Background

The scarcity of human resources for health is a global problem, which is further amplified by maldistribution between urban and rural areas within countries [1, 2]. In Tanzania, for example, the doctor to population ratio stood at 0.26 per 10,000 in 2012 [3], which is far lower than the recommended 1 to 5,000 [4]. On the other hand, despite the overall low doctor to population ratio, in rural areas, where 75% of the total population lives, only 31% of the available doctors practice in these settings [2]. This shortage impacts health care service provision, especially in places where it is much needed and has greater impact [4].

To reduce healthcare worker shortages in sub-Saharan Africa (SSA), many governments have encouraged the establishment of new medical schools (both public and private) and increasing student enrolment in existing schools. Within a 12-year period (1997 to 2009), Tanzania expanded the enrolment of medical students by five-fold in eight public and private health training institutions [5]. However, this expansion in enrolment has not been accompanied by adequate increases in faculty size or physical infrastructure. The Study on sub-Saharan African Medical Schools identified lack of basic infrastructure (reliable sources of power, teaching laboratories, libraries, teaching space, to mention a few) coupled with critical faculty shortages as impeding factors for the expansion of medical schools in Africa [5]. Consequently, classes were overcrowded, leading to compromise in the quality of training, especially in the clinical clerkships. Furthermore, most of the medical schools in sub-Saharan Africa are located at tertiary hospitals (which also serve as teaching hospitals). These teaching hospitals have limited clinical practice areas and mostly concentrate on curative care [6]. There is, thus, a dire need for innovative approaches to overcome these challenges.

According to a review by Barrett and colleagues, teaching sites outside the tertiary teaching hospitals have been found to be under-utilized. They represent potentially effective areas for medical education in low- and middle-income countries (LMIC) [7], which could solve some of the problems pertinent to increased numbers of medical students. It has been shown that peripheral placements can provide a high-quality clinical learning environment, hands-on practical experience, and at the same time students get acquainted with diseases that are common in communities and that are not referred to the tertiary teaching hospitals [8]. Students get the opportunity to see a wider spectrum of diseases, may have more patient-preceptor interaction, and learn the culture of the communities around the peripheral hospital [9]. Furthermore, peripheral placement inculcates the culture of teamwork within students [10].

Like many other medical schools in SSA which increased enrolment rate over the past five years up to 2010 [1], medical schools in Tanzania faced similar challenges. Thus, in Tanzania with only one medical school enrolling 55 first-year medical students in 1991, by 2015 the 11 Tanzania Commission for Universities (TCU)-accredited medical schools (both public and private) were enrolling a total 1,580 students per year (an average increase of 144 students per school) [11].

In 2010, the United States Government directed unprecedented resources to support African medical schools in meeting their training needs and triggered a number of collaborations, including those between south-south partners in Africa [12, 13]. The Medical Education Partnership Initiative (MEPI) distributed $130 million to 13 medical schools in 12 sub-Saharan African countries. The thematic focus of MEPI was on improving the quality, quantity, and in-country retention of medical graduates along with building research capacity in African medical schools [14]. The Kilimanjaro Christian Medical University College (KCMUCo), located in northern Tanzania and one of the 13 MEPI beneficiary African medical schools, was established in 1997 with 15 students. By 2010, when the MEPI project started the number of students, had increased by almost nine-fold to 129 [15]. By 2012 the KCMUCo had an average annual undergraduate medical student population (year 1 through year 5) of 700.

In the first two years of medical education KCMUCo students learn basic sciences, while during the third through fifth year clinical clerkships are core. To cope with the increased number of students, without compromising the scarce training resources (including faculty) during basic sciences years (year 1 and year 2), MEPI program has utilized, with great success, a learning management system (LMS), team-based learning [16] and other technologies to support medical training.

