Strengthening pediatric surgical readiness: outcomes of a needs-directed short course for non-specialist physicians in Rwanda

Abstract

Introduction

Although the recent expansion of pediatric surgical services in Rwanda has improved patient outcomes, pediatric surgery specialists are available only at referral hospitals. Children with surgical conditions are often first evaluated by non-specialist physicians at district or regional hospitals prior to definitive transfer. Knowledge gaps exist in the initial management of pediatric surgical conditions among non-specialist physicians. This study assesses the impact of a needs-directed short course for non-specialist physicians in Rwanda.

Methods

Nine prioritized pediatric surgical conditions were identified through an expert-led modified two-stage Delphi. Course content was created to fill knowledge gaps, determined through a needs assessment. Pediatric surgeons and local trainees delivered the course to non-specialist physicians at five district and regional hospitals. Participant reactions and knowledge were analyzed using RStudiov1.1.4, and thematic analysis was performed on free text responses.

Results

Fifty-nine non-specialist physicians with heterogeneous experience participated. Participants rated the course highly, and self-rated knowledge and confidence increased on all conditions. Overall knowledge-based scores and initial management for all conditions improved substantially. More participants indicated that they knew how to contact a member of the pediatric surgical team after the course. Positive themes included acquisition of knowledge that was relevant to practice, effective course execution, and the interactive nature of the course. Themes for improvement centered on desire for more content and course availability.

Conclusions

As pediatric surgical capacity improves, safe initial management is critical. A novel, needs-directed short course for non-specialist physicians appears to be an effective way to address knowledge gaps and improve the referral process through direct connection with specialists. Providing such a course on a regular basis to senior medical students and new graduates is a promising avenue to improving the provision of safe, timely pediatric surgical care.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12893-025-03281-1.

Introduction

Limited access to pediatric surgical care in low- and middle-income countries (LMICs) leads to morbidity and mortality that is significantly disproportionate to outcomes in high-income countries [1]. Although efforts to improve pediatric surgical training, access, and impact are increasingly showing improved outcomes, specialized services are often only available at referral or tertiary level hospitals [2–5]. Thus, a large proportion of children in LMICs, including Rwanda, receive care at first-level (district) or second-level (regional) hospitals, which play important roles in diagnosis, resuscitation, and referral to higher levels of care [6–8]. Understanding how these challenges are addressed within specific health systems is critical, as local structures can shape both barriers and opportunities for pediatric surgical care.

Rwanda provides a particularly important case example as its decentralized healthcare model within a universal health insurance program is designed to expand access to care at the population level, especially in rural areas, by relying on district and regional hospitals to triage and refer a proportion of the patients who need specialist care. The majority of children presenting to a district or regional hospital are evaluated and managed by non-specialist physicians (medical doctors with no specialist residency training) prior to definitive transfer [9]. Because the growth of pediatric surgical services in Rwanda has been rapid, the landscape of care may be different than the one in which non-specialist physicians have trained. Improving the ability of non-specialist physicians to care for children needing surgery is critical to reducing delays in care and improving outcomes [10, 11]. Thus, continuing education of physicians at first and second-level hospitals is necessary as care of the pediatric surgical patient and neonate advances. Recent estimates suggest that scaling up surgical care at district hospitals in LMICs could avert over 314,000 deaths annually among individuals under 20 years of age [12]. Studies in high-income and upper-middle income countries have suggested the utility of providing specialist training on specific topics to non-specialist physicians [13–15]. However, there remains a gap in educational strategies tailored specifically to the frontline of pediatric surgical care, especially in a decentralized health system.

To address this gap, we developed and implemented a one-day short course for non-specialist physicians focused on the diagnosis, stabilization, and safe transfer of pediatric surgical patients, piloting it across five district and regional hospitals. A needs assessment was previously conducted to identify gaps in knowledge on contextually-prioritized conditions (Hsu et al., unpublished), allowing us to tailor this course to specific needs. We hypothesized that this needs-directed short course would increase both confidence in caring for children with surgical conditions as well as objective knowledge.

