A model, system, and device for predicting the risk of chronic pain after surgery in adolescents with idiopathic scoliosis.

Abstract

This invention discloses a model, system, and device for predicting the risk of chronic postoperative pain after surgery in adolescents with idiopathic scoliosis, belonging to the field of medical decision support technology. This application provides the first CPSP prediction model integrating psychological and physical factors, and a risk assessment system built based on this model. By acquiring relevant patient parameters, it can calculate an individualized probability of chronic postoperative pain risk and further output risk levels and personalized perioperative management recommendations. This invention is the first to integrate psychological factors (depressive symptoms) with physiological factors (gender, surgical extent, etc.) to construct a prediction model, which has been verified to have excellent discriminative power (AUC up to 0.852), enabling rapid and accurate preoperative identification of high-risk patients and providing a usable decision support tool for clinical implementation of integrated "psychological-pain" interventions.

Description

Technical Field

[0001] This application belongs to the field of medical decision support technology, specifically relating to a model, system, and device for predicting the risk of chronic pain after surgery for adolescent idiopathic scoliosis. Background Technology

[0002] Adolescent idiopathic scoliosis (AIS) is the most common three-dimensional structural deformity of the spine during adolescence, with a global prevalence of approximately 1%–3%, reaching as high as 2%–5% in the 10–16 age group. Large-scale screening data in my country show that the detection rate of AIS among junior high school students is approximately 1.5%–2.7%, with a significantly higher incidence in females than males (male-to-female ratio approximately 1:4–1:7). Although most patients have stable conditions, about 10%–30% of cases progress to the point requiring surgical intervention. With the widespread adoption of early screening and advancements in surgical techniques, the annual number of AIS surgeries in my country has exceeded 10,000 and is showing a year-on-year upward trend.

[0003] For patients with progressively worsening Cobb angles (usually ≥45°-50°) or concomitant cardiopulmonary dysfunction, posterior spinal fusion is currently the primary treatment. While surgery effectively corrects deformities and improves trunk balance and cardiopulmonary function, postoperative pain remains a core issue affecting the quality of rehabilitation. Studies show that up to 60%–80% of AIS patients experience moderate to severe acute pain (NRS ≥ 4) in the early postoperative period, and approximately 15%–30% develop chronic postoperative pain (CPSP, defined as incision area pain lasting more than 3 months), significantly reducing quality of life and even leading to long-term psychological burden and impaired social functioning.

[0004] Depressive symptoms, characterized by persistent low mood, loss of interest, decreased energy, and sleep disturbances, have an overall prevalence of 10-20% in adolescents. In patients with acute myocardial infarction (AIS), the detection rate of depressive symptoms may be even higher due to factors such as body image disturbances, social avoidance, and treatment stress. Current evidence suggests that preoperative depression in the surgical population is closely associated with increased acute postoperative pain, increased opioid consumption, prolonged hospital stays, and delayed functional recovery. However, in the specific AIS population, high-quality prospective evidence regarding the impact of preoperative depression on the dynamic evolution of postoperative pain is lacking, particularly regarding its independent predictive value for CPSP (postoperative pain syndrome), which remains a research gap. Summary of the Invention

[0005] Current technologies for assessing the risk of postoperative chronic pain (CPSP) in adolescent idiopathic scoliosis (AIS) patients largely rely on single or limited physiological indicators such as surgical extent and pain intensity, lacking an integrated consideration of psychological factors. This results in inaccurate risk stratification and hinders early identification and individualized intervention for high-risk patients. Therefore, providing a solution that integrates psychological and physical factors to accurately quantify and predict the risk of postoperative chronic pain in AIS patients is a pressing technical problem that needs to be solved in this field.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A model for predicting the risk of chronic pain after surgery in adolescents with idiopathic scoliosis, the model being a predictive model constructed based on a logistic regression algorithm, the function of which is:

[0008] Logit(P) = β0 + β1X1 + β2X2+ β3X3+ β4X4+ β5X5+ β6X6+ β7X7+ β8X8+ β9X9,

[0009] in,

[0010] P represents the predicted probability of chronic postoperative pain in the target patient;

[0011] X1 represents a preoperative state of depression;

[0012] X2 represents the gender variable;

[0013] X3 represents the number of surgical segment fusions;

[0014] X4 represents the preoperative resting pain score;

[0015] X5 represents the preoperative exercise pain score;

[0016] X6 is an SRS-22 rating.

[0017] X7 is the PSQI score;

[0018] X8 is the rating for the SF-36;

[0019] X9 represents the preoperative hematocrit value;

[0020] β0 is the intercept, and β1, β2, β3, β4, β5, β6, β7, β8, and β9 are the regression coefficients of the model.

