Scientific Abstract Presentations: Research & Technology 4 Friday, October 27 Fri, Oct 27 10:30 a.m.-12:30 p.m. CDT

Purpose The field of Artificial Intelligence (AI) has been defined as the development of machines capable of solving problems and achieving goals in an ever-changing and unpredictable environment. AI involves automating complex algorithms that rely heavily on statistics, allowing computers to perform calculations and modeling that would be too complicated and time-consuming for humans to do. In recent years, there has been a significant increase in investments in AI in healthcare applications, such as diagnosing patients, researching pharmaceuticals, facilitating communication, and tracking and monitoring patient progress. Open AI has launched its GPT-3 models, which can generate and understand natural language, and use AI to create written texts mimicking that of humans. The present study aims to assess the capability of GPT-3 to generate an abstract from a baseline article and compare it to the original abstract, hypothesizing that current AI technology is capable of equal or superior abstract writing compared to human authorship.

Methods We searched the database of a leading plastic surgery journal for the latest articles on lymphedema and selected an article published between November 2022 and January 2023. We then used OpenAI's GPT-3 model to generate an abstract for the selected article, instructing the AI to match the number of words in the original abstract. Ten reviewers with M.D. or Ph.D. degrees were asked to evaluate both the original abstract and the AI-generated abstract blinded from abstract authorship. An Independent Student's T test was performed to determine statistical difference in the scores given to the two abstracts.

Results Abstract #1 was written by human authors, while abstract #2 was written by AI. Four out of the five reviewers incorrectly identified abstract #1 as being written by AI, resulting in the lowest scores on our grading scale. Although abstract #2 received higher scores in both total score and individual question score, the difference was not statistically significant (p=0.2).

Conclusion Our results suggest that the AI-generated abstract was preferred by reviewers and received higher scores, despite being incorrectly identified as the human-authored abstract by some reviewers. The results of our study also support and confirm our initial hypothesis that AI tools, such as GPT-3, can be beneficial for abstract writing and may even be superior to human authorship when used appropriately. The use of AI technology in medical research is expected to grow, and future studies should explore the potential of AI in this area.

Background: Depression and anxiety have been shown to affect clinical outcomes following surgery, impacting the immune response and lowering the pain threshold.1 As there is a paucity of mental health studies in pediatric surgical patients, the goal of our study is to describe the prevalence of depression and anxiety in this population. Methods: EPIC SlicerDicer was used to facilitate retrospective chart review of all patients seen by a surgical care specialty at Lurie Children's Hospital from January 2012 to December 2022. Percentage of patients within each specialty with concurrent diagnosis of depression, anxiety, and both were identified. Descriptive and statistical analysis was performed for differences in demographics and prevalence of depression and/or anxiety within surgical subspecialties. Results: 494,160 patients were included from 7 surgical subspecialties: Neurosurgery (n = 26,428, 5.35%), Ophthalmology (n = 52,593, 10.64%), Orthopedic Surgery (n = 143,567, 29.05%), Otolaryngology (n = 110,621, 22.39%), Pediatric Surgery (n = 64,620, 12.08%), Plastic Surgery (n = 23,348, 4.72%), and Urology (n = 72,983, 14.77%). Concurrent diagnoses of depression, anxiety, and both occur at significantly higher rates in pediatric patients seen by plastic surgery (0.24%, 0.32%, 0.18%, respectively) than those seen by most other surgical subspecialties (p<0.05). Median age of plastic surgery patients was 17 years (IQR = 15, 19) for those with concurrent anxiety and 18 years (IQR = 17, 20) for concurrent depression. Females made up greater proportions than males for both anxiety (69% vs. 31%) and depression (70% vs. 30%). Those who identified racially as White made up the greatest proportions out of all races for both anxiety (70.3%) and depression (73.2%), and those who identified ethnically as Non-Hispanic/Latino made up greater proportions than those who identified as Hispanic/Latino for both anxiety (75.7% vs 21.6%) and depression (73.2% vs 26.8%). There were no significant differences in rates of anxiety in pediatric plastic surgery (0.32%) compared with pediatric surgery (0.22%, p = 0.561) or neurosurgery (0.27%, p = 0.357). Conclusions: Our single institution study demonstrates that rates of depression, anxiety, and both are higher in pediatric plastic surgery patients compared to patients from most other surgical subspecialties. As depression and anxiety have been associated with greater postoperative morbidity following surgery, pediatric surgical patients should undergo mental health screening using tools such as the PHQ-2/9 and GAD-7 with subsequent mental health resources provided as appropriate. Future studies examining postoperative outcomes in the setting of concurrent mental health diagnosis may better elucidate the relationship between psychosocial factors and surgical outcomes.

References 1. Ghoneim, M.M., O'Hara, M.W. Depression and postoperative complications: an overview. BMC Surg 16, 5 (2016). https://doi.org/10.1186/s12893-016-0120-y

Background: Surgical drains, used in a variety of procedures, are placed prior to closing an operative wound and allow for drainage of bodily fluids that would otherwise accumulate in potential spaces created by operations.1 To prevent premature drain removal, drain tubing is secured to the patient at the exit point by tying suture in a "Roman garter" technique.2 Suture-based drain securement is known to cause discomfort, particularly at the site of skin fixation as drain tubing movement is transmitted to a focal point. Patients may experience skin irritation, pain, skin tugging, loosening, and unintended loss of the drain. A novel, sutureless drain securement device (SDSD) was created to address these factors which impact patient experience while maintaining the functionality and purpose of surgical drains.

Methods: An IRB-approved randomized control trial enrolled 21 patients of 7 different surgeons. Patients 18 years or older who underwent placement of 2 or more drains as part of either breast reconstruction, breast reduction, abdominoplasty, or body contour surgery were included. Patients under the age of 18 and/or those with an allergy to skin adhesives or Tegaderm dressing and unilateral drain placement were excluded. Subjects were then randomized to receive a right or left SDSD with the contralateral site undergoing traditional suture-based drain securement. Outcomes evaluated included: time taken to secure and dress SDSD versus traditional suture-based technique, surgeon assessment of SDSD usability, adverse outcomes, patient satisfaction, and a blinded evaluation of each skin site at the time of drain removal. Analysis was performed using paired t-tests.

