Distance Traveled and Disparities in Patients Undergoing Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy.

BACKGROUND: The impact of distance traveled on cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) outcomes needs further investigation. METHODS: This retrospective study reviewed a prospectively managed single-center CRS/HIPEC 1992-2022 database. Zip codes were used to calculate distance traveled and to obtain data on income and education via census data. Patients were separated into three groups based on distance traveled in miles (local: ≤50 miles, regional: 51-99 miles, distant: ≥100 miles). Descriptive statistics, Kaplan-Meier method, and Cox regression were performed. RESULTS: The 1614 patients in the study traveled a median distance of 109.5 miles (interquartile range [IQR], 53.36-202.29 miles), with 23% traveling locally, 23.9% traveling regionally, and 53% traveling distantly. Those traveling distantly or regionally tended to be more white (distant: 87.8%, regional: 87.2%, local: 83.2%), affluent (distant: $61,944, regional: $65,014, local: $54,390), educated (% without high school diploma: distant: 10.6%, regional: 11.5%, local: 13.0%), less often uninsured (distant: 2.3%, regional: 4.6%, local: 5.2%) or with Medicaid (distant: 3.3%, regional: 1.3%, local: 9.7%). They more often had higher Peritoneal Carcinomatosis Index (PCI) scores (distant: 15.4, regional: 15.8, local: 12.7) and R2 resections (distant: 50.3%, regional: 52.2%, local: 40.5%). Median survival did not differ between the groups, and distance traveled was not a predictor of survival. CONCLUSION: More than 50% of the patients traveled farther than 100 miles for treatment. Although regionalization of CRS/HIPEC may be appropriate given the lack of survival difference based on distance traveled, those who traveled further had fewer health care disparities but higher PCI scores and more R2 resections, which raises concerns about access to care for the underserved, time to treatment, and surgical quality.

Defining Volume Targets for Regionalization of Pancreaticoduodenectomy, Esophagectomy, and Major Lung Resection Centers to Improve Oncologic and Surgical Outcomes.

BACKGROUND: Performance of complex cancer surgeries at high-volume (HV) centers has been shown to reduce operative mortality. However, the case volume threshold that should be used to define HV centers is unknown. In this study, we determined thresholds to define HV pancreaticoduodenectomy, esophagectomy, and major lung resection centers based on clinical parameters. Then, we assessed the association of hospital volume with oncologic outcomes and overall survival. METHODS: We identified adult NCDB patients undergoing pancreaticoduodenectomy, esophagectomy, and major lung resections between 2004 and 2015. Multivariable models with restricted cubic splines were built to predict 5-year overall survival for each surgery group according to average yearly case volume, adjusting for demographic and clinicopathologic factors. The change point procedure was then used to identify volume cut-points for each surgery type. RESULTS: We identified the following thresholds to define HV status: 25 cases/year for pancreaticoduodenectomy; 18 cases/year for esophagectomy; and 54 cases/year for major lung resections. For all surgery types, treatment at a HV center was associated with an increased likelihood of R0 resection and adequate lymph node evaluation. HV centers had significantly decreased 30- and 90-day, postoperative mortality after adjusting for age, sex, race, comorbidities, histology, and stage. An overall survival benefit also was observed for patients undergoing resections at HV centers. CONCLUSIONS: Using novel methodology, our study identified volume thresholds for HV pancreaticoduodenectomy, esophagectomy, and major lung resection centers that were associated with improved oncologic outcomes and overall survival. These definitions of HV centers should be considered when evaluating regionalization of complex cancer care.

Impact of cerebrovascular accidents on lung transplant survival.

BACKGROUND: Cerebrovascular accidents (CVA) are a source of postoperative morbidity. Existing data on CVA after lung transplantation (LT) are limited. We aimed to evaluate the impact of CVA on LT survival. METHODS: A retrospective analysis of LT recipients at the University of Texas Southwestern Medical Center was performed. Data was obtained from the institutional thoracic transplant database between January 2012 and December 2018, which consisted of 476 patients. Patients were stratified by the presence of a postoperative CVA. Univariate comparisons of baseline characteristics, operative variables, and postoperative outcomes between the cohorts were performed. Survival was analyzed by Kaplan-Meier method. Aalen's additive regression model was utilized to assess mortality hazard over time. RESULTS: The incidence of CVA was 4.2% (20/476). Lung allocation score was higher in the CVA cohort (46.2 [41.7, 57.3] vs. 41.5 [35.8, 52.2], p = 0.04). There were no significant differences in operative variables. CVA patients had longer initial intensive care unit (ICU) stays (316 h [251, 557] vs. 124 [85, 218], p < 0.001) and longer length of stay (22 days [17, 53] vs. 15 [11, 26], p = 0.007). CVA patients required more ICU readmissions (35% vs. 15%, p = 0.02) and had a lower rates of home discharge (35% vs. 71%, p < 0.001). Thirty-day mortality was higher in the CVA cohort (20% vs. 1.3%, p < 0.001). Overall survival was lower in the CVA cohort (log rank p = 0.044). CONCLUSIONS: Postoperative CVA following LT was associated with longer ICU stays, more ICU readmissions, longer length of stay, and fewer home discharges. Thirty day and long-term mortality were significantly higher in the CVA group.

Machine Learning Improves Prediction Over Logistic Regression on Resected Colon Cancer Patients.