[0021] In some embodiments, the regression coefficients of the model are specifically: β0 = -7.31, β1 = 3.72, β2 = 1.85, β3 = 0.27, β4 = 0.08, β5 = 0.19, β6 = -0.02, β7 = -0.11, β8 = 0.02, β9 = 0.02.

[0022] In some embodiments, the X1 variable is defined as follows: X1=1 when the total score of the Child Depression Scale is ≥19, otherwise X1=0.

[0023] In some embodiments, X2=2 when the patient is female and X2=1 when the patient is male.

[0024] A storage medium having the prediction model described in any one of the claims stored thereon.

[0025] A risk assessment system for chronic pain after surgery in adolescent idiopathic scoliosis includes:

[0026] Data input unit, used to acquire data of the target patient's X1 to X9 variables as defined in claim 1;

[0027] A model processing unit, which is loaded with any of the aforementioned prediction models, is used to calculate a risk probability value P based on the data;

[0028] The result output unit is used to output the risk probability value P and related risk assessment information.

[0029] In some embodiments, the result output unit is further configured to:

[0030] Based on the preset threshold range where the risk probability value P is located, output a high-risk, medium-risk, or low-risk level determination; output personalized perioperative management suggestions corresponding to the risk level.

[0031] In some embodiments, the preset threshold range includes:

[0032] When P < 0.2, it is considered low risk;

[0033] When 0.2 ≤ P < 0.5, it is classified as medium risk;

[0034] When P ≥ 0.5, it is considered high risk.

[0035] An electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the functions of any of the systems.

[0036] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is used to implement the functions of any of the systems.

[0037] Compared with the prior art, the beneficial effects of this application are as follows:

[0038] This application is the first to construct and validate a risk prediction model integrating psychological and physical factors. This model, by inputting the patient's gender, the number of surgical fusion segments, and preoperative depressive symptom scores, outputs an individualized probability value for chronic postoperative pain (CPSP). Its predictive power is excellent, with an area under the receiver operating characteristic (AUC) of 0.852, significantly superior to single-factor prediction, providing an objective and reliable quantitative tool for accurate preoperative identification of high-risk patients. Through rigorous analysis, this invention not only confirmed that preoperative depressive symptoms are the strongest independent predictor of CPSP (OR=41.06), but also clarified the independent risk value of female sex and an increased number of fusion segments. This finding transforms previously fragmented clinical observations into predictive variables with clear effect sizes, providing a scientific basis for risk stratification. This application further found that the dynamic changes in perioperative depressive symptoms are significantly and gradient-wise associated with CPSP risk. Compared with patients with persistent symptoms, those with remission of depressive symptoms had a significantly lower risk of CPSP. This result suggests that the predictive system of this invention can not only perform static risk assessment, but its output can also provide a key time window and decision-making basis for dynamically monitoring psychological state and implementing early intervention, thereby potentially preventing the transformation of acute pain into chronic pain. Based on the above results, this application provides direct evidence-based support and technical means for shifting from the traditional "analgesia-centered" model to a "biopsychosocial integrated intervention" model in the perioperative management of adolescent idiopathic scoliosis. By combining standardized psychological assessment with risk prediction models, more comprehensive and personalized patient management pathways can be developed in clinical practice. Attached Figure Description

[0039] Figure 1 Flowchart for a prospective observational study;

[0040] Figure 2 The trajectory of chronic postoperative pain changes in two groups of patients at rest;

[0041] Figure 3 The trajectory of chronic postoperative pain changes in two groups of patients under active conditions;

[0042] Figure 4-1 Figures for establishing and validating a predictive model for chronic postoperative pain: (a) Nonograph (showing some variables), (b) ROC curve;

[0043] Figure 4-2Figures for the establishment and validation of a chronic postoperative pain prediction model: (c) Calibration curve, (d) Decision curve;

[0044] Figure 5 A comparison of ROC curves between combined indicators and single depressive symptom indicators for predicting chronic postoperative pain.

[0045] Figure 6 Bar chart showing the incidence of chronic postoperative pain grouped by different changes in depressive symptoms. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention is further described below with reference to specific embodiments. Unless otherwise described in detail, the technical means used in the following embodiments are all conventional means well known to those skilled in the art, or are performed according to the kit and product instructions. Unless otherwise specified, the materials and reagents used in the following embodiments are commercially available.

[0047] Example 1

[0048] 1. Patients and experimental methods

[0049] 1.1 Experimental Design and Ethical Approval

[0050] The experimental protocol was reviewed and approved by the Ethics Committee of Nanjing University Affiliated Gulou Hospital (Approval No.: 2023-458-02) and registered with the Chinese Clinical Trial Registry (ChiCTR; https: / / www.chictr.org.cn / ) (Registration No.: ChiCTR2300077637). All subjects and their legal guardians signed written informed consent forms after being fully informed.