Results: No drain securement failure was encountered. The average duration of drain placement was 8.24 days (range = 5-14 days). Securement of the SDSD was significantly faster (p= 0.0008) when compared to traditional suture. Blinded skin site evaluations concluded no significant difference in erythema and blistering, (p=0.655) and (p=0.8936) respectively. Patients significantly favored the SDSD in all survey categories including activities of daily living, irritation, itching, tugging, and pain.

Conclusions: The novel SDSD enhances the overall patient experience with surgical drains. The SDSD offers a reliable, standardized, patient-friendly alternative to traditional suture-based drain securement. Utilization of the SDSD reduces the time to secure drains and has the potential to reduce the risk of sharps injury. Implementation of the SDSD technology may be adapted to other suture-secured medical devices.

References: 1. Khansa, Ibrahim M.D.; Khansa, Lara Ph.D.; Meyerson, Joseph M.D.; Janis, Jeffrey E. M.D.. Optimal Use of Surgical Drains: Evidence-Based Strategies. Plastic and Reconstructive Surgery 141(6):p 1542-1549, June 2018. | DOI: 10.1097/PRS.0000000000004413 2. Heskin L, Cahill V, Filobbos G, Regan P, O'Sullivan ST, Bryan K. A new adaptation for a secure surgical drain placement and a comparison with four common drain fixation methods. Ann R Coll Surg Engl. 2019;101(1):60-68. doi:10.1308/rcsann.2018.0177

Background: Adipose-derived stem cells (ADSCs) have been widely used in the management of wounds because of their ability to release a variety of cytokines. Owing to their superior migratory ability, ADSCs are rapidly recruited into wounded sites, where they possibly undergo differentiation towards dermal fibroblasts, endothelial cells, and keratinocytes. However, the molecular mechanisms of the paracrine effect, including the types of cytokines released from ADSCs, are largely unknown. We previously reported that basic fibroblast growth factor (b-FGF) enhances the proliferative potential of ADSCs, and we identified an increased release of CXCL1 and IL-8 at 4 hours after stimulation of ADSC with b-FGF, and found that these cytokines promoted angiogenesis and lymphangiogenesis.1 In this study, we investigated the effects of CXCL1 and IL8 on the migratory ability of fibroblasts, fibrosarcoma cells, and ADSC cells.

Methods: HT1080 (fibrosarcoma), TIG114, and ADSCs obtained from a 45-year-old African American male were seeded onto 48-well dishes in DMEM containing 10% FBS and cultured until they reached confluence. After scratch, HT1080, TIG114, and ADSC cells were incubated in DMEM containing 2% and 1% FBS, respectively, in the presence or absence of 10 ng/ml human CXCL-1 or IL-8 for 6 hours. Images of each well were taken before and after incubation, and the area of the scratch was measured to calculate the reduction in the scratched area after 6 hours. Initially, the entire well was first photographed and recorded with a fluorescence microscope BZ-X800, and then a 350 x 350-pixel square area at the center of the image was measured using ImageJ software. Cell migration rate (%) was calculated by dividing the difference between scratch area (pixels) before and after 6-hour incubation by 350 x 350 pixels.

Results: The cell migration rate of TIG114 cells without supplemental cytokines was 3.499% after a 6-hour incubation. This rate was significantly increased in the presence of CXCL-1 (8.830%, P=0.001) and IL-8 (9.304%, P=0.007). The cell migration rate of HT1080 cells without supplemental cytokines was 14.714%, which was significantly increased in the presence of CXCL-1 (20.115%, P=0.001) and IL-8 (19.934%, P=0.002). The cell migration rate of ADSCs without supplemental cytokines was 5.779%, which was significantly increased in the presence of CXCL-1 (15.325%, P<0.001) and IL-8 (16.044%, P<0.001).

Conclusion: In this study, by scratch assay analysis, we demonstrated that CXCL-1 and IL-8 can promote the migration of fibroblasts and ADSCs, which is required for wound healing. The scratch assay also demonstrated that CXCL-1 and IL-8 promote the migration ability of fibrosarcoma cells. Our results suggest that CXCL-1 and IL-8 secreted by ADSCs may be involved in the ADSC-promoted migration of cancer cells. Therefore, it is necessary to bear in mind the role of mesenchymal cells during fat grafting after excision of adipo-cutaneous or glandular neoplastic processes.

1. Matsui C, Koide H, Ikeda T, et al. Cytokines Released from Human Adipose-tissue–Derived Stem Cells By b-FGF Stimulation: Effects on Angiogenesis and Lymphatic Vessels Formation By IL-8 and CXCL1. Plast Reconstr Surg Glob Open. 2021;20;9(10Suppl):124-125.

PURPOSE – Studies have demonstrated that negative pressure wound therapy (NPWT) reduces hematoma and seroma formation and improves healing compared to a standard wound dressing. However, the specific mechanisms of action by which NPWTs improve healing remain unclear. We have developed a multimodal implantable biosensor which is capable of measuring cutaneous blood flow velocity, tissue oxygenation, and pressure. In this work, we utilize this novel tool to evaluate the effect of PrevenaTM NPWT on local tissue oxygenation (StO2), cutaneous blood flow, and pressure.

METHODS – We present a flexible, wireless, Bluetooth-enabled, percutaneously introducible subcutaneous multimodal device that directly and continuously measures StO2, blood flow velocity, and pressure within the target tissue. The multimodal device uses near-infrared technology to quantify tissue O2 saturation and precision thermal conductivity to quantify blood flow. A high sensitivity pressure sensor is also included. PrevenaTM therapy was placed on intact swine skin which had been implanted with a multimodal device which had previously been introduces into the subcutaneous plane. Four cycles of 15-min PrevenaTM on, 15-min PrevenaTM off were completed followed by one cycle of 60-min PrevenaTM on, 30-min PrevenaTM off.

RESULTS – The multimodal device recorded blood flow changes that correlated with negative pressure. When the PrevenaTM therapy was turned on, local blood flow rate increased and when the therapy was turned off, blood flow rate decreased. The multimodal device recorded negligible variation in StO2 and pressure values when the PrevenaTM was turned on.

CONCLUSIONS – This novel intramuscular multimodal device identifies increased local tissue blood flow, without changes in pressure or StO2, in the presence of NPWT. This porcine study provides initial evidence that PrevenaTM increases blood flow within the tissue that underlies the sponge. This offers some support for our clinical practice which includes the use of PrevenaTM for support of high risk and ischemic wounds. Future studies will include utilizing the multimodal device in both non-survival and survival excisional wound models with NPWT.

