Scientific Abstract Presentations: Residents Craniomaxillofacial, Hand and Research & Technology Abstracts Session 6 Friday, October 16 Fri, Oct 16 10:30 a.m.-12:30 p.m. CDT

Purpose: Craniosynostosis affects approximately 1 in 2,000 infants. Earlier diagnosis may expand management options, including consideration of minimally invasive strip craniectomy in appropriate candidates, and can facilitate timely monitoring of potential complications such as papilledema. As the smartphone photograph is often the first piece of imaging available to clinicians evaluating an infant's head shape, we explored whether off-the-shelf commercial multimodal large language models (MLLMs) could enhance screening at the primary-care level. We benchmarked four state-of-the-art systems on ability to detect craniosynostosis from clinical photographs, and subtype classification.

Methods: A set of 189 hand-cropped infant photographs was assembled from 68 published open access peer-reviewed articles representing 101 patients: 59 with non-syndromic craniosynostosis (16 sagittal, 15 metopic, 28 coronal), 9 positional plagiocephaly, and 33 normal controls. Each patient's photograph set was submitted as one query to four flagship MLLMs (GPT-5.5, Claude Opus 4.7, Gemini 3.1 Pro, Grok 4.20). Two analyses were performed: binary affected-versus-normal classification (n=101), and subtype classification (sagittal vs. metopic vs. coronal vs. positional, n=68). Sensitivity, specificity, accuracy, and 95% Wilson confidence intervals were computed. Pairwise model differences were tested with McNemar's exact test, with Bonferroni correction.

Results: For binary classification, model performance split into two distinct profiles. Claude (sensitivity 92.6%, specificity 51.5%) and Gemini (79.4%, 69.7%) over-called pathology, prioritizing case-detection at the cost of false positives. GPT (36.8%, 75.8%) and Grok (51.5%, 72.7%) defaulted toward "normal", prioritizing specificity at the cost of missed cases. For subtype classification, Gemini achieved 54.4% accuracy [95% CI 42.7-65.7], Claude 44.1%, GPT 38.2%, and Grok 33.8% (Cochran's Q p=0.034). 19.8% of GPT submissions were refused despite repeated retries likely due to content moderation. For binary classification, Claude and Gemini significantly outperformed GPT and Grok in overall accuracy (p<0.001). No model achieved both sensitivity and specificity above 85%.

Conclusions. Commercial MLLMs are not ready for unsupervised craniosynostosis screening. The best-balanced model misses one in five affected infants and false-alarms three in ten healthy infants. Inter-model variation is substantial; clinicians and parents should recognize that an LLM's response is highly platform-dependent, with some favoring false positives (Claude, Gemini) and others to false negatives (GPT, Grok). Content moderation around infant facial photography is an underappreciated barrier for OpenAI-based applications. A craniofacial-specific fine-tuned model with human review is required before photographic AI triage can shorten time-to-diagnosis.

Positive Screening Rate for Obstructive Sleep Apnea in the Cleft Palate Population after Buccal Flap Reconstruction

Background and Purpose Patients with craniofacial anomalies are considered high risk for developing obstructive sleep apnea. As such, patients with cleft palates are encouraged to be screened for signs and symptoms of obstructive sleep apnea by their pediatrician. Given our center's surgical expertise with the Furlow palatoplasty using double-opposing z-plasty buccal flap reconstruction, we seek to report on the positive screening rate for sleep apnea in this patient population in comparison to other surgical techniques.

Method/Description Inclusion criteria involves patients seen at our multidisciplinary cleft clinic who have undergone operative repair of their cleft palate and have not been diagnosed with sleep apnea prior to 2019. Patient stratification was based on the initial surgical technique used. Starting in 2019, all patients were screened for obstructive sleep apnea using the 2012 guidelines from the American Academy of Pediatrics. Patients who screened positive for signs and symptoms of obstructive sleep apnea were subsequently referred for polysomnogram. Chi-squared analysis was used to test for significant differences between rates of positive screening results using a p-value of 0.05.

Results A total of 169 patients met inclusion criteria for our study. 61.54% (n=104) of these patients received Furlow palatoplasty using a double-opposing z-plasty with buccal flap reconstruction. The positive screening rate for obstructive sleep apnea in this population was 22.12% (n=23). This is in comparison to 24.62% (n=16) of patients who received other surgical repairs for their cleft palate. While patients undergoing Furlow palatoplasty with double-opposing z-plasty buccal flap reconstruction screened positive for signs and symptoms of obstructive sleep apnea more frequently, this difference was not statistically significant (p=0.7075).

Tonsillar and adenoidal hypertrophy was present in 27.27% (n=15) and 16.36% (n=9) of patients who underwent buccal flap reconstruction, respectively. This is contrasted with 14.29% (n=5) and 08.57% (n=3) of patients who received repair of their cleft palate utilizing a different surgical technique.

Conclusion Based on our initial results, there is no significant difference in postoperative signs and symptoms of obstructive sleep apnea based on operative technique. However, additional investigation comparing polysomnogram proven obstructive sleep apnea is warranted.

Introduction: Mandibular fractures can be highly prone to complication, with rates in the literature ranging from 13% to as high as 58% in some studies. Unhoused and incarcerated individuals are at higher risk for trauma and face unique challenges after mandibular fracture, including poor access to soft or pureed food necessary after intermaxillary fixation, limited access to antibiotics and dental hygiene which may increase the rate of infection, and potential for discharge to an unsafe environment where there is risk for re-fracture. Trauma centers often treat many unhoused and incarcerated patients, but currently little data exists to guide treatment or delineate the risks in these populations. In this study, we examine the relationship between housing status and mandibular fracture outcomes to inform treatment guidelines and better support these patients in their recovery.

Methods: A retrospective chart review was conducted of patients presenting to our institution between January 2010 and December 2025 with a diagnosis of acute mandibular fracture. Charts were analyzed for demographic information including mention of homelessness or active incarceration. Information was collected regarding fracture details, concurrent injuries, mechanism, comorbidities, interventions, complications, reinterventions, and follow-up. Statistical analysis was performed using ANOVA and chi-squared analysis.

