Cesarean section postoperative acute pain prediction model

Abstract

A group of marker sets capable of effectively predicting acute pain of a patient after cesarean section is screened out, and the marker sets comprise whether the patient has preoperative anxiety or not, whether the patient is gestational diabetes mellitus or not, the age of the patient, the Pittsburgh sleep quality index (PSQI) of the patient, the gestational week age of the patient and the abdominal circumference of the patient. On the basis, a visual and quantifiable column diagram prediction model is constructed, the method can be used for evaluating the risk probability of acute pain occurring after a cesarean section patient is operated clinically before an operation, targeted and individualized pain management is provided for the patient subsequently, the pregnancy safety of the patient is improved, and physical and psychological pain of the patient after the operation is relieved.

Description

Technical Field

[0001] This invention belongs to the field of medical diagnostic technology, specifically relating to the construction and application of a prediction model. Background Technology

[0002] Postoperative acute pain refers to severe pain in and around the surgical incision that occurs within 48 hours after the completion of surgery. Postoperative pain after cesarean section is a common physiological reaction in postpartum women, mainly including somatic pain and visceral pain. Somatic pain primarily originates from the abdominal surgical incision, originating from mechanoreceptors in the skin and abdominal wall. This pain is transmitted via Aδ neurons to the dorsal horn of the spinal cord and then ascends to the spinothalamic tract. Aδ fiber nociceptors are small-diameter spinal sensory neurons, characterized by rapid transmission speed, sharp pain sensation, and relatively clear localization. Visceral pain mainly arises from temporary compression and occlusion of blood vessels above the uterus, causing temporary hypoxia and ischemia in surrounding tissues. This pain is transmitted via C fibers along the sympathetic pathway to the pelvic cavity and the superior hypogastric plexus, then to the lumbar sympathetic chain and enters the T nerve. 10 The white communicating branches of L1 and the posterior roots of these nerves enter the spinal cord and eventually ascend to the thalamus. C-fiber nociceptors are small-diameter, unmyelinated sensory neurons that encode persistent pain, but with relatively ambiguous localization.

[0003] In the process of pain signal transmission, there are multiple regulatory links that process the signals. Among them, the enhancement of nociceptive afferent signals is called "sensitization." This process is affected by various substances, cytokines, and pathways, leading to significant differences in pain perception among different individuals. In addition, different external factors also have a significant impact on the degree of pain. Currently, opioids are the main drugs for intravenous analgesia after cesarean section, and epidural analgesia is considered the most suitable method for postoperative analgesia in women who have undergone cesarean section. However, the adverse reactions associated with opioid use have attracted widespread attention. Studies have shown that among patients who use opioids after cesarean section, 1 in 300 patients exposed to opioids become long-term opioid users (Posteromedial quadratus lumborum block versus woundinfiltration after caesarean section: A randomised, double-blind, controlled study[J]. Eur J Anaesthesiol. 2021;38(Suppl 2):S138-S144.). Intrathecal morphine is considered by international researchers to be an ideal choice for treating acute pain after cesarean section. However, related side effects such as nausea, vomiting, sedation, and itching may affect the mother and the mother-infant bond. The ideal pain management after cesarean section is to achieve maximum analgesia with the minimum dosage while ensuring the safety of both mother and baby. Therefore, individualized assessment and prediction of postpartum pain after cesarean section are becoming increasingly important for better pain management and the development of targeted pain management plans.

