Kassam, Faazil; Damji, Karim F. MD, MBA; Kiage, Dan MBChB, MMed; Carruthers, Chris MD, MBA; Kollmann, K.H Martin MBChB, MMed, MD, Diploma in Tropical Medicine and Medical Parasitology, MBA
As academic hospitals strive to excel in patient care, teaching, and research, they are committed to the ongoing professional development of medical staff, including physicians, nurses, technicians, and others. Through ever-expanding opportunities for international medical education experiences, medical staff can learn much from institutional and patient encounters in the developing world while sharing their time and competencies with counterparts in a mutually beneficial manner. Underlying such exchanges is the fact that academic hospitals in the developing world continue to struggle under enormous human and financial resource constraints, with capacity shortages in countries that have been hit hard by wars, tyranny, or natural disasters.
The Millennium Development Goals are the world's time-bound and quantified targets for addressing extreme poverty in its many dimensions—income poverty, hunger, disease, lack of adequate shelter, and exclusion.1 As the international community has adopted these commitments, it has become clear that health and development can only be attained together by paying attention to resolving human resources constraints. A key challenge to achieving these goals is the low supply of trained health workers in developing regions, worsened by rising migration. Sub-Saharan Africa, for example, has the lowest ratios of health workers to population anywhere in the world. In recent years the global health policy environment has begun to acknowledge the value that strong human resources have for the health sector as well as for development in resource-poor areas of the world.2,3
Health care systems in many developing countries are undergoing dramatic transformation characterized by rapid urbanization, aging populations, and a gradually changing epidemiological profile. Health care providers are under pressure to meet rising demands for basic and evolving specialized health care services. Such circumstances can encourage the development of high-quality subspecialty services if the right resources are available.
To address these challenges and opportunities over the long term, it is critical that institutions in developing nations have access to appropriate human resources and global perspectives on higher education and research, as well as a sense of complete dedication to the highest educational standards.4 It is within this context that institutions in developed nations could engage in partnerships with institutions in the developing world to enable sustainable capacity development of health care institutions and professionals based on the mutual sharing of knowledge and skills. Systematically linking professional development of an individual fellow from a developing world institution with capacity development at the fellow's home institution further enhances empowerment and motivation as it reduces the risk of frustration and brain drain. The beneficiaries of this process are the hospitals, their staff, and, most important, patients at both institutions.
An emerging component of patient care in the developing world is adequate access to subspecialty care. To address a rapidly changing epidemiological profile along with new challenges and patients' demand for quality care in the developing world, it is high time to develop models for sustainable quality subspecialty services that are appropriate to the local environment while achieving global best practice standards.
We describe the “Sandwich fellowship,” an educational program piloted in collaboration with The Ottawa Hospital at the University of Ottawa in 2007–2008, that can facilitate the goals outlined above. Because similarly structured programs already exist, we wish to describe our experience with a model that, if largely implemented across academic institutions, could have profound effects on the quality of care received by many in areas of the world currently struggling to meet the needs of their patients.
During the last 25 years, the postgraduate fellowship has developed into a mentorship model that enables trainees to augment their knowledge, skills, and approaches in a highly subspecialized area. Fellowships also develop leaders in clinical, educational, and research realms that can guide larger departmental or institutional direction.5 Traditionally, fellowships take place between a trainee and one or more mentors at a single institution. Training fellows for the developing world has conventionally entailed candidates from a developing country coming to host institutions in the developed world for 12 to 24 months of training and subsequently returning to their home countries to practice their newly acquired skills. However, a significant limitation to the traditional model has been the lack of an enabling environment and/or institutional capacity in the fellow's home country. Such environmental limitations prevent the new fellow from practicing in a meaningful way the new skills he or she has learned, to the disadvantage of both the patients and the fellow. This can result in the fellow electing to license and practice elsewhere if there are no other pathways of career development in place in the fellow's home country.