Results: A total of 400 patients with mandibular fractures were analyzed, of whom 20 (5%) were unhoused and 17 (4.25%) were incarcerated at the time of their injury. Unhoused patients were more likely to use tobacco (75% vs 29% incarcerated and 43% secure) and IV drugs (15% vs 0% incarcerated and 3% secure; p < 0.05). Incarcerated patients were more likely to present with other concomitant facial fractures (58% vs 24% secure and 30% unhoused; p < 0.05). There were no differences between the populations in terms of race, BMI, injury mechanism, fracture pattern, comminution, or malocclusion. There were no differences in how fractures were initially managed in these populations, with similar rates of antibiotic prescription, inpatient admission, operative intervention, MMF, and ORIF. Although the three populations had similar rates of treatment with MMF, incarcerated and unhoused patients were less likely to undergo formal MMF removal (p < 0.05). Unhoused patients were less likely to attend follow-up appointments (10% vs 59% incarcerated and 67% secure; p < 0.05) and both incarcerated and unhoused patients were less likely to achieve discharge after sufficient treatment (7% and 35% respectively, vs 65% secure p < 0.05). Although there was no difference in overall complication rate between the three populations, no incarcerated individuals underwent revision surgery versus 40% of secure and 42% of unhoused patients (p < 0.05).

Conclusion: Unhoused individuals are less likely to attend follow-up appointments after mandible fracture and less likely to achieve discharge after sufficient treatment. Incarcerated individuals, who rely on prison coordination for transportation and medical care, attended initial appointments but were less likely to have enough follow up to achieve discharge after sufficient treatment. Despite similar overall complication rates, incarcerated patients did not undergo surgical intervention for complications at comparable rates to other patients, possibly as a result of this truncated follow up.

Background: Plastic surgeons frequently work together with pediatric dentists, orthodontists, and speech language pathologists in multidisciplinary teams to care for pediatric craniofacial patients. This multidisciplinary approach has promoted proficiency in speech screening among providers on these teams. However, such resources are not part of routine dental care. Moreover, traditional dental training provides limited education surrounding speech screening. We hope to generalize the knowledge gained from multidisciplinary teams to fill this gap for general dental practitioners, who are often asked to comment on speech development by their patients. This project aims to assess the knowledge, comfort level, and current practices of pediatric dental and orthodontic specialists regarding speech screenings and to determine the need for a standardized speech screening tool in routine pediatric dental care.

Methods: An 18-question survey was disseminated across multiple networks, including the New York State Third District Dental Society, two pediatric dental residency programs, three state dental societies and dental groups on Facebook, and later via a national email blast to all members of the American Academy of Pediatric Dentistry, using both Survey Monkey and Qualtrics surveys. 

Results: Results across 132 responses indicate that 67.1% of dental professionals feel that speech screening is within the scope of their practice, with newer graduates endorsing higher rates of obligation to screen for speech. However, on a ten-point scale, mean self -rated provider knowledge of speech development is reported on average at 4.3 on a ten-point scale. Over 87% of providers may be interested in utilization of a standardized speech screening tool, 69.9% of which prefer this tool to utilize no more than 1–2-minutes of a routine pediatric dental visit. 

Conclusions: Based on study results, pediatric dentists and orthodontists recognize the importance of speech screening in their practice. However, they report low comfort levels with conducting screening, which could be improved with a standardized speech screening tool. These providers would favor a method that takes up to two minutes during their routine examination. 

Introduction: Selective serotonin reuptake inhibitors (SSRIs) are among the most used antidepressants owing to their efficacy and safety profile. Serotonin has been implicated in platelet aggregation and vasoconstriction as well as immune cell activation and cytokine production. Serotonergic antidepressants may impair these pathways, potentially increasing the risk of perioperative bleeding. While this association has been characterized in other surgical populations, data regarding the safety profile of SSRIs specifically in head and neck free flap reconstruction remain limited.

Methods: A retrospective cohort study was conducted using TriNetX, a global federated EHR database. Patients aged ≥ 18 years undergoing head and neck free flap reconstruction for malignant neoplasms of the lip, oral cavity, or pharynx were identified using ICD-10 and CPT codes. Patients with trauma-related ICD-10 codes within one year of reconstruction were excluded. Cohorts were stratified by SSRI exposure (documented prescription within 90 days prior to surgery). Propensity score matching was performed for comorbidities influencing bleeding risk. Primary outcomes included hematoma, hemorrhage, and blood transfusion. Secondary outcomes included wound dehiscence and surgical site infection (SSI) at 30, 60, 90, and 365 days.

Results: After propensity matching, 2,094 patients were included (n = 1,047 per cohort). At 30 days, no significant differences were observed between the SSRI-exposed and unexposed cohorts regarding hematoma/hemorrhage (RR = 1.08; P = .84), transfusion (RR = 1.04; P = .83), SSI (RR = 1.32; P = .25), or wound dehiscence (RR = 1.21; P = .25). These findings remained consistent across all timepoints up to one year postoperatively.

Conclusion: In patients undergoing head and neck free flap reconstruction for malignancy, SSRI use was not associated with an increased risk of hemorrhage or adverse wound outcomes. These data do not support the routine preoperative discontinuation of serotonergic antidepressants in this population

Background: Free anterolateral thigh flaps are widely used for pharyngoesophageal reconstruction; however, monitoring these buried flaps remains challenging. Externalized skin paddles are commonly used but are prone to false-positive signals that often reflect localized perfusion insufficiency rather than true pedicle failure, leading to unnecessary surgical re-exploration. To address this limitation, we implemented a dual-monitor skin paddle design in which two independent paddles serve as internal diagnostic controls. This study evaluates whether this strategy improves diagnostic accuracy and reduces unwarranted re-exploration. Methods: We retrospectively reviewed patients undergoing pharyngoesophageal reconstruction with anterolateral thigh flaps between June 2018 and March 2025 at a single institution. Patients were categorized into single-monitor and dual-monitor groups. The primary outcome was unplanned re-exploration for suspected vascular compromise. Secondary outcomes included monitoring alarm events, flap survival, and hospital length of stay. Results: A total of 43 patients were included, with 18 in the single-monitor group and 25 in the dual-monitor group. Baseline and operative variables were comparable between groups. Clinical alarm events occurred in 34.9% of patients. In the single-monitor group, two patients required emergent re-exploration despite patent anastomoses, indicating false-positive signals; one resulted in partial necrosis and one in total flap loss requiring salvage reconstruction. In contrast, no patients in the dual-monitor group required re-exploration. All alarm events were confined to a single paddle, while the second remained well perfused. These discordant findings, termed the "black–pink" phenomenon, were successfully managed with observation, and all isolated paddle necrosis healed with secondary intention. Overall flap survival was 97.7%, with no significant difference between groups. The dual-monitor group demonstrated a significantly shorter hospital stay, with a median of 22 days compared with 28 days. Conclusion: Dual monitoring skin paddles provide a clinically actionable framework for interpreting postoperative flap signals. Discordant findings, termed the "black–pink" phenomenon, indicate localized perfusion compromise rather than global pedicle failure, allowing safe observation and avoiding unnecessary re-exploration. Analogous to the traditional use of divination blocks, this discordant pattern represents an affirmative signal that the primary flap remains viable. By integrating two independent observations, this approach enhances diagnostic confidence and refines clinical decision-making in buried free flap reconstruction.