[0004] One widely used method for predicting postoperative pain is through preoperative experimental pain assessment. However, studies have shown no significant correlation between preoperative pain threshold and postoperative pain scores after cesarean section (Postcesarean section pain prediction by preoperative experimental pain assessment[J]. Anesthesiology. 2003;98(6):1422-1426.). In a study conducted by Hsu et al. using preoperative stress stimulation and anxiety questionnaires, the mean VAS pain score of patients with high anxiety was significantly higher than that of patients with mild anxiety immediately after surgery. However, at 24 hours after surgery, there was no significant difference in the mean VAS pain scores between the two groups, and there was no significant correlation between preoperative anxiety state and morphine consumption immediately after surgery or at 24 hours after surgery (Predicting postoperative pain by preoperative pressure pain assessment[J]. Anesthesiology. 2005;103(3):613-618.). A comprehensive analysis incorporating data from multiple countries showed that four preoperative risk factors—age, preoperative chronic pain, female gender, and preoperative opioid intake—could predict the occurrence of severe postoperative pain. However, the accuracy of this predictive model was low, with an AUC of 0.607 (Predicting poor postoperative acute pain outcome in adults: an international, multicentre database analysis of risk factors in 50,005 patients[J]. Pain Rep. 2020;5(4):e831.). Existing techniques have confirmed that certain factors can predict postoperative acute pain. In a prospective study of 47 women undergoing elective cesarean section, pain catastrophizing and response to experimental heat stimuli were found to predict post-cesarean section pain (Pain catastrophizing, response to experimental heat stimuli, and post-cesarean section pain[J]. J Pain. 2007;8(3):273-279.).A 2019 review of psychological factors affecting pain concluded that pain catastrophizing, optimism, pain anticipation, neuroticism, anxiety, negative emotions, and depression may all be associated with acute postoperative pain, with pain catastrophizing being the most closely related to acute postoperative pain (Preoperative predictors of poor acute postoperative pain control: a systematic review and meta-analysis[J]. BMJ Open. 2019;9(4):e025091.).

[0005] In conclusion, predicting postoperative pain in women undergoing cesarean section is a prerequisite for targeted and individualized pain management, and an essential step in improving maternal pregnancy safety and reducing postoperative physical and psychological suffering. Summary of the Invention

[0006] This invention discovers that preoperative anxiety, gestational diabetes, age, Pittsburgh Sleep Quality Index (PSQI), gestational age, and abdominal circumference influence the occurrence of acute postoperative pain after cesarean section, and the probability of acute postoperative pain after cesarean section can be predicted by the values ​​of these indicators. Based on this, this invention was completed.

[0007] In a first aspect, the present invention provides a set of biomarkers for predicting the risk of acute pain after cesarean section, the set of biomarkers including a combination of preoperative anxiety, gestational diabetes mellitus, age, PSQI, gestational age, and abdominal circumference; the set of biomarkers is determined by a p-value representing the probability of acute pain after cesarean section, the p-value being calculated using the following formula: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6, X1 represents whether or not there is preoperative anxiety, which is a binary variable. In the calculation formula, if the patient has preoperative anxiety, then X1=80; if the patient does not have preoperative anxiety, then X1=0. X2 represents whether the patient has gestational diabetes, which is a binary variable. In the calculation formula, if the patient has gestational diabetes, then X2 = 60; if the patient does not have gestational diabetes before the operation, then X2 = 0. X3 is age, a continuous variable, which can be directly substituted into the patient's actual age (in years). X4 is the PSQI, which is a continuous variable and is directly substituted into the patient's PSQI score; X5 represents gestational age in weeks, which is a continuous variable. It is directly substituted into the patient's actual gestational age in weeks (unit: days). X6 represents waist circumference, a continuous variable, which is directly substituted into the patient's actual waist circumference (unit: cm).

[0008] Furthermore, when the patient's P value is greater than 0.5, the patient has a high risk of experiencing acute pain after cesarean section.

[0009] In a second aspect, the present invention provides the application of the biomarker set as described in the first aspect in the preparation of a reagent for predicting acute pain after cesarean section, wherein the reagent is capable of detecting the biomarker indicators described in the first aspect.