The Sandwich fellowship model aims to link the fellow's step-by-step development with the development of institutional capacity to create an enabling environment for professional growth, ultimately benefiting patient care. The program involves the fellow completing a series of rotations spent in both a developed world institution as well as a home-based institution that shows a commitment to the development of subspecialty areas and a capacity relevant for the practice of that subspecialty. The unique name of the fellowship model is a nod to the fact that the candidate's training experience occurs in cumulative layers at diverse geographic sites, including at the fellow's home institution where subspecialty capacity is enhanced. Initiatives in advancing subspecialty care can be mutually beneficial to both institutions and can be facilitated by agreements such as a memorandum of understanding or any other formal partnership agreement. This overarching concept of joint institutional training has been used elsewhere in “Sandwich PhD” programs like the ones at the University of California, San Francisco; John Hopkins University; and others.6–9 Other Sandwich programs in the clinical context occur through partnerships shared by the University of Munich and Herzog Carl Theodor Clinic with University of Nairobi and Kenyatta National Hospital, whereby the University of Munich and partners offer education in pediatric ophthalmology and vitreoretinal training occurring at both sites. To our knowledge, the Medical University of South Carolina; University of California, San Francisco; and University of California, Los Angeles have offered similarly structured programs for ophthalmology fellowship training (Yonas Tilahun, MD, assistant professor of opthalmology, Addis Ababa University, Ethiopia, e-mail communication, December 12, 2008).
The Sandwich model intentionally links institutional capacity development of a committed local partner with access to quality training opportunities for the fellow in diverse contexts. Preceptors also gain from continued professional growth and development in a developing world setting that results in exposure to a spectrum and severity of cases not typically seen in their home populations.
In the context of this model we use the common pairing of an institution in the “developed world” with an institution in the “developing world” to frame the Sandwich fellowship, but it should be noted that the most important outcome of any pairing is that institutions in need of subspecialty development collaborate with institutions that have the educational resources to help meet those needs, irrespective of the classification of the countries to which they belong.
Fellowship steering committees
Before any program details can be determined, the initial step in creating the Sandwich educational model is selecting a partner institution through consideration of various criteria and guiding principles (List 1). Subsequently, integral to the program is the establishment of fellowship steering committees at both institutions who oversee the program and are committed to its success. The steering committees are charged with identifying relevant frame conditions, defining the rationale for the partnership, and developing a structured approach to meeting mutually defined institutional needs and goals. Such needs may include operational plans involving rotation scheduling of individuals, budget development, orientation and logistics, and establishing an evaluation framework. The fellowship committee is also responsible for handling matters pertaining to licensing and credentialing, which take place through the normal college and university processes. An important factor for licensing at the fellow's home institution is ensuring adequate time for processing, as delayed licensure can potentially create many obstacles to rotation scheduling.
To meet the goal of capacity building at the partner institution in the developing world, a suitable candidate should be selected in a specialty chosen on the basis of local needs. Successful candidates must exemplify strong clinical skills, motivation to pursue teaching and research, potential to lead a department, and the ability to recruit and retain individuals to form a high-performing team. There is no single way in which a candidate is selected. Preceptors may identify and pursue potential fellows during ongoing international collaboration. Alternatively, the application process may be initiated by a potential fellow with expressed institutional support. In such cases, the partners assess the application and preceptors conduct an interview process. The application package includes a letter from the candidate indicating motivation, current skill set, future career path in academic institution, and how the fellowship will help further such a career path. Additional application materials include the candidate's medical school transcripts, a curriculum vitae, a copy of the candidate's current medical license, an evaluation from the residency training program director, peer evaluations from those who can attest to the character and skill set of the candidate (normally a peer in the same field as well as department head), and a letter from an individual in senior management at the candidate's institution indicating support and clear commitment to resource allocation for the fellowship and subsequent development of the subspecialty. The successful candidate should also have a strong command of English, as this has become the international language of science and teaching in most developed regions. If the preceptor requires, this may involve various forms of testing, including a formal Test of English as a Foreign Language exam. The fellowship committee, led by a fellowship training director (List 2), subsequently approves the recommended candidate. We encourage candidates' commitment to return to their home institutions; however, this does not preclude the use of return-of-service agreements between the home institution and the potential fellow.