Background: Perforator flaps have offered new locoregional options to reconstruct soft tissue defects. Perforator flaps from supraclavicular and upper thoracic regions have been valuable for head and neck (H&N) reconstruction in particular. Their vicinity to the H&N confers two major advantages: they can reach the reconstructive site without microsurgery and are a great color match in resurfacing indications. The delto-acromial artery perforator (DAAP) flap is a less well known, recently described option. This study aims to explore its utility in various indications for reconstruction. Methods: Patients who received a DAAP flap for H&N reconstruction across two institutions from 2018-2023 were reviewed retrospectively. Demographics, indications for surgery, defect size, operative time, and complications were noted. For external flaps with adjacent native skin, we performed a colorimetric analysis comparing the DAAP flap skin paddle to the surrounding skin by calculating the International Commission on Illumination Delte E 2000 (dE2000) score, with a lower score indicating better color match. Results: There were 19 patients included in this case series. In 13 cases the flap was pedicled, and used for closure of trachea-esophageal or pharyngo-esophageal fistula (n=11) as well as neck resurfacing (n=2). However, the DAAP flap was also used as a free flap in 6 cases for nasal (n=3), cheek (n=1), eyelid (n=1), and forehead (n=1) skin replacement. At least one reliable perforator was found proximally emerging from the delto-acromial pedicle under the clavicle (n=8) or distally lateral to the delto-pectoral groove (n=11). The average defect size was 28.3±18.1 cm2, with the DAAP flap providing 34.6±17.6 cm2 of coverage on final inset. Operative time ran from 123 minutes to 829 minutes with pedicled flaps averaging 369.4±123.6 minutes and free flaps averaging 438.8±223.0 minutes. Most flap donor sites were closed primarily (n=17). There were no cases of flap loss; a second flap was necessary in one case for fistula recurrence. Photographs were able to be obtained for four patients who were at least two years out from surgery. Analysis showed low dE2000 values for forehead (2.54), nasal (2.94 and 3.00), and eyelid (5.58) reconstruction. Conclusion: The DAAP flap is a thin, pliable skin flap with reliable perforators, acceptable pedicle length, and great color match in H&N reconstruction. The flap is robust enough to address fistulas in H&N cancer patients and is especially valuable in vessel-depleted necks. It is also delicate enough for eyelid and nasal reconstruction and presents the enormous advantage of being the same color as the facial skin. It can be combined with other flaps (e.g. radial forearm flow-through flap) for reconstruction following total rhinectomy. This study represents the largest cohort of DAAP flap use in H&N reconstruction to date. As patients with H&N cancers often require adjuvant chemotherapy and radiation that can jeopardize first-line reconstructive choices, reconstructive surgeons need other options for H&N reconstruction, and the DAAP flap serves as a versatile soft tissue coverage option in the H&N region.

Introduction: In low-middle income countries (LMICs), limited access to speech-language pathologists (SLPs) and delayed diagnosis of velopharyngeal insufficiency (VPI) contribute to language delays, communication barriers, and stigma for patients with cleft lip and palate (CLP). Artificial intelligence (AI) offers a potential solution by analyzing speech recordings from a smartphone application, trained on cleft-focused SLP evaluations, to detect errors and guide triage for therapy or surgical referral when previously inaccessible. This pilot program explores the feasibility of implementing an AI-enabled speech assessment tool across Operation Smile programs in Latin America.

Methods: Standardized speech recordings and palatal photographs were collected from children aged 4 or older with repaired or unrepaired CLP and from controls. Using a smartphone application, participants repeated 16 standardized words targeting key phonemes, preselected by SLPs, to identify key phonetic metrics associated with VPI. Demographics and cleft characteristics were recorded. Five cleft-trained SLPs from the patient's country independently evaluated each sample using ASHA-based criteria to grade VPI and identify articulation errors. Based on these evaluations, SLPs recommended triage needs, including speech therapy, surgical referral, or both. Annotated recordings with SLP assessments will be used to train an AI algorithm to detect VPI, grade severity, and guide triage.

Results: To date, 140 participants aged 4-18 with repaired or unrepaired CLP and one control have been enrolled. Standardized speech recordings, palatal photographs, and demographic data were collected, and cleft-focused SLPs are independently reviewing recordings to create annotated datasets for algorithm training.

Conclusion: This pilot project demonstrates the feasibility of using a smartphone application to collect standardized speech samples for AI-based evaluation in low-resource settings. Ongoing work will support development of an AI algorithm to detect VPI, assess severity, and guide triage for speech therapy intervention or surgical referal across diverse global care settings.

Introduction: Orbital decompression is most often performed to relieve thyroid exophthalmos. Orbitalization of the ethmoid and maxillary sinuses is a mainstay of this surgical procedure. The volume that is gained, however, may be insufficient when proptosis exceeds 5.0-6.0 mm or when the sinus cavities are diseased or underdeveloped. Additional orbital volume has been obtained by techniques that remove or reposition constituents of the lateral orbital wall. However, the effectiveness of these measures in achieving predictable volume increases is controversial. A predictable method for increasing orbital volume is needed to suit the varying needs of patients having extreme proptosis and those with recurrent proptosis after decompression. We studied the effectiveness of translocation of the contiguous temporalis muscle combined with removal of various components of the lateral wall as a means of adding volume to the orbit.

Methods: Ten orbits of 7 fresh cadavers were exenterated, and in each case, the resulting cavity was filled with modeling clay. The clay was removed and placed into a calibrated graduate that was partially filled with water. The increase in water volume was determined, and the orbital volume was recorded. Orbital volume was measured after three different steps of decompression.