Background The Ewing Amputation is a modified approach to below-knee amputation (BKA) that incorporates the construction of Agonist-Antagonist Myoneural Interfaces (AMI) at the time of limb amputation. The AMI is a surgical construct in which naturally opposed, neurotized muscles are biomechanically linked in order to recreate the neural feedback loops present in intact human joints. We here present the long-term clinical outcomes amongst our lower extremity AMI amputee cohort and demonstrate a mitigation of phantom and residual limb pain, preservation of residual limb volume, and functional excursion of AMI constructs across the post-operative period.

Methods We performed 36 Ewing procedures in a cohort of patients at Brigham & Women's Hospital (BWH) and Brigham and Women's Faulkner Hospital (BWFH) between July 2016 – March 2022. Outcomes were assessed prospectively over time and were clinical (demographics, surgical variations, complications), functional (construct excursion, prosthetic usage), and sensorial (residual limb pain, phantom pain, proprioception) in nature.

Results Thirty-Six (36) Ewing patients were included in this study; 17 left (47.20%), 15 right (41.70%), and 4 bilateral (11.10%) for a total of 40 limbs. Patients were split between male (20, 55.60%) and female (16, 44.40%). Mean age at index amputation was 40.19 ± 11.11 years. Mean residual limb volume preservation at 12-months was 97.58% ± 4.13% of the preoperative state. Average construct excursion evidenced by ultrasound at 12-months post-operatively was 4.31mm ± 2.18mm, with strain relationships paralleling those of normal muscle dynamics. 72.20% (n=18) of patients reported no residual limb pain, 77.80% (n=18) reported no phantom limb pain, and 100.00% (n=15) reported anatomically correct phantom limb sensation at 12-months post-operatively.

Conclusion The Ewing amputation presents as a promising surgical intervention that can mitigate phantom and residual limb pain, preserve proprioception, and preserve residual limb volume.

Background Free tissue transfer stands at the apex of the reconstructive ladder, however, despite decades of surgical innovation, free flaps remain limited by physiologic need for rapid and sustained reperfusion to maintain viability. This need can cause a particular limitation in cases of prolonged ischemia periods and restricted availability of recipient vessels. To overcome this limitation, paradigm shifts towards device-assisted and/or ex vivo perfusion supported flap technology are necessary to expand capabilities of free flap surgery. However, in developing these technologies, there is a need for a replicable, docile, and anatomically relevant large-animal model from which we could easily assess and modify free tissue transfer with external analytic and/or treatment devices. Here we aim to describe the novel ovine model of fasciocutaneous free flap transfer, which we have developed for this purpose.

Methods Female Suffolk sheep cadavers weighing 50-65 kg were studied to identify a saphenous system-based fasciocutaneous free flap model. Anatomical measurements and photography were performed to standardize flap harvest and feasibility of cadaveric flap perfusion was assessed via ex vivo fluorescein angiography. Following flap identification, the autologous microsurgical transfer of the flap to the neck was performed in a female Hampshire sheep weighing 65 kg. The sheep was followed for 2 weeks with daily photography. Extracellular tissue lactate and glucose levels were characterized during the transfer and weekly with a custom microdialysis probe placed in the flap tissue. Doppler ultrasonography was performed weekly to assess pedicle viability. Fluorescein angiography was used to assess flap viability at the initial surgery as well as the endpoint. Tissues were collected for histology at the endpoint.

Results Saphenous vessels branching off the femoral system were identified to have up to 2.5 mm vein and 2.0 mm artery diameter, with the total pedicle length reaching up to 6 cm. The cadaveric flap tissue demonstrated feasible inflow and outflow with fluorescein angiography. Upon autologous transfer, the transferred flap of 9x6 cm provided complete coverage of the neck defect and demonstrated viability for the entire duration of the experiment. Viability of the free flap was confirmed with doppler ultrasonography and fluorescein angiography. The donor site was covered with a skin graft from the neck with bolster placement.

Conclusion We identified a novel model of free flap transfer to serve as a testbed for clinically translatable approaches to improve outcomes in reconstructive microsurgery. Ovine docility, relative to other quadrupeds, improves daily flap accessibility and supports survival. These initial studies serve to set the foundation for the use of complex extracorporeal membrane oxygenation (ECMO)-like devices for enhancing uses of the free flaps.

Background: Over time, patient-reported outcomes (PROs) have progressed from rudimentary surveys to validated questionnaires that gauge health-related quality of life. These measures hold great significance in clinical trials, as they empower patients to relay their experiences and enable clinicians to ascertain the efficacy of treatments from a patient-centric viewpoint. The objective of this systematic review was to evaluate the prevalence of PROs in plastic and reconstructive surgery (PRS) clinical trials (CTs) conducted in the United States and determine the proportion of studies that utilize validated instruments.

Methods: A comprehensive systematic review of several databases was performed. The search strategy was designed and conducted by an experienced librarian using controlled vocabulary with keywords. Inclusion criteria encompassed all clinical trials in PRS from 2012 to 2022. A two-stage screening process was conducted to identify CTs then assess for the use of PROs. Summary Statistics were calculated based on underlying variable distribution and a subgroup analysis was then performed using Fisher Excact testing.

Results: A total of 3,609 studies were identified in an initial search with 154 later determined to be CTs in PRS. Overall, 80 studies (52%) used PROs and were included in the analysis. A total of 13,190 participants were present in the included studies with 95% (12,229) females. Moreover, 37 (46%) of the CTs were in the field of reconstruction, while 25 (31%) were cosmetic. Further, 35% of the CTs reported pain as the primary outcome, followed by 24% assessing patient's satisfaction. However, only 61% of the 80 included CTs had a validated PRO as the primary outcome of the study. Visual Analog Scale (VAS) (19%) was the most frequently used validated questionnaire, followed by Breast-Q (15%). Regarding funding, 34% (27) of the CTs were funded by a private institution while 49% (39) did not report any type of funding. A trend analysis did not show a statistically significant increase in the use of validated PROs in CTs between 2012 and 2022.

Conclusions: The use of PROs is very relevant for healthcare delivery and improvement as they provide insight into the efficacy of treatments from a patient-centric viewpoint. PROs are reported in just over half of PRS CTs, and within those CTs, the use of validated questionnaires is inconsistent. Therefore, emerging clinical trials should strive to incorporate PROs measures and utilize the existing validated tools to assess novel interventions and ensure that the data reported is objectified.