Methods

Nine prioritized neonatal and pediatric surgical conditions were identified through two separate modified Delphi processes involving international and Rwandan pediatric surgeons, anesthesiologists, and general practitioners, modified to use a literature-based confirmatory questionnaire and streamlined rounds with new panelists [16]. The nine conditions were gastroschisis, omphalocele, intestinal atresia, incarcerated inguinal hernia, anorectal malformations, intussusception, undescended testes, testicular torsion, and abdominal solid tumors. Knowledge gaps regarding the diagnosis, stabilization, and safe transfer of patients with these conditions were identified through a needs assessment conducted at five district and regional hospitals spanning the geography of rural Rwanda (Hsu et al., unpublished). Briefly, we utilized the results of the above Delphi processes to create a needs assessment evaluating participants’ ability to diagnose, initially manage, and perform timely transfer of children with these conditions. Content validity and face validity were confirmed by a Rwandan pediatric surgeon. The needs assessment was refined using cognitive interviewing with Rwandan general practitioners and surgical trainees. All non-specialist physicians, which included medical interns (doctors in their first year of practice) and general practitioners (GPs), at five district and regional hospitals spanning the geography of rural Rwanda were invited by their hospital director general to complete the needs assessment via REDCap.

A day-long course on the initial management of the above nine conditions was developed and delivered by Rwandan pediatric surgery fellows in collaboration with an American general surgery trainee (Supplemental File 1). Competencies and course objectives were determined based on the identified knowledge gaps. Course content was developed to meet the competencies and objectives, referencing existing literature and adapting it to local needs [17–19]. The course included a pre-test, nine interactive lectures, nine interactive case discussions, two hands-on simulations (coverage of gastroschisis and manual reduction of incarcerated inguinal hernia), and a post-test (which used the same items as the pre-test). The course was reviewed by multiple individuals with extensive experience in the field, including pediatric surgery fellows, an American pediatric surgeon, and a Rwandan pediatric surgeon. The senior author, a Rwandan pediatric surgeon, ensured that recommended practices were optimized according to the resources and referral systems present in Rwanda.

The Rwanda Medical and Dental Council approved of the course and accredited it for five hours of Continuing Professional Development credit. IRB approval was obtained from the Rwanda National Ethics Committee (No.504/RNEC/2024), the University of Michigan (HUM00261603), and all district and regional hospital ethical boards (permission letters). Participation was voluntary, and participants were able to withdraw without penalty. All participants provided written consent to participate.

Distribution, participants, analyses

The course was delivered at five district and regional hospitals spanning the geography of rural and urban Rwanda, selected by convenience sampling (Fig. 1). Each is staffed by 10–20 non-specialist physicians, as typical among many hospitals nationally, though some hospitals (not studied) only employ 2–3 non-specialist physicians. Hospital characteristics otherwise varied, representing a range of capacities and patient populations served by other hospitals across Rwanda. The district hospitals are staffed by one general surgeon each, while the regional hospitals are staffed by at least two general surgeons. Only one hospital has advanced imaging technology including a CT scanner.

Fig. 1.

Location of hospitals at which needs assessment was conducted

All non-specialist physicians, which included medical interns (doctors in their first year of practice) and GPs at these five sites were invited to attend the course, which lasted one day. Each course was taught by at least one Rwandan pediatric surgeon or pediatric surgery trainee, as well as an American general surgery trainee.

The pre- and post-test, delivered via REDCap form, assessed self-rated confidence and knowledge, as well as knowledge taught in the course (Supplemental File 2). These correlate with the first two levels of Kirkpatrick’s Four Levels of Training Evaluation (reactions to the course, assessment of knowledge) [20]. Self-rated knowledge and confidence were assessed on a 5-point Likert scale (1 = not at all confident, 5 = extremely confident), Wilcoxon Signed-Rank Tests were used to determine differences in self-reported confidence and knowledge before and after the course, as well as differences in pre- and post-test scores. Chi-squared tests were performed to determine differences in the frequency of correct answers before and after the course. A statistical significance threshold of P < 0.05 was used. Analyses were performed using RStudio v1.1.4.

To better understand participant reactions to the course, we analysed free text written responses using grounded theory, an inductive methodology that generates conceptual categories from qualitative data [21, 22]. Transcripts were analyzed using a constant comparative method with no a priori hypothesis to generate emerging and recurring themes [23]. We used group open coding involving independent coding two investigators (PJH and PDR), with iterative comparison and refinement of codes, and resolved disagreements by discussion until consensus was achieved. This was performed manually without qualitative analysis software. Responses were grouped by theme, and frequencies and percentages of responses were calculated.