Step 1: The lateral orbital wall, consisting of the zygoma and greater wing of the sphenoid to the inner table of the middle cranial fossa was removed. Step 2: A portion of the full thickness of the temporalis muscle was removed. The defect in the muscle corresponded in width and height to the defect in the lateral orbital wall. Step 3: The lateral orbital rim was removed from the fronto-zygomatic suture to the inferior orbital fissure.

Results: The volume of the orbital cavity was 27-30 ± 1.2cc. Medial decompression (ethmoidectomy) produced a gain in volume of 3.0 ± 0.4 cc. Simple removal of the lateral orbital wall (Step 1) produced a gain of 1.4 ±0.4 cc. Temporalis myectomy (Step 2) increased the gain in volume to 3.3±0.6 cc. Orbital rim removal (Step 3) increased the gain to 6.0±1.2 cc.

Discussion: In a previous pressure-volume study of decompression, lateral wall decompression was found to produce the least volume gain and was stated to be an inferior technique. To create a more useful gain in orbital volume, we removed the portion of the temporalis muscle lying immediately lateral to the bony defect (Step 2). Our findings show that this step produces a combined gain in volume equal to that produced by standard medial wall decompression. To create a further gain in volume, we removed the lateral orbital wall (Step 3). The combination of all three steps gave a volume gain as great as 6.0 cc. An understanding of the usable space contributed by each of the methods should allow better matching of proptosis and tissue elasticity with the available decompressive options. The technique that we describe has the potential to provide a greater gain of volume, with less operative risk, as we are seeing in the clinical setting at our institution already.

Dr. Robert Acland passed away before the publication of this work; we gratefully acknowledge his contributions.

Purpose Adequate treatment of hand abscesses is paramount to avoid poor outcomes and permanent deficits (1, 3, 4). Treatment typically involves debridement, antibiotics, elevation, and splinting, yet no consensus exists regarding the optimal setting for debridement (2, 5). This study analyzes patient outcomes following initial debridement of hand abscesses at the bedside versus in the operating room. We hypothesize that initial bedside debridement results in increased complication rates and need for further debridement.

Methods A seven-year retrospective chart review was conducted between 2017 and 2023. Patients 18 years and older presenting with a soft tissue abscess distal to the elbow requiring plastic, orthopedic, or general surgery consultation at a single tertiary care center were included. Data included patient demographics, abscess characteristics, setting of initial debridement, length of stay, number of debridements, complications, and readmission. Bedside procedures were performed by residents; attendings supervised all operative procedures. The primary outcome was need for further debridement. Secondary outcomes included length of stay, readmission, and complications including osteomyelitis and amputation. Chi-squared and t-tests were used for comparisons; p<0.05 was considered significant.

Results 279 patients were included; 95 (34.1%) underwent initial operative debridement and 184 (65.9%) underwent initial bedside debridement. 50 patients (17.9%) required repeat debridement. Only 5 patients (5.3%) in the operative group required repeat debridement, compared to 45 patients (24.5%) in the bedside group - a statistically significant difference (p<0.0001). Patients who underwent initial bedside debridement were more likely to develop osteomyelitis (4.35% vs. 0%), though this did not reach statistical significance (p=0.054). Of those who developed osteomyelitis, 75% ultimately required amputation. Readmission rates were similar between groups (8 bedside vs. 4 operative, p=1.00). Mean length of stay was slightly longer in the operative group (3.99 days) versus the bedside group (3.58 days), though this was not significant. Among high-risk subgroups, diabetic patients and IV drug users who underwent initial bedside debridement were significantly more likely to require repeat debridement compared to those initially debrided in the operating room. This pattern was not observed in smokers.

Conclusion Initial bedside debridement of hand abscesses is associated with significantly increased rates of repeat debridement and higher rates of complications such as osteomyelitis and amputation compared to operative debridement. Initial operative debridement may therefore improve outcomes, particularly in high-risk patients such as diabetics and IV drug users. While logistical constraints may limit universal access to the operating room, patient risk factors should guide management decisions.

References 1. Koshy JC, Bell B. Hand Infections. J Hand Surg Am. 2019 Jan;44(1):46-54. 2. Jennings JD, Vroome C, Ly JA, Thoder J. Initial Debridement of Dorsal Hand Abscesses in the Operating Room Does Not Improve Outcomes. Hand (N Y). 2020 Nov;15(6):858-862. 3. Teo WZW, Chung KC. Hand Infections. Clin Plast Surg. 2019 Jul;46(3):371-381. 4. Gundlach BK, Sasor SE, Chung KC. Hand Infections: Epidemiology and Public Health Burden. Hand Clin. 2020 Aug;36(3):275-283. 5. Watkins RR, David MZ. Approach to the Patient with a Skin and Soft Tissue Infection. Infect Dis Clin North Am. 2021 Mar;35(1):1-48.

Introduction Open metacarpal fractures are often fixed with Kirschner wires (K-wire) due to their contaminated or dirty nature. K-wires are associated with infection rates of 7-25%1 and typically do not allow early motion protocols, as they are non-rigid constructs. Intramedullary (IM) screw fixation provides minimal soft-tissue disruption and favorable functional outcomes in closed injuries2, but its role in open fractures remains unclear. This study compares infection, union, and functional outcomes between percutaneous K-wire pinning and IM screw fixation in open metacarpal fractures.

Methods A retrospective study was conducted on patients who underwent fixation of open metacarpal neck or shaft fractures at a single Level I trauma center between January 2024 and January 2026. Demographics, injury, surgical, and clinical data were collected for each fracture. Thumb fractures, intra-articular fractures, fracture-dislocations, and patients with less than 6 weeks of follow-up were excluded. Univariate logistic regression assessed associations between patient and injury characteristics with fixation method.

Results A total of 18 patients with 30 metacarpal fractures were included; 23 were treated with IM screws and 7 with K-wires. Clinical union was achieved in 100% of IM screw fractures compared to 85.7% (6/7) of K-wire fractures (p=0.23). Total active motion (TAM) did not differ significantly between IM screw (n=8, mean 141.6°, SD 29.4°) and K-wire (n=2, mean 155.0°, SD 35.4°) fixation (p=0.56). Three complications occurred: one superficial infection in the IM screw group (successfully treated with oral antibiotics), one non-union in the K-wire group (revised to IM screw with subsequent union), and one extensor tendon tenolysis. Univariate logistic regression showed no significant associations with fixation method for age (OR 1.05, 95% CI 1.00-1.10), sex (OR 0.47, 95% CI 0.05-4.47), smoking (OR 0.28, 95% CI 0.05-1.65), mechanism of injury (OR 1.20, 95% CI 0.66-2.18), contamination level (OR 0.28, 95% CI 0.05-1.42), or mangled hand status (OR 3.25, 95% CI 0.52-20.37).

