Results will be presented as follows: the study site, perceptions of the health policy (content and implementation), the implementation process, and the interactional factors that were found to have an influence on this maternal health policy.
Study site
This district with 390.000 residents is located in the province of Soum in Burkina Faso [
30].
Composed of rural departments, it is one of the poorest regions in Burkina Faso. The 193 villages in this region are unevenly distributed. There are 31 health centres (one per 12.500 people vs. a national average of one per 10.000) and one district hospital. Twenty-seven health centres do not meet infrastructure standards and 10% of health centres do not meet human resource standards [
30]. The average distance between villages and health centres is estimated at 11.4 km [
30]. In 2009, the rate of assisted delivery was estimated at 53.2% [
30].
Perceptions of the health policy
Health workers welcomed this national health policy. First, they see it as a good strategy for improving access to health care, both to increase the rate of antenatal care and as an opportunity to further extend the intended results into other areas, such as a higher vaccination rate. However, health staff stated that this national health policy should cover the whole process of pregnancy, not only the delivery, in order to prevent complications. They indicated that they are aware that this national health policy is not enough to effectively decrease maternal mortality. From their perspective, other measures should be added. Comment from a district health manager: “This health policy alone cannot meet the challenges of access to health care. We need to reduce distances between villages and health centres, and change people’s behaviour.”
The implementation process was perceived to be difficult because the guidelines were not clear in terms of targets, extension (period of time) and drugs covered by the policy. Therefore, at the beginning of the implementation there was some variation in the understanding of the national policy, as noted by a nurse: “Everybody is doing it his own way.” (health centre 5).
Implementation
The implementation analysis section was described according to the components of the health policy (Figure
1). For each component, the congruency of what was planned, what was implemented as per official records, and what was actually done as reported by participants was examined.
The reimbursement system
According to the policy guidelines, health centres can be reimbursed for two types of assisted deliveries: normal deliveries (no complications) and complicated deliveries (dystocic labour). For normal labour, health centres were reimbursed 3600 CFA (5.5 Euro) for each delivery performed. For complicated deliveries, health centres were reimbursed 14.400 CFA (22 Euro). According to the district health manager, the rate of dystocic labour in health centres has doubled since the beginning of the implementation of this health policy. To reduce these costs, the health district manager ordered staff to stop writing dystocic labour on the reimbursement sheet. Therefore, health centres are reimbursed only for uncomplicated deliveries. “They re-defined dystocia for us. It is not the dystocia that we learnt at medical school. Money is involved in that” (nurse, health centre 4). “We are told not to do any more dystocia labour. However, in health centres we deal with dystocia labour” (nurse, health centre 2).
Two years after the launch of the policy, the reimbursement system was changed. Since 2009, health centres are no longer reimbursed at a fixed rate, but on the basis of actual expenses. Reimbursement sheets for the policy were consulted in the 6 health centres that were studied. The cost of normal deliveries ranged from 1200 to 1500 CFA (2 to 2.3 Euro) in health centres. Health district managers perceived the national level estimate of the cost of a normal delivery as being too high. Changes in the modalities of reimbursement in health centres were key turning points in the implementation of this health policy. According to the district managers, health workers asked for reimbursement of dystocic labour in order to have higher refund payments and a more generous budget: “We found that it was for a greater balance.”
The estimated cost per patient for a caesarean section, as planned in the policy, was 11.000 CFA (17 Euro). In 2008, health district managers decided to pay 5000 CFA (7.6 Euro) of this cost by drawing on the budget line for drugs provided by the state. In 2009, the remaining 6000 CFA (9 Euro) was paid by each COGES each month. Therefore, parturients do not pay for caesarean sections (which cost 11.000 CFA). This initiative came into effect in July 2010 and the impact remains to be studied.
The grant for the worst-off
The policy planned to fully exempt the poorest patients from all costs associated with assisted deliveries and emergency obstetric care [
22]. A budget from the national subsidy obstetric care policy of 50000 million CFA (76.2 million Euro) was planned to fund the poorest patients. In the district of Djibo the subsidy for the worst-off was not implemented.
“The application is very hard, starting with the definition: what is a worst-off?” “Which criteria should we use to say that this person is a worst-off and this one is not?” (district health
manager)
.