The following is an example of a rotation schedule for a fellow that involves rotations at the institution in the developed world as well as preceptor visits to the fellow's home institution in the developing world. The duration of rotations depends on the area of subspecialty, and the fellowship duration can be extended reasonably with support of the training institution. Rotation durations may vary and must account for factors such as sufficient family time, periods of low clinical activity at the training institution, orientation, etc. However, benchmarks of actual rotations (typically 9–12 months) should be developed in keeping with the overall spirit of the program and to ensure that quality is not compromised. Throughout the rotations overseas, fellows maintain an electronic education link to promote ongoing education and research and to access the mentor's social and professional networks. Such links are developed in conjunction with comprehensive and ongoing support for the overseas setting including equipment, midlevel staff training, consumables, etc.
The preceptor visits the selected fellow for one to two weeks at his or her home institution with a twofold purpose: first, to work with the fellow and his or her institutional administration to engage in a needs assessment to develop a mutually agreeable plan. The assessment should ideally target a specific area of new knowledge or a specific service. The plan should also include details of support for the fellow in addition to identifying the investments the institution requires to implement subspecialty services (e.g., space, equipment, nurses, technicians). The second purpose of the visit is to work with the fellowship candidate to ensure that he or she merits further training and also possesses the desired attributes to work well in the health care system of the training institution.
The fellow spends three to four months at the institution in the developed world. This involves observing and assisting in the clinic and/or operating room depending on the area of training. The fellow should also develop an understanding of the environment necessary for the development and practice of the subspecialty. It is within this rotation that subspecialty development is best paired with the development of management and leadership skills. Integration of clinical exposure with sessions in health management is an important component of the program, and this new knowledge can be applied during subsequent rotations at home.
The fellow returns home for a period of three to four months. During this rotation, the fellow implements and continues to build on knowledge and skills acquired during the previous rotation and collects patients to be seen by the preceptor during a structured visit. E-consultation and learning, combined with the preceptor's visit, allow the fellow to learn with guidance on patients with pathology common to that geographic location, and also to discuss all key areas of institutional capacity development with key decision makers including equipment, space, and human resource development.
Following Rotation 1 and between Rotations 2 and 3 are important periods of building and strengthening the home setting comprehensively to achieve the desired level of support agreed on and documented during Rotation 1. Once this has been achieved against an agreed protocol with measurable outcomes, the preceptor is invited for Rotation 3. This is essential in order to practically implement the areas learned from the previous rotation.
The fellow returns to the specialized training institution for the final four- to six-month rotation. This involves clinical and/or surgical exposure as well as ongoing seminars to ensure the fellow's knowledge base is at the level expected for graduation.
The fellow resumes work at his or her home institution while maintaining collaboration with the training institution. Collaboration primarily occurs electronically between the fellow and preceptor. The preceptor should serve as a long-term mentor to the fellow, providing guidance and acting as a resource for ongoing consultation, research, case discussions, etc., but the preceptor can also pursue follow-up visits for monitoring and further teaching.
There are several possible funding sources for fellows and mentors participating in a Sandwich fellowship: institutional funding from universities, grant agencies, industry, public–private partnerships, and, potentially, the government of the developing nation. It should be noted that the Sandwich model has a potential cost reduction compared with the traditional model because the fellow's family may remain at home instead of joining the fellow overseas, although strong nonfinancial family support is necessary.
It is important for the fellow's home institution to financially support various aspects of the fellowship in order to give the institution greater interest in the fellow's professional progress, and also to make the fellow aware that institutional support is present during training and will remain afterward. This knowledge can encourage the fellow to remain at his or her home institution after completing the fellowship. The detailed costs should be negotiated between the two institutions.
In the program model we suggest, the fellow's home institution should be responsible for fellow and preceptor travel, local preceptor accommodation, and the fellow's tuition, salary, and living expenses while away. Alternatively the tuition, salary, and living expenses could be assumed by external funding if obtained.
There are three essential areas of evaluation in the proposed fellowship model: the fellow, the institution in the developed world, and the institution in the developing world. The evaluation of the fellow largely follows that of a traditional fellowship: clinical acumen, surgical skill, patient rapport, and objective evidence of knowledge in the basic and clinical sciences in the subspecialty area. These are all taken into account during the evaluation process through various assessments and examinations. Evaluation of the developed world institution primarily takes place through evaluating the delivery of the educational program, because other areas, such as preceptor professional growth, are difficult to assess. Fundamental to the evaluation of the educational program itself is direct feedback from the fellow and his or her home institution, both during and after the educational experience, to allow the fellowship steering committee and preceptors to improve in subsequent iterations. Finally, evaluators must monitor outcomes and impact at the home institution in terms of subspecialty service development (i.e., number of patients seen for a service both prior and subsequent to the fellowship, quality of services/outcomes, community impact, and evidence of sustained institutional capacity-building against agreed indicators).