Background: 91% of aesthetic plastic surgeons tend to prescribe opioids for postoperative pain after abdominoplasty.[1] We aimed to compare the postoperative opioid usage with the local addition of Bupivacaine (29.25mg/ml)/Meloxicam (0.88mg/ml)(ZynrelefTM, Heron Therapeutics, (Z)) to the current multimodal regimen after a panniculectomy. Methods: With informed consent, 34 patients scheduled for panniculectomy were divided into the Z group(n=17) and the Non-Z group(n=17). Both groups had the same intraoperative anesthesia. At the end of panniculectomy, both groups received 1% lidocaine with epinephrine. Z-group received an additional 14ml of ZynrelefTM at the panniculectomy site. ZynrelefTM was applied over the fascia using an applicator and a subcutaneous fat flap was created from the cephalad incision to separate it from skin closure. All the patients were discharged on oral acetaminophen. Opioids were prescribed based on pain control at discharge. Demographics, analgesic use, operative time (minutes), and duration of hospital stay (days) were recorded. Outcomes included hourly postoperative pain scores using a Visual Analog Scale, in-hospital opioid use in oral morphine equivalents (OME), and post-discharge opioid prescriptions. This study did not involve any off-label use of the medication. Results: Mean age (45.5±9.8 vs. 47.3±10.8 years), BMI (26.3±1.8 vs. 25.3±2.37 kg/m2), and operative time (254.2±80.9 vs. 219.4±83.6 minutes) (p>0.05) were similar in Z and Non-Z groups respectively. History of opioid use was present in 11.8% and 23.5% of the patients in the Z and Non-Z groups (p>0.05), respectively. 35.3% and 41.1% underwent brachioplasty, thighplasty, or incisional ventral hernia repair with panniculectomy in Z and Non-Z groups. Z compared to the Non-Z group had a shorter hospital stay (0.17±0.39 vs. 0.88±0.48 days, p=0.0002) and fewer overnight stays (17.7% vs. 82.3%, p=0.0004). Z group had a lower mean OME on postoperative day#1 (0.005±0.002 vs. 2.071±4.7, p=0.03) and post-discharge opioid prescriptions (0% vs. 35.3%, p=0.02). All Z-group postoperative pain scores were lower but not statistically significant. There were no side effects with the use of ZynrelefTM. Conclusions: Adding ZynrelefTM decreases hospital stay and opioid usage after a panniculectomy. The local inflammation contributes to an acidified environment which hinders the ability of the local anesthetic to remain active.[2] The lower opioid usage observed could be attributed to the addition of meloxicam to decrease the local inflammation, which potentiates better and longer action of bupivacaine on the panniculectomy site. References: [1] Sherif RD, Lisiecki J, Waljee J, Gilman RH. Opioid Prescribing Habits and Pain Management Among Aesthetic Plastic Surgeons. Aesthetic Plast Surg. 2022;46(2):965-971. [2] Ottoboni T, Quart B, Pawasauskas J, Dasta JF, Pollak RA, Viscusi ER. Mechanism of action of HTX-011: a novel, extended-release, dual-acting local anesthetic formulation for postoperative pain [published online ahead of print, 2019 Dec 16]. Reg Anesth Pain Med. 2019;rapm-2019-100714.

Purpose Skeletal muscle injury (SMI) caused by trauma or surgery can result in permanent disability and loss of function. Physical therapy is the current standard of care for SMI, but long-term recovery of muscle strength has shown to be inadequate in severe SMI. Initial evidence suggests that in animals invasive mechanical stimulation can increase up to 3-fold the tetanic torque after SMI. Here, we hypothesize that mechanical stimulation can improve functional recovery after SMI by stimulating skeletal muscle regeneration and by mitigating fibrosis.

Methods A standard excisional muscle injury (8mm ∅) was created in the left Tibialis Anterior (TA) muscle of female adult (200-250 grams) Sprague Dawley rats (n= 10/group). Post-injury, animals were either followed up with no treatment (control group) or subjected to controlled mechanical stimulation of injured muscles (experimental group). Functional recovery was measured at 14 and 28 days post-injury (PID) by measuring the TA tetanic torque and the animals' endurance on a treadmill run. In addition, as a measure of not challenging physical activity, the distance traveled was measured by fixed-point observation using a camera. At PID 28, samples of injured TAs were processed for histology (Masson) and immunohistochemistry (CD31 marker for angiogenesis) to measure myocyte/fibrosis percentage composition.

Results At PID 28, the tetanic torque in the experimental group was significantly higher than in controls (73.1±18.7% of pre-injury baseline vs. 47.0±23.7%, p=0.014). The resistance to fatigue in the experimental group was significantly higher than in controls (88.7±24.9% of pre-injury baseline vs. 56.6±27.5%, p=0.014). Endurance on a treadmill run was also significantly higher in the experimental group compared to controls (97.8±58.5% of pre-injury baseline vs. 40.6±35.4%; P=0.016) at PID 28. Histological analysis of the TA muscle showed a smaller percentage of fibrosis with increased CD31 immunohistochemical staining.

Conclusion In rats, mechanical stimulation promotes improvement in functional recovery after skeletal muscle injury. Validation of these findings in large animal models and in humans might help develop novel treatments for patients with muscle injuries caused by trauma or surgery.

Introduction: Radiation-induced skin fibrosis is one of the main adverse effects of radiation therapy for cancer treatment. Radiation fibrosis syndrome is caused by the overactivation of TGF-B that promotes fibroblast that induces collagenases and breaks down type III collagen and replaces it with type I collagen. The inhibition of P53 and the accumulation of ROS are the main mechanisms of radiation-induced damage. To fully understand these mechanisms, we tested the use of our novel human skin perfusion model to recreate the damage caused by radiation and as a platform to test possible therapeutic and/or prophylactic treatments to prevent fibrosis.

Methods: We use our perfusion model, which consists of a human tissue sample recovered from abdominoplasty. We dissect the superficial inferior epigastric artery and cannulated it and perfuse it with special culture media. On the first day after cannulation, we exposed the skin to a single targeted irradiation dose of 20gy and 40gy and took punch biopsies on days 3, 6, 12, and 16 for histological and gene expression analyses. The histological samples were stained with H&E, and Masson´s Trichrome stain to determine the morphology changes and extracellular matrix deposition respectively. TUNEL and DAPI immunofluorescent staining was performed to analyze for apoptotic changes in the epidermis/dermis. Expression of inflammatory, fibrotic and apoptotic genes was analyzed by real-time quantitative PCR.