Results

Fifty-nine non-specialist physicians participated in the course, most of whom were intern doctors (42.4%) or GPs in their first three years of practice (39.0%) (Table 1). The majority of participants (66.1%) had a pediatric surgery rotation in medical school. Fifty-nine non-specialist physicians participated in the course, most of whom were intern doctors (43.3%) or GPs in their first three years of practice (38.3%). Participants rated the course highly; 98.3% agreed or strongly agreed that the course was beneficial and relevant to their practice, and that they would recommend it to a colleague (Fig. 2). Participants expressed that the interactive lectures, practice case scenarios, and instructors were all effective.

Table 1.

Demographics of participants and their practice sites

Respondent and Practice Site Characteristics	n (%)
Hospital
Hospital 1	5 (8.5)
Hospital 2	15 (25.4)
Hospital 3	11 (18.6)
Hospital 4	13 (22.0)
Hospital 5	15 (25.4)
Type of Practitioner
Intern	25 (42.4)
Junior GP (< 3 years experience)	23 (39.0)
Senior GP (≥ 4 years experience)	11 (18.6)
Had pediatric surgery rotation in medical school
Yes	39 (66.1)
No	20 (33.9)

Fig. 2.

Participant reactions to the course

Self-rated knowledge increased significantly across all conditions (mean increase: 1.1 ± 0.3 points, P < 0.001) (Fig. 3A). Similarly, self-rated confidence increased across all conditions (mean increase: 1.3 ± 0.2 points, P < 0.001)(Fig. 3B).

Fig. 3.

Pre- and post-course subjective and objective findings. A Self-rated knowledge, B self-rated confidence, C overall score, D percentage of participants who identified the correct initial management of six conditions, and E percentage of participants who felt that they could easily contact the pediatric surgery team. Error bars show standard deviation for numeric variables. Significance values: * p < 0.05

Overall objective knowledge-based scores increased by 18.9%, (69.2% to −87.1%,P < 0.001) (Fig. 3C). Initial management improved substantially for all conditions (P < 0.05) (Fig. 3D). 36.7% more participants indicated that they knew how to contact a member of the pediatric surgical team after the course (P < 0.001) (Fig. 3E).

Participants were also asked about their reactions to the course, including the best parts of the course and areas of improvement. All participants contributed free text responses. Positive themes included acquisition of knowledge that was relevant to practice, effective course execution, and the interactive nature of the course. Themes for improvement centered on desire for more content (specifically, case scenarios and simulations) and course availability (including course frequency and duration) (Table 2). Notably, availability of the course as a standard component of medical education was also identified as a theme.

Table 2.

Thematic analysis of course participants’ free-text reactions to the course. N indicates participants

	Themes	n	(%)	Example phrases
Best parts of the course	Acquisition of knowledge that was relevant to practice	27	45.8	"We discussed the most common surgical conditions we encountered in our practice, and we have learned the initial management we have to perform as GPs before considering transfer, which is key to better management of the patient"
	Effective execution of the course	22	37.3	"The instructors were the best. If the instructor is good, the course is awesome, and the content is also enticing"; The instructors made it comfortable enough for us to learn;" "The sessions were short and straightforward"
	Interactive nature of the course	28	47.5	"The instructors were the best. If the instructor is good, the course is awesome, and the content is also enticing"; The instructors made it comfortable enough for us to learn;" "The sessions were short and straightforward"
Areas of improvement	Increase in number of case scenarios and simulations	12	20.3	"More scenarios would be better so that we can understand it better"; "More simulation trainings are needed"
	Increase in course frequency	11	18.6	"Increase training sessions because most of us will not see these conditions so trainings could serve as a reminder"; "Sutainability in periodic training like every 3 months"
	Increase in course duration and having protected educational time	8	13.6	"Extend time of studying"; "Incrrease time"; "Maybe there next time there could be a day dedicated for training alone where doctors do not have other responsibilities"
	Availability of course as standard component of medical education	4	6.8	"Include in our curriculum during medical school training"; "We need more courses like this before starting practice"; "To be applied in every district hospital"

