RESUMO
BACKGROUND: Clinically-relevant postoperative pancreatic fistula (CR-POPF) following pancreaticoduodenectomy (PD) is a major postoperative complication and the primary determinant of surgical outcomes. However, the majority of current risk calculators utilize intraoperative and postoperative variables, limiting their utility in the preoperative setting. Therefore, we aimed to develop a user-friendly risk calculator to predict CR-POPF following PD using state-of-the-art machine learning (ML) algorithms and only preoperatively known variables. METHODS: Adult patients undergoing elective PD for non-metastatic pancreatic cancer were identified from the ACS-NSQIP targeted pancreatectomy dataset (2014-2019). The primary endpoint was development of CR-POPF (grade B or C). Secondary endpoints included discharge to facility, 30-day mortality, and a composite of overall and significant complications. Four models (logistic regression, neural network, random forest, and XGBoost) were trained, validated and a user-friendly risk calculator was then developed. RESULTS: Of the 8666 patients who underwent elective PD, 13% (n = 1160) developed CR-POPF. XGBoost was the best performing model (AUC = 0.72), and the top five preoperative variables associated with CR-POPF were non-adenocarcinoma histology, lack of neoadjuvant chemotherapy, pancreatic duct size less than 3 mm, higher BMI, and higher preoperative serum creatinine. Model performance for 30-day mortality, discharge to a facility, and overall and significant complications ranged from AUC 0.62-0.78. CONCLUSIONS: In this study, we developed and validated an ML model using only preoperatively known variables to predict CR-POPF following PD. The risk calculator can be used in the preoperative setting to inform clinical decision-making and patient counseling.
RESUMO
BACKGROUND: The aim of this study was to investigate the associations between individual surgeon's intraoperative nerve monitoring (IONM) practice and factors associated with vocal cord (VC) dysfunction in patients with thyroid cancer undergoing thyroidectomy. METHODS: Using Collaborative Endocrine Surgery Quality Improvement Program (CESQIP) 2014-21 data, multivariable logistic regression analyses investigated variables associated with short- and long-term VC-dysfunction, associations of routine use of IONM with postoperative outcomes, and patient characteristics associated with IONM use. RESULTS: Among 5,446 patients (76.7% female, mean age 49 years), 68.5% had surgery by surgeons using IONM in ≥ 90% of cases (63% of surgeons, n = 73). Post-operative VC-dysfunction was diagnosed by laryngoscopy in 3.0% of patients in the short-term and 2.7% in the long-term. When surgeons routinely used IONM, the incidence of VC-dysfunction was 2.4% in the short-term and 2.2% in the long-term, compared to 4.4% and 3.7%, respectively, when surgeons did not routinely use IONM (p < 0.01). After adjustment, routine use of IONM was independently associated with reduced risk of short- (OR 0.48, p < 0.01) and long-term (OR 0.52, p < 0.01) VC-dysfunction, a lower risk of postoperative hypoparathyroidism in the short- (OR 0.67, p < 0.01) and long-term (OR 0.54, p < 0.01), and higher likelihood of same-day discharge (OR 2.03, p < 0.01). Extrathyroidal tumor extension and N1-stage were factors associated with postoperative VC-dysfunction in the short- (OR 3.12, p < 0.01; OR 1.92, p = 0.01, respectively) and long-term (OR 3.11, p < 0.01; OR 2.32, p < 0.01, respectively). CONCLUSION: Routine use of IONM was independently associated with a lower risk of endocrine surgery-specific complications and greater likelihood of same-day discharge.