Results: The morphological changes in the skin were significant. The radiation-exposed skin started peeling compared to the control group. The H&E staining showed an increase of inflammation in the dermis along with epidermis/dermis separation as well as papillary dermis containing fibrin deposition, accumulation of inflammatory cells, reactive changes in the endothelial cells, and abundant necrotic keratinocytes. Masson's Trichrome staining revealed a increased deposition of extracellular matrix between the papillary and reticular dermis in the irradiated skin. The TUNEL and DAPI stain shows an increase in apoptotic cells in the radiation group that correlates with the damage induced by radiation. Gene expression analyses revealed upregulation of inflammatory and anti-apoptotic genes expression.

Conclusions: Our perfusion model was stable for 19 days and was able to recreate the radiation-induced deposition of collagen shown in the H&E and Masson´s trichrome as well as the radiation-induced damage in the TUNEL and DAPI staining. Our system is reliable and can be used as a platform for developing therapeutics to mitigate the effects of radiation on the skin

Purpose: In October 2012 an open-access multi-media digital cleft simulator was released in partnership with the philanthropic sector, academia, and industry. The purpose of this resource is to address global disparities in cleft surgery education and to provide an easily accessible surgical atlas for trainees nationally. Ranging from operative videos, self-assessment quizzes, and interactive diagrams, the platform provides the user with several learning opportunities. Disparities in Plastic Surgery training between high and low-to-middle-income regions impact operative outcomes and equitable access to proper craniofacial care. While the availability of operative resources may not be promptly mitigated, electronic educational programs in other surgical sub-specialties have been proven clinically effective. This report assesses the demographics, usage, and global impact of our simulator, in its 10th year since inception.

Methods: Usage data of the simulator over 10 years were retrospectively collected and analyzed. Data parameters included the number of users, sessions, countries reached, and content access.

Results: The total number of simulator new and active users reached 7,687 and 12,042. Since its inception, the platform has reached 146 countries on all continents. The countries with the most numbers were the United States, Brazil, India, Mexico, and China. The simulator was accessed an average of 172.9.0 ± 197.5 times per month. Low-to-middle-income regions accounted for 43% of these sessions - an increase of 8% from 2017. The mean session duration was 11.4 ± 6.3 minutes yielding a total screen time of 181,320 minutes (3,022 hours). The most accessed learning modules were unilateral cleft anatomy followed by the Mohler cleft lip procedure, Furlow (primary palate), and bilateral cleft anatomy. The most common device used was the desktop (83.4%), followed by mobile (12.3%), and tablet (4.3%).

Conclusion: A freely available, digital, multi-media simulator can provide valued educational resources in low-resource as well as high-resource regions of the world. Global utilization has been sustained after 10 years from inception with an increased presence in low-to-middle-income nations. Access to this resource was not restricted by socioeconomic barriers and was recurrently used. This suggests that future surgical simulators of this kind may provide sustainable training platforms to surgeons working in areas with limited access to resources as well as US trainees.

Introduction: Chatbots can be leveraged to execute a variety of tasks and answer questions in seconds. Recently, ChatGPT, an open access chatbot, has been used for various applications including writing essays, responding to emails, and helping with research questions. We had OpenAI's ChatGPT chatbot take the American Society of Plastic Surgeons (ASPS) In-Service Examination. The performance of the natural language processing artificial intelligence was evaluated and summarized.

Methods: OpenAI's 2023 ChatGPT February 13 release was used to answer questions from the 2022 ASPS In-Service examination using the answer key. Each question was asked in three separate structures using a new chat to remove any possible memory retention in the following standardized formats: 1) question in an open ended (OE) format that removed all answer choices, 2) question in multiple choice (MC) format with answer choices after the phrase "please select the correct answer:", and 3) question in multiple choice format with answer choices and prompted reasoning (MCE) after the phrase "please select the correct answer and provide an explanation:". A standardized rubric was applied to evaluate each strategy. The 2022 ASPS Norm Table was utilized to compare ChatGPT's performance on the examination to subgroups of plastic surgery residents. Question prompts with graphics or images in addition to free text were included. All questions that the examination committee subsequently removed prior to scoring due to possible content ambiguity or poor statistical performance were excluded from the analysis. Questions that solely required analyzing a graphic or image to correctly answer the question were also excluded. Each question was assigned to ASPS Education Network (EdNet) subcategories. Analysis included metrics on accuracy, concordance, and insight that were assessed by two independent reviewers and tabulated.

Results: A total of 238 questions were included in the final analysis. A total of 17 (7%) questions were indeterminate amongst all strategies employed (14 OE and 3 MC). Accuracy for OE questions, MC, and MCE were 57.6%, 54.9%, and 56.3%, respectively. Output and concordance were best for OE questions 100% (238/238) followed by 97.1% (231/238) for MCE and 96.2% (229/238) for MC. ChatGPT was insightful on 83% (197/238) of questions across all structured formats. When faced with a question relating to graphics or images (32/238) ChatGPT was accurate from 46.9% (OE) to 59.4% (MCE) of the time.

Conclusion: ChatGPT's performance on the 2022 ASPS In-Service Examination was assessed. Based on ChatGPT's overall accuracy using the OE methodology, it scored better than more than half of first year integrated residents (52 percentile). ChatGPT performed the best using the OE strategy. This is an early release of the artificial intelligence software. Similar to a typical resident undergoing training who benefits from additional expertise gained from each year of residency, it will be interesting to trend the performance of ChatGPT on the In-Service Examination with improvements in future updates. With the release of new chatbots and advances in the technology, plastic surgeons should explore their capabilities with caution.