[0051] 1.2 Experimental Subjects

[0052] This study included adolescent patients scheduled for elective scoliosis correction surgery at our center between November 2023 and August 2024. Inclusion criteria: 1) Age ≥10 years and ≤18 years; 2) Diagnosed with adolescent idiopathic scoliosis and scheduled for elective spinal fusion surgery; 3) American Society of Anesthesiologists (ASA) classification I-III; 4) Informed and signed consent forms from the patient and their guardian. Exclusion criteria: 1) Refusal to participate in this application; 2) Long-term use of analgesics (>7 days) or history of opioid abuse before surgery; 3) Comorbid severe hepatic insufficiency (Child-Pugh C), renal failure (requiring dialysis), or expected survival ≤24 hours; 4) Preoperative diagnosis of schizophrenia, bipolar disorder, epilepsy, or myasthenia gravis; 5) Inability to complete scale assessment due to coma, severe cognitive impairment, language or hearing impairment, etc.; 6) Preoperative uncontrolled severe hypertension (systolic blood pressure >180 mmHg), significant sinus tachycardia (heart rate >120 bpm), acute coronary syndrome or severe intracranial hypertension within the past week; 7) Use of ketamine or esketamine during surgery; 8) Currently participating in other interventional clinical studies.

[0053] 1.3 Data Collection

[0054] Baseline information was collected from patients upon admission, including: gender, age, body mass index (BMI), education level, previous surgical history, medication history, smoking and alcohol consumption history, drug or food allergy history, ASA classification, laboratory test results (complete blood count, liver and kidney function, electrolytes), and auxiliary examinations (electrocardiogram, echocardiography, etc.). Perioperative indicators included: operation time, anesthesia time, total fluid infusion volume, intraoperative urine output, blood loss, and blood transfusion status. All data were independently entered into the electronic database by two uniformly trained researchers and cross-checked to ensure data integrity and accuracy.

[0055] 1.4 Pain Assessment

[0056] Pain intensity at rest and during movement was assessed using the Numerical Rating Scale (NRS) on 1 day preoperatively, 1-5 days postoperatively (twice daily: 8:00-10:00 and 18:00-20:00), 30 days postoperatively, and 90 days postoperatively. The NRS score ranged from 0 to 10 (0 representing "no pain" and 10 representing "the most imaginable pain"). Resting pain refers to pain reported by the patient while lying supine and at rest; movement pain refers to pain assessed immediately after completing a standardized movement (such as turning over or sitting up). The Advanced Pain Support Scale (APSP) was defined as an NRS ≥4, a threshold widely used to identify clinically significant moderate to severe pain. The Pain Persistent Pain Scale (CPSP), defined according to the International Classification of Diseases, 11th Revision (ICD-11), is defined as pain that persists or newly develops after the healing of surgically related tissues, located in the original surgical incision area, and lasts for ≥3 months. This application defines CPSP as persistent incision site pain with a NRS ≥ 4 at 90 days post-operation. Pain data were collected face-to-face in the ward by standardized trained researchers before surgery; data at 30 and 90 days post-operation were obtained through structured telephone interviews. All interviews were recorded in their entirety and independently reviewed by a second researcher to ensure consistency of assessment and data reliability.

[0057] 1.5 Depression Assessment

[0058] The Children's Depression Inventory (CDI) was used to assess patients' depressive symptoms 1 day before surgery, 30 days after surgery, and 90 days after surgery. The CDI consists of 27 items covering multiple dimensions including depressed mood, anhedonia, feelings of worthlessness, interpersonal difficulties, and somatic symptoms. The total score ranges from 0 to 54 (0 indicates "no depressive symptoms," and 54 indicates "all depressive symptoms are present," with higher scores indicating more severe symptoms). This scale has good reliability and validity in adolescents and has been widely used in clinical practice and research. Based on previous research, this application defines a CDI total score ≥19 as the presence of clinically significant depressive symptoms. All questionnaires were completed independently by patients in a quiet environment. When necessary, trained researchers provided neutral interpretations of the item meanings, but did not guide or interfere with the responses.

