Knowledge gains and retention following the rural emergency trauma system strengthening (RETSS) programme: a longitudinal observational study in rural Nepal Authors: Theresa Farhat, Santosh Paudel, Tarek Razek, Gabriela Sanchez, Yagya R Shakya, Nancy Tze, Salim Sohani, Jeremy Grushka, Rachel Nadeau, Bal K Sedai, Rohit Shrestha, Sanu Krishna Shrestha, Sangina Ranjit, Dan Deckelbaum, Samjhana Basnet Published: 2026 ABSTRACT AbstractIntroductionFollowing the 2015 Nepal earthquake, the Nepal Red Cross Society, Canadian Red Cross, Centre for Global Surgery at McGill University Health Centre, and local leaders launched the Rural Emergency Trauma System Strengthening (RETSS) programme to strengthen first-responder capacity in trauma care, cardiac arrest management and disaster preparedness during the early phase of emergencies. RETSS employed a district-level train-the-trainer (ToT) model to establish a sustainable local training network.MethodsWe conducted a retrospective analysis of RETSS programme data (2017–2020), including participant demographics, geographic distribution and knowledge outcomes. Pre- and post-training scores were compared using Wilcoxon signed-rank tests. Per cent improvement and retention decline were calculated using standardised formulas. Knowledge retention was assessed among repeat basic life support (BLS) participants using time-stratified analyses.ResultsIn 2017, 24 healthcare providers (HCPs) (70.8% male; mean age 35.6 years) attended ToT sessions. Between 2017 and 2020, a total of 1170 HCPs were trained across 158 government facilities in Sindhupalchok (59.7%) and Kavrepalanchok (40.3%). Most trainees were female (61.4%) and aged 18–29 (52.5%). BLS scores improved from a median of 9.0 (IQR 6 to 12) to 22.0 (IQR 20 to 23) (+122.2%, p<0.0001), and Primary Trauma Care scores from 11.0 (IQR 7.2 to 14.0) to 18.5 (IQR 16 to 20) (+55.6%, p<0.0001). Knowledge declined between sessions, with retention at first repeat of 59.1% overall. Retention was highest among participants repeating within ≤6 months (75.5%) and varied across longer intervals (7–12 months: 57.1%; 13–18 months: 61.8%; >18 months: 72.2%, n=4).ConclusionsRETSS demonstrates the feasibility of a scalable, locally led training model to strengthen emergency and trauma care in rural, disaster-prone regions and was associated with substantial short-term gains in provider knowledge. Substantial knowledge decay over time supports the implementation of structured refresher training at approximately 6-month intervals. WHAT IS ALREADY KNOWN ON THIS TOPIC Disaster-affected low- and middle-income countries such as Nepal have limited emergency and trauma care capacity, and although short trainings are often delivered after emergencies, there is little evidence on district-wide train-the-trainer (ToT) models that reach rural providers and assess knowledge retention over time. Disaster-affected low- and middle-income countries such as Nepal have limited emergency and trauma care capacity, and although short trainings are often delivered after emergencies, there is little evidence on district-wide train-the-trainer (ToT) models that reach rural providers and assess knowledge retention over time. WHAT THIS STUDY ADDS Using routine data from the Rural Emergency Trauma System Strengthening programme in two earthquake-affected districts of Nepal, this study found that a district-based ToT model reached a large, predominantly young and female rural workforce and was associated with substantial immediate increases in emergency and trauma knowledge. It also documents important patterns of knowledge decline between refresher trainings. Using routine data from the Rural Emergency Trauma System Strengthening programme in two earthquake-affected districts of Nepal, this study found that a district-based ToT model reached a large, predominantly young and female rural workforce and was associated with substantial immediate increases in emergency and trauma knowledge. It also documents important patterns of knowledge decline between refresher trainings. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings support integrating structured refresher trainings and locally led ToT networks into district health planning for emergency care and offer a practical model for designing and scaling similar capacity-building programmes in other resource-limited, disaster-prone settings. These findings support integrating structured refresher trainings and locally led ToT networks into district health planning for emergency care and offer a practical model for designing and scaling similar capacity-building programmes in other resource-limited, disaster-prone settings. Background Natural disasters pose escalating health threats globally, especially in low- and middle-income countries (LMICs), where