[0010] Furthermore, the biomarker set is determined by the probability P-value of acute pain after cesarean section, and the formula for calculating the P-value is as follows: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6, X1 represents whether or not there is preoperative anxiety, which is a binary variable. In the calculation formula, if the patient has preoperative anxiety, then X1=80; if the patient does not have preoperative anxiety, then X1=0. X2 represents whether the patient has gestational diabetes, which is a binary variable. In the calculation formula, if the patient has gestational diabetes, then X2 = 60; if the patient does not have gestational diabetes before the operation, then X2 = 0. X3 is age, a continuous variable, which can be directly substituted into the patient's actual age (in years). X4 is the PSQI, which is a continuous variable and is directly substituted into the patient's PSQI score; X5 represents gestational age in weeks, which is a continuous variable. It is directly substituted into the patient's actual gestational age in weeks (unit: days). X6 represents waist circumference, a continuous variable, which is directly substituted into the patient's actual waist circumference (unit: cm).

[0011] Furthermore, when the patient's P value is greater than 0.5, the patient has a high risk of experiencing acute pain after cesarean section.

[0012] Thirdly, the present invention provides a risk prediction system for acute pain after cesarean section, the system comprising a data input module, a data processing module and a result output module; The data input module allows users to input data such as whether the patient has preoperative anxiety, whether they have gestational diabetes, their age, PSQI, gestational age, and abdominal circumference. The data processing module processes the data input from the data input module to obtain the probability P-value of acute pain after a cesarean section. The formula for calculating the P-value is as follows: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6; The output module refers to the prediction results obtained from the output data processing module.

[0013] Furthermore, in the data processing module, X1 represents whether there is preoperative anxiety, which is a binary variable. In the calculation formula, when the patient has preoperative anxiety, X1=80; when the patient does not have preoperative anxiety, X1=0. X2 represents whether the patient has gestational diabetes, which is a binary variable. In the calculation formula, if the patient has gestational diabetes, then X2 = 60; if the patient does not have gestational diabetes before the operation, then X2 = 0. X3 is age, a continuous variable, which can be directly substituted into the patient's actual age (in years). X4 is the PSQI, which is a continuous variable and is directly substituted into the patient's PSQI score; X5 represents gestational age in weeks, which is a continuous variable. It is directly substituted into the patient's actual gestational age in weeks (unit: days). X6 represents waist circumference, a continuous variable, which is directly substituted into the patient's actual waist circumference (unit: cm).

[0014] Furthermore, when the patient's P value is greater than 0.5, the patient has a high risk of experiencing acute pain after cesarean section.

[0015] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the contents of the system described in the third aspect.

[0016] Beneficial effects This invention identifies a set of biomarkers that can effectively predict acute pain after cesarean section. These biomarkers include preoperative anxiety, gestational diabetes, patient age, PSQI, gestational age, and abdominal circumference. Based on this, this invention constructs a visualized and quantifiable nomogram prediction model, which can be used for preoperative clinical assessment of the risk probability of acute pain after cesarean section. This provides targeted and individualized pain management for patients, improving pregnancy safety and reducing postoperative physical and psychological suffering. Attached Figure Description

[0017] Figure 1The value represents the incidence of acute postoperative pain in patients undergoing cesarean section, where a: training set; b: validation set.

[0018] Figure 2 The results of LASSO regression analysis on acute postoperative pain in patients undergoing cesarean section are shown. A: Lasso path coefficient plot; B: Lasso cross-validation plot.

[0019] Figure 3 This is a line graph.

[0020] Figure 4 Calibration curves for a prediction model of postoperative acute pain in cesarean section patients, where A: training set; B: validation set. Figure 5 The Hosmer-Lemeshow goodness-of-fit test is used to evaluate the prediction model of acute postoperative pain in cesarean section patients, where A: training set; B: validation set.

[0021] Figure 6 ROC curve for predicting acute postoperative pain in cesarean section patients.

[0022] Figure 7 Clinical decision curve analysis for a predictive model of acute postoperative pain in cesarean section patients.

[0023] Figure 8 Clinical impact curve analysis of a predictive model for acute postoperative pain in cesarean section patients, where A: training set; B: independent validation cohort. Detailed Implementation

[0024] The specific embodiments of the present invention will be further described below. It should be noted that these descriptions are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the embodiments described below can be combined with each other as long as they do not conflict with each other.

[0025] Unless otherwise specified, the experimental methods used in the following embodiments are conventional methods, and the experimental materials used in the following embodiments can be purchased through conventional commercial channels unless otherwise specified.