[0059] 1.6 Other clinical and quality of life assessments

[0060] The Pittsburgh Sleep Quality Index (PSQI) was used to assess patients' sleep quality 1 day before surgery and 1-5 days after surgery (8:00-10:00). A PSQI total score > 5 was defined as impaired sleep quality. Spinal function and health-related quality of life were assessed using the Scoliosis Research Society-22 (SRS-22) scoring system and the Short Form Survey-36 (SF-36) on 1 day before surgery, 30 days after surgery, and 90 days after surgery. The SRS-22 includes five dimensions: functional activity, pain, self-image, psychological state, and treatment satisfaction; the SF-36 covers both physical and mental health. Both systems use standardized scoring systems, with higher scores indicating better function or quality of life.

[0061] 1.7 Surgical and Anesthetic Management

[0062] All patients fasted and abstained from fluids for 8 hours preoperatively. Upon admission, routine monitoring of electrocardiogram (ECG), pulse oximetry (SpO2), non-invasive blood pressure (NIBP), and bispectral index (BIS) was performed. Anesthesia induction was administered intravenously: midazolam 0.1 mg / kg, sufentanil 0.05 μg / kg, propofol 2 mg / kg, and vecuronium bromide 0.1 mg / kg. After endotracheal intubation, mechanical ventilation was initiated using an anesthesia machine with the following parameters: Inspired fraction of oxygen (FiO2) 60%-80%, tidal volume (VT) 8-10 ml / kg, respiratory rate (RR) 12-15 breaths / min, and inspiration-to-expiration ratio (I / E) 1:2. Continuous monitoring of invasive arterial pressure (ABP), central venous pressure (CVP), and blood oxygen saturation (BIS) was performed intraoperatively. Propofol was continuously infused at 4-10 mg / kg during the maintenance phase. -1 .h -1 Remifentanil 0.5-1.0 μg / kg -1 .min -10.2 mg / kg of cisatracurium -1 .h -1 and dexmedetomidine 0.2 μg.kg -1 .h -1 Medication regimens were adjusted based on BIS values ​​(target 40-60) and hemodynamic responses. Circulatory management followed standard procedures: if systolic blood pressure <80 mmHg, ephedrine 5 mg was administered or norepinephrine was administered intravenously or via infusion pump; if heart rate <50 bpm, atropine 0.2 mg was administered intravenously. Muscle relaxants were discontinued 30 minutes before the end of surgery. After the patient regained spontaneous breathing and consciousness, the tube was extubated, and a patient-controlled intravenous analgesia (PCIA) device was connected. Rescue analgesia was initiated when the patient complained of NRS ≥4. To standardize opioid exposure levels, all opioid medications used in the first 5 days post-surgery (including PCIA consumption and rescue medications) were converted to oral morphine milligram equivalents (MME) using internationally accepted conversion factors. Total MME = PCIA cumulative MME + Rescue medication MME.

[0063] 1.8 Data Analysis (Statistical Analysis)

[0064] All data analyses were performed using SPSS 26.0 software (IBM Corp., Armonk, NY, USA). Continuous variables were described as mean ± standard deviation based on their distribution characteristics. The mean or median (interquartile range, IQR) was used for categorical variables; frequencies and percentages [n (%)] were used for categorical variables. Independent samples t-tests (normal distribution), Mann-Whitney U tests (non-normal distribution), or χ² tests (categorical variables) were used for intergroup comparisons. Repeated measures covariance analysis (ANCOVA) was used to assess the impact of preoperative depressive symptoms on the dynamic evolution trajectory of APSP. In the model, "time" (days 1-5 postoperatively) was used as the internal factor, "depressive symptom grouping" as the intergroup factor, and preoperative NRS scores, PSQI, SRS-22 total scores, and SF-36 mental health dimension scores were included as covariates to correct for potential confounding effects. For CPSP (binary outcome variable: yes / no), candidate variables (including age, sex, BMI, preoperative depressive symptoms, number of surgical fusion segments, preoperative pain intensity, sleep quality, and quality of life scores) were first screened using univariate logistic regression. Variables with p < 0.1 were then included in a multivariate logistic regression model, and the backward stepwise method was used to determine independent risk factors for CPSP. Results were reported as odds ratios (OR) and their 95% confidence intervals (95% CI). Furthermore, based on the dynamic changes in CDI scores preoperatively and 90 days postoperatively, patients were divided into three groups: persistent depressive symptoms group (preoperative CDI ≥ 19 and postoperative CDI ≥ 19), depressive symptom relief group (preoperative CDI ≥ 19 but postoperative CDI < 19), and no depressive symptoms group (preoperative CDI < 19 and postoperative CDI < 19). The incidence of CPSP in each group was calculated, and the χ² test was used to compare differences between groups. All statistical tests were two-tailed, and a p-value < 0.05 was considered statistically significant.