INTRODUCTION: Despite advances, readmission and mortality rates for surgical patients with colon cancer remain high. Prediction models using regression techniques allows for risk stratification to aid periprocedural care. Technological advances have enabled large data to be analyzed using machine learning (ML) algorithms. A national database of colon cancer patients was selected to determine whether ML methods better predict outcomes following surgery compared to conventional methods. METHODS: Surgical colon cancer patients were identified using the 2013 National Cancer Database (NCDB). The negative outcome was defined as a composite of 30-d unplanned readmission and 30- and 90-d mortality. ML models, including Random Forest and XGBoost, were built and compared with conventional logistic regression. For the accounting of unbalanced outcomes, a synthetic minority oversampling technique (SMOTE) was implemented and applied using XGBoost. RESULTS: Analysis included 528,060 patients. The negative outcome occurred in 11.6% of patients. Model building utilized 30 variables. The primary metric for model comparison was area under the curve (AUC). In comparison to logistic regression (AUC 0.730, 95% CI: 0.725-0.735), AUC's for ML algorithms ranged between 0.748 and 0.757, with the Random Forest model (AUC 0.757, 95% CI: 0.752-0.762) outperforming XGBoost (AUC 0.756, 95% CI: 0.751-0.761) and XGBoost using SMOTE data (AUC 0.748, 95% CI: 0.743-0.753). CONCLUSIONS: We show that a large registry of surgical colon cancer patients can be utilized to build ML models to improve outcome prediction with differential discriminative ability. These results reveal the potential of these methods to enhance risk prediction, leading to improved strategies to mitigate those risks.

Frame-wise detection of surgeon stress levels during laparoscopic training using kinematic data.

PURPOSE: Excessive stress experienced by the surgeon can have a negative effect on the surgeon's technical skills. The goal of this study is to evaluate and validate a deep learning framework for real-time detection of stressed surgical movements using kinematic data. METHODS: 30 medical students were recruited as the subjects to perform a modified peg transfer task and were randomized into two groups, a control group (n=15) and a stressed group (n=15) that completed the task under deteriorating, simulated stressful conditions. To classify stressed movements, we first developed an attention-based Long-Short-Term-Memory recurrent neural network (LSTM) trained to classify normal/stressed trials and obtain the contribution of each data frame to the stress level classification. Next, we extracted the important frames from each trial and used another LSTM network to implement the frame-wise classification of normal and stressed movements. RESULTS: The classification between normal and stressed trials using attention-based LSTM model reached an overall accuracy of 75.86% under Leave-One-User-Out (LOUO) cross-validation. The second LSTM classifier was able to distinguish between the typical normal and stressed movement with an accuracy of 74.96% with an 8-second observation under LOUO. Finally, the normal and stressed movements in stressed trials could be classified with the accuracy of 68.18% with a 16-second observation under LOUO. CONCLUSION: In this study, we extracted the movements which are more likely to be affected by stress and validated the feasibility of using LSTM and kinematic data for frame-wise detection of stress level during laparoscopic training. The proposed classifier could be potentially be integrated with robot-assisted surgery platforms for stress management purposes.

Determining the Significant Kinematic Features for Characterizing Stress during Surgical Tasks Using Spatial Attention.

It has been shown that intraoperative stress can have a negative effect on surgeon surgical skills during laparoscopic procedures. For novice surgeons, stressful conditions can lead to significantly higher velocity, acceleration, and jerk of the surgical instrument tips, resulting in faster but less smooth movements. However, it is still not clear which of these kinematic features (velocity, acceleration, or jerk) is the best marker for identifying the normal and stressed conditions. Therefore, in order to find the most significant kinematic feature that is affected by intraoperative stress, we implemented a spatial attention-based Long-Short-Term-Memory (LSTM) classifier. In a prior IRB approved experiment, we collected data from medical students performing an extended peg transfer task who were randomized into a control group and a group performing the task under external psychological stresses. In our prior work, we obtained "representative" normal or stressed movements from this dataset using kinematic data as the input. In this study, a spatial attention mechanism is used to describe the contribution of each kinematic feature to the classification of normal/stressed movements. We tested our classifier under Leave-One-User-Out (LOUO) cross-validation, and the classifier reached an overall accuracy of 77.11% for classifying "representative" normal and stressed movements using kinematic features as the input. More importantly, we also studied the spatial attention extracted from the proposed classifier. Velocity and acceleration on both sides had significantly higher attention for classifying a normal movement (p <= 0.0001); Velocity (p <= 0.015) and jerk (p <= 0.001) on non-dominant hand had significant higher attention for classifying a stressed movement, and it is worthy noting that the attention of jerk on non-dominant hand side had the largest increment when moving from describing normal movements to stressed movements (p = 0.0000). In general, we found that the jerk on non-dominant hand side can be used for characterizing the stressed movements for novice surgeons more effectively.

Identifying Kinematic Markers Associated with Intraoperative Stress during Surgical Training Tasks.

Increased levels of stress can impair surgeon performance and patient safety during surgery. The aim of this study is to investigate the effect of short term stressors on laparoscopic performance through analysis of kinematic data. Thirty subjects were randomly assigned into two groups in this IRB-approved study. The control group was required to finish an extended-duration peg transfer task (6 minutes) using the FLS trainer while listening to normal simulated vital signs and while being observed by a silent moderator. The stressed group finished the same task but listened to a period of progressively deteriorating simulated patient vitals, as well as critical verbal feedback from the moderator, which culminated in 30 seconds of cardiac arrest and expiration of the simulated patient. For all subjects, video and position data using electromagnetic trackers mounted on the handles of the laparoscopic instruments were recorded. A statistical analysis comparing time-series velocity, acceleration, and jerk data, as well as path length and economy of volume was conducted. Clinical stressors lead to significantly higher velocity, acceleration, jerk, and path length as well as lower economy of volume. An objective evaluation score using a modified OSATS technique was also significantly worse for the stressed group than the control group. This study shows the potential feasibility and advantages of using the time-series kinematic data to identify the stressful conditions during laparoscopic surgery in near-real-time. This data could be useful in the design of future robot-assisted algorithms to reduce the unwanted effects of stress on surgical performance.