Management of Complications in Robotic Thoracic Surgery.

The rapid adoption of robotic-assisted thoracic surgery has led to increased interest in the management of complications. Overall rates of complication during robotic-assisted thoracic surgery are low. Reported complications include pulmonary vascular injury; great vessel injury; thoracic duct injury; erroneous transection; tracheobronchial injury; and esophageal, diaphragmatic, and abdominal organ injury. A robotic thoracic surgeon should understand and have a management plan for any potential complication. When a complication occurs, the priority is to stabilize the patient. Then, after stabilization, an assessment of the situation will determine whether the procedure can be continued robotically or whether conversion to thoracotomy or sternotomy is required.

A Deep Deterministic Policy Gradient Approach to Medication Dosing and Surveillance in the ICU.

Medication dosing in a critical care environment is a complex task that involves close monitoring of relevant physiologic and laboratory biomarkers and corresponding sequential adjustment of the prescribed dose. Misdosing of medications with narrow therapeutic windows (such as intravenous [IV] heparin) can result in preventable adverse events, decrease quality of care and increase cost. Therefore, a robust recommendation system can help clinicians by providing individualized dosing suggestions or corrections to existing protocols. We present a clinician-in-the-loop framework for adjusting IV heparin dose using deep reinforcement learning (RL). Our main objectives were to learn a new IV heparin dosing policy based on the multi-dimensional features of patients, and evaluate the effectiveness of the learned policy in the presence of other confounding factors that may contribute to heparin-related side effects. The data used in the experiments included 2598 intensive care patients from the publicly available MIMIC database and 2310 patients from the Emory University clinical data warehouse. Experimental results suggested that the distance from RL policy had a statistically significant association with anticoagulant complications $(p< 0.05)$, after adjusting for the effects of confounding factors.

An Interpretable Machine Learning Model for Accurate Prediction of Sepsis in the ICU.

OBJECTIVES: Sepsis is among the leading causes of morbidity, mortality, and cost overruns in critically ill patients. Early intervention with antibiotics improves survival in septic patients. However, no clinically validated system exists for real-time prediction of sepsis onset. We aimed to develop and validate an Artificial Intelligence Sepsis Expert algorithm for early prediction of sepsis. DESIGN: Observational cohort study. SETTING: Academic medical center from January 2013 to December 2015. PATIENTS: Over 31,000 admissions to the ICUs at two Emory University hospitals (development cohort), in addition to over 52,000 ICU patients from the publicly available Medical Information Mart for Intensive Care-III ICU database (validation cohort). Patients who met the Third International Consensus Definitions for Sepsis (Sepsis-3) prior to or within 4 hours of their ICU admission were excluded, resulting in roughly 27,000 and 42,000 patients within our development and validation cohorts, respectively. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: High-resolution vital signs time series and electronic medical record data were extracted. A set of 65 features (variables) were calculated on hourly basis and passed to the Artificial Intelligence Sepsis Expert algorithm to predict onset of sepsis in the proceeding T hours (where T = 12, 8, 6, or 4). Artificial Intelligence Sepsis Expert was used to predict onset of sepsis in the proceeding T hours and to produce a list of the most significant contributing factors. For the 12-, 8-, 6-, and 4-hour ahead prediction of sepsis, Artificial Intelligence Sepsis Expert achieved area under the receiver operating characteristic in the range of 0.83-0.85. Performance of the Artificial Intelligence Sepsis Expert on the development and validation cohorts was indistinguishable. CONCLUSIONS: Using data available in the ICU in real-time, Artificial Intelligence Sepsis Expert can accurately predict the onset of sepsis in an ICU patient 4-12 hours prior to clinical recognition. A prospective study is necessary to determine the clinical utility of the proposed sepsis prediction model.

Early sepsis detection in critical care patients using multiscale blood pressure and heart rate dynamics.

Sepsis remains a leading cause of morbidity and mortality among intensive care unit (ICU) patients. For each hour treatment initiation is delayed after diagnosis, sepsis-related mortality increases by approximately 8%. Therefore, maximizing effective care requires early recognition and initiation of treatment protocols. Antecedent signs and symptoms of sepsis can be subtle and unrecognizable (e.g., loss of autonomic regulation of vital signs), causing treatment delays and harm to the patient. In this work we investigated the utility of high-resolution blood pressure (BP) and heart rate (HR) times series dynamics for the early prediction of sepsis in patients from an urban, academic hospital, meeting the third international consensus definition of sepsis (sepsis-III) during their ICU admission. Using a multivariate modeling approach we found that HR and BP dynamics at multiple time-scales are independent predictors of sepsis, even after adjusting for commonly measured clinical values and patient demographics and comorbidities. Earlier recognition and diagnosis of sepsis has the potential to decrease sepsis-related morbidity and mortality through earlier initiation of treatment protocols.