[0065] 2. Experimental Results

[0066] 2.1 Baseline Characteristics

[0067] This application recruited 200 patients, of whom 157 completed all follow-ups and were included in the final analysis. For details of loss to follow-up and exclusion, please refer to [link to relevant documentation]. Figure 1Based on the preoperative CDI score (≥19 points defined as clinically significant depressive symptoms), patients were divided into a depressive symptom group (n = 32) and a non-depressive symptom group (n = 125). As shown in Table 1, there were no statistically significant differences between the two groups in perioperative indicators such as demographic characteristics (age, sex, BMI), ASA classification, number of surgical fusion segments, anesthesia time, operation time, intraoperative blood loss, total fluid infusion volume, and transfusion rate (P > 0.05), indicating good baseline comparability between the two groups. However, in terms of psychological and functional status, the depressive symptom group had significantly higher preoperative NRS pain scores at rest and during movement (P < 0.05), higher PSQI scores (indicating poorer sleep quality), and significantly lower SRS-22 and SF-36 total scores than the non-depressive symptom group (P < 0.05), indicating poorer preoperative spinal function and health-related quality of life. The outcome indicators are shown in Table 2. There were no statistically significant differences between the two groups in the incidence of APSP, MME in the first 5 days post-surgery, delirium in the first 5 days post-surgery, all-cause mortality in the 30-day post-surgery period, and SRS-22 score in the 90-day post-surgery period (P > 0.05). However, the incidence of CPSP was significantly higher in the depressive symptom group (P < 0.001), and the RS-22 and SF-36 scores in the 30-day post-surgery group and the SF-36 score in the 90-day post-surgery group were significantly lower than those in the non-depressive symptom group (P < 0.05), suggesting that their postoperative functional recovery and quality of life were more significantly impaired.

[0068] Table 1. Demographic and baseline characteristics of the depressive and non-depressive groups.

[0069]

[0070]

[0071] Note: BMI (Body Mass Index), NYHA (New York Heart Association) functional classification, ASA (American Society of Anesthesiologists), PSQI (Pittsburgh Sleep Quality Index), MMSE (Minor Mental State Examination), SF-36 (Short Scale for Preoperative Health), SRS-22 (Society for the Study of Scoliosis-22), Alb (albumin), NRS (Numerical Rating Scale). Bold p-values ​​indicate statistical significance.

[0072] Table 2 Comparison of outcome indicators between the depressive symptom group and the non-depressive symptom group

[0073]

[0074] Note: Acute pain after APSP, chronic pain after CPSP, MME (morphine milligram equivalent), SRS-22 Scoliosis Study-22 scale, SF-36 Short Form Health Survey. Postoperative time is abbreviated as "Aftersur". "D" represents days. Bold P-values ​​indicate statistical significance.

[0075] 2.2 The impact of preoperative depressive symptoms on APSP

[0076] The Shapiro-Wilk test showed that pain scores at all time points followed a normal distribution. The Mauchly test of sphericity indicated that the covariance matrix did not meet the football shape assumption (W = 0.088, P < 0.001). Therefore, repeated measures covariance analysis (ANCOVA) after adjusting for degrees of freedom using the Greenhouse-Geisser method was used for statistical inference, and the adjusted results were considered valid. Preoperative NRS pain score, PSQI total score, SRS-22 total score, and SF-36 mental health dimension score were included as covariates in the model to control for baseline confounding.

[0077] 2.3 Changes in pain at rest (see...) Figure 2 After adjusting for covariates, the main effect of the depressive symptoms group was significant (F=7.144, P<0.05, partial η). 2 = 0.045), indicating that, after controlling for baseline differences, the overall postoperative resting pain level in the depressive symptom group was significantly higher than that in the non-depressive symptom group. However, the main time effect (F=1.597, P=0.150, partial η) was not significant. 2 = 0.011) and the interaction effect between group and time (F=1.26, P=0.276, partial η) 2 = 0.008) were not statistically significant. This suggests that the resting pain of all patients did not show a significant trend over time during the first 1-5 days after surgery, and the pain evolution trajectories of the two groups were similar; although the depressive symptom group maintained a high level of pain throughout, its recovery pattern was not substantially different from that of the non-depressive symptom group.

[0078] 2.4 Changes in pain during exercise (see...) Figure 3 Similarly, after adjusting for covariates, repeated measures ANCOVA showed that the main effect of time (F=2.176, P<0.05, partial η) was significant. 2 = 0.014) and the main effect of the depressive symptoms group were both significant (F=10.960, P<0.05, partial η) 2 All values ​​were statistically significant (F=0.068), while the interaction effect did not reach statistical significance (F=9.209, P=0.061, partial η). 2=0.013). This result indicates that motor pain in all patients showed a significant decreasing trend over time post-surgery; simultaneously, the overall intensity of motor pain was significantly higher in the depressive symptom group than in the non-depressive symptom group. Notably, the slopes of pain reduction over time were similar in both groups; that is, although the initial pain was higher in the depressive symptom group, its dynamic relief pattern was parallel to that of the non-depressive symptom group, showing no delayed recovery.

