Medical schools in rural areas – necessity or aberration ?

Introduction: There is major maldistribution of physicians in the Democratic Republic of the Congo (DRC). While 70% of Congolese live in rural areas, relatively few doctors practice there. Of the 25 medical schools in the DRC (14 public and 11 private) only one private medical school is located in a rural area. The purpose of this article was to analyse and compare the graduates of an urbanand a rural-located university in DRC. Methods: The 6 first classes of the Catholic University of Graben (UCG) Medical School at Butembo (rural) and Université Evangélique en Afrique (UEA) at Bukavu (urban) (43 and 120 graduates, respectively) were compared according to their rural or urban employment, among other variables. Results: In total, 97.7% of graduates from the rural-located medical school were employed in the province where they trained, the majority (81.4%) in rural areas. In contrast, 40.0% of graduates from the urban-located school were employed in the province where they trained, with 23.7% working in a rural area. Over 55% of all graduates worked 78 km of their training medical school. Only 2.5% of the rural-school graduates entered residency programs, compared with 15.2% for urban-school graduates. Conclusions: The results support the policy of establishing medical schools in rural areas, and also provide indications of approaches likely to increase the number and expertise of rural-located physicians.


Introduction
There is major maldistribution of physicians in the Democratic Republic of the Congo (DRC).While 70% of Congolese live in rural areas, relativley few doctors practice there 1,2 .Until recently all three Congolese medical schools were publicly funded and located in large cities.In 1989 the government authorized the establishment of additional medical schools throughout the country, partially to address the need for physicians in rural areas.This policy also allowed for private orgainisations to develop medical schools, especially religious organizations in rural areas.The Catholic University of Graben (UCG) Medical School at Butembo was one of the first rural medical schools established in the DRC.
Of the 1989 total of 25 DRC medical schools, 11 are privately funded and only one is located in a rural area.All 14 publicly funded schools are located in urban areas.This is consistent with the Congolese system where religious (catholic and protestant) services and activities in education and medicine are generally localized in rural areas, while public enterprises are dominant in urban areas.
Uneven urban-rural distribution of healthcare services and medical personnel is a concern in many countries 3,4 , often leading to poorer access to good health care and services for rural populations.The situation is magnified in Africa 5,6 and compounded by the massive migration of medical personnel to other well-resourced countries [7][8][9] .Because medical studies are fully self-funded in DRC, graduation practice location is self-determined, and the government is unable to prevent the loss of new graduates to foreign countries.
Although in developing countries approximately 80% of the population live in rural areas (as opposed to 20% in such countries as the USA and Canada 10,11 ), entire-country decisions are often determined by urban policy-makers, to the disadvantage of rural populations.Multiple factors increase the uneven distribution of resources to rural areas, including difficult rural living conditions, a lack of amenities such as shops and educational opportunities, and inadequate access to medical equipment 12 .
Bowman 13 suggested four factors are pivotal in reversing this maldistribution: (i) selecting medical students who have rural birth origins; (ii) selecting medical students who are older than 29 years on graduation; (iii) changing curriculum and policies to increase the likelihood that graduates will choose family practice; and (iv) increasing the number of 'distributional' medical schools located in rural and underserved areas.The latter factor supports the hypothesis that the uneven distribution medical personnel is related to an urban-location bias for faculties of medicine.
To overcome this problem, several strategies have been suggested [14][15][16][17] , and Hays suggested the following 17 : • Organize rotation programs for medical students to work in resource poor and neglected rural areas.
• Develop a rural hospital or clinic for each faculty of medicine and rotate all students there.
• Create faculties of medicine in rural areas where a complete curriculum of adapted training is available.
At the end of the 1980s, DRC policy development led to the establishment of a rural-located medical school, as recommended by Hays 17 .In this article, the initial outcomes of this newly established rural-located medical school are reported, and compared with an urban-located medical school, both located in east DRC.The movement of early career doctors from each institution was tracked.

Definition of rural and urban areas
In DRC the term 'urban' is related to a town environment which has a legally designated mayor.However, rural or non-urban areas include small settlements, villages and other areas distant from towns, lacking in human and material resources (eg no running water, electricity).Currently in DRC, both urban and rural areas lack facilities so the presence of a mayor (urban) is an important distinction, as is the reduced availability of drinking water, electricity, medical services, public transportation and sanitation of rural areas.

