United States

AMA Research Challenge 2022

Education Without Walls: Using a Street Medicine Program to Provide Real-World Interprofessional Learning

poster

Welcome to the largest national, multi-specialty research event **featuring research from 750 medical students, residents and fellows, and international medical graduates.**


#### Event Details
The 2022 American Medical Association Research Challenge includes a poster symposium hall and a semifinals hall with the top scored research. Log in at any time throughout the event to:
* **View posters and presentations** on a variety of topics and specialties
* **Score posters in the semifinals** to help decide the top five finalists (AMA member exclusive opportunity)
* **Connect with participants** and provide feedback to help advance their research

At the conclusion of this event, the AMA will announce the top five finalists who will advance to the AMA Research Challenge finals, present their research to an elite panel of judges and compete to win a $10,000 grand prize sponsored by Laurel Road.
 

#### Attend Live Forums
**Oct. 20, 6 — 7 p.m. Central**
 **How to plan a career in academic medicine** 
 Learn how to plan a career in academic medicine and get your questions answered. [Access Here](https://underline.io/events/363/lounge)
 
 **Speakers**
![](https://assets.underline.io/uploads/markdown_image/1/image/d5adb133dfc58eb0a6d469e9cedcd5d4.jpg)
Ricardo Correa, MD, EdD
![](https://assets.underline.io/uploads/markdown_image/1/image/5f9437ab87861538bf02b9ca7c0d926e.jpg)
Stuart M. Greenstein, MD
![](https://assets.underline.io/uploads/markdown_image/1/image/05b1b637c99286d558c371f8e8b185ce.jpg)
Stephanie Strohbeen, MD
 
 **Oct. 22, 11 a.m. — Noon Central**
 **How to get published**
 Learn tips to get your research published and get your questions answered. [Access Here](https://underline.io/events/363/lounge)
 
 **Speakers**
![](https://assets.underline.io/uploads/markdown_image/1/image/8ce6235e957d31cb5e6f9097cc6f6972.jpg)
Jessica Li, MD/PhD Candidate
![](https://assets.underline.io/uploads/markdown_image/1/image/9950ef2e0237f54cb3d8280334d42491.jpg)
Charles Lopresto, DO

### Sponsor
AMA would like to thank Laurel Road for sponsoring a $10,000 grand prize for the winner of the AMA Research Challenge. Laurel Road is not involved in the judging or scoring process to select the winner. The winner will be announced on Dec. 7 during the AMA Research Challenge finals. Payment of the grand prize will be directly from Laurel Road. See the AMA Research Challenge Official Rules for terms and conditions.

![](https://assets.underline.io/uploads/markdown_image/1/image/611e331d4a0eba0920f240cfc04b4d77.png)

**There is no registration fee for this event! **

Register to explore research in a variety of topics. AMA members get exclusive access to score posters in the semifinals.

**Please use the link below to register for free if you have not already done so:**

https://www.ama-assn.org/about/events/ama-research-challenge

Registration Is Required

Welcome to the largest national, multi-specialty research event featuring research from 750 medical students, residents and fellows, and international medical graduates.


The development and implementation of artificial intelligence and machine learning (AI/ML) into healthcare offers exciting promises such as enhanced detection, prevention, and response to disease, automation of mundane tasks, promotion of healthy lifestyles, and disease self-management. While AI/ML can certainly be an asset, it is imperative to proactively confront potential problems throughout the development process. This presentation provides considerations across the health spectrum, including changes in medical training and knowledge, healthcare equity, the standard of care, and the patient-physician relationship. To maximize benefit from AI/ML, ethical and policy challenges that accompany the technology must be addressed so that AI/ML genuinely improves patients’ health and physician workplaces. This presentation provides an overview of such challenges and emphasizes how creating fair, beneficial AI/ML starts in its development process.