[0079] 2.5 The impact of preoperative depressive symptoms on CPSP

[0080] A total of 157 patients were included in the study, of whom 36 (22.93%) were diagnosed with CPSP 90 days postoperatively. First, univariate logistic regression analysis was used to screen for potential factors associated with the occurrence of CPSP. The results showed that preoperative depressive symptom grouping, gender, preoperative hematocrit (HCT), number of surgical fusion segments, resting pain score, movement pain score, SRS-22 total score, SF-36 score, and PSQI score were all significantly associated with CPSP (P < 0.10) (see Appendix Table 2). These variables were then included in a multivariate logistic regression model, and a forward stepwise method was used for variable selection. The final model showed (Table 3) that: female sex (OR = 6.36, 95% CI: 1.02-39.57, P < 0.05), increased number of fusion segments (OR = 1.31, 95% CI: 1.07-1.60, P < 0.01), and preoperative depressive symptoms (OR = 41.06, 95% CI: 8.68-194.19, P < 0.001) were independent risk factors for CPSP.

[0081] 2.6 CPSP Prediction Model

[0082] Based on the aforementioned experimental data and results, a personalized risk prediction model for CPSP was constructed and transformed into a visual nodal chart.

[0083] Based on the significant variables identified by the aforementioned multivariate logistic regression analysis, the final logistic regression model for CPSP risk prediction was constructed. Its complete mathematical expression and parameters are as follows:

[0084] Logit(P) =β0 + β1X1 + β2X2+ β3X3+ β4X4+ β5X5+ β6X6+ β7X7+ β8X8+ β9X9,

[0085] Variable meaning:

[0086] P represents the predicted probability of a patient developing chronic postoperative pain (CPSP).

[0087] X1 represents preoperative depressive state: a binary variable. When the total score of the Childhood Depression Scale (CDI) is ≥ 19, X1 = 1 (clinically significant depressive symptoms exist); otherwise, X1 = 0.

[0088] X2 represents gender: a binary variable. Female = 2; Male = 1.

[0089] X3 represents the number of surgical fusion segments: a continuous variable, input with the actual number of segments planned for the surgery.

[0090] X4 represents the preoperative resting pain score: a continuous variable, assessed using a numerical rating scale for resting status.

[0091] X5 represents the preoperative motor pain score: a continuous variable, which is a motor status score assessed using a numerical rating scale.

[0092] X6 represents the SRS-22 total score: a continuous variable, the Scoliosis Research Society-22 scale total score.

[0093] X7 represents the PSQI total score: a continuous variable, the total score of the Pittsburgh Sleep Quality Index.

[0094] X8 represents the SF-36 total score: a continuous variable, the total score of the Health Survey Summary.

[0095] X9 represents hematocrit: a continuous variable, entered as a percentage of preoperative laboratory test results.

[0096] Model regression coefficients (obtained by fitting data obtained in this application):

[0097] Β0 (intercept) = -7.31, β1 = 3.72, β2 = 1.85, β3 = 0.27, β4= 0.08, β5 = 0.19, β6= -0.02, β7 = -0.11, β8 = 0.02, β9 = 0.02.

[0098] Substituting the nine variables of the target patient into this equation, the Logit(P) value is calculated. Then, using the formula P = 1 / (1 + e^(-Logit(P))), the individualized CPSP risk probability P for that patient can be obtained. The nomogram generated based on this equation is shown below. Figure 4-1 As shown in a.

[0099] The area under the curve (AUC) for the receiver operating characteristic (ROC) of this model was 0.852 (95% CI: 0.773–0.932, P < 0.01) (see [link to relevant documentation]). Figure 4-1 (b) Using the Youden index, the optimal cutoff value was determined to be 0.213, corresponding to a sensitivity of 0.750 and a specificity of 0.868. The calibration curve shows a high degree of consistency between the predicted probability and the actual observed probability (slope close to 1, intercept close to 0) (see...). Figure 4-2 c); The Hosmer-Lemeshow test results (χ² = 13.132, P = 0.107) further indicate a good model fit. Decision Curve Analysis (DCA) shows ( Figure 4-2 d) When the probability of the clinical intervention threshold is between 10% and 70%, the application of this prediction model can achieve significant net clinical benefit, indicating that it has good clinical applicability.