Discussion

In this study, we created and assessed the effectiveness of a short course for non-specialist physicians covering the diagnosis, initial management, and transfer of children with surgical conditions. Participants responded very positively to the course and reported improved knowledge and confidence. Objective measures of knowledge improved after the course as well. Thematic analyses identified a desire to increase the availability and frequency of the course. Our results suggest that even a short, targeted training can make a meaningful difference in how prepared non-specialist physicians feel when managing pediatric surgical cases. By improving care at the first point of contact, this kind of training could help reduce delays and lead to better outcomes for children. It also offers a practical, scalable way to strengthen surgical care in rural and decentralized health systems.

Previous work has demonstrated that short courses can be an effective way to improve the surgical care available to children in LMICs [24]. In Uganda, structured short courses on pediatric trauma and emergency surgery have demonstrated lasting improvements in provider knowledge and care processes [25–27]. Similarly, in Vietnam, a three-day short course helped develop national capacity in pediatric surgical emergencies through team-based training and simulation [28, 29]. These courses have primarily focused on pediatric trauma care and pediatric surgical emergencies, with an emphasis on providing training in definitive surgical care along with the initial diagnosis and triage, often with specialist physicians as the target audience [25–29]. Our work expands upon this foundation by describing the effectiveness of an educational strategy aimed at addressing the initial diagnosis and management of pediatric surgical conditions by non-specialists prior to transfer for definitive care. This work is particularly useful for decentralized healthcare systems, which are present in many LMICs as a way to improve responsiveness to local needs and triage patients who need subspecialty care [30, 31]. Our course improved non-specialist physicians’ confidence and objective knowledge in caring for pediatric surgical conditions at a level that is within their scope: providing timely diagnosis and stabilizing patients in a way that allows safe transfer to pediatric surgeons for definitive care. Importantly, collaboration with the Rwanda Medical and Dental Council to obtain Continuing Professional Development credit adds sustainability and validity to the course.

Our results suggest that, by focusing on knowledge gaps identified through a Delphi-based needs assessment, the course was aligned with the practical challenges and educational priorities faced by non-specialist physicians. We believe that the strong enthusiasm shown by participants highlight their recognition of the need for further pediatric surgical education. Thus, the course may improve the care delivered to children with surgical conditions, leading to faster diagnosis, proper stabilization of patients, and fewer avoidable delays in transfer. While comprehensive pediatric surgical care is critical for enabling the survival of otherwise fatal conditions in Rwanda, appropriate initial care may reduce complications and improve survival rates [2, 3]. Moreover, as over 80% of the participants were interns or GPs in their first three years of practice, if the immediate impact is sustainable, the course may continue to positively impact their work over the course of their career. However, the long-term impact of the course is currently unknown, and should be investigated with further study. More broadly, the approach of identifying specific knowledge gaps in non-specialist physicians and addressing them through context-appropriate short courses might offer a replicable model for strengthening the first line of care in other medical and surgical specialties. As many LMICs also rely on decentralized healthcare systems, there is potential for this model to be used outside of Rwanda as well.

The course was designed to communicate with participants that education about improved pediatric surgical care does not end at the conclusion of the course. As the course improved the ease with which participants were able to contact pediatric surgery, we speculate that follow-up communication can enable a continuous learning cycle, where challenges in management can be identified and addressed collaboratively. Although we did not formally study this, in the months following the course, participants have indeed used personal messaging with the course instructors (who are also authors) to ask for recommendations regarding non-urgent complex cases, a finding we were thrilled to see. While this follow-up communication was informal and not systematically tracked, they illustrate how the education could be enforced or augmented, refining how non-specialist physicians approach subsequent cases. Moreover, although we did not evaluate its impact, we heard in personal communication that participants expressed interest in sharing key takeaways with non-physician team members. A large number of patients who present to district hospitals are referred in by nurses who staff local health centers or brought in by nurses who staff ambulances. This enthusiasm suggests a potential future role for a “train-the-trainer” model, which could be explored to improve referral pathways and reduce delays in patient care prior to their arrival at the district hospitals.