[0026] Example 1: Case Selection This study ultimately included 350 eligible full-term, low-risk patients scheduled for elective cesarean section, who were randomly assigned to the training set (n=245) and the validation set (n=105) in a 7:3 ratio. "Postoperative acute pain" was defined as a VAS score ≥4 (moderate or severe pain) at any time point, based on the visual analog scale (VAS) scores at rest, during movement, and during uterine contractions at 6 h, 12 h, 24 h, and 48 h postoperatively.

[0027] 1. Discharge Standards Patient inclusion criteria: full-term pregnancy, low risk and elective cesarean section; singleton pregnancy; age ≥18 years.

[0028] Exclusion criteria for patients: history of mental illness, central nervous system disease, liver disease, abnormal kidney function, abnormal cardiopulmonary function; obese patients with a body mass index (BMI) > 35; severe obstetric complications, such as placenta previa or placental abruption; contraindications to combined spinal-epidural anesthesia, including coagulation disorders, anticoagulation therapy, severe hypovolemia or hemodynamic instability; patients who refuse to participate.

[0029] 2. Data Acquisition Baseline data collection: Collect general clinical information of patients, including age, pre-pregnancy BMI, gestational BMI, gestational age, abdominal circumference, number of previous surgeries, number of previous hysteroscopy, number of previous laparoscopy, number of previous deliveries, number of previous cesarean sections, whether cerclage is present, gestational diabetes, and gestational thyroid disease.

[0030] Preoperative questionnaire indicators: PSQI; Amsterdam Preoperative Anxiety and Information Scale (APAIS).

[0031] Intraoperative indicators: Intraoperative hypotension. The mean blood pressure of two measurements after entering the operating room is used as the baseline blood pressure. Intraoperative hypotension is considered to be a decrease of ≥20% in mean arterial pressure after anesthesia, systolic blood pressure <90 mmHg or diastolic blood pressure <60 mmHg. Other indicators include whether uterine fibroids are present, whether uterine fibroids have been removed, postoperative hypotension, time from skin incision to delivery of the fetus, whether carboprost tromethorphan was used, uterine artery ligation, total fluid resuscitation, intraoperative blood loss, and neonatal weight.

[0032] Postoperative pain indicators: VAS scores at rest, during movement, and during uterine contractions at 6 h, 12 h, 24 h, and 48 h postoperatively; number of active analgesia pump presses; and the proportion of indomethacin suppositories used.

[0033] Acute pain index: resting pain >4 or movement pain >5 or uterine contraction pain >6 or simultaneously actively pressing the analgesia pump and using indomethacin suppositories.

[0034] 3. Preoperative auxiliary examinations: Anesthesia assessment: Anesthesia assessment is conducted 24 hours before surgery by a senior anesthesiologist at the anesthesia clinic.

[0035] Preoperative preparation: 500 ml of lactated Ringer's solution was infused intravenously 1 hour before the operation and the infusion was completed within 10 minutes.

[0036] After the patient enters the operating room: Continuous monitoring of vital signs (including blood pressure, pulse, heart rate, pulse oxygen saturation, and respiratory rate) is conducted. The patient is placed in the right lateral decubitus position, and an epidural puncture is performed at the L2-3 or L3-4 intervertebral space. Once the epidural space is accessed, a spinal puncture needle is inserted. Successful puncture is confirmed by the observation of cerebrospinal fluid outflow. Subsequently, 12-15 mg of 0.5% ropivacaine solution is injected, and a 4-5 cm epidural catheter is inserted.

[0037] Anesthesia level: If the sensory block level reaches T4-T6 before the start of surgery, the anesthesia is considered successful. No additional epidural medication was administered during the operation.

[0038] Intravenous administration: After delivery, 3 mg remimazolam besylate, 8 mg ondansetron, and 5 mg dexamethasone were administered intravenously. Patients requiring intraoperative salvage analgesia or switching to general anesthesia were excluded from the study.