The Sandwich fellowship provides trainees from the developing world with improved access to training opportunities in diverse learning environments, enhancing the educational experience for fellows. The graduate by the end of training will have significantly expanded the breadth and depth of his or her competencies with internationally recognized levels of knowledge, skills, and attitude in subspecialty patient care, research, teaching, and management methods. The fellow's new competencies, in conjunction with a better understanding of policy and administration, will not only directly benefit patients but also advance the department and institution at which he or she serves. At the same time, the Sandwich model systematically builds local capacity for the subspecialty, meeting international standards and enhancing long-term sustainability. Engaging in this type of partnership is mutually beneficial, as the preceptors gain from continued professional growth in a challenging developing world setting that allows individuals to encounter a unique pathology pattern not commonly seen in their local populations. In addition, there is tremendous opportunity to improve cultural awareness and sensitivity, which would be of benefit to faculty, staff, students, and patients at the preceptor's academic institution. The academic hospital in the developed nation benefits from an improved social responsibility profile for supporting global outreach and also has the opportunity to establish similar programs for residents and medical students, who gain from diverse learning opportunities in the developing world as well.
Risks and Mitigation Strategies
A significant concern regarding the Sandwich fellowship is the possible compromise of training opportunities for residents, which is largely dependent on the educational capacity of the developed world institution. However, in a Sandwich model, the fellow is spending less time at this institution as compared with a traditional fellowship model, because the fellow returns early on to his or her home institution to begin applying new knowledge. Involving the fellow in a manner that adds value to resident education (e.g., via teaching in clinical setting as well as in the wet lab) and having the fellow present at more than one local hospital for the training program can also help manage capacity issues.
Another potential risk is that the fellow will seek licensure and remain in the developed nation. The Foundation for Advancement of International Medical Education (FAIMER) acknowledges that developing skills, creating professional networks, and enhancing opportunities for career advancement may be important elements in diminishing such migration.10 However, the institutional capacity-building component of the Sandwich model intends to systematically provide the fellow with a pathway for professional growth and development at home, making migration less likely. Moreover, the fellow maintains professional and socioeconomic links with his or her home setting as he or she continues to see patients and spends more time at home compared with traditional fellowships, also making it less likely for the fellow to remain abroad.
Other considerations include the possibility of the fellow not receiving adequate hands-on training in the developed country. In this case, Rotations 2 and 4 above can be minimized and more emphasis placed on the fellow's home institution rotation. Finally, it is possible that the preceptor is unable to travel to the destination of the fellowship candidate because of factors such as political instability. Under this circumstance, rotations may be delayed or the program could revert to the more traditional model and increase rotation time at the developed world institution.
The University of Ottawa Departments of Ophthalmology and Orthopedic Surgery have piloted the Sandwich model in 2007–2008, partnering with the Aga Khan University Hospital in Nairobi, Kenya (AKUHN). The head of the section of ophthalmology at AKUHN has participated in a glaucoma fellowship with an associate professor of glaucoma, and an orthopedic surgeon from AKUHN was a joint surgery fellow with an assistant professor of surgery. Although we offered a general description of our model above, we describe the case of the head of the section of ophthalmology at AKUHN as a specific example.
The candidate (D.K.) was initially encountered by the preceptor (K.F.D.) during international work being conducted at the AKUHN and displayed many of the desired qualities for fellowship training. Both the fellow and the institution expressed strong interest in the development of glaucoma-related subspecialty services to serve patients of the hospital and the region of East Africa more broadly. A number of individuals had applied for glaucoma fellowship training in Ottawa, and these applications were reviewed with the fellowship program director. The concept of a Sandwich fellowship was raised during the review, as well as with the postgraduate training office, and support was garnered from both the Department of Ophthalmology as well as the University of Ottawa. The proposal for this program was presented to the Residency Training Committee at the University of Ottawa to ensure that the fellow's clinical and surgical experience would complement rather that detract from achieving glaucoma-related clinical and surgical objectives for residents in training. One key aspect of this was the understanding that the fellow would cosupervise (along with the main preceptor) residents in weekly glaucoma clinics, and that the fellow would participate in surgery as appropriate when residents were not present, or in cases that were more complex and felt to be outside the scope of what would be expected for the resident in training to participate in. It was also agreed that the fellow would host a few wet labs for residents on basic trabeculectomy skills.