Transportation
The cost of transportation between a health centre and the district hospital is covered by the policy. The transportation funding was supposed to be managed separately from the budget for deliveries and emergency obstetric care. The main problem perceived by the district is the long delay in reimbursement for transportation from health centres to the district hospital. The outstanding amount, supported by the district hospital, is estimated at 6 million CFA (91.500 Euro). According to the health district managers, this is leading to financial difficulties. The delay is due to misunderstanding the process of reimbursement by the district health manager. The district health manager does not keep receipts from the purchase of gasoline; however, this ticket is the basis on which the reimbursement is made.
The district of Djibo covers an area of 12.700 km
2[
30]. Only one ambulance is functional for the whole district, which leads to significant delays in referrals, as outlined by one health worker:
“You can call for the ambulance and find that it is elsewhere, so you must wait for the ambulance to come back out. It is a real problem” (health centre 3). The ambulance is used frequently and it is a multi-purpose pick-up vehicle, which is also used for other emergency transportation.
Quality of care
The implementation guidelines for this policy recommend improving the quality of care [
22]. The quality of care component in the guidelines was composed of various activities, such as continuity in conducting maternal death audits, and the implementation of a quality assurance service (QAS) team in the district hospital in charge of safety and enforcement of hygiene standards. Health workers are supposed to conduct surveys of patients to assess quality of care in health centres. For each delivery, health staff is supposed to use a partograph to assess the process of delivery. Health workers are informed and trained to write maternal death audits, these audits are carried out in the district of Djibo, and the records of these were available for study. Sessions to discuss maternal death audits were organized at the district hospital. However, the surveys of the beneficiary population and the QAS team were not implemented.
“In our health centre we did not implement these activities. The survey that we should do, we learnt that from you [the researcher]. They did not give us any guidelines or document that talks about that” (nurse at health centre 1).
In health centres where there are no midwives or skilled birth attendants, the partograph is not used. These centres represent 13.3% of health centres in the district of Djibo [
30]. Problems using the partograph reported by health workers are related to both social context and working conditions in health centres. The patients often come to the health centre when they are fully dilated or in the expulsion phase of labour, while the start of writing a partograph should begin at four centimetres dilation. As noted by a skilled birth attendant (health centre 1)
: “Women usually come to the health centre when they are about to give birth. It is then difficult to open a partograph to monitor.” The lack of infrastructures (no electricity) also makes using partographs difficult. A skilled birth attendant said that she stopped opening partographs because she refused to work with a flashlight (health facility 1); 24% of health centres do not have electricity [
30].
Supplies and equipment
One of the key components of the policy guidelines is to make drugs and supplies available for performing deliveries. Although there have not been any significant disruptions in drugs and supplies, items are not sufficiently available in the delivery room, as indicated by a midwife: “In health centres there were no gloves for revision [of the uterus], while you are often asked to make revisions, you are forced to tinker” (Health centre 6). In addition, equipment such as birth boxes and delivery tables is limited and heavily used, resulting in problems with providing good quality care. As a health district manager has indicated: “the material is not enough. This material wears out quickly.”
Incentive measures
The policy as planned did not provide a specific incentive system for health workers in addition to the usual (20% bonus per procedure). The system of bonuses was added later in the district with the approval of the central level: “Health workers thought that this policy came as a huge work load. They asked for a financial motivation for this policy. But it is written nowhere. We asked the central level, and we were told that is only on the medical act that you can take 20% bonus” (district manager). In the district of Djibo, this system is applied differently from one health centre to another. Some health centres have applied this 20% bonus only to medical procedures. Others have applied it to the contribution of the patients, and others have not applied it at all. According to health workers, this policy should have been planned to offer a formal financial incentive measure to encourage them in their work in addition to the 20% bonus.
Communication
In the district of Djibo, the process of informing the population of this health policy was vertical. The information came first to the district health manager and went through health workers, the local political community and finally to the communities. The main communication channels that were used were antenatal care consultations, routine vaccinations in villages, and the radio. However, many women were still not informed of this health policy: “We did not know that the price of deliveries has changed. Nobody came to our village to tell us that” (women’s focus group, health centre 1).
Monitoring system
Health workers have to fill in two documents related to this policy: individual patient records and the birth registry. At the beginning of the implementation, health workers had problems completing the individual patient record. “Filling forms was a real problem because we did not understand some elements of the form to fill properly” (nurse at health centre 3).