Neuromuscular control of bionic arms has constantly improved over the past years, however, restoration of sensation of the lost limb remains elusive. Previous approaches to reestablish sensory feedback include tactile and electrical stimulation of remnant skin as well as peripheral nerve stimulation. However, none of these modalities have reached daily clinical practice in part due to their inability to create natural, intuitive sensation. Here, we establish an experimental biological sensorimotor interface and demonstrate its potential use for prosthetic control and sensory feedback (Figure 1). In rats, a mixed nerve was transferred to a skeletal muscle combined with the transplantation of a glabrous dermal skin graft on top of it, thus forming a bi-directional communication unit. Morphological analyses indicated reinnervation of the transplanted skin, neuromuscular junctions and muscle spindles. Furthermore, sequential retrograde labeling revealed specific sensory reinnervation of the skin at the level of the dorsal root ganglia, supporting the morphological findings. The results indicate the possibility of surgically creating an interface for both decoding efferent motor control as well as encoding afferent tactile and proprioceptive feedback in a compact reinnervated muscle-skin unit using reinnervated end organs as robust translators. This pre-clinical evidence may indicate the way forward regarding clinical translation of biological bi-directional communication pathways for neuroprosthetic applications.

Purpose: Miniature soft breast implants for pre-clinical research are essential for advancing the field of plastic surgery. However, obtaining such custom implants from commercial implant manufacturers has becoming increasingly difficult in the current environment of company cost-cutting and decreased resources for sponsored research. As an alternative, researchers can fabricate their own custom implants allowing for faster turnaround times and reducing the reliance on external entities. Here we present a straightforward and inexpensive method for the fabrication of custom miniature implants with a separate silicone shell filled with silicone gel for high throughput pre-clinical investigation.

Methods: Polydimethylsiloxane (PDMS) was mixed in a 20:1 silicone to curing agent ratio, placing it in a CAD-designed inverted-mandrel negative mold. A vacuum degassed the PDMS at 37°C and cured until reaching a honey-like viscosity. The mold was spun using a Dremel to create a 2.5-3.5mm smooth implant shell, which immediately underwent a second curing step utilizing a heat gun fanning ~455°C air for 5s. A 1mm indented mold of the flat implant shell segment was covered with non-cured PDMS, and the rounded implant segment was placed on top of the PDMS. The PDMS cured around the rounded implant shell at 37°C overnight, creating an airtight soft pliable implant. The curing time and curing temperature determined the shell firmness. A 16G catheter was slipped through the flat implant segment and breast implant gel harvested from a commercial smooth implant was injected. The implant was sealed with a layer of PDMS and cured overnight at 37°C. Textured implants can be created using a salt-loss technique, and other implant features can be added to the design including suture tabs. Implants were stress tested using a lab shaker and a 46g sphere striking the implant 300,000 times at 37°C over the course of 7 days. One teardrop-shaped implant was placed under the panniculus carnosus muscle on the dorsa of Sprague-Dawley rats bilaterally. Three anchoring Polysorb sutures were placed through the suture tabs anchoring the implant to the underlying tissue to ensure short-term implant stability counteracting the loose skin rat model. Implants were explanted at 1-month.

Results: Implants recapitulated the small mold structures and stress testing did not cause shell cracking or gel extrusion. Implant shell firmness was tunable by changes in curing temperature, curing time, and filling materials used. Breast implant profile was modifiable by amount of implant gel injected. To ensure the safety and effectiveness of the new implants, testing was conducted on rats. At 1 month, implants were found to be safe and retained its shape during the testing period.

Conclusions: This method for fabricating custom miniature implants allows researchers to quickly modify implant designs based on their findings, thereby enabling high throughput pre-clinical investigation. This approach allows for the tunable fabrication of implant shell thickness, firmness, and implant forward projection while recapitulating intricate surface topographies as small as 1mm in size. This reduces the reliance on external sources for implant fabrication and implant design, which can be completed and implanted within a few days.

BACKGROUND Medical students who attend institutions without plastic surgery residency programs are at a significant disadvantage in the plastic surgery match. We developed an educational program for medical students without home programs called "Explore Plastic Surgery" to provide an overview of the steps towards a career in plastic surgery. The purpose of this study was to assess the impact, utility, and success of the novel program.

METHODS Pre- and post-event surveys were distributed to participants. Survey data were analyzed including participant demographics, perceptions of barriers unique to those without home programs, and the overall event utility.

RESULTS Two hundred eighteen medical students registered for the Explore Plastic Surgery program. Ninety-five participants completed the pre-event survey (44%) and of those, 57 participants completed the post-event survey. There was a significant increase in understanding of the steps towards a career in plastic surgery (p < 0.001), confidence in overcoming barriers unique to those without a home program (p = 0.005), and level of comfort in reaching out to faculty for opportunities (p = 0.01). There was a significant decrease in the perceived negative impact that attending a medical school without a home program will have on their abilities to pursue careers in plastic surgery (p = 0.006).

CONCLUSIONS Following the event, participants demonstrated an increase in their confidence in overcoming barriers and a decrease in their perceptions that attending an institution without a home program will negatively impact their ability to pursue plastic surgery. Initiatives focused on early exposure and recruitment of medical students are important to promote accessibility and diversity within plastic surgery.

Background: Artificial intelligence has entered many aspects of medicine, with applications growing daily. ChatGPT, a text-based artificial intelligence (AI) has been shown to solve natural language problems with surprising accuracy with preliminary studies demonstrating ChatGPT narrowly passing USMLE 1, 2, and 31. In this study, we challenged ChatGPT to take the annual 2022 ASPS In-Service Examination for Plastic Surgery residents. Methods: All 250 items in the 2022 American Society of Plastic Surgeons In-Service Exam were answered by listing the prompt: "answer the multiple choice question:" followed by the question listed verbatim. Answers were verified with the official answer key. Results: ChatGPT successfully answered 126 of 250 (50.6%) questions. However, seven questions were thrown out in the answer key, resulting in an adjusted score of 126/243 (51.9%). Many prompts included an image; for six of these prompts, ChatGPT requested the image in order to answer the question properly (even though it cannot process input images). With removal of these six questions, ChatGPT scored 53.2% (126/237). ChatGPT scored highest in Section 3: Craniomaxillofacial with 59.1% (29/49) and lowest in Section 5: Core Surgical Principles with 45.8% (22/48). Discussion & Conclusion: According to the published 2022 Norm Table, a 53% total test score ranked ChatGPT in the 26th percentile of first year integrated plastic surgery residents. This is a remarkable result given the ChatGPT's lack of specialized resources and training in plastic surgery. This study speaks to the power and abilities of current artificial intelligence, as well as the limitations and shortcomings of current resident evaluation methods. Furthermore, ChatGPT does not utilize the internet to answer questions and formulates answers by providing rationales in the text generated. Further studies are needed to determine the effects and applications of AI in collaboration with current physicians and physicians in training for future medical practice.