limited infrastructure and workforce capacity amplify morbidity and mortality.1 In these settings, strengthening emergency care systems is critical to improving resilience and response capacity.13 Nepal, a low-income country with a Human Development Index of 0.601 (ranked 146th globally), faces frequent climate-related hazards due to its geographic location and diverse topography.4 More than 80% of the population is exposed to natural disasters, including earthquakes, floods, landslides, windstorms, fires, droughts and glacial lake outburst floods.5 6 Nepal ranks as the fourth most vulnerable country to climate change and the eleventh most susceptible to earthquakes globally.7 In April 2015, Nepal was struck by a series of devastating earthquakes and aftershocks, the largest measuring 7.8 on the Richter scale.8 The quakes severely affected rural regions, impacting nearly half of Nepal’s districts, and resulting in nearly 9000 deaths, more than 22 000 injuries and the displacement of 2.8 million people.9 10 Healthcare infrastructure was heavily damaged, with over 1200 health facilities destroyed or rendered non-functional; some districts lost up to 90% of their facilities.1012 Even before the earthquake, Nepal’s healthcare system faced significant challenges, including limited prehospital services and emergency preparedness, especially in rural areas.1316 The disaster exposed critical gaps in trauma care and emergency response systems, including inadequate referral mechanisms and prehospital transport. It also worsened the burden of disability, mental health issues and poverty due to the collapse of essential health and social infrastructure.11 12 17 In response, the Rural Emergency Trauma System Strengthening (RETSS) programme was launched to address the urgent need for effective emergency care in rural Nepal. Spearheaded by the Nepal Red Cross Society (NRCS) and key national institutions, including the Ministry of Health and Population’s National Health Training Centre, Bir Hospital, National Academy of Medical Sciences (NAMS), and Dhulikhel Hospital Kathmandu University Hospital (DHKUH), the programme received technical support from the Centre for Global Surgery (CGS) at McGill University Health Centre (MUHC) and the Canadian Red Cross (CRC) and financial support from CRC. The programme ran from January 2017 to July 2020, delivering training through basic life support (BLS), Primary Trauma Care (PTC), the Rural Trauma Team Development Course (RTTDC) and the CGS-Trauma and Disaster Team Response (TDTR) courses, with a focus on the severely earthquake-affected districts of Sindhupalchok and Kavrepalanchok. In this study, we evaluated the reach, knowledge gains, and retention patterns associated with the RETSS programme from 2017 to 2020. Our goal is to demonstrate RETSS as a potentially replicable model for strengthening emergency care systems and resilience in resource-limited, disaster-affected and vulnerable settings worldwide. Methods Study design We conducted a retrospective, observational pre-post design using programme data from 2017 to 2020, including participant demographics, geographic distribution of training sessions and knowledge outcomes. Study setting Sindhupalchok, one of the hardest-hit districts, was selected as the pilot site during the programme’s first year. The district, characterised by mountainous terrain and limited road access, had 3 Primary healthcare Centres (PHCCs) and 75 Health Posts, evenly distributed across its territory. Kavrepalanchok, a densely populated district with 4 PHCCs and 89 Health Posts, was also included due to its accessibility and the proximity of Dhulikhel Hospital, a key programme partner and referral centre. Rural emergency trauma system strengthening (RETSS) programme courses The RETSS programme was developed following the 2015 Nepal earthquake through a series of collaborative workshops involving local clinical and institutional partners, including Bir Hospital/NAMS, Dhulikhel Hospital, the NRCS, the CRC and the CGS at McGill University. The training model was adapted to the rural Nepal context by incorporating common emergency scenarios, resource constraints at peripheral facilities and existing referral pathways. BLS and PTC materials were translated into Nepali to improve accessibility, uptake and reproducibility among local healthcare providers (HCPs). The RETSS programme operated in Sindhupalchok and Kavrepalanchok from 2017 through July 2020 to strengthen HCP capacity in emergency and trauma response. The programme delivered tiered training based on provider roles and facility capacity, focusing on the following courses: Basic life support (BLS): delivered across all facility types, from Health Posts to District Hospitals, this training was designed to support HCPs in providing