Ethical and Policy Challenges Associated with Machine Learning and Artificial Intelligence in Healthcare

Background: The Front Line Resilience (FLR) program was developed to provide mental resiliency tools to firefighters, health care professionals, and medical students serving Asheville, NC during the COVID-19 pandemic. Participants attended 15–30-minute weekly Zoom sessions to learn mindfulness topics. Interested participants were paired with a practice partner. The purpose of this study was to evaluate participants’ perceptions of the effectiveness of a virtual curriculum on their mindfulness practices and resiliency. 

Methods: FLR participants were invited to evaluate program effectiveness via online feedback surveys. Survey respondents were then invited to participate in an interview for more in-depth evaluation. Surveys were analyzed using descriptive statistics. Interviews were transcribed and independently assessed by two researchers via manifest content analysis. 

Results: Of all 46 participants who attended at least one FLR session, ten respondents completed the survey. Of those respondents, five (50%) participated in interviews. Respondents indicated that FLR had a moderate (70%) or significant (30%) effect on their mindfulness practice, and a moderate (60%) or significant (30%) effect on resilience. Thematic analysis of interviews revealed participants learned specific techniques to expand their resilience toolbox and valued connection and accountability with other frontline professionals; however, their experiences did not always meet expectations. 

Conclusion: FLR positively impacted mindfulness practices and resiliency of frontline workers during the COVID-19 pandemic. This model could be used to support mindfulness practices for other professions.

Front Line Resilience: A virtual curriculum with medical providers and firefighters during COVID-19

The aim of this study is to investigate the impact of the coronavirus pandemic on the rates of major lower extremity amputations related to the delayed treatment of patients with peripheral vascular disease and diabetes mellitus at Eastern Maine Medical Center (EMMC) in Bangor, Maine.

Impact of the COVID-19 Pandemic on the Rates of Peripheral Vascular Disease and Diabetes-Related Lower Extremity Amputations in a Level II Trauma Center in Northern Maine

Improving Health Professionals' and Learners' Attitudes Towards Homeless Individuals Through Street-Based Outreach

Background
Opioid-Use Disorder is characterized by the problematic use of opioids leading to significant impairment and distress. The disorder affects approximately 2.1 million people within the United States alone, and 16 million people worldwide as of 2016. Fortunately, there are several methods poised to aid in the prevention of OUD. One method is preemptive multimodal analgesia to reduce subjective post-operative pain outcomes as well as oral morphine equivalents. Another method utilizes a cheap, cost-effective at-home opioid disposal kit as a method to discard left-over opioid prescriptions after recovery.
Methods
The preemptive MMA study consisted of a retrospective chart analysis, which found 1416 eligible patients who underwent THA or TKA. The patients received acetaminophen, celecoxib, and gabapentin between 30-60 minutes pre-operatively. Post-operatively, oral morphine equivalents were monitored along with subjective pain scores. In reference to the at-home disposal kit study, a RCT compared patients receiving the usual standard of care, to the group obtaining an at-home opioid disposal kit post-op. A text messaging service was implemented to obtain self-reported data and disposal rates.
Results
The preemptive MMA led to a significant decrease in oral morphine equivalents at day zero and day two post-op, as well as a significant reduction in subjective pain scores at almost every time-period regardless of the operation.
In the at-home disposal kit study, 42.7% of patients in the control group self-reported disposal of their remaining opioids, while 60.0% of patients who received a kit disposed of their left-over medication. 
Conclusion
Preemptive MMA significantly reduced the patient’s subjective pain levels post-operatively at almost every time-period tested. It is important to emphasize that in this study the medications administered were non-opioids, as perioperative opioid prescriptions lead to an increased risk of developing OUD. Additionally, in a randomized clinical trial, mild opioids (Codeine/Acetaminophen) were found to be non-inferior to strong opioids in respect to pain after fractures requiring surgical intervention; therefore, a mild treatment plan is preferred and preventive in the development of OUD. Furthermore, at-home disposal kits could be used in conjunction with preemptive MMA to reduce the rate of OUD in the greater population by motivating patients with a proper disposal method.