The Kilimanjaro Christian Medical Centre (KCMC), one of four referral hospitals in Tanzania, which has an official bed capacity of 418 in major clinical wards (paediatrics, internal medicine, surgery, obstetrics and gynaecology) (KCMC Annual Report, 2014), is the principal clinical training site for KCMUCo clerkships. In this regard, with an average of 450 medical students in the third to fifth year, the teaching hospital capacity was inadequate to accommodate the increased number of students. To tackle the problem, through MEPI, KCMUCo introduced the eight-week peripheral hospital placement program for third year medical students (MD3) within Kilimanjaro and Arusha regions during the 2012/13 academic year to undertake their junior clerkships. This was aimed at reducing student overcrowding at their main teaching hospital and to expose the students to a real working environment with hands-on practice experience. This was also geared toward training medical students in the rural settings as one of the ways to enhance retention of medical doctors where they are needed most. This study provides a three-year (2014–2016) description of the implementation strategies employed to enhance the quality of clinical clerkships in the resource-limited context through the utilization of peripheral hospitals and to assess students’ perceptions of peripheral hospital clinical clerkships using three different student cohorts.

Results

Seventy-five percent (342 out of 456) of MD3 students took part in the three consecutive surveys. The majority were male (69.6%), younger than 25 years (72.5%) and placed in peripheral hospitals located in urban centres (59.1%). The yearly distribution of students according sex, age and location of peripheral hospital is shown in Figure 1.

Out of 342 MD3 students who completed surveys, 319 (93.3%) answered the question on overall satisfaction with peripheral hospital placement. Of these 245 (76.8%) were satisfied with the rotation with strength of consensus measure (sCns) of 76.8%. The overall satisfaction exhibited a declining trend in strength of consensus measure (from 80.5% in 2013/2014 to 65.1% in 2014/2015) and then in 2015/2016 rose to 84.8%. Analysis of different domains probed for satisfaction indicated that, satisfaction with clinical practice competencies/skills acquired increased in consecutive cohorts (sCns rising from 81.3% in 2013/2014 to 84.5% in 2015/2016) indicating strong agreement. Though slightly declining in 2014/2015, attitude towards peripheral hospital placement in general showed a rising trend, which was also indicative of strongly positive attitudes. Similar to attitude, general satisfaction with peripheral hospital rotation also exhibited a rising general trend though with weak strength of consensus measure (sCns < 80%). On the other hand, satisfaction with preceptors/supervisors at the peripheral hospital, while slightly rising in 2014/2015 to sCns = 70.4% from sCns = 68.1% in 2013/2014, it declined to 67.5% in 2015/2016 showing weak agreement on satisfaction with preceptors/supervisors (Figure 2).

A detailed analysis was performed on the three-year trends of strength of consensus to specific constructs related to general satisfaction with peripheral hospital placement and attitude towards peripheral hospital placement. Regarding general satisfaction with peripheral hospital placement, some constructs exhibited increasing trends through the years; others a fluctuating trend, and one construct showed a declining trend. Increasing trends were noted for constructs related to clear initial orientation prior to placement; learning opportunities being consistent with learning goals; availability of opportunities to apply basic sciences knowledge; longing to return to the placement hospital to practice clinical skills; and acknowledgement of the importance of basic sciences in clinical work. Fluctuating trends (mainly showing a downward trend in academic year 2014/2015) were observed in statements related to prior clear plans to accomplish placement objectives; adequacy of variety and number of patients seen; adequacy of variety of clinical services at the placement hospital; supportive learning environment; utilization of planned time at the placement hospital; and being happy to have had a chance to utilize clinical skills. The one aspect, which showed a declining trend, was that related to hospital administration and staff being open and welcoming.

On attitudes towards peripheral hospital placement, analysis of individual constructs demonstrated an increasing trend only with regard to the construct that “Placing students in the peripheral hospital for clinical care is a good idea”. Insignificant fluctuating trends were observed with regard to liking the idea of peripheral hospital placement; having favourable attitudes towards peripheral hospital placement and likelihood to accept deployment to rural hospitals after completion of medical education.

Table 1 shows in-depth analysis according to constructs with respect to satisfaction with clinical practice competencies/skills acquired and satisfaction with preceptors/supervisors at the peripheral hospital during the three-year period. The findings revealed an increasing trend with regard to enjoyment in practicing and gaining new clinical clerkship skills at the peripheral hospital. Fluctuating trends were noted on issues regarding ease of practicing clinical skills and importance of teaching clinical skills at the peripheral hospital.

