Search

Background: Surgical video recording provides the opportunity to acquire intraoperative data that can subsequently be used for a variety of quality improvement, research, and educational applications. Various recording devices are available for standard operating room camera systems. Some allow for collateral data acquisition including activities of the OR staff, kinematic measurements (motion of surgical instruments), and recording of the endoscopic video streams. Additional analysis through computer vision (CV), which allows software to understand and perform predictive tasks on images, can allow for automatic phase segmentation, instrument tracking, and derivative performance-geared metrics. With this survey, we summarize available surgical video acquisition technologies and associated performance analysis platforms. Methods: In an effort promoted by the SAGES Artificial Intelligence Task Force, we surveyed the available video recording technology companies. Of thirteen companies approached, nine were interviewed, each over an hour-long video conference. A standard set of 17 questions was administered. Questions spanned from data acquisition capacity, quality, and synchronization of video with other data, availability of analytic tools, privacy, and access. Results: Most platforms (89%) store video in full-HD (1080p) resolution at a frame rate of 30 fps. Most (67%) of available platforms store data in a Cloud-based databank as opposed to institutional hard drives. CV powered analysis is featured in some platforms: phase segmentation in 44% platforms, out of body blurring or tool tracking in 33%, and suture time in 11%. Kinematic data are provided by 22% and perfusion imaging in one device. Conclusion: Video acquisition platforms on the market allow for in depth performance analysis through manual and automated review. Most of these devices will be integrated in upcoming robotic surgical platforms. Platform analytic supplementation, including CV, may allow for more refined performance analysis to surgeons and trainees. Most current AI features are related to phase segmentation, instrument tracking, and video blurring.

Background: Many surgical adverse events, such as bile duct injuries during laparoscopic cholecystectomy (LC), occur due to errors in visual perception and judgment. Artificial intelligence (AI) can potentially improve the quality and safety of surgery, such as through real-time intraoperative decision support. GoNoGoNet is a novel AI model capable of identifying safe (“Go”) and dangerous (“No-Go”) zones of dissection on surgical videos of LC. Yet, it is unknown how GoNoGoNet performs in comparison to expert surgeons. This study aims to evaluate the GoNoGoNet’s ability to identify Go and No-Go zones compared to an external panel of expert surgeons. Methods: A panel of high-volume surgeons from the SAGES Safe Cholecystectomy Task Force was recruited to draw free-hand annotations on frames of prospectively collected videos of LC to identify the Go and No-Go zones. Expert consensus on the location of Go and No-Go zones was established using Visual Concordance Test pixel agreement. Identification of Go and No-Go zones by GoNoGoNet was compared to expert-derived consensus using mean F1 Dice Score, and pixel accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Results: A total of 47 frames from 25 LC videos, procured from 3 countries and 9 surgeons, were annotated simultaneously by an expert panel of 6 surgeons and GoNoGoNet. Mean (± standard deviation) F1 Dice score were 0.58 (0.22) and 0.80 (0.12) for Go and No-Go zones, respectively. Mean (± standard deviation) accuracy, sensitivity, specificity, PPV and NPV for the Go zones were 0.92 (0.05), 0.52 (0.24), 0.97 (0.03), 0.70 (0.21), and 0.94 (0.04) respectively. For No-Go zones, these metrics were 0.92 (0.05), 0.80 (0.17), 0.95 (0.04), 0.84 (0.13) and 0.95 (0.05), respectively. Conclusions: AI can be used to identify safe and dangerous zones of dissection within the surgical field, with high specificity/PPV for Go zones and high sensitivity/NPV for No-Go zones. Overall, model prediction was better for No-Go zones compared to Go zones. This technology may eventually be used to provide real-time guidance and minimize the risk of adverse events.