Pre- and Post-Operative Practices to Reduce the Rate of Opioid-Use Disorder

Background: In the US, the majority of Class I medical device recalls—FDA’s most serious designation, indicating high probability of adverse health consequences or death—are attributed to devices cleared under the 510(k) pathway. The role of predicates, which are previously authorized devices used to clear new devices under 510(k), in Class I recalls has not been systematically studied. We sought to determine whether 510(k) devices subject to Class I recalls used predicates that had a previous history of recalls, and to assess whether recalled devices were used as predicates in future device authorizations.

Methods: In this cross-sectional study, all 510(k) devices subject to Class I recalls between January 2020 through December 2021 (“index devices”) were identified from the FDA’s annual recall listing. Information about predicate devices was extracted from index device decision memos in the Devices@FDA database. Devices authorized using index devices as predicates (“descendants”) were identified using a commercially available database. Devices were characterized by their regulatory history (medical specialty, product code, special designations, use of clinical evidence) and recall history (date of recall relative to index device recall or clearance, FDA classification of recall severity).

Results: Between 2020-21, 75 devices were subject to Class I recalls, with each recall affecting nearly ten thousand devices. Of these devices, 31 (41.3%) had prior recalls, including 13 (41.9%) with Class I recalls. These 75 devices had 109 unique predicates, of which 62 (56.9%) had prior recalls, including 43 (69.4%) with at least one Class I recall. Recalls for 36 predicates (30.3%) preceded the predicate’s use to support clearance of the index device, and recalls for 51 predicates (42.9%) preceded the index device’s Class I recall. Fifty-two (69.3%) index devices were predicates for subsequent device authorizations. Of the 94 unique descendants identified, 49 (52.1%) had recalls, including 39 (79.6%) with at least one Class I recall.

Conclusion: Many 510(k) devices subject to Class I recalls were authorized using predicates with a known history of safety concerns and were used to clear additional devices that were later recalled. Stronger safeguards for the 510(k) pathway are needed to prevent such problematic predicate selection.

Predicate Lineages and Medical Device Recalls in the 510(k) Pathway

Abstract Title: Targeting reduced mitochondrial apoptotic signaling in relapsed acute myeloid leukemia

Background
The therapeutic landscape for acute myeloid leukemia (AML) has expanded impressively over the last few years, with 8 new drug approvals by the Food and Drug Administration. However, 40-60% of AML patients still relapse, and the majority of these patients succumb to their disease within 5 years due to limited therapeutic options. Since development of acquired drug resistance contributes significantly to treatment failure in AML, we aimed to investigate emerging mechanisms of resistance against targeted therapies and to identify potential strategies overcoming resistance.

Methods
We utilized dynamic BH3 profiling (DBP), a functional assay that measures drug-induced mitochondrial apoptotic signaling to derive pharmacologic landscapes for 17 AML patient-derived xenograft (PDX) mouse models, together with genomic and transcriptomic profiles. First, we performed ex vivo DBP on treatment naïve and relapsed PDX myeloblasts to predict drugs active in vivo. Next, we modeled in vivo acquired resistance to single agents (venetoclax, quizartinib, birinapant, LCL-161, JQ-1, and S63845) with different mechanisms of action to elucidate a mechanism of pan-resistance. Lastly, we reproduced drug-response patterns identified in PDX models in primary (n=81) and relapsed (n=53) AML patients, underscoring the clinical application of our methods.

Results
By applying ex vivo DBP assays on PDX myeloblasts, we showed that ex vivo mitochondrial drug sensitivity at 14 hours correlated with in vivo activity at day 14 post-treatment, thereby validating DBP as a functional predictive biomarker for personalizing AML therapy. We discovered that PDXs with acquired drug resistance to 6 agents targeting distinct pathways emerged via the common mechanism of reduced mitochondrial apoptotic signaling and developed broad chemoresistance. RNA sequencing revealed gene expression signatures corresponding to common relapse phenotypes conserved across distinct targeted agents. Importantly, we observed that AML patients at relapse showed similar pan-resistance phenotypes as PDX models with acquired resistance.