Conclusions Intramedullary screw fixation appears safe and effective for open metacarpal fractures, with a lower complication rate (4.4%) compared to K-wire fixation (14.3%). All IM screw fractures achieved clinical union, and the single non-union in the K-wire group was successfully revised with IM fixation. Despite trends towards more severe injuries in the IM screw group - including higher-energy mechanisms, greater contamination, and mangled hand status - outcomes remain favorable.

Purpose: Hand injuries are common, yet their macroeconomic impact is rarely quantified. This study quantifies the macroeconomic burden of hand injuries in Canada and globally, identifying high-impact injury subtypes and populations most relevant to plastic surgery workforce planning and trauma care delivery.

Method: Disability estimates from the Global Burden of Disease (GBD) 2023 study were linked with World Bank Gross Domestic Product (GDP) data to calculate the Value of Lost Welfare (VLW) from 2009–2023 using income-adjusted value-of-statistical-life-year (VSLY) functions (1,2). Analyses were conducted globally and for Canada, stratified by injury subtype, age, and sex. Percent-of-GDP burden enabled international comparison. Multivariable linear regression assessed adjusted associations (3).

Results: Globally, VLW attributable to hand injuries increased by 95% over the study period (4.0% annually, p<0.05), with the steepest rise occurring between 2021-2023. Thumb amputations imposed a 9–10% higher economic burden than finger amputations, while distal hand injuries generated substantially lower losses. Males and adults aged 20–59 years accounted for most global economic burden. In Canada, VLW increased by 75% (4.1% annually, p<0.05). Thumb amputations imposed a 2.1-fold higher economic burden than finger amputations-highlighting the disproportionate national impact of thumb loss. Men experienced a 1.6-fold higher burden, with working-age Canadians driving most losses. Canada's percent-of-GDP burden (0.02–0.04%) was comparable to other high-income regions but varied markedly by injury subtype.

Conclusions: Hand injuries impose a growing and economically significant burden globally and in Canada; specifically, thumb amputations generate greater welfare-based economic loss, underscoring the importance of thumb preservation. The disproportionate impact of thumb amputations and injuries affecting working-age Canadians supports targeted investment in hand trauma systems and microsurgical reconstruction capacity to inform Canadian and global surgical workforce and health system planning.

REFERENCES (1) Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2023 (GBD 2023) Results. Seattle, WA: Institute for Health Metrics and Evaluation; 2025. (2) Ranganathan K, Singh P, Raghavendran K, et al. The global macroeconomic burden of breast cancer: implications for oncologic surgery. Ann Surg. 2021;274(6):1067-1072. doi:10.1097/SLA.0000000000003662 (3) Aldy JE, Viscusi WK. Adjusting the value of a statistical life for age and cohort effects. Rev Econ Stat. 2008;90(3):573-581. doi:10.1162/rest.90.3.573

Purpose

High-voltage electrical injuries (HVEI) of the upper extremity are uncommon yet devastating injuries that often require complex reconstruction or limb amputation.1 Despite improved burn care over recent decades, no standardized surgical protocol exists in the literature, and the optimal timing and type of intervention remains controversial. The aim of this review study is to describe the current literature on high-voltage electrical injuries involving the upper extremity and evaluate different management strategies for HVEI in order to identify a stepwise protocol to optimize limb salvage and minimize amputations rates.

Methods

A systematic review was conducted across three major databases: in accordance with PRISMA guidelines. Databases searched include PubMed, Scopus, and CINAHL from database date of inception through December 19, 2025. References were screened in Covidence. Extracted variables include demographics, timing, and types of interventions performed, treatment outcomes, and complications.

Results

Of the fourteen studies included, the mean age among those reported was 30.74 years (95% CI 20.78–40.69; SD 155.80). The mean total body surface area (TBSA) reported was 18.04% (95% CI 8.97–27.12; SD 79.52). Across all studies, 67.65% were male (95% CI, 32.781 to 93.935), 2.94% were female (95% CI, 0.950 to 5.968), and 24.84% had unspecified gender (95% CI, 1.459 to 63.531). Among studies reporting TBSA distribution, 96.34% of injuries involved 0–10% TBSA (95% CI, 92.353 to 98.906). The most common anatomical locations of high-voltage electrical injuries in the upper extremities were the hand (35.13%, 95% CI, 13.472 to 60.680), forearm (32.56%, 95% CI, 8.602 to 62.987), and wrist (15.62%, 95% CI, 2.107 to 38.259). Frequently reported interventions included free flaps (74.05%; 95% CI, 35.782 to 98.068), nerve grafting (68.74%; 95% CI, 11.095 to 99.801), fasciotomies (59.90%; 95% CI, 19.249 to 93.645), amputation (42.64%; 95% CI, 27.682 to 58.331), and local tissue rearrangement (34.98%; 95% CI, 24.460 to 46.702). Studies that included data regarding complications, the most common ones include mortality (9.99%; 95% CI, 1.412 to 25.028), infection (7.10%; 95% CI, 0.251 to 22.099), arterial thrombosis (4.91%; 95% CI, 0.934 to 11.771), myoglobinuria (3.84%; 95% CI, 0.0625 to 16.568), and contractures (3.71%; 95% CI, 0.0142 to 13.477). Additionally, 2.44% of patients required further procedures during follow-up (95% CI, 0.153 to 7.339). Using Begg's rank correlation test, there was no significant publication bias (Kendall's= -0.096, p=0.59), which was confirmed using Egger's regression test (intercept = 2.71, p = 0.63).

Conclusion

HVEI's to the upper extremity remain complex with variable management. Evidence on timing of interventions is limited, demonstrating the need for further data collection and evaluation of treatment for this population. A standardized, stepwise protocol for providing these interventions may improve limb salvage, reduce life-altering sequelae such as amputations and mortality, and promote consistent, outcome-driven care.