urgent, high-quality care.18 The curriculum followed American Heart Association and Heart and Stroke Foundation standards and was adapted for zonal-level delivery, in partnership with CGS-MUHC. Basic life support (BLS): delivered across all facility types, from Health Posts to District Hospitals, this training was designed to support HCPs in providing urgent, high-quality care.18 The curriculum followed American Heart Association and Heart and Stroke Foundation standards and was adapted for zonal-level delivery, in partnership with CGS-MUHC. PTC: a 2 day course focused on trauma management, conducted for providers at PHCCs and District Hospitals. It followed the PTC Foundation’s curriculum and emphasised hands-on trauma assessment and stabilisation skills.19 PTC: a 2 day course focused on trauma management, conducted for providers at PHCCs and District Hospitals. It followed the PTC Foundation’s curriculum and emphasised hands-on trauma assessment and stabilisation skills.19 RTTDC: delivered in district and zonal hospitals in collaboration with CGS-MUHC, this course aimed to strengthen team-based trauma care, emphasising communication, early recognition of referral needs and coordinated stabilisation to improve outcomes.20 RTTDC: delivered in district and zonal hospitals in collaboration with CGS-MUHC, this course aimed to strengthen team-based trauma care, emphasising communication, early recognition of referral needs and coordinated stabilisation to improve outcomes.20 TDTR: developed by CGS, this multidisciplinary course focused on standard trauma care and disaster response. It combined online modules with hands-on, in-person sessions to reinforce applied learning in austere environments.21 TDTR: developed by CGS, this multidisciplinary course focused on standard trauma care and disaster response. It combined online modules with hands-on, in-person sessions to reinforce applied learning in austere environments.21 Data collection and variables Participant data were collected at course registration and included name, age, gender, speciality, facility location and course dates. Knowledge was assessed using course-specific written tests administered before and after training: BLS (25 multiple-choice items; maximum score 25) and PTC (20 multiple-choice items; maximum score 20). RTTDC knowledge test data were not consistently recorded and are therefore not included in this analysis. Identical test forms were used for pre-course, post-course and repeat-session assessments. Practical skills were assessed using standardised checklists for BLS (adult [six items] and infant cardiopulmonary resuscitation (CPR)/automated external defibrillator (AED) [five items]) and for PTC and RTTDC (core trauma competencies). However, these data were recorded on paper and not electronically captured and were therefore not included in the analysis. Data analysis Descriptive statistics summarised participant demographics and training coverage. Knowledge score changes were evaluated by comparing pre- and post-course scores using Wilcoxon signed-rank tests. Per cent improvement was calculated as ([post-score−pre-score]/pre-score)×100. Knowledge retention (%) was calculated as ([pre-score of repeat session/post-score of prior session])×100. For repeat BLS participants, time intervals between sessions were computed in months (days/30.44) and categorised as: ≤6 months, 7–12 months, 13–18 months and >18 months. Participants who attended multiple sessions were included in each session’s pre-post analysis independently. To assess potential selection bias among repeat trainees, a sensitivity analysis was performed comparing repeat and non-repeat BLS participants on baseline characteristics, including initial pretest score, age, sex and cadre distribution. Mann-Whitney U tests were used for continuous variables and χ2 tests for categorical variables. Analyses were conducted using GraphPad Prism V.10.4. Results Training of trainers (ToT) In 2017, the RETSS programme conducted Training of Trainers (ToT) enrolling 24 HCPs, of whom 70.8% (n=17) were men and 29.2% (n=7) women, with a mean age of 35.6 years (SD±4.3). Of the cohort, 95.8% (n=23) were trained in BLS, 66.7% (n=16) in PTC, 54.2% (n=13) in the RTTDC and 50.0% (n=12) in the CGS-TDTR course (table 1). Five participants (20.8%, n=5/24) completed the Nepali BLS course. Two participants, one from the NAMS and one from DHKUH, were certified as Master Trainers in BLS, qualifying them to conduct instructor-level training, develop new trainers and deliver provider-level training. Table 1Demographics of trained trainers in the Rural Emergency Trauma System Strengthening programme in 2017CharacteristicAll (n=24)RTTDCPTCBLSTDTR ToT (n=12)Course (n=13)ToT (n=18)Course (n=16)ToT (n=16)Course (n=23)ToT (n=13)Age, years—Mean±SD35.57±4.3———————Gender—n Female73554623 Male171013111217119Organisation—n DHKUH131212111112812 NAMS1016551040 NRCS10000110Speciality—n Anesthesiologist1147971053 ER Physician22211222 General Practitioner22222222 Nursing Officer10100100 Orthopaedic Surgeon43424423 Paramedic32222322 Staff Nurse10000100Multiple courses (2017)—n Two1420201000 Three or more80030500Five participants additionally completed the Nepali BLS course (1 female, 4 males; 4 from DHKUH, 1 from NAMS; specialities: 2 anesthesiologists, 1 general practitioner, 1 orthopaedic surgeon and 1 paramedic). Participants may appear in multiple columns because individuals attended more than one training in 2017; the “All” column reflects 24 unique participants.BLS, Basic Life Support; DHKUH, Dhulikhel Hospital Kathmandu University Hospital; NAMS, National Academy of Medical Sciences; NRCS, Nepal Red Cross Society; PTC, Primary Trauma Care; RTTDC, Rural Trauma Team Development Course; TDTR, Trauma and Disaster Team Response; ToT, Train-the-Trainer. Demographics of trained trainers in the Rural Emergency Trauma System Strengthening programme in 2017 Five participants additionally completed the Nepali BLS course (1 female, 4 males; 4 from DHKUH, 1 from NAMS; specialities: 2 anesthesiologists, 1 general practitioner, 1 orthopaedic surgeon and 1 paramedic). Participants may appear in multiple columns because individuals attended more than one training in 2017; the “All” column reflects 24 unique participants. BLS, Basic Life Support; DHKUH, Dhulikhel Hospital Kathmandu University Hospital; NAMS, National Academy of Medical Sciences; NRCS, Nepal Red Cross Society; PTC, Primary Trauma Care; RTTDC, Rural Trauma Team Development Course; TDTR, Trauma and Disaster Team Response; ToT, Train-the-Trainer. Participants were primarily anaesthetists (45.8%, n=11/24), followed by orthopaedic surgeons (16.7%, n=4/24), paramedics (12.5%, n=3/24), emergency physicians (8.3%, n=2/24) and general practitioners (8.3%, n=2/24). Organisational representation included DHKUH (54.2%, n=13/24), NAMS (41.7%, n=10/24) and the NRCS (4.2%, n=1/24). Repeated attendance at instructor courses was common: two participants completed both RTTDC and PTC twice; three completed PTC three times and five completed BLS three times (table 1). Demographic trends and programme reach Between 2017 and 2020, RETSS delivered training to 1170 HCPs in BLS, PTC and RTTDC, with BLS comprising the majority (81.5%, n=953), followed by PTC (14.3%, n=167) and RTTDC (4.3%, n=50) (figure 1A). Training was delivered in 158 government health facilities across 28 municipalities in two severely earthquake-affected districts: Sindhupalchok (59.7%, n=699) and Kavrepalanchok (40.3%, n=471) (figures1B 2). The highest concentration of sessions occurred in Chautara Sangachokgadi municipality (13.6%, n=160), followed by Namobuddha (8.7%, n=102), Panauti (7.4%, n=87) and Melamchi (6.7%, n=78). Rural Emergency Trauma System Strengthening programme reach and participant demographics (2017–2020). (A) Total number of participants by course type, (B) Geographical distribution of participants by districts, selected as the most severely earthquake-affected areas and (C) Age-sex distribution of participants with available demographic data. BLS, basic life support; PTC, Primary Trauma Care; RTTDC, Rural Trauma Team Development Course. Distribution of participants by course type across municipalities in Sindhupalchok (n=699) and Kavrepalanchok (n=471) districts, 2017–2020. Bars represent the total number of trained healthcare providers per municipality, stacked by course. Municipalities are ordered by total participants. BLS, basic life support; PHCC, Primary healthcare Centre; PTC, Primary Trauma Care; RTTDC, Rural Trauma Team Development Course. Most participants were female (61.4%, n=669/1089) and aged 18–29 years (52.5%, n=606/1154). Additional age distribution included 25.2% aged 30–39 (n=291/1154) and 22.3% aged 40 and above (n=257/1154) (figure 1C). Mean age was 31.7 years (SD±9.5; median 29, IQR 24 to 38). The most common cadre was auxiliary health workers (AHWs) (65.8%, n=769/1168), followed by health assistants (13.5%, n=158/1168) and medical officers (8.9%, n=105/1168) (figure 3). Course repetition was common: 332 participants completed BLS twice, 86 three times and 4 four times. Eleven participants completed PTC twice. Distribution of participants by speciality/cadre and course type in the Rural Emergency Trauma System Strengthening programme (2017–2020). BLS, basic life support; PTC, Primary Trauma Care; RTTDC, Rural Trauma Team Development Course. Knowledge gains Both BLS and PTC courses were associated with statistically significant increases in knowledge scores; RTTDC knowledge test scores were not available in the programme database and are not