Shorter time to emergency department aortic occlusion did not improve survival in noncompressible torso hemorrhage. “Scoop and control” with hemostatic resuscitation closer to the point of injury needs special consideration.Introduction:Time is an essential element in outcomes of trauma patients. The relationship of time to treatment in management of noncompressible torso hemorrhage (NCTH) with resuscitative endovascular balloon occlusion of the aorta (REBOA) or resuscitative thoracotomy (RT) has not been previously described. We hypothesized that shorter times to intervention would reduce mortality. Methods:A review of the Aortic Occlusion for Resuscitation in Trauma and Acute Care Surgery registry from 2013 to 2022 was performed to identify patients who underwent emergency department aortic occlusion (AO). Multivariate logistic regression was used to examine the impact of time to treatment on mortality. Results:A total of 1,853 patients (1,245 [67%] RT, 608 [33%] REBOA) were included. Most patients were male (82%) with a median age of 34 years (interquartile range, 30). Median time from injury to admission and admission to successful AO were 31 versus 11 minutes, respectively. Patients who died had shorter median times from injury to successful AO (44 vs. 72 minutes, P< 0.001) and admission to successful AO (10 vs. 22 minutes, P< 0.001). Multivariate logistic regression demonstrated that receiving RT was the strongest predictor of mortality (odds ratio [OR], 6.6; 95% confidence interval [CI], 4.4–9.9; P< 0.001). Time from injury to admission and admission to successful AO were not significant. This finding was consistent in subgroup analysis of RT-only and REBOA-only populations. Conclusions:Despite expedited interventions, time to aortic occlusion did not significantly impact mortality. This may suggest that rapid in-hospital intervention was often insufficient to compensate for severe exsanguination and hypovolemia that had already occurred before emergency department presentation. Selective prehospital advanced resuscitative care closer to the point of injury with “scoop and control” efforts including hemostatic resuscitation warrants special consideration.

Patients receiving immune checkpoint inhibitors at the end of life had poor performance status and low palliative care and hospice enrollment

Background: Genomic classifiers (GC) have been shown to improve risk stratification post prostatectomy. However, their clinical benefit has not been prospectively demonstrated. We sought to determine the impact of GC testing on postoperative management in men with prostate cancer post prostatectomy. Methods: Two prospective registries of prostate cancer patients treated between 2014 and 2019 were included. All men underwent Decipher tumor testing for adverse features post prostatectomy (Decipher Biosciences, San Diego, CA). The clinical utility cohort, which measured the change in treatment decision-making, captured pre- and postgenomic treatment recommendations from urologists across diverse practice settings (n= 3455). The clinical benefit cohort, which examined the difference in outcome, was from a single academic institution whose tumor board predefined “best practices” based on GC results (n= 135). Results: In the clinical utility cohort, providers’ recommendations pregenomic testing were primarily observation (69%). GC testing changed recommendations for 39% of patients, translating to a number needed to test of 3 to change one treatment decision. In the clinical benefit cohort, 61% of patients had genomic high-risk tumors; those who received the recommended adjuvant radiation therapy (ART) had 2-year PSA recurrence of 3 vs. 25% for those who did not (HR 0.1 [95% CI 0.0–0.6], p= 0.013). For the genomic low/intermediate-risk patients, 93% followed recommendations for observation, with similar 2-year PSA recurrence rates compared with those who received ART (p= 0.93). Conclusions: The use of GC substantially altered treatment decision-making, with a number needed to test of only 3. Implementing best practices to routinely recommend ART for genomic-high patients led to larger than expected improvements in early biochemical endpoints, without jeopardizing outcomes for genomic-low/intermediate-risk patients.

Supplemental digital content is available in the text.

The objective of the study was to characterize house staff time to response and intervention when notified of a patient care issue by pager vs. smartphone. We hypothesized that smartphones would reduce house staff time to response and intervention.

We sought to identify a simple bedside method to predict successful extubation outcomes that might be used during rounds. We hypothesized that a direct 2-minute unassisted breathing evaluation (DTUBE) could replace a longer spontaneous breathing trial (SBT). Data were pro-spectively collected on all patients endotracheally intubated for >48 hours nearing extubation in a tertiary center's mixed trauma/surgical intensive care unit from August 2012 to August 2013. The SBT was performed for at least 30 minutes at 40 per cent FiO2, PEEP 5, and PS 8. DTUBE was performed by physically disconnecting the intubated patient from the ventilator circuit for a 2-minute period of direct observation on room air. Successful extubation was defined freedom from ventilator for greater than 72 hours. Both SBTand DTUBE were performed 128 times, resulting in 90 extubations. The DTUBE correctly predicted success in 75/79 (94.9%) extubations versus 82/89 (92.1%) via SBT. No adverse effects were directly attributed to the DTUBE. The DTUBE is a rapid method of evaluating patients for extubation with prediction accuracy similar to the SBT.

