for malnutrition after anorectoplasty according to claim 2, characterized in that: The types of patients with anorectal malformation include recto-vesical bulbar fistula or recto-prostatic fistula; the feeding methods include breast milk, artificial feeding or mixed feeding; the malnutrition at different ages includes malnutrition at a first age, a second age, and a third age; the first age is the age corresponding to the stoma, the second age is the age corresponding to the PSARP operation, and the third age is the age corresponding to the stoma closure.

4. The method for predicting risk factors of malnutrition after anorectoplasty according to claim 1, characterized in that: The step of screening out candidate factors through univariate regression analysis includes: using R Studio software to perform statistical analysis on the predictive factors; categorical data are represented by n and compared using chi-square test; numerical data are evaluated for normality by Shapiro-Wilk test: if they conform to normal distribution, they are represented by mean ± standard deviation and Student's t test is used; if they do not conform to normal distribution, they are represented by interquartile range and Mann-Whitney test is used.

5. The method for predicting risk factors of malnutrition after anorectoplasty according to claim 4, characterized in that: The candidate factors screened included: feeding pattern, caregiver, malnutrition in the third age, congenital heart disease, systemic malformation surgery, and bowel function.

6. The method for predicting risk factors of malnutrition after anorectoplasty according to claim 1, characterized in that: The independent influencing factors screened out included caregivers, malnutrition in the third age, systemic malformation surgery, and bowel function.

7. The method for predicting risk factors of malnutrition after anorectoplasty according to any one of claims 1 to 6, characterized in that: The step of evaluating the effectiveness of the Nomo model includes: drawing a receiver operating characteristic curve and calculating the Harrell consistency index based on the characteristic curve to evaluate the discrimination of the Nomo model; using the HL goodness of fit test and the calibration curve to evaluate the calibration of the Nomo model, and drawing a decision curve to evaluate its clinical effectiveness.

8. The method for predicting risk factors of malnutrition after anorectoplasty according to claim 7, characterized in that: The Harrell consistency index of the Nomo model was 0.

960.

9. The method for predicting risk factors of malnutrition after anorectal plasty according to claim 8, characterized in that: The corrected Harrell consistency index of the Nomo model was 0.

957.

10. The method for predicting risk factors of malnutrition after anorectoplasty according to claim 7, characterized in that: The HL goodness of fit test result was χ2=1.827, P=0.873.

Images