reported. For BLS (n=932 with paired data), median pre-course scores increased from 9.0 (IQR 6.0 to 12.0) to 22.0 (IQR 20.0 to 23.0), with a median within-participant relative score increase of 122.2% (p<0.0001). For PTC (n=166), scores rose from a median of 11.0 (IQR 7.2 to 14.0) to 18.5 (IQR 16.0 to 20.0), corresponding to a 55.6% relative increase (p<0.0001) (figure 4). Pre- and post-course knowledge scores by session in the Rural Emergency Trauma System Strengthening programme (2017–2020). Box plots show the median, IQR, and 1.5×IQR for pretest and posttest scores at each training session for BLS and PTC. Rural Trauma Team Development Course knowledge test scores were not available in the programme database. Statistical comparisons between pre and posttest scores within each session were performed using the Wilcoxon signed-rank test (**p<0.001; ***p<0.0001). BLS, basic life support; PTC, Primary Trauma Care. After the first BLS repeat (n=328), scores increased from a median of 13.0 (IQR 10.0 to 16.0) to 21.0 (IQR 20.0 to 23.0), a 60.0% gain (p<0.0001). The second BLS repeat (n=82) yielded a 28.6% gain, with scores rising from 17.0 (IQR 14.0 to 20.0) to 23.0 (IQR 21.0 to 24.0) (p<0.0001). A small number of participants who attended a third BLS repeat (n=4) showed a 157.4% increase, from 9.5 (IQR 6.5 to 11.0) to 24.0 (IQR 23.0 to 25.0), but it was not statistically significant (p=0.125). For PTC repeat (n=11), median scores improved by 53.8%, from 13.0 to 20.0 (p=0.001) (figure 4). Knowledge retention Before the first repeat training (n=330), median scores declined from 22.0 (IQR 20.0 to 23.0) to 13.0 (IQR 10.0 to 16.0), representing a 40.9% decline (p<0.0001). Before the second repeat (n=85), scores declined from 21.0 to 17.0 (IQR 14.0 to 20.0), a 17.4% decline (p<0.0001) (figure 4). Before the third repeat (n=4), scores declined from 21.5 to 9.5, a 55.0% decrease. PTC participants showed relatively stable retention, with a median decline of 7.1% before the first repeat (p=0.147), although the sample size was small (n=11). Sensitivity analysis: repeat versus non-repeat BLS participants Repeat BLS participants (n=332) and non-repeat participants (n=621) were compared on available baseline characteristics. Baseline knowledge scores were similar between groups (repeat: median pretest 10.0, IQR 6.0 to 13.0, n=323; non-repeat: median 9.0, IQR 6.0 to 12.0, n=609; p=0.46), as were sex distribution (59.8% vs 65.2% female; p=0.12) and cadre composition (AHW: 72.9% vs 70.7%; p=0.47). Repeat participants were modestly younger (median 26 vs 30 years; n=332 and n=620, respectively; p<0.0001). All repeat BLS sessions were delivered in Sindhupalchok, reflecting programme delivery structure rather than participant self-selection (online supplemental figure 1). Retention across repeat training intervals Among BLS repeat participants (n=331), the median interval from initial to first repeat training was 11.4 months (IQR 8.9 to 11.8 months; range 2.3–21.0 months). The median interval from first to second repeat was 6.5 months (IQR 5.0 to 8.8 months). Time-stratified retention analysis for BLS revealed a clear interval-dependent pattern (figure 5, table 2). For the initial to first repeat interval: participants who repeated ≤6 months after initial training (n=50) retained 75.5% of knowledge (within-group decline 24.5%, p<0.0001); those repeating at 7–12 months (n=223) retained 57.1% (within-group decline 42.9%, p<0.0001); those at 13–18 months (n=52) retained 61.8% (within-group decline 38.2%, p<0.0001) and the >18 month group (n=4) showed 72.2% retention, though this was not statistically significant (p=0.25) due to the small sample. BLS knowledge retention by time interval between training sessions (2017–2020). Bars represent the median percentage of knowledge retained at the pretest of each repeat session relative to the posttest of the prior session. Error bars represent the IQR/2. The dashed line at 100% indicates full retention of prior posttest scores. (A) Retention from initial training to first repeat session, stratified by time interval and (B) Retention from first to second repeat session, stratified by time interval. BLS, basic life support. Table 2Basic life support knowledge retention by time interval between sessionsIntervalnPrior post (median)Pre at repeat (median)Retention (%)Decline (%)P valueInitial → 1st repeat ≤6 months5021.015.075.5%24.5% <0.0001 7–12 months22322.013.057.1%42.9% <0.0001 13–18 months5222.013.561.8%38.2% <0.0001 >18 months421.015.072.2%27.8%0.25 Overall32922.013.059.1%40.9% <0.0001 1st → 2nd repeat ≤6 months4119.017.083.3%16.7% <0.0001 7–12 months3822.018.083.0%17.0% <0.0001 13–18 months622.510.044.5%55.5%0.094 Overall8521.017.082.6%17.4% <0.0001 Bold values indicate statistical significance (p<0.05). Basic life