[0100] To further evaluate the value of model gain, the joint indicator model was compared with a univariate model that only included preoperative depressive symptoms (see [link to model]). Figure 5 The results showed that the AUC of the combined indicator prediction model was 0.852 (95% CI: 0.773–0.932, P < 0.01), significantly better than the univariate model (AUC = 0.782, 95% CI: 0.683–0.882, P < 0.01), and the difference in AUC between the two models was statistically significant (P < 0.05). This result strongly confirms that integrating multidimensional predictive factors can significantly improve the accuracy of CPSP risk prediction.

[0101] Table 3. Univariate and multivariate analyses of factors influencing postoperative chronic pain (CPSP) P

[0102]

[0103] Note: NRS Numerical Rating Scale, SRS-22 Preoperative Scoliosis Study Society-22 Scale, PSQI Preoperative Pittsburgh Sleep Quality Index, SF-36 Preoperative Short Form Health Survey, Hct Hematocrit. Bold p-values ​​indicate statistical significance.

[0104] 2.7 The impact of dynamic changes in depressive symptoms on CPSP

[0105] Based on the dynamic changes in CDI scores preoperatively and 90 days postoperatively, patients were divided into three groups: persistent depressive symptoms group (preoperative CDI ≥ 19 and postoperative CDI ≥ 19, n = 4), depressive symptom relief group (preoperative CDI ≥ 19 but postoperative CDI < 19, n = 28), and no depressive symptoms group (preoperative CDI < 19 and postoperative CDI < 19, n = 125). The incidence of CPSP in the three groups was: no depressive symptoms group 10.4% (13 / 125), depressive symptom relief group 67.86% (19 / 28), and persistent depressive symptoms group 100% (4 / 4). Figure 6 As shown, the difference in the incidence of CPSP between groups was highly statistically significant (χ² = 86.32, P < 0.001). Further pairwise comparisons revealed that the risk of CPSP was significantly higher in the group with persistent depressive symptoms than in the group with remission of depressive symptoms or the group without depressive symptoms; while the incidence of CPSP in the group with remission of depressive symptoms was still significantly higher than in the group without depressive symptoms, it was significantly lower than in the group with persistent depressive symptoms. These results suggest that the dynamic evolution of perioperative depressive symptoms is gradedly associated with the risk of CPSP, and early improvement of depressive state may effectively reduce the long-term burden of postoperative chronic pain, providing direct evidence for the timing and goals of perioperative psychological intervention.

[0106] In summary, this prospective cohort design, for the first time, systematically reveals the dual impact of preoperative depressive symptoms on postoperative pain in adolescents with idiopathic scoliosis (AIS): on the one hand, while depression does not alter the incidence of acute postoperative pain (APSP), it significantly exacerbates pain intensity at rest and during movement, exhibiting a unique trajectory of "high baseline, parallel remission"; on the other hand, preoperative depression is the strongest independent predictor of chronic postoperative pain (CPSP), with an effect size far exceeding that of traditional clinical variables. Crucially, the dynamic evolution of perioperative depressive symptoms showed a significant gradient association with CPSP risk—the incidence of CPSP reached 100% in those with persistent depression, while the risk was significantly reduced in those with symptom remission, suggesting that the reversibility of psychological state provides a window of opportunity for intervention. These findings not only deepen our understanding of the comorbidity of pain and depression but also provide evidence-based support for incorporating standardized psychological assessment into routine perioperative management of AIS, promoting a shift in pain management models from "analgesia-centered" to "biopsychosocial integrated intervention."

[0107] This study demonstrates a robust and independent association between preoperative depressive symptoms and CPSP. In this cohort, the incidence of CPSP was 71.88% in patients with clinically significant depressive symptoms (CDI ≥ 19) preoperatively, significantly higher than the 10.4% in the group without depressive symptoms (P < 0.001). Multivariate logistic regression analysis showed that preoperative depressive symptoms, female sex, and the number of surgical fusion segments were risk factors for CPSP. Among these, the effect size of preoperative depressive symptoms was the most prominent (OR = 41.06), suggesting its central role in the pathogenesis of CPSP. This strong association may stem from the neurobiological mechanisms of pain-depression comorbidity. Depression can impair descending pain inhibition pathways in the central nervous system, weaken endogenous analgesia, and lead to stronger pain perception under the same noxious stimuli, making individuals more prone to developing chronic pain. This is highly consistent with the theory of "central sensitization"—that is, strong or repeated noxious input combined with psychological vulnerability leads to persistent changes in the processing of pain signals in the spinal cord and higher centers.