Background: Muscle co-contraction resulting from motor aberrant reinnervation sequelae (MARS) is a debilitating outcome of brachial plexus birth injury (BPBI), leading to significant shoulder and/or elbow deformities. Surgical management remains controversial with no established consensus on the optimal strategy. Methods: A retrospective review was conducted on patients treated for BPBI-related MARS between 1986 and 2022. A total of 62 patients with shoulder deformities and 95 patients with elbow deformities were included in the study. Two-thirds of the patients had overlapping deformities. The surgical strategy was based on the hypothesis of aberrant reinnervation syndrome, which involved multiple muscle transpositions to convert antagonists to synergists. The strategy for shoulder deformity was prioritized with shoulder abduction and external rotation restoration by transferring the shoulder adductors to the abductors. The strategy for the elbow deformity was prioritized with the elbow extension restoration by transferring elbow flexors (biceps/ brachialis) to the triceps. Results: Postoperatively, the mean shoulder abduction improved from 70° to 148° (p<0.001), and the mean external rotation improved from 22° to 72° (p<0.001). For the elbow, the mean extension strength improved from M2 to M3-4 (p<0.0001). While initial elbow flexion strength decreased post-transfer, 78% of patients required second-stage gracilis functioning free muscle transplantation for augmentation, and the end result was successfully maintained at M3 or greater in all patients. Elbow flexion contracture also significantly improved from a mean of 40° to 10° (p=0.0293). Conclusion: The described surgical strategies by converting antagonists to synergists by multiple muscle transfers are proven effective, durable and long-lasting for treating MARS in BPBI.

References: 1. Sumner AJ: Aberrant reinnervation. Muscle Nerve 1990; 13: 80 1-803. 2. Weiss, P., and M. V. Edds. Sensory-motor nerve crosses in the rat. J. Neurophysiol. 1945; 30: 173-193. 3. Dey JK, Boahene KDO. Facial Aberrant Reinnervation Syndrome Following Facial Nerve Injury and Recovery. Facial Plast Surg Aesthet Med. 2024 Jul 1. doi: 10.1089/fpsam.2023.0351. Epub ahead of print. https://ww6.aievolution.com/asps/Abstracts/viewAbs?subView=1&abs=14401 PMID: 38949952. 4. Chuang DCC. Management of traumatic brachial plexus injury in adults. Hand Clinics 15(4): 737-755, 1999. 5. Chuang DCC, Ma HS, Wei FC. A new evaluation system to predict the sequelae of late obstetric brachial plexus palsy. Plast Reconstr Surg 101:673-685, 1998.

Aims: There is a plethora of research into the efficacy and safety of open and endoscopic surgery for Carpal Tunnel Syndrome (CTS), including a growing number of systematic reviews in the field. This overwhelming research evidence makes it difficult for decision makers and clinicians to utilise the best available evidence to inform their decision making. This novel scoping review aims to provide a comprehensive overview of the synthesised evidence available on this topic, and compare and contrast the findings of previous systematic reviews.

Methods: Following the JBI Manual for Evidence Synthesis a systematic search of multiple databases was conducted from inception to June 2024 for systematic reviews +/- meta-analyses and scoping or umbrella reviews comparing endoscopic and open carpal tunnel release surgery (OCTR and ECTR respectively). 558 papers were independently screened by two researchers leading to inclusion of 21 studies. Data were extracted using evidence gap analysis and formation of evidence map, with outcomes including type of intervention, complications, return to work time, functional outcomes and symptomatic outcomes. Data were analysed using frequency analysis and categorical techniques for calculation and comparison of relative risk.

Results: Seven functional and seven symptomatic outcomes were compared between ECTR and OCTR. Most outcomes showed no difference, except return to work which favoured ECTR in 15 of 21 studies. Grip strength was equivocal (6 favouring ECTR, 6 no difference), while pinch strength favoured ECTR (6/8). Point discrimination (3/4), Boston Carpal Tunnel Syndrome Questionnaire - Functional Outcome (2/3), scar tenderness (4/4) and pillar pain (2/3) also favoured ECTR however had only small reported sample sizes. Pain (4/6) and paraesthesia (3/4) showed no difference. Complication risks were comparable (RR 0.82–2.2), though transient nerve injury was higher with ECTR (RR 1.18–5.8).

Conclusions: ECTR may be preferred when early recovery and reduced scar morbidity are priorities, while OCTR remains a reliable and comparable option with potentially lower nerve injury risk. Given the largely equivalent outcomes, the choice of technique should be guided by surgeon experience and familiarity, as well as patient preference and individual clinical context.

Purpose: Partial hand amputations are the most common form of upper extremity amputation, yet factors associated with post-injury rehabilitation utilization are poorly characterized. We hypothesized that injuries involving functionally important structures (thumb or dominant hand) and with greater amputation severity (multi-digit amputations) would predict hand therapy engagement and referral to a prosthetist.

Methods: An IRB-approved retrospective chart review at a single Level I trauma center identified partial hand amputations using CPT codes. Inclusion required final amputation at/or proximal to the thumb IP, proximal to the digit DIP, or trans-metacarpal level. Demographics and amputation characteristics were tabulated. Univariable logistic regression assessed associations between amputated digit/level, hand dominance, and the following outcomes: hand therapy referral, hand therapy attendance, prosthetist referral, and prosthetic device use.

Results: 402 patients (649 amputated digits) were included; mean age 48±16 years; 301 male, 101 female. 170 patients (42.3%) had an amputation on their dominant hand. 65.2% of patients participated in hand therapy, and only 10.2% and 16.6% received care from physical medicine and rehabilitation physicians (PM&R) and prosthetists, respectively. Multi-digit injury was associated with higher odds of hand therapy referral (OR 1.77, 95% CI 1.13–2.80, p=0.012), prosthetist referral (OR 4.26, 95% CI 2.49–7.43, p<0.001), and prosthetic device use (OR 5.29, 95% CI 2.38–12.80, p<0.001), though it did not significantly predict actual hand therapy attendance (OR 1.60, 95% CI 0.76–3.58, p=0.22). Injury to the dominant hand did not predict hand therapy referral (OR 1.45, 95% CI 0.95–2.21, p=0.08), therapy attendance (OR 1.46, 95% CI 0.72–3.08, p=0.30), or prosthetist referral (OR 1.22, 95% CI 0.72–2.05, p=0.47). When looking at single amputated digits, thumb amputation predicted prosthetist referral (OR 2.92, 95% CI 1.15–7.00, p=0.026), however index-finger amputation was associated with lower odds of attending hand therapy (OR 0.26, 95% CI 0.11–0.62, p=0.002) and lower odds of prosthetic use (OR 0.11, 95% CI 0.00–0.89, p=0.036). Ray level amputations predicted increased hand therapy referral (OR 3.28, 95% CI 1.60–7.36, p<0.001) but not prosthetist referral (OR 0.45, 95% CI 0.09–1.45, p=0.20). Middle phalanx-level (OR 0.52, 95% CI 0.30–0.92, p=0.024) and MCPJ-level amputations (OR 0.42, 95% CI 0.19–0.91, p=0.027) were associated with lower odds of hand therapy referral.