[0039] Transversus Abdominis Plane Block (TAP): The TAP block is performed by a senior anesthesiologist within 30 minutes post-surgery. The patient is then transported back to the ward with an analgesia pump.

[0040] 4. Statistical Methods For continuous variables that follow a normal distribution, a t-test is used, and the results are expressed as mean ± standard deviation. For continuous variables that do not follow a normal distribution, a Mann-Whitney U-test is used, and the results are expressed as median [Q1, Q3]. For categorical variables that meet the criteria for a chi-square test, a chi-square test is used, and the results are expressed as frequency and percentage. For categorical variables that do not meet the criteria for a chi-square test, a Fisher exact test is used, and the results are expressed as frequency and percentage.

[0041] Example 2: Screening for risk factors for acute pain The incidence of postoperative acute pain was statistically analyzed in 350 enrolled patients. The results are as follows: Figure 1 As shown, the incidence of postoperative acute pain in the entire cohort was 48.57%; the incidence of postoperative acute pain in the training set was 46.94%, and the incidence of postoperative acute pain in the validation set was 52.38%. There was no statistically significant difference in the incidence between the two groups (χ²=0.87, P=0.351), indicating that the incidence of postoperative pain in the two groups was roughly the same and that the two groups were comparable.

[0042] Based on the diagnosis results of acute postoperative pain in 245 patients in the training set, they were divided into a pain group (n=115) and a non-pain group (n=130). The general information, preoperative indicators, intraoperative indicators and postoperative related indicators of the two groups were compared between the two groups.

[0043] Intergroup comparisons showed that, in terms of count data, the incidence of preoperative anxiety was significantly higher in the pain group than in the non-pain group (60.00% vs 28.46%, P<0.001); the incidence of gestational diabetes mellitus was also higher in the pain group than in the non-pain group (27.83% vs 12.31%, P=0.004). In terms of continuous data, the median age of patients in the pain group was lower than that in the non-pain group [30 (27, 34) years vs 32 (28.25, 35) years] (P=0.048); the median PSQI in the pain group was higher than that in the non-pain group [6 (4, 9) points vs 5 (3, 8) points] (P=0.013), indicating poorer preoperative sleep quality; the median gestational age in the pain group was higher than that in the non-pain group [272 (269, 277) days vs 270 (266, 276) days] (P=0.044). In addition, there were no statistically significant differences between the two groups in general information (height, pre-pregnancy BMI, gestational BMI, and abdominal circumference), preoperative comorbidities (thyroid disease and whether uterine fibroids were present), surgical-related indicators (whether cerclage was performed, whether lesions were removed, post-anesthesia hypotension, whether carboprost tromethorphan was used, uterine artery ligation, time from skin incision to delivery, duration of surgery, total intraoperative fluid resuscitation and intraoperative blood loss), past medical history (number of surgeries, number of hysteroscopy sessions, number of laparoscopy sessions, number of deliveries and number of cesarean sections), vital signs (mean arterial pressure upon arrival at the ventricular chamber, lowest mean arterial pressure after anesthesia, and mean arterial pressure upon exiting the ventricular chamber), laboratory test indicators (D-dimer, fibrinogen, white blood cells, neutrophils, lymphocytes, monocytes, eosinophils, basophils, red blood cells, hematocrit, hemoglobin, and platelets), and neonatal weight (all P>0.05), as shown in Table 1.

[0044] Table 1 Comparison of general and perioperative data between the postoperative acute pain group and the non-pain group in the training set. To screen for potential factors associated with acute postoperative pain in cesarean section patients, this invention included 42 variables for analysis based on previously reported risk factors for postoperative pain. Using "occurrence of acute postoperative pain" as the dependent variable, LASSO regression analysis was employed, and the optimal regularization penalty parameter λ was determined through cross-validation. Under the condition of the smallest error λ parameter value (0.039), LASSO regression identified 6 non-zero variables: preoperative anxiety, gestational diabetes mellitus, age, PSQI, gestational age, and abdominal circumference. The remaining 36 variables were excluded. (See attached table). Figure 2 As shown.

