Anesthesiology

Point of care gastric ultrasound in ICU patients before and after initiation of post-pyloric enteral feeds

Faculty Mentor’s Name: Dr. Meghan Brennan
Email: mbrennan@anest.ufl.edu
Phone Number: (357) 872-8017
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Background: Aspiration of gastric contents during airway manipulation is a risk when undergoing general anesthesia and associated with significant morbidity and mortality. Preoperative fasting recommendations developed by the American Society of Anesthesiologists may not apply in the critically ill patient population. Currently, no consistent preoperative fasting guidelines exist for critically ill patients receiving post-pyloric enteral feeds. Our study aims to evaluate the use of point of care gastric ultrasound (POCUS) to examine gastric content before and after initiation of post-pyloric enteral feeds in intensive care unit (ICU) patients and potentially clarify how fasting guidelines should be applied to this patient population.

Hypothesis: We hypothesize gastric contents in ICU patients receiving enteral feeds through a post-pyloric feeding tube are identifiable on point of care gastric ultrasound and are similar to patients who have consumed medium to high volume clear liquids preoperatively.

Methods: This is a single center prospective observational cohort study of adult patients admitted to the surgical ICUs receiving post-pyloric enteral feeds. POCGUS was performed prior to initiation of enteral feeds and after feeds had been ongoing at goal rate for at least 6 hours. Ultrasound images will be obtained in both the supine and right lateral decubitus positions to identify the gastric antrum and qualitatively characterize its content.

Role of Medical Student: Medical students must complete IRB training. Once that is completed they will be trained in identification of patients that meet study criteria, enrolling patients in the IRB approved study, use of ultrasound, point of care gastric ultrasound with the aim of being able to independently collect study images before the 10-week program ends. Medical students will also assist in data collection from the EPIC EHR for patients enrolled in the study, data analysis, abstract, and manuscript preparation. IRB approval has been obtained for this study, this study is currently enrolling patients, with the participation of a medical student in the 2021 MSRP.

Funding: Jerome H. Modell Endowed Professorship, through the department of anesthesiology.
Publications: Point of care gastric ultrasound in ICU patients before and after initiation of post-pyloric enteral feeds. Abstract submission to the International Anesthesia Research Society Conference (IARS) 2022.

Constructing a Universal Dataset for Patient Blood Management

Faculty Mentor’s Name: Dr. Keith Howell
Email: khowell@anest.ufl.edu
Phone Number: (804) 399-3988
Project Category: Literature Review
International Component or Travel: No

Universal data sets such as the Minimum Data Set for clinical assessment of all residents in Medicare and Medicaid certified nursing homes (https://www.cms.gov/Research-Statistics-Data-and-Systems/Computer-Data-and-Systems/Minimum-Data-Set-3-0-Public-Reports) and NACC database for Alzheimer’s disease (https://www.nia.nih.gov/research/dn/national-alzheimers-coordinating-center-nacc) facilitate collaborative research and advance both patient care and medical knowledge. Universal datasets are comprised of a well-defined, documented, and agreed upon set of data elements. The work of constructing a universal dataset is arduous yet essential.

UF Health/Shands stands in front of a defining opportunity for itself as a leader in regional and national healthcare. We have been approached by teams within the World Health Organization to form the International Institute for Patient Blood Management (IIPBM). If embraced, this opportunity should catapult UF Health to becoming the most important education, clinical and research resource for the State of Florida, The South East, the Nation, and potentially the World. In this project, we will work to construct a Universal Dataset for Patient Blood Management (PBM) by conducting reviews of the literature and developing consensus of PBM experts on data elements for inclusion. Students who will be part of our team will gain experience in conducting literature reviews, using the Delphi method, constructing a survey in REDCap, and statistical analysis of survey data.