Conclusions: In this cohort, the study team found that more severe, multi-digit amputations more consistently predicted referral to hand therapy and prosthetic utilization than dominant hand amputations. While thumb amputation increased prosthetist referral, index finger and distal level amputations were associated with lower therapy engagement and device use. These findings suggest that referral and utilization patterns may reflect perceived clinical benefit or clinician practice patterns rather than digit functional ranking alone. The gap between referral and attendance for some groups highlights potential unmeasured barriers. Retrospective design, potential selection bias, incomplete variables (e.g. mechanism), and reliance on univariable analyses limit causal inference. Future prospective, multivariable studies are needed to confirm independent predictors and to identify modifiable barriers between referral and participation.

Background Cellulite affects up to 90% of post-pubertal women and results from fibrous septal tethering, adipose protrusion, and dermal laxity. Traditional subcision methods-manual, vacuum-assisted, and laser-assisted-show variable durability and lack real-time confirmation of septa release. Targeted Verifiable Subcision (TVS, Avéli) advances this approach by enabling the operator to identify, tension, and release septa with tactile and visual confirmation, potentially improving precision and consistency. Early studies show promising short-term outcomes, but long-term real-world data remain limited. Objective To evaluate clinical improvement, durability, patient satisfaction, and safety at 12–24 months following TVS in a practice-based cohort. Methods A retrospective review was conducted of 15 women treated with TVS with available 12–24-month follow-up. Data sources include clinical records, standardized photographs, and procedural documentation. Outcomes include blinded Cellulite Severity Scale (CSS) scoring, Global Aesthetic Improvement Scale (GAIS), a 5-point Likert satisfaction metric, and adverse events (AEs). AEs classified by severity. The number of dimples treated per region will also be documented. Descriptive statistics used; CSS changes compared using paired testing. Presently under IRB review. Preliminary Findings Early review shows sustained improvement in dimpling and contour at ≥12 months, with most patients scoring "Improved" or "Much Improved" on GAIS. Most would undergo the procedure again. AEs appear limited to expected transient effects such as bruising and edema, with no major complications identified. Discussion Durability is essential when evaluating cellulite treatments. Earlier subcision methods rely on indirect confirmation, whereas TVS provides verifiable septa release that may reduce operator variability and enhance long-term results. This dataset, though limited, offers real-world evidence on extended outcomes and may help refine patient selection and counseling. Conclusion This study represents one of the earliest evaluations of 12–24-month TVS outcomes, demonstrating durable improvement, high satisfaction, and a favorable safety profile. These results support the role of TVS as a minimally invasive, reproducible option for cellulite correction.

Background: Private equity (PE) consolidation of physician practices has accelerated across multiple medical specialties and has been associated with changes in clinical behavior, healthcare utilization, and academic productivity. Prior studies have demonstrated reductions in scholarly output following PE acquisition in dermatology and other procedural fields [1–3]. As PE investment in plastic surgery practices continues to expand, concerns have emerged regarding its potential impact on investigator-initiated clinical research. However, the role of PE-affiliated practices in plastic surgery clinical research remains poorly defined. This study aimed to characterize institutional participation in plastic surgery clinical trials registered on ClinicalTrials.gov. Methods: A cross-sectional analysis of ClinicalTrials.gov was performed. Registered studies were screened and duplicates removed. Studies were included if a plastic surgeon was listed as the principal investigator (PI). Trials without a listed PI or sponsored exclusively by industry without physician investigator involvement were excluded. Studies were classified by institutional affiliation as academic or private practice. PE affiliation was determined using publicly available business records and industry reports, consistent with prior health services research [1,2]. Analyses were descriptive, with results reported as counts and proportions. Results: A total of 1,732 registered studies were identified, of which 241 duplicates were removed, yielding 1,491 unique trials. After applying inclusion criteria, 159 studies with plastic surgeon PIs were included. Academic institutions led 140 studies (88%), while private practices led 19 studies (12%). Among private practice-led trials, 4 (21.1%) were affiliated with PE-backed organizations. No PE-affiliated academic institutions were identified. Overall, PE-affiliated practices accounted for 2.5% (4/159) of registered plastic surgery clinical trials. Conclusion: Clinical research activity in plastic surgery remains predominantly concentrated within academic institutions, consistent with prior analyses of research productivity in the specialty [4]. Contributions from private practice are limited, and PE-affiliated practices currently represent a small proportion of investigator-initiated research registered on ClinicalTrials.gov. These findings establish a meaningful baseline for the specialty; however, the growth of PE ownership necessitates ongoing surveillance. While not yet proven, it has been speculated that PE-backed practices may prioritize research focused on operational efficiencies and cost reduction rather than higher-risk clinical innovation. If research activity in PE-backed settings remains limited to financial or operational performance metrics, a widening innovation gap may emerge between academic centers and the expanding private sector. Continued monitoring will be important to identify shifts in research participation and leadership over time [1,2].

References 1. Brenner S, Dayal A, Duffy E, et al. Association of private equity acquisition of physician practices with changes in health care spending and utilization. JAMA. 2021;325(17):1720–1731. 2. Gondi S, Song Z, Upadhyay D, et al. Private equity ownership of physician practices and changes in health care delivery: A systematic review. Health Affairs. 2022;41(7):1028–1036. 3. Resneck JS Jr, Pletcher MJ, Lohr RH. Long-term effects of private equity acquisition on dermatology practices: Changes in physician output and scholarly productivity. J Am Acad Dermatol. 2020;83(6):1608–1614. 4. Chao AH, Yaney A, Brill E, et al. Trends in clinical research leadership in plastic surgery: An analysis of ClinicalTrials.gov. Plast Reconstr Surg. 2020;146(5):1145–1153.