An acute care surgery rotation contributes significant general surgical operative volume to residency training compared with other rotations.

BACKGROUND: Surgical resident rotations on trauma services are criticized for little operative experience and heavy workloads. This has resulted in diminished interest in trauma surgery among surgical residents. Acute care surgery (ACS) combines trauma and emergency/elective general surgery, enhancing operative volume and balancing operative and nonoperative effort. We hypothesize that a mature ACS service provides significant operative experience. METHODS: A retrospective review was performed of ACGME case logs of 14 graduates from a major, academic, Level I trauma center program during a 3-year period. Residency Review Committee index case volumes during the fourth and fifth years of postgraduate training (PGY-4 and PGY-5) ACS rotations were compared with other service rotations: in total and per resident week on service. RESULTS: Ten thousand six hundred fifty-four cases were analyzed for 14 graduates. Mean cases per resident was 432 ± 57 in PGY-4, 330 ± 40 in PGY-5, and 761 ± 67 for both years combined. Mean case volume on ACS for both years was 273 ± 44, which represented 35.8% (273 of 761) of the total experience and exceeded all other services. Residents averaged 8.9 cases per week on the ACS service, which exceeded all other services except private general surgery, gastrointestinal/minimally invasive surgery, and pediatric surgery rotations. Disproportionately more head/neck, small and large intestine, gastric, spleen, laparotomy, and hernia cases occurred on ACS than on other services. CONCLUSIONS: Residents gain a large operative experience on ACS. An ACS model is viable in training, provides valuable operative experience, and should not be considered a drain on resident effort. Valuable ACS rotation experiences as a resident may encourage graduates to pursue ACS as a career.

Amphetamine self-administration in light and moderate drinkers.

BACKGROUND: Light and moderate drinkers respond differently to the effects of abused drugs, including stimulants such as amphetamine. The purpose of this study was to determine whether light and moderate drinkers differ in their sensitivity to the reinforcing and subjective effects of d-amphetamine. We hypothesized that moderate drinkers (i.e., participants that reported consuming at least seven alcohol-containing beverages per week) would be more sensitive to the reinforcing and positive subject-rated effects of d-amphetamine than light drinkers. METHODS: Data from four studies that employed similar d-amphetamine self-administration procedures and subject-rated drug-effect measures were included in the analysis. Light (n = 17) and moderate (n = 16) drinkers sampled placebo, low (8 to 10 mg), and high (16 to 20 mg) doses of oral d-amphetamine administered in eight capsules. Following sampling sessions, participants worked for a maximum of eight capsules, each containing 12.5% of the previously sampled dose, on a modified progressive-ratio schedule of reinforcement. RESULTS: Both active doses of d-amphetamine functioned as a reinforcer in the moderate drinkers, while only the high dose did so in the light drinkers. The moderate drinkers worked for significantly more capsules that contained the high dose of d-amphetamine than did the light drinkers. d-Amphetamine produced prototypical stimulant-like subjective effects (e.g., dose-dependent increases in ratings of Good Effects; Like Drug and Willing to Take Again). Moderate drinkers reported significantly greater subjective effects than the light drinkers. CONCLUSION: These results are consistent with those from previous laboratory experiments and suggest that moderate alcohol consumption may increase vulnerability to the abuse-related effects of stimulants.

Chemical mechanism and specificity of the C5-mannuronan epimerase reaction.

C5-mannuronan epimerase catalyzes the formation of alpha-L-guluronate residues from beta-D-mannuronate residues in the synthesis of the linear polysaccharide alginate. The reaction requires the abstraction of a proton from C5 of the residue undergoing epimerization followed by re-protonation on the opposite face. Rapid-mixing chemical quench experiments were conducted to determine the nature of the intermediate formed upon proton abstraction in the reaction catalyzed by the enzyme from Pseudomonas aeruginosa. Colorimetric and HPLC analysis of quenched samples indicated that shortened oligosaccharides containing an unsaturated sugar residue form as transient intermediates in the epimerization reaction. This suggests that the carbanion is stabilized by glycal formation, concomitant with cleavage of the glycosidic bond between the residue undergoing epimerization and the adjacent residue. The time dependence of glycal formation suggested that slow steps flank the chemical steps in the catalytic cycle. Solvent isotope effects on V and V/K were unity, consistent with a catalytic cycle in which chemistry is not rate-limiting. The specificity of the epimerase with regard to neighboring residues was examined, and it was determined that the enzyme showed no bias for mannuronate residues adjacent to guluronates versus those adjacent to mannuronates. Proton abstraction and sugar epimerization were irreversible. Existing guluronate residues already present in the polysaccharide were not converted to mannuronates, nor was incorporation of solvent deuterium into existing mannuronates observed.