Pleural effusions occur frequently in mechanically ventilated patients, but no consensus exists regarding the clinical benefit of effusion drainage. We sought to determine the impact of thoracentesis on gas exchange in patients with differing severities of acute lung injury (ALI). A retrospective analysis was conducted on therapeutic thoracenteses performed on intubated patients in an adult surgical intensive care unit of a tertiary center. Effusions judged by ultrasound to be 400 mL or larger were drained. Subjects were divided into groups based on their initial P:F ratios: normal >300, ALI 200 to 300, and acute respiratory distress syndrome (ARDS) <200. Baseline characteristics, physiologic variables, arterial blood gases, and ventilator settings before and after the intervention were analyzed. The primary end point was the change in measures of oxygenation. Significant improvements in P:F ratios (mean ± SD) were seen only in patients with ARDS (50.4 ± 38.5, P= 0.001) and ALI (90.6 ± 161.7, P= 0.022). Statistically significant improvement was observed in the pO2(31.1, P= 0.005) and O2saturation (4.1, P< 0.001) of the ARDS group. The volume of effusion removed did not correlate with changes in individual patient's oxygenation. These data support the role of therapeutic thoracentesis for intubated patients with abnormal P:F ratios.

Massive transfusion protocol (MTP) is used to resuscitate patients in hemorrhagic shock. Our goal was to review MTP use in the elderly. All trauma patients who required activation of MTP at an urban Level I trauma center from January 1, 2011 to December 31, 2013 were reviewed retrospectively. Elderly was defined as age = 60 years. Sixty-six patients had MTP activated: 52 non-elderly (NE) and 14 elderly (E). There were no statistically significant differences between the two cohorts for gender, injury severity score, head abbreviated injury scale, emergency department Glasgow Coma Scale, initial hematocrit, intensive care unit length of stay, or hospital length of stay. Mean age for NE was 35 years and 73 years for E (P< 0.01). Less than half (43%) of E patients with activation of MTP received 10 or more units of blood products compared with 69 per cent of the NE (P= 0.07). Mortality rates were similar in the NE and the E (53% vs50%, P= 0.80). After multivariate analysis with Glasgow Coma Scale, injury severity score, and blunt versuspenetrating trauma, elderly age was not a predictor of mortality after MTP (P= 0.35). When MTP is activated, survival to discharge in elderly trauma patients is comparable to younger patients.

Pain associated with rib fractures impairs respiratory function and increases pulmonary morbidity. The purpose of this study was to determine how epidural catheters alter mortality and complications in trauma patients. We performed a retrospective study involving adult blunt trauma patients with moderate-to-severe injuries from January 1, 2004 to December 31, 2013. During the 10-year period, 526 patients met the inclusion criteria; 43/526 (8%) patients had a catheter placed. Mean age of patients with epidural catheter (CATH) was higher compared with patients without epidural catheter (NOCATH) (54 vs 48 years, P = 0.021), Injury Severity Score was similar (26 CATH vs 27 NOCATH, P = 0.84), and CATH had higher mean rib fractures (7.4 vs 4.1, P < 0.001). Mortality was lower in CATH (0% vs 13%, P = 0.006). Deep vein thrombosis (DVT) rate was higher in CATH (12% vs. 5%, P = 0.036). After regression analysis, we found catheter placement to be a predictor for DVT (adjusted odds ratios 2.80, P = 0.036). Our center noted increased use of epidural catheters in patients who present with moderate-to-severe injuries. Patients with catheters were older and had a mean of 7.4 ribs fractured. The epidural cohort had longer hospital LOS and decreased mortality. In contrast to other studies, DVT rates were increased in patients who received epidural catheters.