[0045] To verify whether collinearity exists among the key variables selected by LASSO, a collinearity diagnosis was performed on the six non-zero coefficient variables (preoperative anxiety, gestational diabetes mellitus, age, PSQI, gestational age, and abdominal circumference), and the variance inflation factor (VIF) was calculated. The results are shown in Table 2. The VIF values ​​of each variable are all much less than 10: preoperative anxiety 1.029, gestational diabetes mellitus 1.021, age 1.112, PSQI 1.025, gestational age 1.108, and abdominal circumference 1.029, indicating that there is no serious collinearity among the above variables, which meets the prerequisite for subsequent predictive model construction.

[0046] Table 2. Collinearity diagnostic results of LASSO screening variables for postoperative acute pain in cesarean section patients. Example 3: Constructing a Prediction Model for Acute Postpartum Pain Based on Selected Variables 1. Model construction Based on the six independent risk factors and their corresponding regression coefficients identified in Example 2, a risk probability regression equation was constructed for a risk prediction model of postoperative acute pain in cesarean section patients: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6; Where X1 represents preoperative anxiety, X2 represents gestational diabetes mellitus, X3 represents age, X4 represents PSQI, X5 represents gestational age in weeks, and X6 represents abdominal circumference. Each unit of the linear predictor of this model corresponds to 60.07 points.

[0047] Table 3. Multifactorial Logistic Regression Model for Acute Postoperative Pain in Cesarean Section Patients Based on the above regression equation, a nomogram is plotted to visualize the model. The nomogram is shown below. Figure 3 As shown in the diagram, the top layer of the nomogram represents the "Points" score. The middle layer sequentially presents six independent risk factors and their corresponding categories and scores: (e.g., for preoperative anxiety, "none" corresponds to 0 points, and "present" corresponds to 80 points; for gestational diabetes, "none" corresponds to 0 points, and "present" corresponds to 60 points). The bottom layer of the nomogram represents the "probability of postoperative acute pain risk (0.1-0.9)," where a total score of 88 points corresponds to a risk probability of 0.1, 169 points to 0.3, 220 points to 0.5, 271 points to 0.7, and 352 points to 0.9. When a patient's risk probability at the bottom layer of the nomogram is greater than 0.5, the patient is considered to have a high risk of experiencing acute pain after a cesarean section.

[0048] In clinical application, corresponding scores can be determined based on the patient's various indicators. After summing the scores to obtain the total score, the predicted risk probability of postoperative acute pain can be read from the bottom layer of the nomogram. For example, a 26-year-old cesarean section patient has preoperative anxiety (corresponding score 70), no gestational diabetes (corresponding score 0), age 24 (corresponding score 77), PSQI score 13 (corresponding score 86), gestational age 264 days (corresponding score 18), and abdominal circumference 90cm (corresponding score 10). The scores of each indicator were summed, and the total score was 70+0+77+86+18+10=261 points. According to the risk probability correspondence at the bottom layer of the nomogram, the predicted risk probability of postoperative acute pain for this patient was close to 0.7, which means that the patient had a high risk of postoperative acute pain. Targeted analgesic interventions should be implemented for this patient in the perioperative period, such as strengthening preoperative psychological counseling to alleviate anxiety and optimize sleep quality. At the same time, based on the patient's high-risk prediction results, an individualized multimodal analgesia plan should be developed, including preoperative prophylactic analgesia and postoperative stepwise analgesia management, to reduce the incidence and severity of postoperative acute pain and improve the patient's postoperative recovery quality.

[0049] 2. Evaluation of the model's calibration capability (calibration curve and Hosmer-Lemeshow test) This invention comprehensively evaluates the calibration capability of the model through calibration curves, Hosmer-Lemeshow goodness-of-fit tests, and the Brier score, that is, the consistency between the model's predicted probability of postoperative acute pain and the actual incidence.

