This study describes the profile of staff in rural health zones in South Kivu, eastern Democratic Republic of Congo. It takes into account the crisis context that has characterized this province for several decades. We conducted a cross-sectional study in four districts selected according to the crisis context following the classification of health zones by the Provincial Health Department of South Kivu. The Idjwi health district was chosen as stable, Katana and Walungu as intermediate and Mulungu as in crisis.
Our findings showed that agents over 34 years of age were the most represented. Women were the least represented in all the health districts but the lowest proportion was in the unstable district. In all districts, administrative staff was the most represented comparing to nursing and medical staff. 96% of the health staff did not receive a government salary and 64% did not receive a government allowance for risk. 63% of staff do not have a state registration number. With regard to the characteristics of the health districts, our results show that the characteristics of the agents found in the stable districts are 34 years old or younger, those with a registration number, those who have benefited from central and/or intermediate recruitment, those who receive continuous training and those who receive the local bonus. Our results also show that it is in the Idjwi health district (stable) where we have a high proportion of health workers with a higher monthly income than in the other districts. In the unstable district; men, staff with less than or equal to 4 years of seniority and those with a lower level of education are more likely to be in the unstable districts than in all the other districts,.
This study, which is one of the rare studies that addresses this issue in the context of crisis in the DRC, will be a reference for future analyses of health personnel and a standard commodity for support programs. This will also enhance decision-makers’ actions in order to ensure equitable and efficient health personnel coverage despite the crisis context.
Socio-demographic characteristics of the respondents according to the Health districts studied
Our results showed some difference between Study Areas in terms of age of staff and seniority. Health care workers over 34 years of age were the most represented, especially in unstable districts .This shows the homogeneous nature of the agents in the African context where retirement policies are not applied in all countries. This shows that young people coming directly from school are not quickly enrolled in the employment system. Our data also showed that it is in unstable and/or intermediate areas where we have older people than elsewhere. As a result of the persistent armed conflicts, since young people are generally better able to leave the unstable area in search of places offering the best security conditions, older people find themselves competitive in this area. This is agreement with findings of Rohini Jonnalagadda Haar and Leonard S. Rubinstein  and Patrick Ilboudo G. et al , in the Burkina Faso Health Workforce Survey 2014, regarding the age of health care workers. According to them, the majority of workers were between 27 and 57 years old. Gautier and Wane in Chad corroborated our results and found that the majority of workers were between 31 and 45 years old .
Moreover, our data showed that agents in unstable districts, although they are the oldest, they have less years of experience (least seniority) compared to the other districts (less than 4 years). This shows that it is difficult to keep the same agents longer in high-risk areas. In the Democratic Republic of the Congo (DRC), a study conducted by Rishma M. et al. showed that the majority of agents have a median age of 6 years, and the study conducted by Patrick Ilboudo G. et al. showed that the majority of agents had about 10 years of experience . However, these results are consistent with those found in Papua New Guinea,  where a significant proportion of agents had less than six years experience (44%). It is also important to note that in developing countries, state institutions are very weak in terms of their capacity to manage and control their own resources.
For gender, our data showed that men are about twice as likely to be in the unstable district as women. In the African context in general, some cultures do not encourage women to work, especially in conflict areas. In the context of insecurity and instability, women are exposed to more violence and harassment than men. This may explain the low proportion of women in at-risk areas  .
Our study showed that it is in the unstable district that we find more agents with a lower level of education. They are about twice as likely to be in the unstable district as those with higher education. This can be explained by the fact that as an unstable district in terms of security, the selection criteria will not be the same as in stable ones. It should be noted that there are other areas where there are several private facilities or humanitarian organizations that attract the best agents because they often offer the best salaries. A study conducted in Papua New Guinea on health care workers in rural areas  is not in agreement with our study. This study showed that 82% of the workers interviewed achieved a higher level of education before beginning the training related to their health sector position. However, our results are close to those found by a cross-sectional study of the sources of income of frontline health workers in the Democratic Republic of Congo , which shows that the majority of health workers are secondary school graduates, with only 30% having a higher level of education and university. The age range was between 30 and 40 years, 90% of the staff were nurses and only 4% doctors.
The study shows that the majority of the health workers in our study area were married and this in all the health districts. Results of a study on rural health workers conducted in Papua New Guinea  do not differ from our results in terms of civil status, where the majority of the workers were married (81%).
With regard to the status of the structures, it can be seen that 100% of the facilities in the Mulungu HD (unstable) are state-owned. This could be explained by the fact that private and religious individuals always choose to invest in stable, secure environments where the population can afford health care costs. Joyasuriya R et al. in their study show that 50% of respondents worked in public facilities and 40% in church-owned facilities, compared to only about 3% in the private sector. The study shows that the majority of staff live with their respective families in all the health districts. This is explained by the fact that recruitment is done locally in most cases, regardless of the hiring authority. For Joyasuriya R et al., almost all of the agents work in their districts. However, accommodation is provided to 68% of them and 57% lived in the same complex as the health facility.
The results of the study reveal that 47% of the agents are administrative staff. Doctors and nurses are more in the stable district than in the intermediate one. The results of our study differ slightly from the results of the 2015 Country Profile Study on Human Resources for Health in the Democratic Republic of Congo  which revealed that at the national level, administrative staff represents 38.4%.