The two medical schools
Two recently established private medical schools located in east DRC (Fig1) were studied: the rural Catholic University of Graben (UCG) and the urban located Université Evangélique en Afrique (UEA).Both follow the national curriculum for DRC medical schools and there is no emphasis on primary health care at the rural school.
According to Congolese higher education policy, student admission is according to the performance of students in their last year of high school and rural background characteristics are not considered.
The Faculty of Medicine of UCG is located in an eastern rural area in Butembo in North-Kivu (NK) Province.This province covers 59 000 km 2 and has a population of 5 million (Fig1).In 1989 when the medical school was

Number of the matriculating and graduating physicians
At both universities, less than 6% of matriculated students finished their studies within the planned seven-year curriculum period.At the rural UGC medical school, 43 doctors graduated over the 6 classes from 1997 to 2004.
This represents an annual average of 7.2 graduating doctors or 3.9% of all matriculated students.At the urban UEA medical school, a total of 120 doctors graduated over the 6 classes, representing an annual average of 20 doctors but only 5.7% of all matriculated students.

North Kivu
South Kivu

Location of employment for the graduating physicians from the Catholic University of Graben rural medical school and Université Evangélique en Afrique urban medical school
Among the 43 rural UCG graduates, almost all (97.7%) worked in the NK Province where they were trained (Table 2).Of these, 81.4% were located in a rural area (Fig2 Of the 107 UEA graduates, 25 (23.4%) were employed in other African countries.In contrast with the rural-trained doctors, a substantial number of these urban-school graduates (15.9%) were pursuing advanced study in the DRC (n = 1), in other African countries (14) or outside Africa (3).
For location of employment or advanced study, the difference was highly significant between rural and urban areas (χ 2 = 653.28,df = 5, p<0.00001).
All 42 doctors who graduated from rural UCG and practising in NK were located between 0 km (in Butembo) and 355 km from Butembo (Table 3).The mean distance from their place of training was 88 km.The majority (71.4%) worked within a radius of 78 km of their medical school, or 3 hours travel by motorcycle.A similar pattern was found for the 46 urban graduates from UEA who were practising in the South Kivu Province.All were located between 0 km (in Bukavu) and 340 km from Bukavu, the majority (68.7%) within a radius of 78 km of their medical school.

Type of employment for the rural and urban graduating doctors in the province
The public sector was the largest single employer of graduate doctors practising in the province: 35.7% of rural UCG graduates (15) and 50% of urban UEA graduates (23%) (Table 4).The second largest employing sector was private healthcare in facilities sponsored by religious organisations, at 33.3% of UCG graduates ( 14) and 39.1% of UEA graduates (18).The remainder were employed in other private facilities.