Introduction : Radiation-induced capsular contracture (RICC) limits implant-based breast reconstruction when radiotherapy is required. We tested whether TNF-α–primed human adipose-derived stem cell (hASC) coatings, supported by hydrophilic implant surface modification, can preserve cell function under irradiation and prevent RICC in vivo. Methods : Hydrophilic silicone membranes were generated by itaconic-acid conjugation to polydimethylsiloxane (PDMS; IA-PDMS). hASCs (early-passage cells; subcultured 3–6 times) were seeded on PDMS/IA-PDMS and primed at ~80% confluence with TNF-α (25 ng/mL) for 48 hours to generate T-hASCs; non-primed cells served as controls (C-hASCs). In vitro radiotolerance was assessed after 0/10/20 Gy. For in vivo RICC, 8-week-old female BALB/c mice (20–25 g) received four subcutaneous dorsal implants per mouse: C-PDMS (control), C-IA-PDMS, T-PDMS, and T-IA-PDMS. One week post-implantation, localized dorsal irradiation (10 Gy, single fraction) was delivered under intraperitoneal anesthesia . Capsules were harvested at 4 and 8 weeks (n=4 per time point) for histology and RT-qPCR. Results : In vitro, TNF-α priming induced significant actin cytoskeleton rearrangement and enhanced inherent radiotolerance, maintaining high hASC viability even after high-dose irradiation (20 Gy). Furthermore, T-hASCs exhibited a robust paracrine effect, releasing massively elevated levels of vascular endothelial growth factor (VEGF), IL-6, and IL-8, both before and after radiation exposure. In vivo, T-hASC–coated implants (most notably the T-IA-PDMS group) formed significantly thinner capsules and demonstrated reduced inflammatory cell infiltration compared to C-hASC controls at 4 and 8 weeks. Histological and molecular analyses confirmed that T-hASCs profoundly suppressed fibrogenesis by drastically reducing the number of vimentin-positive fibroblasts and α-SMA-positive myofibroblasts, while also downregulating key pro-fibrotic factors (COL1A1, COL3A1, α-SMA, SMAD3, and TGF-β1). Crucially, T-hASCs modulated the local immune microenvironment by suppressing pro-inflammatory M1 macrophage polarization (decreased iNOS, IL-1R1) and promoting tissue-regenerating M2 macrophage polarization (increased MMR/CD206, Arg1, MMP12, IL-10, and IL-13). Additionally, TUNEL assays revealed that T-hASC-coated implants actively protected surrounding normal tissues by significantly reducing radiation-induced apoptosis. Conclusion : Our findings suggest that T-hASCs represent a promising intervention for preventing RICC. Hydrophilic surface support combined with TNF-α priming reduced RICC-associated capsule formation and promoted a less fibrotic, anti-inflammatory peri-implant tissue response with the potential to enhance post-reconstruction outcomes. Further studies are warranted to support clinical translation.

Purpose: High-quality research depends on financial support, skilled personnel, administrative , and efficient workflows (1,2). In an era where time is valuable and costly, wielding the power of artificial intelligence (AI) may reduce low-value tasks and manual labor, creating a new generation of efficient researchers (3,4). This study investigates Claude (Anthropic, San Francisco, CA) (5), a large language model (LLM)–based AI system commonly used for research due to its interconnectivity to PubMed/Medline. The aim is to explore the usability of AI to identify all the possible variables in hand trauma research to construct a comprehensive hand data dictionary.

Methods: This mixed-methods study combines AI-assistance with investigator verification to extract available full metadata variables on hand trauma literature abstracts/full texts. Inclusion criteria included 1) English peer-reviewed published articles on adult or pediatric hand trauma 2) any study design 3) articles reporting ≥3 extractable variables. Isolated wrist injuries without hand involvement, single case reports, and non-traumatic hand conditions were excluded. Following a PRISMA-flow generated prompt, articles were screened by title/abstract before extracting available metadata from the available full text articles. Extracted variables from the generated dataset were ranked according to prevalence across the reviewed literature: high (≥40%), moderate (20–39%), low (5–19%), rarely (<5%). Variables were then compared to publicly available database dictionaries (i.e. NIS, NRD, TriNetX, NSQIP) and assigned a coverage score (0-4) with 1 point given for each database if extracted variable was present.

Results: Seven automated search queries yielded 270 articles. After automatic duplicate removals and title/abstract screening, full text reviews were identified. Of these, 28 contained extractable metadata. A total of 69 variables were identified under 10 domains: patient demographics, injury characteristics, anatomical detail, operative variables, clinical outcomes, functional assessment, patient reported outcomes, hand therapy and rehabilitation, social and systemic outcomes, follow up and recurrence. 30.4% had high prevalence, 43.5% moderate, 21.7% low, and 4.3% were rare. All variables had definitions and additional subtheme variables outlined. Comparison to the 4 large network databases demonstrated about 70% of variables had a coverage score <1.

Conclusion: This study utilizes AI to create a comprehensive data dictionary that eliminates the time-intensive process of creating a database, ultimately improving research workflow inefficiencies. This may aid with planning, organization, and creating cost-effective solutions where research funding is limited.

References

1. Ewing JN, Lemdani MS, Gala Z, et al. A longitudinal evaluation of collaboration in plastic surgery clinical research. Plast Reconstr Surg Glob Open. 2024;12(8):e6023. doi:10.1097/GOX.0000000000006023. PMID:39534075. PMCID:PMC11557106.

2. Rohrich RJ, Robinson JB Jr, Adams WP Jr. The plastic surgery research fellow: revitalizing an important asset. Plast Reconstr Surg. 1998;102(3):895-898. doi:10.1097/00006534-199809030-00046. PMID:9727462.

3. Mir MA. Artificial intelligence revolutionizing plastic surgery scientific publications. Cureus. 2023;15(6):e40770. doi:10.7759/cureus.40770.

4. Liang X, Yang X, Yin S, et al. Artificial intelligence in plastic surgery: applications and challenges. Aesthetic Plast Surg. 2021;45(2):784-790. doi:10.1007/s00266-019-01592-2. PMID:31897624.

5. Anthropic. Claude (large language model) [Internet]. San Francisco, CA: Anthropic; 2026. Accessed April 30, 2026. https://claude.ai