This prospective, single-center observational study will investigate chronic postsurgical pain after cardiac implantable electronic device (CIED) implantation and develop an artificial intelligence (AI)-based risk prediction model. The study will be conducted at the Istanbul University-Cerrahpaşa Cardiology Institute, Electrophysiology Outpatient Clinic, and will include 180 consecutive patients who received their first CIED within the past 12 months.
Study Design and Population Eligible participants are adults (≥ 18 years) undergoing first-time CIED implantation (pacemaker, ICD, or CRT device). Exclusion criteria include preexisting neuropathic disorders, neurodegenerative or autoimmune diseases, malignancy, trauma, infection, or previous generator replacement. All participants will provide written informed consent.
Data Collection and Measurements Patients will be evaluated at 3, 6, and 12 months post-implantation. Subjective pain assessment: Visual Analog Scale (VAS) and DN-4 questionnaire (neuropathic pain screening).
Objective pain threshold: Pressure Pain Threshold (PPT) measured with a digital pressure algometer on both the implantation and contralateral sides.
Quality of life: SF-12 (Physical and Mental Component Scores). Risk factors: Demographic, preoperative, intraoperative, and postoperative variables (age, sex, BMI, comorbidities, anesthesia type, surgical time, device type, lead number, battery position, acute postoperative pain, rehabilitation, analgesic use).
Data Management and Quality Control All patient data will be anonymized and stored on password-protected institutional servers. Data accuracy and completeness will be checked regularly. Source data verification will be performed by comparing electronic records and case report forms. Range and consistency checks will be applied to identify out-of-range or inconsistent entries.
Sample Size and Statistical Analysis Based on previous literature reporting an estimated CPSP prevalence of 24% after CIED implantation, a sample size of 163 was calculated (α = 0.05, power = 0.95, OR = 2.0). Allowing for a 10% dropout rate, the target enrollment is 180 participants.
Descriptive statistics will summarize demographic and clinical variables. Between-group comparisons will use independent-sample t-tests or Mann-Whitney U tests for continuous variables and χ² or Fisher's exact tests for categorical data. Logistic regression will identify independent predictors of CPSP (VAS \> 3 or DN-4 ≥ 4). Correlations between continuous variables will be analyzed with Pearson or Spearman coefficients. A two-tailed p \< 0.05 will denote statistical significance.
Secondary Analysis Description
Multivariable logistic regression analysis will be performed to evaluate the association between preoperative, intraoperative, and postoperative risk factors and the presence of chronic postsurgical pain (CPSP) at 12 months after cardiac implantable electronic device implantation.
Candidate variables will include age, sex, body mass index (BMI), preexisting chronic pain, psychological status, comorbidities (ischemic heart disease, heart failure, diabetes mellitus, hypertension, chronic kidney disease, thyroid disease), type of anesthesia, surgical duration, device type (pacemaker, ICD, CRT), pocket side and location, number of leads, postoperative pain management and analgesic use, and rehabilitation.
Results will be reported as adjusted odds ratios (ORs) with 95% confidence intervals (CIs).
Artificial Intelligence Model Development Collected data will be split into training and validation sets. Supervised machine-learning algorithms (logistic regression, random forest, XGBoost, artificial neural networks) will be trained to predict CPSP occurrence at 12 months. Model performance will be evaluated using ROC-AUC, accuracy, sensitivity, and specificity metrics. Feature importance analysis will identify the most influential predictors.
The best-performing model will be incorporated into a prototype Clinical Decision Support System (CDSS), accessible via a secure web interface and mobile application. This tool will allow clinicians to input patient parameters and obtain individualized CPSP risk estimates to guide preventive analgesic strategies.