Though agreement on adequacy of interaction with preceptors/supervisors showed a slight fluctuating trend during the three-year period, the strength of consensus measure was consistently above 80%, indicating strong agreement that the interaction with preceptors/supervisors was adequate. Very weak agreement was demonstrated on the aspect of regular review of logbooks by preceptors/supervisors (sCns < 50% for all three cohorts). Agreement on receiving timely and constructive feedback from preceptors/supervisors was also weak but sCns was above 76% for all three cohorts.

Comparison between students placed in rural and urban hospitals showed that attitude towards peripheral hospital placement was significantly positive for students placed in urban-located hospitals in all four constructs investigated, and in the adequacy of variety and number of clients seen (p < 0.05). However, students placed in rural-located hospitals were significantly satisfied with respect to administration and staff at the peripheral hospital being open and welcoming, and preceptors/supervisors regularly reviewing their logbooks (p < 0.05). Other constructs in the three domains did not differ significantly according to location of the peripheral hospital (p > 0.05).

Table 2 shows the challenges/suggestion for improvement during the 2014 and 2015 rotations and actions taken to mitigate them. We analysed the effect of provision of Rapid Diagnostic Test (RDT) toolkit to the pilot hospitals on the satisfaction of students with and attitude towards peripheral hospital placement by comparing students who received the toolkits and those who did not. Significant differences (p < 0.05) in the satisfaction were observed in favour of students provided with toolkits, mainly in the general satisfaction with the rotation and ease of practicing clinical skills in the peripheral hospital. The proportion of students using MiFi at least once per week increased from 55.4% for the 2015 cohort to 69.5% for the 2016 cohort. On the other hand, the mean strength of consensus measure on the usefulness of MiFi almost remained stable between the two cohorts (sCns = 53.2% and 54.5% respectively) (Figure 3).

Discussion

Our study has shown that the majority of MD3 students were satisfied with the clinical rotations in peripheral hospitals, similar to findings reported in other studies [18, 19]. Detailed analysis of trends of individual domains in our study showed that satisfaction with clinical practice competencies/skills acquired and attitudes toward peripheral hospital placement increased in consecutive cohorts (sCns > 80%), indicating strong agreement. Specifically, there was an increasing trend in the constructs of enjoyment of the placement, learning new clinical clerkship skills, and belief that it was a good idea to place students in peripheral hospitals.

Previous work has shown that peripheral hospital placement provides a chance for medical students to learn and practice clinical procedures by integrating basic sciences, clinical knowledge and skills into practice in the provision of patient care [20, 19]. Our observations are also consistent with findings reported by Wilson et al. [21] where the majority of students reported enjoying their rural rotations, and that they acquired important procedural skills, and the rural experience enhanced their theoretical knowledge. Couper [22] and Mc Neill & Campbell [23] argued that successful rural placements offer greater opportunity to learn and practice procedures than at metropolitan teaching hospitals.

Regarding general satisfaction with the placements, increasing trends through the years were related to provision of clear initial orientation prior to placement; learning opportunities being consistent with learning goals; availability of opportunities to apply basic science knowledge; longing to return to the placement hospital to practice clinical skills; and acknowledgement of the importance of basic sciences in clinical work; most of them with strong agreement (sCns > 80%). It has been suggested that successful rural placements must give consideration to the interaction between learning needs, social needs and academic accountability [24] and also provide a supportive clinical environment [25]. Furthermore, Hampshire [26] attested that rural clinical placements link theory with clinical practice, thus offering experience in patient interviewing skills (communication skills).

Interactions of medical students during peripheral hospital placement with preceptors/supervisors were felt to be adequate with strong agreement (sCns > 80%). It has been argued that adequate supervision and feedback from supervisors/preceptors boosts the likelihood of the students toward peripheral rotations by assisting them to acquire the desired clinical competencies [27]. Andrews and Roberts [28] emphasize the importance of the relationships between staff and students. Hsue, Wilkson and Bills [29] point out that students learn and gain skills/competencies through interpersonal relationships between preceptors and students since preceptors are mentors and role models and also influence student career choices [29, 30].