Conclusion
We showed that DBP-based assays can direct personalized therapy for mouse and human AML in diagnosis and relapsed settings, which is notable given that pan-resistance phenotypes limit treatment options for relapsed tumors. Driven by these compelling pre-clinical findings, we are initiating a phase 2 DBP-guided clinical trial to selectively and personally choose therapies for patients with relapsed or refractory AML.

A New Achille’s Heel: Targeting Reduced Mitochondrial Apoptotic Signaling in Relapsed Acute Myeloid Leukemia

INTRODUCTION: 
Femoral shaft fractures are significant injuries with a predominately traumatic etiology. Intramedullary nailing is the preferred surgical intervention, and reduction of the fracture can be accomplished through closed or open methods. Closed reduction is often considered the gold standard as it is theorized that an intact fracture hematoma has osteogenic properties that will result in superior healing with fewer complications. Open reduction procedures persist in cases of polytraumatic injury, when the fracture proves difficult to reduce by closed methods, or in geographical areas with limited resources. Studies comparing outcomes between the treatments have conflicting results. The purpose of this systematic review and meta-analysis was to compare the outcomes and complications of intramedullary nailing of femoral shaft fractures between open and closed reduction. 
METHODS: 
Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, PubMed (MEDLINE), Embase, Scopus, and CENTRAL were searched to identify comparative studies up to May 2022. Additional studies were identified through hand and citation searching. Exclusion criteria included pediatric patients younger than 17 years of age, studies not published in English, and studies that reported fractures of the proximal or distal femur. The primary outcome was the rate of union. Secondary outcomes included time to union, operative time, and complication rate. 
RESULTS:
Closed reduction had a higher pooled union rate (93.93%, 92.02%), an increased odds ratio for union (OR =1.624 [95% CI: 1.004, 2.624]; p =.048), and a faster time to union (SMD = -0.292 [95% CI: -0.549, -.0.035]; p =.026). There were no differences in operative time (SMD = .128 [95% CI: -0.700, .956] p = .762) or overall complication rate (OR = 1.314 [95% CI: .966, 1.787] p = .082).
DISCUSSION AND CONCLUSION: 
Closed reduction has the advantage of higher union rates, quicker time to union, and lower overall infection compared to open reduction for intramedullary nailing of femoral shaft fractures. Open reduction remains a reasonable alternative with acceptable union rates when closed reduction is not feasible.


A Systematic Review and Meta-analysis of Open Versus Closed Reduction and Intramedullary Nailing of Femoral Shaft Fractures

BACKGROUND: About 30-40% of PDAs can be missed on CTs due to subtle imaging findings. Unfortunately, PDA can advance from being undetectable to widely metastatic in a matter of months which underscores the critical need to develop novel tools such as AI to improve its detection. Most current AI-based PDA detection approaches require manual segmentations that are not clinically scalable. Our objective was to develop a 3D Convolutional Neural Network (CNN) for fully automated detection of PDA on standard-of-care CTs.

METHODS: A fully-automated 3D-CNN was trained on a dataset of 696 portal venous phase diagnostic CTs with PDA and 1080 control CTs with normal pancreas. Model was evaluated on (a) test subset of 409 PDA and 829 control CTs; (b) simulated cohort with high pre-test probability of PDA (1-5% over 3-years), which matched the case-control distribution in high-risk groups (glycemically-defined new-onset diabetes and high ENDPAC score); (c) multi-institutional public dataset of 194 PDA and 80 control CTs; and (d) a cohort of 102 prediagnostic CTs (i.e. CTs with normal-appearing pancreas incidentally acquired 3–36 months before clinical diagnosis of PDA) and 134 control CTs.