After a review of candidates, the candidate from AKUHN was then suggested as the best candidate, based on expertise and merit (clinical knowledge and skills, aptitude, cultural sensitivity, etc.) as well as support provided by AKUHN, and was subsequently invited for further training.
Funding was obtained from multiple sources, including AKUHN, industry, and the University of Ottawa. The AKUHN provided airfare for the fellow, accommodation for the preceptor during visits, and acquisition of the relevant equipment and development of human resources. An unrestricted educational grant from Allergan Canada was used to support preceptor travel as well as the purchase of educational materials. The University of Ottawa postgraduate department provided funding for the fellow's salary and living expenses.
This one-week rotation took place at AKUHN with the preceptor and candidate evaluating glaucoma patients and offering treatment as appropriate. Meetings were also held with the head of surgery, the postgraduate medical education dean, and hospital administration to secure support for development of a glaucoma subspecialty service at AKUHN.
The 10-month period between Rotations 1 and 2 was used to map out details of clinical, educational, and research experiences, to obtain approval for the program at both institutions, and to secure an educational license for the fellow and funding for the program.
The candidate spent three months at the University of Ottawa Eye Institute with the primary preceptor and two secondary preceptors in glaucoma. This included seeing patients in the clinic and assisting and performing a variety of glaucoma procedures (using lasers as well as incisional procedures). A few days were also spent on the glaucoma diagnostic service observing and learning about visual field machines, optical coherence tomography, Heidelberg retinal tomography, ultrasound biomicroscopy, etc. Seminars took place one to two times per week for one hour at a time. Shields' Textbook of Glaucoma was covered chapter by chapter, and additional topics were discussed depending on interesting or complex cases seen in the clinic or operating room setting.
Objectives were also established for clinical research training, including attending workshops on scientific writing, research ethics board guidelines, development of grant proposals, research funding opportunities, epidemiological and statistical considerations in research study design and analysis, University of Ottawa Department of Ophthalmology research meetings, and publishing at least one paper involving a small-scale clinical trial and/or review of an important topic related to glaucoma in East Africa.
Additional time was spent shadowing the primary preceptor in his role as residency training director and chair of the Ottawa Eye Institute Working Group on National and International Partnerships, as well as attending a conference for physician executives in management training. This yearly conference organized by Canadian physicians provided useful insights and models for health care leadership and management, and costs for this conference were waived.
This six-month rotation took place at AKUHN and involved the primary preceptor seeing patients and supervising the fellow conducting surgery. This period was used to strengthen and build the AKUHN setting for future teaching and subspecialty care. This included the acquisition of necessary diagnostic, laser, and surgical equipment as well as providing extra space and hiring ophthalmic technicians and nurses.
This five-month rotation took place at the Ottawa Eye Institute and involved the fellow spending time in both adult and pediatric glaucoma with continued seminar participation. The fellow also participated in resident glaucoma clinics, organized a few wet labs for residents, and assisted with glaucoma emergencies. The fellow also assisted in organizing a session for residents, “Glaucoma Bell Ringer,” in which rare but important cases of glaucoma were brought in. During the session, patients were put in examining rooms as unknowns, and residents went around simply looking for clinical findings and came up with differential diagnoses. The fellow and preceptor then reviewed each case with the residents. The fellow also presented at grand rounds, at professors' rounds teaching sessions for residents, and at the glaucoma journal club. Additionally, the fellow had an opportunity to conduct research and present findings at the Association for Research in Vision and Ophthalmology 2008 meeting in Fort Lauderdale, Florida, at the Canadian Ophthalmology Society Annual meeting in Whistler, British Columbia, and at the annual Eye Institute research day in Ottawa. Short elective experiences were also organized with preceptors at other institutions to complement experiences from the Ottawa Eye Institute. For example, the fellow spent one week with the director of the Glaucoma and Advanced Anterior Segment Surgery Fellowship Program at the University of Toronto learning about newer approaches to glaucoma surgery, such as nonpenetrating glaucoma surgery, canaloplasty, etc., and the fellow also spent one week with the residency training director at the University of Alberta learning about a teleophthalmology model.