References: 1. Performance of ChatGPT on USMLE: Potential for AI-Assisted Medical Education Using Large Language Models. Tiffany H. Kung, Morgan Cheatham, ChatGPT, Arielle Medenilla, Czarina Sillos, Lorie De Leon, Camille Elepaño, Maria Madriaga, Rimel Aggabao, Giezel Diaz-Candido, James Maningo, Victor Tseng medRxiv 2022.12.19.22283643; doi: https://doi.org/10.1101/2022.12.19.22283643

Purpose: In recent years, there has been growing interest in the history of Black American plastic surgeons and their contributions to the field of Plastic and Reconstructive Surgery (PRS). Several recent studies have evidenced extreme disparities in representation of African Americans among plastic surgeons.(1) While significant forward-looking efforts are being made to identify and intervene in the factors perpetuating this demographic inequality, the history of how plastic surgery became one of the most ethnically segregated surgical specialties remains unexplored.

Methods: This project uses a variety of primary and secondary sources including archival documents, historical newspapers, obituaries, and oral histories, to identify the political and cultural factors that excluded Black practitioners from the first 50-years of American PRS professional societies. The historical framework explores how the political campaigns employed by early societies to legitimize plastic surgery as an independent specialty relied on the explicit association between PRS and the rehabilitation of disfigured soldiers. This strategy had the indirect consequence of barring Black surgeons from centers of PRS innovation. Until 1948, when President Truman signed Executive Order 9981 and officially desegregated the US Armed Forces, African American medical professionals were essentially ineligible to participate in the surgical reconstruction of injured veterans.(2) Ironically, integration of military veteran hospitals on the Homefront similarly disqualified Black surgeons from participating in PRS because the military refused to place Black physicians in positions of authority over White patients.(2)

Results: Despite these barriers, a handful of pioneers practiced plastic surgery without the support of the early professional societies. Dr. DeHaven Hinkson, the first Black surgeon to run a US Army station hospital, served in World War I and II and became an early advocate for plastic surgery in the psychological rehabilitation of veterans.(2,3) In 1930, Dr. H. Dodford Dismukes and his assistant, Dr. Joseph E. Brown, opened the largest private African American hospital in the nation with an emphasis on the surgical reconstruction of local workers injured in coal mining accidents.(4) In WW II, Dr. Walter Scott Brown became the first Black man to serve as flight surgeon for a primarily White military unit before dedicating his career to plastic surgery.(5)

Conclusion: The achievements of these groundbreaking figures are almost entirely absent from the PRS literature. By telling their stories, we can address the legacy of inequity and recognize these individuals as part of American PRS history.

1. Butler PD, Britt LD, Longaker MT. Ethnic Diversity Remains Scarce in Academic Plastic and Reconstructive Surgery. Plast Reconstr Surg. 2009;123(5):1618-1627. doi:https://doi.org/10.1097/prs.0b013e3181a07610
2. Marble S. "Separate, But Almost Equal": The Army's Negro Medical Field Units in World War II. J Natl Med Assoc. 2012;104(1-2):96-103. doi:https://doi.org/10.1016/s0027-9684(15)33513-6
3. Hinkson D. Problems in the Rehabilitation of Victims of War. J Natl Med Assoc. 1922;14(4):229.
4. Calvin FJ. Dr. Dismukes tells of his $75,000 Hospital. The Pittsburgh Courier. October 25, 1930:A1.

Background: Patient-reported outcomes (PROs), collected via patient-reported outcome measures (PROMs), are assessments of health status made directly by the patient. PROs are particularly important to capture in plastic surgery, as many procedures are aimed to improve patients' quality of life. More commonly used measures such as mortality, functional outcomes, or radiographic outcomes are inadequate as they do not directly measure quality of life. Although self-administration of PROMs is cost-effective and avoids potential for interviewer bias, it is unknown whether plastic surgery PROMs are written at or below the 6th-grade reading level per recommendation by the American Medical Association (AMA). This is important because an estimated one-fifth to almost half of Americans have low levels of print literacy. The aim of this study is to evaluate the readability of plastic surgery-specific PROMs.

Methods: We conducted a literature review to identify the most commonly used PROMs in plastic surgery. We included English PROMs that underwent both qualitative and psychometric testing. We extracted the text and instructions from these PROMs and analyzed readability. Our primary outcome of interest was the Simple Measure of Gobbledygook (SMOG) index, which detects word complexity. Our secondary outcome measures included the Flesch-Kincaid (FK), the Coleman-Liau index (CLI), and the Automated Readability index (ARI) which capture different aspects of readability. All readability measures report a score that corresponds to the grade level required to understand the text. We used paired t-tests to compare readability of the PROM questions and responses to readability of the instructions.

Results: We identified a total of 62 PROMs; 39 met our inclusion criteria. Of those, 12 (31%) were hand/peripheral nerve, 10 (26%) craniofacial, 5 (13%) aesthetic, 5 (13%) transgender, 4 (10%) lower extremity reconstruction, 2 (5%) burn, and 1 (2%) breast. The average SMOG index was 8.22 (SD=1.28), indicating a reading level above the 8th grade, and average FK=4.75 (SD=1.60), CLI=6.03 (SD=2.17), and ARI=3.34 (SD=1.65). Only 3 (8%) PROMs had a SMOG index at or below the 6th-grade level. For PROMs that included instructions (n=32), the text of the directions had a SMOG index of 9.88 (SD=2.53), while the text of the questions and responses had a SMOG index of 8.04 (SD=1.24). The PROM instructions had a significantly higher reading level than the questions/responses for all readability indexes (p<0.01 for SMOG, FK, CLI, and ARI).

Conclusions: PROMs used in plastic surgery, and especially the instructions for these PROMs, are written at a higher reading level than is recommended by the AMA. This may limit comprehension and accurate completion of PROMs, and therefore PROMs' validity and reliability for a large proportion of plastic surgery patients. PROMs should be better designed and developed to address the needs of low and limited-literacy patients.