Document storage and entering the data in the electronic database mandated by the policy were the main administrative problems perceived by district health managers. The absence of an archiving system and the lack of space in the district caused more problems with storing all the reimbursement sheets for the 31 health centres that reach the district each month, as outlined by a health district manager: “This policy has reiterated the problem of archiving in the district.” There can be as many as 60 record sheets each month for a busy health centre.
The database software came one year after the implementation of the policy, creating a great delay in entering data in the database. In addition, the software is only available on one computer and cannot be duplicated on others. To overcome this problem, the health district manager hired a secretary from the district’s own budget to perform data entry. During the period of data collection for our study (November 2010), she was entering data from November 2009.
Interactional factors
In this section, interactions between groups of participants at the community level were identified as having an influence on the implementation process, relation between patients and providers, health workers and TBAs, health workers and members of COGES. Poor relations between these groups may reduce the effectiveness of this policy.
An ambivalent relationship between health workers and patients
The implementation of this policy influences relations between health workers and patients. On the one hand, health workers perceived an improvement in relations, because they no longer have to negotiate medical costs with their patients. On the other hand, the misunderstanding surrounding the items covered by the policy leads to strained relations: “The villager is what he is; he thinks that everything is free.”(nurse, health centre 1) “When patients come and there is nothing to pay, the atmosphere is more relaxed.” (district manager).
However, some communities have not observed any changes in relations with health workers. In some villages the population is satisfied with the quality of care, whereas in others relations with their health workers are perceived as more negative. Informal payments, the rate of absenteeism, the unavailability of health workers, and the expression of authority over patients are the main issues highlighted by communities: “When you go to pay for drugs at the pharmacy, they cheat all your money. I have had to pay drugs out there that rose from 35.000 (53.3 Euro) to 50.000 CFA”(76.2 Euro). “The skilled birth attendant did not assist me to deliver, she came when I had already delivered”(health centre 1). Another women indicated: “I decided not to go to the health centre to give birth because when we went for my co-wife, nobody was there to take care of her” (health centre 2). Another woman reported: “When you do something that did not fit, she [the skilled birth attendant] scolds you” (health centre 1).
Traditional birth attendants and community health workers: forgotten by the policy
The national subsidy for obstetric care as planned did not involve traditional birth attendants, who had no assigned role under this national health policy. However, a decree from the Ministry of Health issued in December 2007 requires districts to redirect the role of TBAs [
26]. Their role was shifted without any change in compensation. Nowadays, TBAs are asked to encourage women to attend health centres. Health workers and district managers have ambivalent perceptions of the role of TBAs and CHWs. On the one hand, they are seen as an obstacle to the policy because they continue to perform deliveries in villages against the aim of the policy. On the other hand, they are considered a potential ally to help with activities:
“Their presence is an advantage but also an inconvenience. They make our life difficult.” (district health manager).
From the perspective of TBAs, they understand their new role but feel frustrated that they are not sufficiently involved in the activities of health centres. Moreover, as the new policy has transferred their former role of assisting during delivery to the health centre, this also means a loss of income for TBAs. “My new role is good. Before we did not go to health centres and many women died after delivery, but since we started to go there is no maternal death but health workers do not invite me to their meetings. They did not give us anything (money).” (TBA, health centre 2).
The perspective of CHWs is similar to that of TBAs. They feel useless and excluded from the activities of the health centre.
Relations between COGES and health workers: allies or enemies?
COGES play an important role in the implementation of the policy. Grant reimbursements are made from the district directly to their bank account. They are required to bear the costs until they receive the subsidy payments. Health workers perceive a lack of cooperation from these agents in the activities of the health centre, such as paying for gas to go to villages to raise awareness. According to health workers, COGES do not perform their role as an interface between the health centre and the community, for example in transmitting information. Problems in understanding the process of reimbursement lead to strained relations: “They do not understand the modality of reimbursement, so when we ask them to pay for medication we have to negotiate with them.” (nurse, health centre 3).
From the perspective of COGES, the long delay in reimbursement is the main difficulty perceived in the implementation of this national health policy, leading to the purchase of drugs on credit. Only one of the COGES members interviewed talked about negative relations between them and health workers, as stated by a COGES president: “Most of the time the relationship doesn’t work. There is a lack of collaboration between health workers and us.” (health centre 2).