Based on clinical evaluations over time as well as an assessment of research experience and productivity submitted to the fellowship training director for ophthalmology at the University of Ottawa, the fellow was presented with certificates from the University of Ottawa as well as the Department of Ophthalmology for successful completion of a “clinical and research fellowship in glaucoma.”
The fellow returned to AKUHN and remains in touch electronically with the primary preceptor to discuss cases. A follow-up visit took place during which a small team visited, including the preceptor, a retina specialist, a visual field specialist (PhD), and a teleophthalmology consultant (MBA). Objectives of the visit were to conduct sessions for professional development of ophthalmologists and allied health personnel, explore the feasibility of a Sandwich fellowship model for medical retina training, and meet with administration to discuss evaluation of the glaucoma fellowship and opportunities for future mutually beneficial institutional collaboration. In the course of this visit, a CME session in glaucoma and retina was hosted by the Aga Khan University Hospital for the ophthalmic community in Kenya. During that session, the preceptor informed those attending that the fellow is now fully trained and that the preceptor has a high degree of confidence in his abilities. The glaucoma service also benefited from training sessions, which took place with technicians (clinical, IT, biomedical) and nurses to develop capacity to perform high-quality visual fields and stereo optic nerve photographs and to review preventive and ongoing maintenance schedules for ophthalmic equipment. The retina specialist attended clinics and laser sessions which had been organized by an ophthalmologist at the AKUHN interested in further training. Meetings also took place with the university and hospital administration at AKUHN to provide support in strategic planning and to discuss the feasibility of e-consultation, e-learning, and e-research via a teleophthalmology program. The teleophthalmology plan, developed jointly by the former fellow and preceptor, proposes that a reading center be established at AKUHN for retinal and optic nerve disease (primarily geared toward diabetic retinal disease and glaucoma), and that pilot satellite centers be established in a few underserved communities with a nurse or technician sending information on patient history and clinical/diagnostic exams (slit lamp and retinal photos; other tests as appropriate) to the hub at AKUHN. The newly trained glaucoma specialist will coordinate reading of images and will be assisted by a colleague eventually trained via a Sandwich program for medical retina training. Difficult cases will be sent to the Canadian preceptors in glaucoma and retina for further evaluation, and, over time, teaching and research programs are envisioned as part of this teleophthalmology collaboration. The program envisions some capital, equipment, and human resource funding in the first year, which would be recovered through user contributions during two years. Additional revenue gained after year 2 would support expansion of the program and development of education and research infrastructure.
The evaluation of the fellow included both overall assessments as well as a focus on glaucoma-related objectives (Appendices 1 and 2). Importance was also placed on evaluating the application of leadership and management skills in relation to institutional capacity development for glaucoma subspecialty services, as well as self-evaluation and creating self-directed learning objectives. Throughout the rotations, the fellow and preceptor exchanged ideas and drafts of proposals and papers on various research topics. This research collaboration was highly successful and resulted in four completed papers, all either submitted or accepted for publication.
Our experience: Successes and challenges
There have been many key factors that were integral to the success of the program. One of the most important contributors to our success was obtaining sufficient funding, both for the infrastructure and human resource development in the developing world institution as well as for the fellowship expenses (i.e., travel, mentor compensation, accommodation, living expenses, educational material, and the fellow's salary stipend). Selecting the most promising candidates is necessary for success, but allowing a long lead time (six months or so) to facilitate accommodations, licensure, visas, etc., ensures that fellows will be able to begin the program on schedule. We also found that offering orientation or hospitality modules at either site helped keep things on track, as did establishing clear educational objectives and hands-on training opportunities for the fellows. Our pilot programs would not have been successful without the highly motivated preceptors in the host institution and appropriate support from administrative personnel in the developing world institution. Success also required support from many other physicians, nurses, and residents at the fellow's home institution, as well as support from additional members of the developed world health care team for training individuals as appropriate on various pieces of equipment. We emphasize the importance of evaluating outcomes for both the fellow and the program and to continuously improve based on those outcomes. Finally, without the willingness of postgraduate medical education offices at both institutions to proceed, the program would not have been possible.