Discussion
This study provides evidence that rural-located medical schools can increase the distribution of practising physicians to rural areas in the DRC.Similar results have been found in other countries.In Canada 30.7% of graduates from a ruralbased medical school practised in rural areas 18 .Similarly, in China 34% of graduates from a rural-based medical school practised rurally 19 .Visiting health academics in Thailand reported that doctors trained in rural areas tended to remain and work there 20 .These 'successful' examples all fell far short of the 81.3% rural practitioners produced by UCG.
In the present study, the contrast between rural UCG and urban UEA students was striking, with 81.4% versus 26.7%, respectively, working in a rural zone on graduation, and this difference was highly significant (χ 2 = 38.41,df = 1, p<0.00001).These data confirm other studies that found graduates from urban settings tend to remain and work there 13,21,22 .Hays 17 and Bowman 13 concluded that a major strategy to increase the distribution of doctors to rural areas is the existence of rural medical schools.While the data on enhanced distribution of doctors to rural postings is encouraging, the present study highlights a number of troubling issues.First, the yield of graduates was low from both schools.This appears to be related to student change of faculty and/or university and is currently being investigated.At the time of the study both schools were newly established in an era of political unrest, and were subject to national and local (community) pressure to admit more students than the faculty could educate.Other likely factors include inadequate preparation of students for higher education; inadequate faculty numbers and preparation for their roles; and inadequate equipment for effective teaching (in common with Indian studies 23,24 ).It may also be necessary to define rural medical schools by including ruralspecific subjects to supplement to the national curriculum 25 .
Second, the already low proportion of graduates from both schools was far lower than that of other African universities for the period 1990-2004 (423   25 . The relatively close location of practising graduates to their parent institutions offers an opportunity to provide continued professional development (CPD) from the medical schools, which other studies have shown to be important in retaining doctors in rural areas 21,26,27 .The contribution of the medical schools to CPD is important in the present setting, where poor roads and communication systems make ongoing education difficult for graduates.Moreover, medical-school based CPR creates the possibility of rural-based practitioners playing a role in undergraduate training 28 .
The limitations of the present study are that only two schools were investigated, and that both were new so only early outcomes of medical practice could be measured.The study focused only on the location of medical schools and there were a low number of graduating students.Other factors reported to increase rural-based medical practitioners (such as the rural origin of medical students, increased age of students, and increased emphasis on family medicine in the curriculum 13 ) were not explored.These other factors, in addition to following newly graduated medical practitioners for longer periods, and investigating other rural and urban located medical schools should be investigated before making major recommendations for policy changes.This and a previous study 9 strongly recommend developing better tracking systems to follow students before, during and after graduation from medical schools in Africa.
The present results shows that all medical sectors within the province (government, private and religious/denominational) benefit from the services these young doctors provide, and that doctors produced by private medical schools associated with religious bodies mainly serve in the public sector where they are desperately needed.With a DRC doctor : population ratio of 1:10 000 2 , it is encouraging to see positive results from policy that promotes the decentralization of faculties of medicine.

Conclusion
This study provides five key messages: (i) a rural-located medical school in the DRC provides the majority of its established, Butembo was a village with less than 74 000 inhabitants and no running water or electricity.The UCG was created by the Catholic church at the request of the local community.It consists of 6 faculties (veterinary science, agriculture, economic sciences, social, political and administrative sciences, law and medicine), as well as several study and research centers.The first doctors graduated from the seven-year curriculum during the academic year 1997-1998.The School of Medicine UEA is located in an urban area at Bukavu in South Kivu, a neighboring province of NK (Fig1).This province covers 65 070 km 2 and has a population of 2 837 779.The chief town Bukavu has 245 000 inhabitants.The UEA is a Protestant university founded in 1991 and has 4 faculties: agriculture, economic sciences, medicine and theology.The first doctors graduated from the seven-year curriculum during the academic year 1997-1998.Records from the academic services offices of both institutions were analysed to establish matriculation and graduation rates of the first 6 classes of both schools.At UCG the first 6 graduation classes included 43 doctors; at UEA of 120 graduating doctors two were lost to death, reducing the follow-up number to 118.Practice location of the graduates was obtained by direct contact with alumni and from provincial medical offices.No information was available for 11 UEA graduates, so the final UEA study population was 107.The data were analyzed using SPSS for Windows v12.0.1 (SPSS; Chicago, IL, USA).Descriptive statistic (percentage, means and standard deviation), t-test and χ 2 were used for comparison.

Figure 1 :
Figure 1: Map showing the locations of the medical schools in the eastern provinces of the Democratic Republic of the Congo.

Figure 2 :
Figure 2: Map showing North Kivu health zones and the location and number of graduates from the rural Catholic University of Graben Medical School.The two North Kivu towns (Beni and Goma) are shaded.
graduates to the rural workforce; (ii) rural graduates work in all sectors where doctors are needed; (iii) attention must be paid to improving the yield of graduates; (iv) very few rural graduates appear to have the opportunity of advanced training; and (v) the majority of graduates work close to their medical school, and this offers universities a relatively easy opportunity to provide CPD, and also increase faculty capacity by recruiting graduates to academic staff.

Table 2 : Location of employment or advanced study of graduates from the Catholic University of Graben rural medical school and Université Evangélique en Afrique urban medical school (2006)
UCG, Catholic University of Graben; UEA, Université Evangélique en Afrique.†Democratic Republic of the Congo, other African countries or outside Africa.

Table 4 : Type of employment of graduates practising in their original province
UCG, Catholic University of Graben; UEA, Université Evangélique en Afrique.