[0108] This application focuses on a single surgical procedure (posterior spinal fusion), effectively controlling surgical heterogeneity and improving the internal validity of causal inferences. Furthermore, the number of fusion segments (OR = 1.31 / segment) serves as an independent risk factor, supporting a dose-response relationship of "trauma load—central sensitization," suggesting that CPSP risk assessment should consider both physical trauma and psychological vulnerability. Based on this, this application constructs the first CPSP prediction model integrating psychological and physical factors, innovatively incorporating preoperative depression into the risk stratification system for AIS patients. The developed nomogram exhibits good discrimination (AUC = 0.852), calibration, and clinical net benefit, supporting rapid preoperative identification of high-risk individuals. This model provides crucial evidence for the precise prevention of CPSP.

[0109] More clinically significant, this study categorized patients into three groups based on longitudinal changes in CDI scores preoperatively and 90 days postoperatively: a persistent depressive symptom group, a depressive symptom relief group, and a non-depressive symptom group. Results showed a significant gradient in the incidence of CPSP—10.4% (13 / 125) in the non-depressive symptom group, 67.9% (19 / 28) in the depressive symptom relief group, and a staggering 100% (4 / 4) in the persistent depressive symptom group (P < 0.001). This dose-response relationship strongly suggests that improvement in perioperative depressive symptoms can significantly reduce the risk of CPSP. Notably, this "reversible" effect is rarely reported in adult surgical populations but is prominent in adolescent AIS patients, possibly due to their greater neuroplasticity and more sensitive window for psychological intervention. This finding provides direct evidence for early psychological intervention.

[0110] Although current guidelines from the Scoliosis Research Society (SRS) and Enhanced Recovery After Surgery (ERAS) emphasize multimodal analgesia, psychological assessment has not yet been included in routine recommendations. This application provides prospective evidence for incorporating standardized depression screening (such as CDI) into the perioperative management pathway for AIS.

[0111] The above description is illustrative only and not restrictive of the present invention. Those skilled in the art will understand that many modifications, variations or equivalents can be made without departing from the spirit and scope defined by the appended claims, and all such modifications, variations or equivalents will fall within the protection scope of the present invention.

Claims

1. A model for predicting the risk of chronic pain after surgery in adolescents with idiopathic scoliosis, characterized in that, The model is a prediction model built based on the logistic regression algorithm, and the function of the prediction model is: Logit(P) = β0 + β1X1 + β2X2+ β3X3+ β4X4+ β5X5+ β6X6+ β7X7+ β8X8+ β9X9, in, P represents the predicted probability of chronic postoperative pain in the target patient; X1 represents a preoperative state of depression; X2 represents the gender variable; X3 represents the number of surgical segment fusions; X4 represents the preoperative resting pain score; X5 represents the preoperative exercise pain score; X6 is an SRS-22 rating. X7 is the PSQI score; X8 is the rating for the SF-36; X9 represents the preoperative hematocrit value; β0 is the intercept, and β1, β2, β3, β4, β5, β6, β7, β8, and β9 are the regression coefficients of the model.

2. The model according to claim 1, characterized in that, The regression coefficients of the model are as follows: β0 = -7.31, β1 = 3.72, β2 = 1.85, β3 = 0.27, β4 = 0.08, β5 = 0.19, β6 = -0.02, β7 = -0.11, β8 = 0.02, β9 = 0.

02.

3. The model according to claim 1 or 2, characterized in that, The variable X1 is defined as follows: when the total score of the Child Depression Scale is ≥19, X1=1; otherwise, X1=0.

4. The model according to claim 1 or 2, characterized in that, The X2 variable is defined as follows: X2=2 when the patient is female, and X2=1 when the patient is male.

5. A storage medium having stored thereon the prediction model of any one of claims 1 to 4.

6. A risk assessment system for chronic pain after surgery in adolescent idiopathic scoliosis, characterized in that, include: Data input unit, used to acquire data of the target patient's X1 to X9 variables as defined in claim 1; A model processing unit, which is loaded with the prediction model according to any one of claims 1 to 4, is used to calculate the risk probability value P based on the data; The result output unit is used to output the risk probability value P and related risk assessment information.

7. The system according to claim 6, characterized in that, The result output unit is also configured to: Based on the preset threshold range where the risk probability value P is located, output a high-risk, medium-risk, or low-risk level determination; output personalized perioperative management suggestions corresponding to the risk level.

8. The system according to claim 7, characterized in that, The preset threshold range includes: When P < 0.2, it is considered low risk; When 0.2 ≤ P < 0.5, it is classified as medium risk; When P ≥ 0.5, it is considered high risk.

9. An electronic device comprising a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the functions of the system according to any one of claims 6 to 9.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it is used to implement the functions of the system according to any one of claims 6 to 9.

Images