[0050] The calibration curve results are as follows Figure 4 As shown, in both the training and validation sets, the model-predicted postoperative acute pain risk curves and the actual observed risk curves show good fit. The mean absolute error of the training set is 0.020, and the mean absolute error of the validation set is 0.069, indicating that the deviation between the predicted and actual values ​​is small.

[0051] The Hosmer-Lemeshow goodness-of-fit test further quantitatively validated this result: the χ² value for the training set was 5.336 (P=0.721), and the χ² value for the validation set was 5.027 (P=0.755). Both P values ​​were greater than 0.05, indicating that there was no statistically significant difference between the model's predicted values ​​and the actual observed values, and the model's fit was ideal (see [link to test results]). Figure 5 Furthermore, the Brillouin index results show that the Brillouin index for the training set is 0.201 (95% confidence interval: 0.178–0.224), and for the validation set it is 0.175 (95% confidence interval: 0.135–0.217). The lower Brillouin index values ​​further confirm the good consistency between the model's predicted probabilities and the actual outcomes. In summary, the model demonstrates good calibration capability.

[0052] Example 4: Performance evaluation of a predictive model for acute pain after cesarean section 1. Evaluation of the discriminative ability of the predictive model (ROC curve analysis) This invention reintroduced 68 patients as an independent validation cohort, generated receiver operating characteristic (ROC) curves, and calculated the area under the curve (AUC) to assess the model's discriminative ability. AUC was used as the core metric (the closer the AUC is to 1, the better the discriminative ability).

[0053] The results are as follows Figure 6 As shown, the AUC of the training set (245 cases) was 0.753 (95% CI: 0.692-0.814), and the AUC of the independent validation cohort (68 cases) was 0.765 (95% CI: 0.647-0.884), which is similar to the results of the training set. This suggests that the discrimination ability of the nomogram model has been well validated in external samples and can effectively distinguish between high-risk and low-risk patients with postoperative acute pain.

[0054] 2. Evaluation of the clinical efficacy of the model To evaluate the clinical applicability of this predictive model, this invention employs Decision Curve Analysis (DCA) for quantitative assessment. Unlike traditional diagnostic performance indicators such as sensitivity and specificity, DCA directly reflects the model's practical application value by integrating the potential benefits and risks of clinical intervention. The analysis uses "threshold probability" as the x-axis and "net benefit" as the y-axis to systematically compare the net benefit differences among three strategies: model intervention, "TreatAll," and "No Intervention." The net benefit of the "No Intervention" strategy is consistently zero.

[0055] Decision curve analysis results are as follows Figure 7 As shown, in the training set (245 cases), when the threshold probability was between 0.04 and 0.87, the net benefit of the model intervention strategy was significantly higher than that of the "full intervention" and "no intervention" strategies, with a maximum net benefit of 0.3. In the independent validation cohort (68 cases), the advantage range of the model intervention further expanded to 0.07 to 0.91, and the maximum net benefit was also 0.3. This suggests that within the aforementioned clinically relevant threshold probability range, identifying high-risk patients for postoperative acute pain using this model and implementing targeted interventions can achieve a clinical net benefit that outweighs the risks, confirming the model's high clinical practical value.

[0056] The Clinical Impact Curve (CIC) further quantifies the clinical effectiveness of this predictive model. Its core function is to assess the model by demonstrating the correlation between the predicted high-risk population and the actual population experiencing the outcome at different thresholds. Results are as follows... Figure 8 As shown, when the threshold probabilities of both the training set (245 cases) and the independent validation cohort (68 cases) are greater than 73%, the high-risk population identified by the model is highly matched with the actual population that experienced postoperative acute pain. This result confirms the model's extremely high effectiveness from the perspective of clinical application.