However, our study revealed a low strength of consensus measure with regard to timely and constructive feedback from supervisors/preceptors, and also regular review of logbooks. A study in Mwanza, Tanzania, it was found that only 75% of the students filled their logbooks during clerkship rotations, indicative of either lack of commitment on the part of students or inadequate supervision [31]. Causby [30] attests that positive feedback and advice from preceptors/supervisors’ influences student satisfaction with their clinical placements.

Attitudes towards peripheral hospital placements and adequacy of variety and number of patients seen were significantly positive for students placed in urban-located hospitals compared to rural-located hospitals. This could have been attributed to inadequacy of learning resources, including shortage of qualified preceptors, variety of patients, erratic availability of medical supplies and equipment, educational infrastructure such as teaching aid, libraries and access to information and communication technology in rurally located compared to urban-located peripheral hospitals. Other issues which impact student experiences include availability of comfortable accommodation and places for recreation in rural settings. Burch & Reid [32] argued that for rural clinical placements to be successful, some major challenges such as availability of teaching space in clinics to accommodate students, and heavy workloads for the few available healthcare workers must be addressed. Strasser & Neusy [33] also acknowledge the resource limitation challenge in rural hospitals, especially with regard to the shortage of teaching staff with the required clinical and teaching skills, poor quality of health services, and lack of good accommodation for medical students and educational resources. A conducive teaching and learning environment has been considered crucial during clinical clerkship rotations [34, 8]. Availability of such conducive environments should result in positive attitudes toward peripheral hospital placement and facilitate student learning [35].

On the other hand, students in rural hospitals were significantly satisfied with respect to administration and staff at the peripheral hospital being open and welcoming, and preceptors/supervisors regularly reviewing their logbooks, even more frequently than their counterparts in urban hospitals. It has been shown that a student’s positive placement experience is influenced by relationships among supervisors/preceptors, hospital staff and other students during rotations [36]. However, rural hospitals are usually small in size with fewer health workers compared to urban hospitals, and hence health workers are more likely to have closer contact and relationships. Furthermore, students usually assist in service provision to ease the burden of fewer available health workers in rural hospitals [37] and also contribute to continuing education of the general practitioners at the host hospital, particularly rural hospitals [37].

Suggestions for improving clinical clerkship rotations mentioned by the students included increasing pocket money during the clinical rotation period, providing laboratory skills training at the clinical clerkship rotation hospitals, and provision of Internet services. Similar suggestions have also been reported in other studies [8, 35]. These are viewed as important ingredients for creating conducive learning environment during peripheral hospital clinical clerkships.

In the context of LMICs, the clerkship rotations are positively viewed as a solution to reduce overcrowding at tertiary teaching hospitals, reducing the rural-urban imbalances of medical professionals by motivating medical students to choose rural practice and in some cases, reducing the burnout of medical personnel at the usually under-staffed peripheral hospitals. However, these rotations are not without drawbacks – such as costs that include accreditation of peripheral hospitals to provide quality learning environment, compensation to preceptors, providing medical supplies that may not be available at the practicing hospital to enable students to learn, transport costs for students and supervisors, accommodation costs and pocket money for upkeep of students. Also, due to varying contexts at different peripheral hospitals, including differences in equipment and medical supplies, availability of qualified staff to serve as preceptors and mentors, variations in the training and learning during rotations may vary among students of the same course-year. All these factors must be weighed before deciding to initiate clinical clerkships at peripheral hospitals.

Conclusion

Overall student experiences in peripheral clerkship rotations were positive with high level of student satisfaction. Opportunities exist for medical schools in sub-Saharan Africa to enhance medical training through the use of peripheral hospitals. From our experience careful implementation with elements of quality improvement and the inclusion of stakeholders is critical. Regular and timely feedback after each rotation serves as an important quality assessment tool for improvement of subsequent clinical clerkship rotations.