Of course, we also encountered unforeseen challenges. Because of an unexpected political conflict, a preceptor's visit to the fellow's home institution was delayed during Rotation 3. Other delays occurred as a result of logistical challenges, such as difficulty obtaining visas, accommodations, licenses, etc., which caused the fellows to begin later than expected. Finally, we encountered difficulty in determining the degree of structure in the program and in determining the amount of responsibility the fellow would be given. These challenges both emphasize the importance of defining a narrow clinical objective. It should be noted that many potential challenges would be specific to the area of subspecialty practice.
Subsequent iterations of the Sandwich fellowship in the same subspecialty area would allow a future fellow to receive guidance from the previous graduate. This model provides a home-based preceptor, thereby creating a more supportive teaching environment. Further, the Sandwich fellowship model can serve as a platform on which to build a formal partnership (Figure 1) that could potentially involve the creation of several other programs. Possibilities include educational and collaborative research opportunities, management and leadership training, elective term rotations for residents and medical students both ways, and telemedicine opportunities for enhanced patient care, education, and research. The long-term goal would be to expand the scope of the fellowship further to include greater emphasis on teaching and management, allowing the fellow to return home and contribute to the development of the subspecialty in his or her health care system. Once the academic hospital sustainably achieves adequate institutional capacity, there is also the promise of creating local or regional fellowship programs that could initially be supported by the overseas partner but, eventually, become fully independent.
The purpose of the Sandwich model is to build institutional capacity in areas in the developing world in which improvements in health care delivery can directly affect the quality of life for many. It also has the potential to serve as a gateway to a long-term partnership that would result in the mutual exchange of knowledge and skills that would benefit both institutions. Through implementation of this model, academic institutions have the opportunity to introduce a creative way for the improvement of subspecialty care in the developing world. The Sandwich fellowship does more than that, however, involving the expansion of human resources while creating an enabling environment at home that allows both personal and professional growth of the fellow, reducing the risk of brain drain (List 3). The model fits the FAIMER framework of human capacity-building for the developing world by identifying young and talented candidates with the potential to become agents for change, delivering an effective learning intervention relevant for their environment, allowing the application of acquired knowledge and skills with support, and promoting development of a career path with opportunity for growth.11 As compared to a traditional fellowship, it permits some continuity of care while individuals are at home, and a higher likelihood of being able to integrate learning acquired into academic practice. The Sandwich model is an innovative approach that combines training in the developed world along with development of institutional capacity in fellows' home environment, but it requires dedicated champions at host and partner institutions to make it a success.
The authors thank Ottawa Hospital for support of interdepartmental partnerships with Aga Khan University Hospital Nairobi; the University of Ottawa postgraduate medical education department for support of this model as well as funding for the glaucoma fellowship; the University of Ottawa Eye Institute for support of the glaucoma fellowship model; Allergan Canada for unrestricted educational grant support toward the glaucoma fellowship model; the University of Alberta, University of Calgary, and University of California, San Diego for donation of ophthalmic equipment and educational material; Dr. Amin Kherani, Dr. Lyne Racette, Dr. Matt Tennant, Dr. Ike Ahmed, and Abshir Moalin for ophthalmic education and/or consultation support; the Rotary Club of Mayfield for a preceptor travel grant; Dr. Mateen Ahmed and New World Medical, Inc., for donation of Ahmed valves; Allergan Canada, Alcon Canada, Pfizer Canada, and Merck Frosst for providing ophthalmic pharmaceutical supplies during trips to Kenya; the Aga Khan University Hospital Nairobi for the support of the glaucoma and orthopedics fellowship programs; The Ottawa Hospital Department of Orthopedics; and the Ludwig–Maximillian–University Munich and Herzog Carl Theodor Clinic, Germany, and the University of Nairobi and Kenyatta National Hospital for demonstrating the effectiveness of Sandwich training in ophthalmology subspecialties.