support knowledge retention by time interval between sessions Bold values indicate statistical significance (p<0.05). For the first to second repeat interval: participants repeating ≤6 months (n=41) retained 83.3% (p<0.0001); those at 7–12 months (n=38) retained 83.0% (p<0.0001) and those at 13–18 months (n=6) retained 44.5%, though this was not statistically significant (p=0.094) (table 2, figure 5). Discussion The RETSS programme demonstrates the feasibility and scalability of a context-adapted, locally led emergency care training model in post-disaster, resource-constrained rural settings. To our knowledge, RETSS represents one of the largest published post-disaster interventions designed to strengthen local emergency-care capacity and build baseline resilience in rural LMIC settings. Initiated after the 2015 Nepal earthquake, the programme brought together national and international partners to address urgent gaps in emergency and trauma care. Similar training initiatives have been implemented in LMIC settings. The WHO Basic Emergency Care (BEC) course uses a ToT approach to strengthen frontline emergency care capacity and has shown knowledge gains in several low-resource settings.2225 RETSS builds on this shared objective but is distinguished by its emphasis on sustained, locally embedded capacity-building rather than short-term emergency support. It also differs from BEC and similar programmes in several important ways: it combines basic life support for cardiac arrest and choking with structured trauma system strengthening, including referral coordination and multi-level provider training; it was adapted through contextual piloting with local institutional partners and translated into Nepali and it was implemented by Nepali authorities/local institutions in a disaster-prone setting where earthquakes, floods and road traffic injuries are major drivers of emergency care needs. These features make RETSS a complementary model with relevance for settings facing recurrent disasters and fragmented prehospital systems. RETSS emerged from collaboration between government bodies and medical institutions, including DHKUH, the CGS-MUHC, CRC and NRCS. This partnership helped identify major gaps in rural emergency care and supported a coordinated response tailored to local needs. One important achievement was the development and integration of a Nepal-specific BLS curriculum into the national training system, which strengthened local ownership and increased the potential for sustainability. A central finding of this study is that a small initial cohort of locally trained instructors enabled broad downstream emergency care training across districts, highlighting the efficiency of the ToT model in resource-constrained rural settings. Repeated participation by some trainers likely reflects both sustained engagement and the limited size of the trainer pool, highlighting the need for broader trainer expansion in future scale-up efforts. Because data beyond July 2020 were not available, we could not assess the long-term sustainability of the trainer network. Future work should examine trainer retention, independently delivered courses and continued district coverage after external support ended. Between 2017 and 2020, locally trained instructors delivered training across district hospitals, PHCCs and health posts. BLS accounted for most training, while PTC, RTTDC and CGS-TDTR were delivered less frequently. This pattern is understandable given the wider applicability of BLS, but broader delivery of advanced courses would help strengthen the trauma care system more fully. Most participants were young, female and from AHW cadres, reflecting the rural health workforce in Nepal, where AHWs and health assistants often serve as frontline providers at health posts and community units. In hospitals and PHCCs, emergency care is more often led by physicians working with multidisciplinary teams. These findings indicate that RETSS effectively reached frontline providers most directly responsible for emergency care delivery in rural settings. Pre and posttest results demonstrated substantial improvements in knowledge scores. The larger gains observed in BLS likely reflect both lower baseline knowledge and the more standardised, algorithm-driven nature of BLS compared with the more complex clinical reasoning required in PTC. Retention findings highlight a clear pattern: knowledge gains decline over time, consistent with prior literature demonstrating rapid decay of resuscitation knowledge following short course training.26 27 The time-stratified analysis suggests that shorter retraining intervals are associated with improved knowledge retention, supporting the implementation of refresher training at approximately 6-month intervals. The slightly higher retention