Claims

1. A set of biomarkers for predicting the risk of acute pain after cesarean section, the set of biomarkers including a combination of preoperative anxiety, gestational diabetes mellitus, age, Pittsburgh Sleep Quality Index (PSQI), gestational age, and abdominal circumference; the set of biomarkers is determined by a p-value representing the probability of acute pain after cesarean section, the p-value being calculated as follows: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6, X1 represents whether or not there is preoperative anxiety, which is a binary variable. In the calculation formula, if the patient has preoperative anxiety, then X1=80; if the patient does not have preoperative anxiety, then X1=0. X2 represents whether the patient has gestational diabetes, which is a binary variable. In the calculation formula, if the patient has gestational diabetes, then X2 = 60; if the patient does not have gestational diabetes before the operation, then X2 = 0. X3 is age, a continuous variable, which can be directly substituted into the patient's actual age (in years). X4 is the PSQI, which is a continuous variable and is directly substituted into the patient's PSQI score; X5 represents gestational age in weeks, which is a continuous variable. It is directly substituted into the patient's actual gestational age in weeks (unit: days). X6 represents the abdominal circumference, which is a continuous variable and should be directly substituted into the patient's actual abdominal circumference (unit: cm).

2. The biomarker set as described in claim 1, wherein the probability P-value indicates that when the patient's P-value is greater than 0.5, the patient has a high risk of experiencing acute pain after cesarean section.

3. The use of the biomarker set as described in claim 1 in the preparation of a reagent for predicting acute pain after cesarean section, wherein the reagent is a reagent capable of detecting the biomarker indicators.

4. In the application described in claim 3, the biomarker set is determined by the probability P-value of acute pain after cesarean section, and the formula for calculating the P-value is as follows: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6, X1 represents whether or not there is preoperative anxiety, which is a binary variable. In the calculation formula, if the patient has preoperative anxiety, then X1=80; if the patient does not have preoperative anxiety, then X1=0. X2 represents whether the patient has gestational diabetes, which is a binary variable. In the calculation formula, if the patient has gestational diabetes, then X2 = 60; if the patient does not have gestational diabetes before the operation, then X2 = 0. X3 is age, a continuous variable, which can be directly substituted into the patient's actual age (in years). X4 is the PSQI, which is a continuous variable and is directly substituted into the patient's PSQI score; X5 represents gestational age in weeks, which is a continuous variable. It is directly substituted into the patient's actual gestational age in weeks (unit: days). X6 represents the abdominal circumference, which is a continuous variable and should be directly substituted into the patient's actual abdominal circumference (unit: cm).

5. In the application as described in claim 3, the probability P value indicates that when the P value is greater than 0.5, the patient has a high risk of experiencing acute pain after cesarean section.

6. A risk prediction system for acute pain after cesarean section, the system comprising a data input module, a data processing module, and a result output module; in, The data input module allows input of data such as whether the patient has preoperative anxiety, whether they have gestational diabetes, age, PSQI, gestational age, and abdominal circumference. The data processing module processes the data input from the data input module to obtain the probability P-value of acute pain after a cesarean section. The formula for calculating the P-value is as follows: logit(P)= -16.817+1.326X1+1.005X2-0.058X3+0.119X4+0.050X5+0.034X6; The output module refers to the prediction result P value obtained from the output data processing module.

7. In the risk prediction system as described in claim 6, in the data processing module, X1 represents whether there is preoperative anxiety, which is a binary variable. In the calculation formula, when the patient has preoperative anxiety, X1 = 80; when the patient does not have preoperative anxiety, X1 = 0. X2 represents whether the patient has gestational diabetes, which is a binary variable. In the calculation formula, if the patient has gestational diabetes, then X2 = 60; if the patient does not have gestational diabetes before the operation, then X2 = 0. X3 is age, a continuous variable, which can be directly substituted into the patient's actual age (in years). X4 is the PSQI, which is a continuous variable and is directly substituted into the patient's PSQI score; X5 represents gestational age in weeks, which is a continuous variable. It is directly substituted into the patient's actual gestational age in weeks (unit: days). X6 represents the abdominal circumference, which is a continuous variable and should be directly substituted into the patient's actual abdominal circumference (unit: cm).

8. In the risk prediction system of claim 6, the probability P value indicates that when the P value is greater than 0.5, the patient has a high risk of experiencing acute pain after cesarean section.

9. A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the contents of the system as described in claim 6.

Images