RESULTS: In the test subset, the model correctly classified 360 (88%) PDA CTs and 783 (94%) normal CTs [accuracy (mean; 95% CI) 0·92 (0·91-0·94); AUROC 0·97 (0·96-0·98), sensitivity 0·88 (0·85-0·91), specificity 0·95 (0·93-0·96)]. Performance was comparable across tumor stages (sensitivity 0·80, 0·87, 0·95 and 1 on T1 through T4 stages, respectively), hypodense versus isodense tumors (sensitivity: 0·90 vs. 0·82), age, sex, CT slice thicknesses & vendors (all p >0·05), and generalizable on both the simulated cohort [accuracy 0·95 (0·94-0·95), AUROC 0·97 (0·94-0·99)] and public dataset [accuracy 0·86 (0·82-0·90), AUROC 0·90 (0·86-0·95), sensitivity 0·88, specificity 0·93] including for T1 and isodense tumors (sensitivity 0·76 and 0·88, respectively). Despite being trained on diagnostic CTs, the model could detect PDA at the prediagnostic stage from normal pancreas [accuracy 0·84 (0·79-0·88), AUROC 0·90 (0·87-0·94), sensitivity 0·75 (0·66-0·83), specificity 0·90 (0·85-0·95)] at a median interval of 475 days (range:93-1082 days) between prediagnostic CTs and the clinical diagnosis of PDA. 

CONCLUSION: A 3D-CNN developed on a large dataset shows high accuracy and generalizability for fully-automated detection of PDA (including small and isodense tumors) on diagnostic CTs as well as for early detection of PDA from normal pancreas on prediagnostic CTs at a substantial lead time prior to clinical diagnosis. Prospective validation with blood-based biomarkers could enable AI-augmented early detection of PDA in high-risk subjects.

AI-Powered Fully automated Early Detection of Pancreatic Ductal Adenocarcinoma on Standard-of-care CT Scans.

Background: The rising prevalence of obesity, diabetes, and smoking among American adults has disproportionately affected female, Black, and younger patients. The purpose of this study was to evaluate for disparity in patient eligibility for Total Ankle Arthroplasty (TAA) based on race, ethnicity, gender, and age by applying Body Mass Index (BMI), smoking, and diabetes criteria to the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database.

Methods: Patients in the ACS-NSQIP database who underwent TAA from 2011- 2020 were retrospectively reviewed. BMI cut-offs of <50kg/m², <45kg/m², <40kg/m², <35kg/m², diabetes, and smoking criteria were then applied. The eligibility rate for TAA was examined for each BMI cut-off, diabetes, and smoking; findings were stratified by race, ethnicity, gender, and age. Independent T-tests, Chi-squared tests, and Fisher’s exact tests were performed to compare differences at an α = 0.05.

Results: 1215/1865 TAA patients (65.1%) were included after applying the exclusion criteria. Black patients had disproportionately lower rates of eligibility at the most stringent BMI cut-off of <35kg/m² (p=0.004). In contrast, Asian American/Pacific Islander patients had higher rates of eligibility at the BMI cut-offs of <35kg/m² (p=0.033) and <40kg/m² (p=0.039). Black patients also had lower eligibility rates after diabetes (p=0.042) and smoking (p=0.008) criteria were applied. Whereas, white non-Hispanic patients had disproportionately higher rates of eligibility after the diabetes criterion was applied (p=0.04). Females had lower eligibility rates across all BMI cut-offs (p<0.05 for all). Younger patients had lower eligibility rates across all BMI cut-offs (p<0.05 for all) and smoking (57 ± 10, p<0.001).

Conclusion: Stringent BMI cut-offs, diabetes, and smoking criteria may disproportionately disqualify women, Black, and younger patients from receiving Total Ankle Arthroplasty. This data suggests that current eligibility criteria may exacerbate disparity in access to arthroplasty care for already vulnerable populations.

Are Body Mass Index (BMI), diabetes, and smoking cutoffs creating racial, ethnic, gender, or age disparities in patient eligibility for Total Ankle Arthroplasty?

Next from AMA Research Challenge 2022

Ethical and Policy Challenges Associated with Machine Learning and Artificial Intelligence in Healthcare