observed at longer intervals in some groups may reflect unmeasured factors such as participant motivation or workplace reinforcement and warrant further investigation. Although direct patient-level outcomes were not systematically measured, several programme-linked indicators suggest possible concurrent system-level changes in emergency care delivery. As a byproduct of the RETSS programme, Dhulikhel Hospital established a 24/7 ambulance dispatch centre with the dedicated three-digit emergency number 102, providing emergency medical services across Kavre, Bhaktapur, Sindhupalchok, Dolakha and Ramechhap districts. RETSS also coincided with strengthened coordination between peripheral hospitals, PHCCs, Dhulikhel Hospital and the central Health Emergency Operation Centre, with more consistent notification of incoming trauma patients, particularly during disaster events. Before the programme, virtually no trauma patients received treatment before arrival; more recently, approximately 15% of patients now receive prehospital care and advance notification, a proportion that continues to increase,28 suggesting evolving system-level changes. Evidence of sustainability beyond the study period has emerged since programme completion. The Nepali BLS curriculum has been institutionalised, with monthly training now conducted at Dhulikhel Hospital using local resources and expertise. The Government of Nepal has begun implementing BLS and PTC training across all seven provinces, with dedicated budget allocation and approximately two to three courses per year in each province. The RTTDC course has been adapted at Dhulikhel Hospital and its 18 outreach centres, where it is delivered annually. TDTR also continues to be administered beyond the original study period. These post-programme developments suggest potential sustainability and increasing local ownership beyond the original programme timeframe, although they were not formally evaluated. In terms of programme resources, the approximate cost per training course was NPR 150 000 for 1-day courses (BLS and RTTDC) and NPR 200 000 for 2-day courses (PTC), with a typical class size of 24–30 participants. This corresponds to an estimated cost of approximately NPR 5000–6250 per participant for 1-day courses and NPR 6667–8333 per participant for 2-day courses. Although formal cost-effectiveness was not assessed, this suggests that once local trainers were established, the decentralised model delivered training at relatively low per-participant cost while minimising travel and service disruption. Strengths and limitations This study has several strengths. It describes a large-scale emergency care training programme implemented during a critical post-disaster recovery period in rural Nepal. RETSS’s scalability, coordination and adaptability underscore its relevance to other low-resource or disaster-affected contexts. The large, diverse sample of HCPs provides a comprehensive view of the programme’s reach and associated knowledge outcomes. This study also has several limitations. First, this was a retrospective analysis based on routine programme data, which may have been incomplete, and some training measures, including practical skills checklist data and TDTR and RTTDC knowledge test scores, were not electronically captured. Second, knowledge outcomes were assessed using written tests and may not fully reflect clinical performance in real-world emergencies. Third, the study period ended in July 2020, preventing evaluation of the long-term sustainability of the trainer network. Fourth, although sensitivity analyses showed no significant differences in baseline knowledge, sex or cadre distribution between repeat and non-repeat participants, repeat participants were younger and all attended sessions in Sindhupalchok, reflecting programme delivery structure rather than random selection. Finally, direct clinical or system-level outcome data, such as survival, complication rates or referral quality, were not collected; proxy indicators are provided but are observational and should be interpreted cautiously. Conclusion The RETSS programme demonstrates that a locally led, tiered training model, supported by strong partnerships, local trainer development and context-adapted delivery, can strengthen trauma and emergency care training capacity in disaster-affected, resource-constrained settings. By reaching a broad frontline workforce and being associated with substantial knowledge gains, RETSS may support emergency care readiness and resilience in highly affected districts of Nepal. Integration of RETSS into national emergency preparedness strategies could support sustainable workforce development and provide a practical, replicable blueprint for other LMICs facing post-disaster recovery challenges. Supplementary material