Our study had several limitations. Although we had measured the first two Kirkpatrick levels (reactions to the course and assessment of knowledge), we did not measure the third and fourth levels (change in behavior and benefit to patients), both of which are important measures that best demonstrate a course’s outcome. To address this, future studies are planned to conduct longitudinal follow-up at 3 months after the course. Kirkpatrick’s third level will be assessed using surveys of participants to assess knowledge retention and behavioral changes, while the fourth level will be assessed by examining patient records to determine whether outcomes after the course were superior to those before the course. Additionally, several participants indicated that they could not dedicate their full attention and time to the course, as they were scheduled for concurrent patient care duties. Thus, it is possible that the full impact of the course may not have been realized. Moreover, because the same test items were used for both pre- and post-course assessments, some improvement may reflect a learning effect from repeated exposure to the questions rather than course content alone. However, the use of identical pre- and post-test items is a standard practice in educational research to allow for direct comparison, and the effect sizes observed suggest that true learning gains occurred beyond any learning effect. Finally, as we delivered the course at a limited number of hospitals, the results may not be generalizable to all non-specialist physicians in Rwanda, as different hospitals may be exposed to different conditions. However, the hospitals spanned the geography of Rwanda, with at least one hospital selected from each of the four provinces, decreasing the likelihood that geographical considerations impacted the outcomes.

Our findings point us toward next steps to maximize the course’s impact and enable its sustainability. Future work should expand the evaluation framework beyond immediate reactions and learning to include longer-term knowledge retention, changes in clinical practice, and effects on patient outcomes. Moreover, as identified by the thematic analysis, participants requested increased frequency of training as well as the availability of the course to all doctors. There are several next steps that could be taken to work toward this goal. One option would be to provide this course to senior medical students who have gone through their rotations and have the sufficient knowledge to make the most of the course materials. The course could also be integrated into continuing professional development activities, offering credits to meet the quota required for renewing medical licensing. Utilizing an interactive online module rather than an in-person course could improve the course’s accessibility. By taking on an asynchronous learning approach, we could allow non-specialist physicians to take the course on their own time, without the need to travel, while also reducing the time required of the pediatric surgery fellows to teach the course. Overall, increased availability of this training could improve the delivery of safe, timely care to pediatric surgical patients, leveraging the individual strengths of each component of the healthcare system.

Conclusions

As pediatric surgical capacity improves, safe initial management is critical. A novel, needs-directed short course for non-specialist physicians appears to be an effective way to address knowledge gaps and provide direct connection to specialists. Providing such a course on a regular basis to senior medical students and new graduates is a promising avenue to improving the provision of safe, timely pediatric surgical care.

Abbreviations

LMICs

Low–and middle–income countries

Authors’ contributions

Conceptualization: PJH, EN, RTP, BTA, RR; Data curation: PJH, DK, FI, DI, JPH, JN, AN, CDN; Formal analysis: PJH, PDR; Funding acquisition: PJH, RR, BTA, RTP, EN; Investigation: PJH, DK, FI, DI, JPH, JN, AN, CDN, ER, CU, PDR, AJN, VO, RR, BTA, RTP, EN; Methodology: PJH, PDR, EN, RTP, BTA, RR; Project administration: PJH; Resources: PJH, DK, FI, DI, JPH, JN, AN, CDN, ER, CU, RR, BTA, RTP, EN; Software: PJH; Supervision: RR, BTA, RTP, EN; Validation: PJH, DK, FI, DI, JPH, JN, AN, CDN, RTP, EN; Visualization: PJH, VO; Writing – original draft: PJH, DK, FI, DI, JPH, JN, AN, CDN, ER, CU, PDR, VO; Writing – review & editing: PJH, DK, FI, DI, JPH, JN, AN, CDN, ER, CU, PDR, AJN, VO, RR, BTA, RTP, EN.

Funding

PJH was supported by NIH T32 CA009672 University of Michigan Surgical Oncology Research Training Program and the HBNU Fogarty Fellowship.

Data availability

All data and materials are available upon reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

IRB approval was obtained from the Rwanda National Ethics Committee (No.504/RNEC/2024), the University of Michigan (HUM00261603), and all district and regional hospital ethical boards (permission letters). Participation was voluntary, and participants were able to withdraw without penalty. All participants provided written informed consent to participate. All research was conducted in compliance with the Helsinki Declaration.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.