Conference Proceedings
Conference Proceedings from the Canadian Global Surgery Trainees' Alliance

Abstract

On May 3rd, 2025, the 13th Annual Global Surgery Conference gathered 150 participants from 23 countries under the theme of Equity in Global Surgery. Organized by the McGill Chapter of the Canadian Global Surgery Trainees' Alliance in collaboration with the Centre for Global Surgery, this hybrid event invited attendees to reimagine global surgery through principles of inclusion, decolonial practice, and shared leadership. The program featured keynote lectures, trainee research, and a disaster-response case competition. Voluntary pre- and post-conference surveys (n=45) revealed significant increases in participants' understanding of global surgery, confidence in pursuing related opportunities, and awareness of equity-targeted strategies. Mean knowledge scores increased from 3.16 to 4.12 (+0.96), with the largest gains in familiarity with global surgery programs (+1.17) and confidence applying for related opportunities (+1.23). This scalable, trainee-led, low-cost conference model is evidence of how educational initiatives can tangibly advance equity and solidarity within the global surgical community.

Abstract

Background: Sub-Saharan Africa (SSA) faces critical surgical and obstetric (SO) workforce shortages. Multinational colleges, including the West African College of Surgeons (WACS) and three SO colleges in the East, Central, and Southern African (ECSA)-Health Community (COSECSA, COECSA, and ECSACOG), build local training capacity through standardized 3–7-year postgraduate programs. Despite the growth of regional colleges, the geographic footprint of programs remains largely unmapped. We aimed to address this by mapping college-accredited SO postgraduate training programs across SSA nations.

Methods: A grey literature review was conducted to extract the distribution and curricula of postgraduate SO programs. Digital maps were generated and integrated with a population-scaled Training Capacity Index (TCI) – calculated as population per accredited program – to visualize and compare national training capacity.

Results: Of 48 SSA countries, 39 (81%) are affiliated with at least one regional college. Of those 39 nations, 28% had no accredited SO fellowship program. Among affiliated nations, we identified 2.61 college-accredited SO programs and 0.98 training hospitals per 5 million people. The SSA-wide surgical TCI is one program per 2.37 million people, while the ObGyn TCI is one per 10.9 million. Moreover, 94% of Anglophone countries host college-accredited programs compared to 25% of Francophone countries.

Conclusion: Our grey literature review, mapping, and population-scaled index unveil an uneven distribution of college-accredited postgraduate training programs and training hospitals across SSA. Asymmetry between Anglophone and Francophone nations was also observed. Addressing the SO workforce deficit requires the strategic expansion of regional college-based training.

Abstract

Background: Achieving fistula healing while preserving continence remains a major challenge in colorectal surgery, encouraging the development of minimally invasive sphincter-sparing techniques.

Aim: To evaluate mid-term outcomes of fistula laser closure (FiLaC) in patients with chronic perianal fistula.

Materials and Methods: A retrospective study including 25 consecutive patients treated between March 2022 and December 2024 was performed. Healing was defined as absence of discharge for at least six months. Recurrence or persistence was defined as ongoing or recurrent discharge after one month.

Findings: Mean age was 54 years, with 52% male. Transsphincteric fistulas were most common (52%). All procedures were performed as day surgery without immediate complications. After a mean follow-up of 12 months, healing rate was 56%, with recurrence in 44%. No patient developed continence impairment.

Interpretation: FiLaC is a safe sphincter-preserving option with favourable continence outcomes. Although healing rates are moderate, its minimally invasive nature supports its use in selected patients.

Abstract

Introduction: Trauma-related mortality is a major global health challenge, disproportionately affecting low- and middle-income countries (LMICs). Trauma registries are critical for strengthening trauma systems, yet their implementation in LMICs remains limited. AMBER—a web-based trauma registry developed by McGill's Centre for Global Surgery—was piloted in the emergency department at Dhulikhel Hospital, Nepal, to assess usability, feasibility, and implementation challenges.