Early trauma-induced coagulopathy (ETIC) is abnormal coagulation detected on presentation, but a clear description is lacking. We used thromboelastography (TEG) to characterize ETIC. Data were prospectively collected on high-acuity trauma activations at an urban Level I trauma center between July 2012 and May 2013. Patients with admission TEG before any blood transfusion were stratified by Injury Severity Score (ISS): mild (less than 16), moderate (16 to 24), severe (25 or greater). TEG parameters were compared between groups. ETIC was defined as any abnormality detected on TEG. Fifty-two patients were included; mean age was 49 years and mean time to the emergency department was 26 minutes. Mean ISS for the cohort was 17 with 28 patients in mild, eight in moderate, and 16 in severe. Glasgow Coma Score was lower and head Abbreviated Injury Scale was higher in severe (P< 0.001). Forty-three (83%) patients had an abnormal TEG. Shortened reaction (R) time was noted in 42 patients. There were no differences in any TEG parameters between the injury severity groups. Hyperfibrinolysis was detected in four (8%) patients. ETIC was present in over 80 per cent of high-acuity trauma activations irrespective of injury severity and characterized primarily by shortened R time, indicating ETIC is initially described by a hypercoagulable state as a result of thrombin generation.

The risk of venous thromboembolism (VTE) for patients taking an antiplatelet agent is largely unknown. This study aimed to investigate the association between antiplatelet agent use before admission with the risk of in-hospital VTE in surgical intensive care unit (ICU) patients. A retrospective review of all patients admitted to the surgical ICU at a Level I trauma center over 30 months was performed. Patients who underwent diagnostic imaging for VTE were selected. Patients were divided based on whether or not antiplatelet agents were used before admission (APTA vsNAPTA). The primary outcome was VTE occurrence. A forward logistic regression model was used to identify factors independently associated with the primary outcome. During the study period, 461 (24%) patients met inclusion criteria: 70 (15%) APTA and 391 (85%) NAPTA. After adjusting for confounding factors, APTA patients were at a significantly higher risk for developing VTE (59 vs40%; adjusted odds ratio, 1.8; 95% confidence interval, 1.0 to 3.0; adjusted P= 0.04). Whether or not antiplatelet agents were resumed during the hospital stay and the day on which they were resumed did not affect VTE risk. In conclusion, surgical ICU patients receiving antiplatelet agents before admission are at a significantly higher risk for development of VTE.

Increased emergency department (ED) length of stay (LOS) has been associated with increased mortality in trauma patients. In 2010, we implemented a 247 open trauma bed protocol in our designated trauma intensive care units (TICUs) to facilitate rapid admission from the ED. This required maintenance of a daily bump list and timely transferring of patients out of the TICU. We hypothesized that ED LOS and mortality would decrease after implementation.

Supplemental Digital Content is available in the text.

Recently, major advances have been made in the identification of antigens from human melanoma which are recognized by T cells. In spite of this, little is known about the optimal ways to use these antigens to treat patients with cancer. Progress in this area is likely to require accurate preclinical animal models, but the availability of such models has lagged behind developments in human tumor immunology. Whereas many of the identified human melanoma antigens are normal tissue differentiation proteins, analogous murine tumor antigens have not yet been identified. In this paper we identify a normal tissue differentiation antigen, tyrosinaserelated protein 2 (TRP-2), expressed by the murine B16 melanoma which was found by screening a cDNA library from B16 with tumor-reactive cytotoxic T lymphocytes (CTL). A peptide conforming to the predicted MHC class I H2-Kb binding motif, TRP-2181-188, was identified as the major reactive epitope within TRP-2 recognized by these anti-B16 CTLs. By site-directed mutagenesis, it was shown that alteration of this epitope eliminated recognition of TRP-2. It was further demonstrated that a CTL line raised from splenocytes by repeated stimulation in vitro with this peptide could recognize B16 tumor and was therapeutic against 3-d-old established pulmonary metastases. The use of TRP-2 in a preclinical model of tumor immunotherapy may be helpful in suggesting optimal vaccination strategies for cancer therapy in patients.