Creation of an intelligent alert to improve efficacy & patient safety in real time during fluoroscopic guided lumbar transforaminal epidural steroid injection

Faculty Mentor’s Name: Dr. Sanjeev Kumar
Email: sanjeevkumar@ufl.edu
Phone Number: (248) 935-7058
Project Category: Clinical
International Component or Travel: No

Research Project Description:

The project has been Funded by IHAF grant and the IRB approval has been given. Starting date is 3/1/22. This is a collaboration between UF Anesthesiology and UF Department of Biomedical Engineering with Dr. Ruogu Fang as the main collaborator/Co PI from UF BME. A UF Engineering post-doc student will be hired to help with designing the convolutional neural network (CNN). UF CTSI will be providing all the deidentified images from patients electronic medical records.

The Medical student’s role could be to assist the UF Engineering grad in testing and refining the CNN on a sample of images and as the CNN gets better then testing and refining it on more images for potentially making the CNN better and better.

This project can be done from anywhere since it mainly involves working with deidentified images from previous procedures done at UF Health.

Artificial Intelligence in Medicine Title: Creating an AI Sandbox

Faculty Mentor’s Name: Dr. Francois Modave
Email: modavefp@ufl.edu
Phone Number: (915) 238-2416
Project Category: Translational
International Component or Travel: No

Research Project Description:

Creation & Validation of Deidentified Perioperative Data Sandbox: Data can be deidentified using either expert determination or the safe harbor method. Deidentified data facilitate data access and speed up knowledge discovery, but can also be useful for teaching and learning without privacy risk. This proiect focuses on understanding deidentification methods and creating a set of perioperative data that can be used to learn perioperative AI.

Prerequisites: none
Number of students: 2-3
IRB-Required

Department: Anesthesiology
Faculty sponsors:

Francois Modave, PhD
modavefp@ufl.edu

Chris Giordano, MD
cgiordano@anest.ufl.edu

Patrick Tighe, MD
ptighe@anest.ufl.edu

Basma Mohamed, MD
bmohamed@anest.ufl.edu

Meghan Brennan, MD
mbrennan@anest.ufl.edu

Admininstrative support:
Carley Hume
chume@anest.ufl.edu

Jeffrey Scott
Jeffrey.scott@ufl.edu

Artificial Intelligence in Medicine Title: Using AI to better appreciate Surgical Site Infections

Faculty Mentor’s Name: Dr. Francois Modave
Email: modavefp@ufl.edu
Phone Number: (915) 238-2416
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Using graph methods (e.g. graph and hypergraph theory, probabilistic Boolean networks, etc.)  to characterize a team-based perspective to surgical site infections: graphs and their extensions can model how teams work and help us understand surgical site infections.

Prerequisites: none
Number of student: 2-3
IRB-Required

Department: Anesthesiology
Faculty sponsors:

Francois Modave, PhD
modavefp@ufl.edu

Chris Giordano, MD
cgiordano@anest.ufl.edu

Patrick Tighe, MD
ptighe@anest.ufl.edu

Basma Mohamed, MD
bmohamed@anest.ufl.edu

Meghan Brennan, MD
mbrennan@anest.ufl.edu

Admininstrative support:
Carley Hume
chume@anest.ufl.edu

Jeffrey Scott
Jeffrey.scott@ufl.edu

Artificial Intelligence in Medicine Title: Using AI to facilitate Clinical Decision-Making

Faculty Mentor’s Name: Dr. Francois Modave
Email: modavefp@ufl.edu
Phone Number: (915) 238-2416
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Greedy methods? Using longitudinal AI methods for sequential clinical decisions: Greedy methods refer to sequential algorithms where at each step you make the “best” choice (e.g. you choose the option that optimizes whatever criterion that was chosen). This project explores such longitudinal AI methods for sequential clinical decisions.

Prerequisites: none
Number student: 2-3
IRB-Required

Department: Anesthesiology
Faculty sponsors:

Francois Modave, PhD
modavefp@ufl.edu

Chris Giordano, MD
cgiordano@anest.ufl.edu

Patrick Tighe, MD
ptighe@anest.ufl.edu

Basma Mohamed, MD
bmohamed@anest.ufl.edu

Meghan Brennan, MD
mbrennan@anest.ufl.edu

Administrative support:
Carley Hume
chume@anest.ufl.edu

Jeffrey Scott
Jeffrey.scott@ufl.edu