Methods: A mixed-methods study was conducted using post-AMBER training surveys of medical students serving as data collectors (n=11), as well as post-information session surveys of residents, interns, medical officers, paramedics, and nurses (n=22). Quantitative measures assessed user-friendliness, satisfaction with the data entry process, and feasibility. Qualitative data were derived from open-ended survey responses and inductive coding of challenges and suggested improvements.

Results: Data collectors rated AMBER highly user-friendly (mean score 4.5/5), with all participants reporting it captured necessary trauma data (n=11). Among hospital staff, 72% were very satisfied (5/5) or satisfied (4/5) with the data entry process, while 77% rated implementation as perfectly feasible (scores 4–5) (n=22). Inductive coding revealed four major challenges: (1) time constraints, (2) lack of dedicated data collectors, (3) data quality concerns, and (4) limited institutional awareness of the registry. Suggestions included emphasizing prospective rather than retrospective data entry, improving personal access to AMBER, and increasing institutional awareness.

Conclusion: AMBER was positively received as a user-friendly tool that captures essential trauma metrics. Implementation challenges centred on time, manpower, and data quality, underscoring the need for ongoing training, broader access, and institutional support to optimize integration into trauma care workflow.

Abstract

Experiential learning with real-time expert instruction is a cornerstone of surgical training. However, it lacks a standardized framework for effective communication, leading to inconsistencies in teaching outcomes. This study identifies speech feature differences between scripted and artificial intelligence (AI)-augmented personalized instructions to assess their effectiveness in improving intraoperative performance during surgical training. The design is a cross-sectional cohort study following a randomized controlled trial at the Neurosurgical Simulation and Artificial Intelligence Learning Centre, McGill University. Fifty-seven medical students underwent simulated subpial resection on the NeuroVR platform while a human expert verbally delivered instructions. All instructions were categorized as scripted or personalized. Thirteen speech features were analyzed using natural language processing (NLP) and a combination of eight large language models (LLMs) to evaluate syntax, semantics, and topic alignment. Analysis included 1,014 occurrences of scripted instructions and 673 personalized instructions. High inter-rater agreement was observed across AI models. Personalized instructions demonstrated higher sentence complexity and information content, specifically reflected by perceptibility and syntax depth features. Significant correlations were identified between linguistic markers and affective-cognitive states, including a positive correlation between syntax depth and intrinsic cognitive load, suggesting that complex sentences require observably greater mental effort. Personalized instructions were associated with higher effectiveness and better long-term surgical training outcomes compared to generic scripted instructions. These results highlight the importance of instructing trainees with context-tailored information rather than generic actions to optimize learning. Future intelligent tutoring system design requires careful evaluation of linguistic markers to guide the development of a standardized framework for education.

Abstract

Background: Pediatric trauma is a significant yet often under-recognized public health issue in low-income countries like Nepal. Factors such as poor infrastructure, lack of safety regulations, and unsupervised children in hazardous environments contribute to high rates of trauma-related morbidity and mortality. Dhulikhel Hospital, a tertiary care centre in Nepal, reports that over 20% of its pediatric emergency visits are trauma-related, with falls being the leading cause.

Objectives: This study explores the challenges families encounter in seeking timely trauma care for children admitted to Dhulikhel Hospital.

Methods: A qualitative phenomenological design grounded in the Three Delays Model was developed, piloted, and deployed. To date, eight of 12 interviews with family members have been conducted, audio-recorded, transcribed, and analyzed using inductive thematic analysis.

Preliminary Results: Preliminary findings indicate that delays in seeking care were influenced by fear of injury severity and long-term consequences, financial insecurity, lack of awareness, and repeated visits to multiple facilities. Conversely, strong family support systems, emotional encouragement from relatives, and health insurance facilitated early decision-making. Delays in reaching care were primarily driven by transportation challenges, including vehicle inaccessibility and difficulty arranging transfers. Delays in receiving care reflected poor prehospital services, limited capacity at local health posts, absence of trained personnel during transport, and procedural anxiety. Timely and efficient emergency care at the tertiary centre emerged as a key facilitator once patients arrived.

Conclusion: Preliminary findings highlight critical gaps in prehospital and referral systems, underscoring the protective roles of social support and emergency care capacity.