Scholarlink

<p>Various factors impact postoperative outcomes among patients who undergo elective abdominal surgery. Malnutrition is a crucial component that has a considerable impact on postoperative complications. Nutrition risk assessment models were developed using the highly advanced Least Absolute Shrinkage Selection Operator (LASSO) regression method. The study aimed to evaluate the predictive accuracy of these models for postoperative complications. This prospective cohort study was carried out in a tertiary care hospital. The study included all patients aged 18&#xa0;years and above who were consecutively admitted for elective abdominal surgery. The primary outcome assessed was the occurrence of postoperative complications till discharge using Clavien-Dindo’s classification. The efficiency of the two LASSO regression-based models was evaluated. The first model included socio-demographic, anthropometric, clinical variables, nutrition risk assessment tools, routine and specific biomarkers. The selected variables of the first model 1 were Nutritional Risk Screening 2002 (NRS 2002), Subjective Global Assessment (SGA), type of surgery, extent of surgery, serum albumin, total proteins, pre-albumin, transferrin, C—reactive protein (CRP) and insulin-like growth factor 1 (IGF1). The second model incorporated all the predictors except specific biomarkers. NRS 2002, SGA, pathology, type of surgery, the extent of surgery, serum albumin, and serum total proteins were the variables chosen for the second model by LASSO regression. The model-based nomogram scores were applied to the samples, and Receiver Operating Characteristic Curve (ROC) was plotted with an Area Under the Curve (AUC) of 0.9437 at 87.30% sensitivity and 86.08% specificity at a cut-off score of 20.3 for model 1. Similarly, the AUC was 0.9294 at 88.89% sensitivity and 82.47% specificity at a cut-off score of 10.5 for model 2. Clinicians can use LASSO regression-based models to predict postoperative complications in patients who undergo elective abdominal surgeries. These models may enable preoperative corrective efforts, which can aid in the reduction of postoperative complications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of Nutritional Risk Assessment Models Using Least Absolute Shrinkage Selection Operator for Predicting Postoperative Complications Among Patients Undergoing Elective Abdominal Surgery

  • 期刊论文

  • 发表日期：

    2025年9月17日

查看全文
  • Raghuraman Hemamalini,
  • Balakrishnan Gurushankari,
  • Mini Rajesh Rajalekshmy,
  • T P Elamurugan,
  • Gomathi Shankar,
  • Nivedita Nanda,
  • Thulasingam Mahalakshmy,
  • Yuvaraj Krishnamoorthy,
  • Vikram Kate

摘要

Various factors impact postoperative outcomes among patients who undergo elective abdominal surgery. Malnutrition is a crucial component that has a considerable impact on postoperative complications. Nutrition risk assessment models were developed using the highly advanced Least Absolute Shrinkage Selection Operator (LASSO) regression method. The study aimed to evaluate the predictive accuracy of these models for postoperative complications. This prospective cohort study was carried out in a tertiary care hospital. The study included all patients aged 18 years and above who were consecutively admitted for elective abdominal surgery. The primary outcome assessed was the occurrence of postoperative complications till discharge using Clavien-Dindo’s classification. The efficiency of the two LASSO regression-based models was evaluated. The first model included socio-demographic, anthropometric, clinical variables, nutrition risk assessment tools, routine and specific biomarkers. The selected variables of the first model 1 were Nutritional Risk Screening 2002 (NRS 2002), Subjective Global Assessment (SGA), type of surgery, extent of surgery, serum albumin, total proteins, pre-albumin, transferrin, C—reactive protein (CRP) and insulin-like growth factor 1 (IGF1). The second model incorporated all the predictors except specific biomarkers. NRS 2002, SGA, pathology, type of surgery, the extent of surgery, serum albumin, and serum total proteins were the variables chosen for the second model by LASSO regression. The model-based nomogram scores were applied to the samples, and Receiver Operating Characteristic Curve (ROC) was plotted with an Area Under the Curve (AUC) of 0.9437 at 87.30% sensitivity and 86.08% specificity at a cut-off score of 20.3 for model 1. Similarly, the AUC was 0.9294 at 88.89% sensitivity and 82.47% specificity at a cut-off score of 10.5 for model 2. Clinicians can use LASSO regression-based models to predict postoperative complications in patients who undergo elective abdominal surgeries. These models may enable preoperative corrective efforts, which can aid in the reduction of postoperative complications.