Purpose: There are over 1,700,000 people living with limb loss in the United States alone and 185,000 new amputations per year. Nevertheless, little information is available regarding the effect of gender on management decisions following devastating injuries or patient care following amputations. Therefore, the primary objective of this study is to identify potential gender bias across a milieu of perioperative milestones for patients who undergo lower extremity amputation (LEA) and prosthesis-based rehabilitation. The secondary aim of this study is to provide a succinct overview of the prosthetic devices available to these patients to equip plastic surgeons with relevant knowledge to best care for their patients longitudinally.

Methods and Materials: A narrative review was conducted within PubMed and several commercial prosthesis websites in February 2023, without any date or geographic restrictions. Studies were screened and reviewed by two reviewers. Studies that evaluated topics in LEA, prosthesis, and rehabilitation were included and underwent descriptive analysis. Aggregate subgroup analyses were performed by gender.

Results: 109 studies met the inclusion criteria. Of these, 34 (31.1%) were retrospective cohort studies, 25 (22.9%) were cross-sectional studies, 14 (12.8%) were systematic reviews or meta-analyses, 10 (9.2%) were prospective cohort studies, 5 (4.6%) were qualitative interview studies, and 3 (2.8%) were randomized controlled trials. Gender bias was identified among several siloes of perioperative care preceding and following LEA and rehabilitation. On average, women undergo higher levels of amputation compared to men: approximately 39.0% of women underwent above-knee amputation compared to 20.0% of male counterparts, despite bearing a lower incidence of comorbid diabetes, in a retrospective review of 5,762 patients conducted by Kamrad et al. [1] Women also experience a 1.1 times higher odds of mortality following major amputation for peripheral vascular disease compared to men, despite similar disease severity. [2] With regards to prosthesis, women are significantly less likely to be successfully fit for prosthesis compared to men in the context of in-patient rehabilitation (29.5% vs. 42.7%, respectively; p <.001). [3] Women also report lower satisfaction with prosthetic fit (p =.030), comfort (p =.070), and appearance (p <.001) compared to male counterparts. [4] Finally, contemporary prosthetic devices are designed based on male models, limiting options that are compatible for women. Several qualitative analyses cite device weight, socket size, shoe size, and adjustable heel-height as limitations in applying the standard prosthetic device to the needs of women users. [5]

Conclusions: Gender bias was isolated across several indices for LEA and prosthesis, elucidating potential areas for reform. There is a need for gender-inclusive study populations and subgroup analyses to elevate the current landscape of patient-centered healthcare for women who experience lower limb loss. We recommend instating checkpoints at the following milestones to introduce intention and mitigate gender disparities: (1) amputation level (including advanced indices for identifying multiple levels of disease, such as Computed Tomography scan), (2) targeted post-operative pain control, (3) multidisciplinary task force with mental health support, (4) shared decision-making for prosthesis selection, (5) novel prosthesis design based on female models and with variable heel heights, and (6) elevated community and occupational re-integration.

Bibliography:

1. Kamrad I, Söderberg B, Örneholm H, Hagberg K. SwedeAmp-the Swedish Amputation and Prosthetics Registry: 8-year data on 5762 patients with lower limb amputation show sex differences in amputation level and in patient-reported outcome. Acta Orthop. Aug 2020;91(4):464-470. doi:10.1080/17453674.2020.1756101
2. Lo RC, Bensley RP, Dahlberg SE, et al. Presentation, treatment, and outcome differences between men and women undergoing revascularization or amputation for lower extremity peripheral arterial disease. J Vasc Surg. Feb 2014;59(2):409-418.e3. doi:10.1016/j.jvs.2013.07.114
3. Davie-Smith F, Paul L, Nicholls N, Stuart WP, Kennon B. The impact of gender, level of amputation and diabetes on prosthetic fit rates following major lower extremity amputation. Prosthet Orthot Int. Feb 2017;41(1):19-25. doi:10.1177/0309364616628341
4. Pezzin LE, Dillingham TR, Mackenzie EJ, Ephraim P, Rossbach P. Use and satisfaction with prosthetic limb devices and related services. Arch Phys Med Rehabil. May 2004;85(5):723-9. doi:10.1016/j.apmr.2003.06.002
5. Manz S, Valette R, Damonte F, et al. A review of user needs to drive the development of lower limb prostheses. J Neuroeng Rehabil. Nov 5 2022;19(1):119. doi:10.1186/s12984-022-01097-1

Background: Underrepresentation of women in plastic surgery remains a concern. Because research has found gender-related gaps in surgical trainee confidence, and because confidence influences others' perception of competence, this study investigates sex-specific differences in self-confidence and correlation with surgical ability in a way that can uniquely be done with simulation. This could inform strategies to mitigate disparity and ultimately work toward gender parity in plastic surgery.

Methods: Plastic surgery trainees were recorded performing up to three cleft lip repairs on a high-fidelity simulator. Demographic information was collected, and questionnaires were completed to assess procedural self-confidence and confidence with specific cleft lip skills. Videos were blindly rated by a single cleft surgeon using the Objective Structured Assessment of Technical Skills (OSATS), a global surgical skill scale, and the Unilateral Cleft Lip Repair competency assessment tool (UCLR), a procedure-specific checklist. Generalized Estimating Equations (GEE) modelling was used to estimate differences in demographic variables, questionnaire scores, and objective ratings between men and women, while controlling for multiple simulations per subject. Correlation between overall procedural self-confidence and performance was examined for men and women using Pearson R.

Results: Twenty-six participants completed 73 simulated procedures. Six participants were women and 20 were men. There was no gender-related difference in training level, volume of prior experience, speed with which participants completed each simulated procedure, comfort level with specific skills, or objective performance measured by OSATS and UCLR. However, a significant difference was found between men and women in overall procedural self-confidence, with female trainees rating their confidence lower (Mean = 16.9, SD = 4.3) than male trainees (Mean = 19.4, SD = 3.8); p = 0.048. Examination of individual questions revealed that women scored significantly lower on questions related to anxiety (p=0.007) and desire to avoid the procedure altogether (p=0.001). Further analysis found that confidence scores correlated more strongly with objective performance on UCLR for females (r = 0.83, p<0.001) than for males (r = 0.45, p<0.001).

Conclusion: This study shows that as a group, female plastic surgery trainees had lower self-confidence than their male counterparts despite demonstrating the same level of skill performing unilateral cleft lip repair on a high-fidelity simulator. However, on an individual level, confidence and ability were more closely linked for women than for men. This suggests a uniquely impactful role for mentoring focused on concrete skill building to help close the confidence gap between men and women in plastic surgery.