The results of the study show that in relation to the status of civil servants (registered agent, New Unit, No status), registered agents have twice the risk of being in the stable district. This is because administrative procedures seem to be more respected. It is probably due to the context of insecurity leading to both the flight of agents and the use of available agents who have not yet completed administrative formalities. A qualitative study among these agents and their employers would be necessary to understand the reason why some of the morkers in transition districts do not have a state registration number.
The Recruitment Process, Capacity Building and Method of Paying Staff
The results of the study show that, compared to the authorities involved in recruiting agents, those who obtained their jobs after an application for employment are about eight times more likely to be in intermediate districts. As we can see in the unstable district, a small seniority of the agents got their job after an applying for it because of the context of instability.
As a district in transition, the intermediate districts would have a greater need for agents to ensure their stability. For the unstable district(Mulungu), 84.1% of the agents are hired per application and test. In the transitional district of Katana (district with a partner), recruitment by recommendation from higher authorities occupies a slightly higher proportion than in other districts. The study on the Human Resources for Health Country Profile for the Democratic Republic of Congo shows that in the DRC, recruitment is either focused on degrees(or diploma) and institutions which issued them or by competitive examination. Recruitment is based on qualifications for candidates who hold a diploma issued or recognized as equivalent by the National Education system and who are specially prepared for that career, as long as the number of candidates does not exceed the number of jobs available. Recruitment is carried out on a competitive basis when the number of vacancies to be filled is less than the number of participants. In this case, only successful candidates who are ranked in order of merit may be appointed. It is within the framework of recruitment by degree (or diploma) that the recommendations are used as we showed in this study. This way of doing things tends to erase the normal recruitment process and to perpetuate familiarity, nepotism, tribalism, for being at a given position in the health sector.
The results of the study reveal that the continuous training of agents has the risk of being in the stable district twice. The Observatory for Human Resources for Health (HRH) in the DRC, in its study on the Human Resources for Health Country Profile of the Democratic Republic of Congo, states that in the DRC the Ministry of Health organizes thematic training within the framework of specific programs. These trainings are also organized in provinces. However, there are central trainers who supervise provincial trainers. In addition to this internal organization, the Ministry of Health uses higher and university education in the training of specialists. Our results show that this training is more beneficial than for agents working in stable areas. The accessibility and the geographical stability and security of stable areas will be factors that would promote the organization of training for agents in unstable or transitional areas as well.
For the local salary (premium or bonus), agents in the stable district are more likely to receive this premium than agents in intermediate and unstable districts. This difference with respect to the collection of the local premium could be explained by the socio-economic disruptions (wars and rural exodus) that the populations of these two areas have experienced over the past few years; these disruptions do not allow the health facilities to maximize incomes to ensure the premium for all agents. In addition, the unstable districts are mostly rural areas where the majority of the population is very poor. Demographic data show that more than 80% of the rural populations in the DRC live on less than a dollar a day and the majority of people who use traditional medicine are more observed in rural areas .
Our results then show that 96% of health personnel do not receive a government salary and 64% do not receive a government bonus. Our results are superior to those found by the cross-sectional study of the sources of income of frontline health workers in the Democratic Republic of Congo , showing that one-third of nurses received no form of government payment and only 18% who received both an occupational hazard allowance and a salary.
Monthly Income of Workers by Occupation and Level of Education
Our results show that physicians are the staff with the highest income compared to other agents. They then show that the incomes of physicians and nurses are distributed differently in the study areas (p < 0.001). Our data corroborate with those of Patrick Ilboudo G. et al 33, also showing a difference in median income between staff according to the position held (p = 0.0321), their study then shows that the median income for all health staff is 295.2 (59.4-797.4) $ while for doctors it is 369 (59.4–646.2) $. Maria Paola B. et al , also show in their study that in the DRC, the median monthly income of doctors is $785 with a maximum of $4,815; administrative staff has $166 and a maximum of $1,396 while nurses have a median income of $101 and a maximum of $2,908. Rishma M., et al , in the cross-sectional study of the sources of income of frontline health workers in the Democratic Republic of Congo, also shows that in the DRC, monthly income from all sources was $85, but the average was almost double( $165). The highest median monthly income was for non-clinical work outside the facility ($119), followed by government wages ($58). The lowest median monthly income came from per diems ($9) and informal payments ($9).
Our data show that there are more agents with more than $151 (21%) in the stable district, while in the intermediate one this proportion is less than 10% and less than one percent in the unstable district. This difference can be explained by the low use of health services by the population due to the rural exodus and wars. This would explain the low income maximization in unstable or transitional district .
Limitations of the study
Despite its multiple advantages, some limitations are recognized. First, the study was conducted in only four of the thirty-four health districts in South Kivu province. Although not representative, the inclusion of three categories of health districts, including emergency, transitional (with and without partners), and stable, allowed us to understand this profile in the province. Second, the analysis on human resources was limited only to their use, which normally required studying production, use, and retirement as well. Thirdly, other aspects that seem to be important for human resources but not taken into account are the level of satisfaction of health workers, retention, working conditions including benefits, housing, allowances, etc., the consideration of other sources of income besides state salary, state and local bonuses, and finally the workers' assessments of security crises and their impact on the performance of their tasks. Finally, this study is a retrospective one which is subject to selection bias.