Seven documents (Burkina Faso, n = 2; Ghana n = 2; Tanzania n = 3) were identified as CPGs for maternal care corresponding to the contents of WHO PCPNC sections B9-E in the three countries (Table ). Guidelines in Burkina Faso and Ghana had been recently updated in 2009 and 2008 respectively, whereas Tanzanian guidelines were undergoing revisions but had not yet been published at the time of the review in 2010. The antenatal care (ANC) cards given to pregnant mothers provide a multi-purpose record form and although it is debatable whether they should be classified as CPGs, they are perceived to be the most available and used protocols by front line health workers in all three countries.
Overview of reviewed national clinical practice guidelines for maternal care including an assessment of their format
Overall, the contents of the national guidelines were similar to those of the WHO PCPNC. The few differences observed included: lack of a specific section on the management of ABC (Airway Breathing Circulation) in the Burkina Faso and Tanzanian guidelines; and deficiencies in guidance on how to respond to problems immediately postpartum in the Burkina Faso and Ghanaian guidelines (Table ).
Content comparison between the WHO PCPNC guidelines and the national clinical practice guidelines for maternal care (not including ANC cards)
The review of guidelines' format revealed variations in the levels of usability and applicability between national CPGs in the three countries. Overall, the Ghanaian 'National Safe Motherhood Service Protocol' was the only CPG that had a high level of both usability and applicability, containing clear algorithms for clinical decision making at all levels of care and for all the stages of pregnancy, delivery, and the postpartum period. The Burkina Faso guidelines and the Tanzanian ANC guidelines had the lowest levels of both usability and applicability with mainly narrative text sections mixing background information and treatment recommendations and containing few algorithms for decision making. In these guidelines, navigation was cumbersome and guidance on problem solving difficult to find. On the other hand, the Tanzanian Emergency Obstetric Job Aid displayed a high level of applicability and medium level of usability, containing clear algorithms and flow charts for decision making, but arranged according to obstetric diagnosis rather than symptoms. All ANC cards had a medium level of usability with the Ghanaian and Tanzanian ones including detailed checklists for antenatal, delivery, and postpartum care. Checklists in the Burkina Faso ANC card were more limited and did not include postpartum care or a Partograph, which were included in the Ghanaian and Tanzanian ANC cards. None of the ANC cards included algorithms for decision making, lowering their level of applicability and use for clinical decision making.
Key informant interviews
Analysis of the 28 interviews from Burkina Faso, Ghana and Tanzania identified three overarching themes: development of national CPGs for maternal health, health workers' access to guidelines, and health workers' use of guidelines. The first theme was part of the initial interview guide, whereas the second two emerged as a result of the semi-structured nature of interviews. Within each theme several categories emerged. Themes, categories, and their definitions are summarised in Table and further described and illustrated with quotes below.
Interview themes, categories and definitions
Development of national clinical practice guidelines for maternal health
In all three countries, the development of maternal health guidelines was reported to be carried out in cooperation between Ministries of Health and key stakeholders including UN organisations, NGOs, clinicians, professional associations, and to some extent health representatives from districts and regional levels as well as universities. A participatory approach seemed to be favoured by the key informants:
'... We do an initial draft; we call key stakeholders who make inputs about two or three times. In addition, the final version, we organize bigger stakeholders meeting to take inputs from key organizations and individuals.' (Medical doctor (MD), government level, Ghana)
In Tanzania and Burkina Faso, opinions however diverged as to whether sufficient participation had been achieved. In Tanzania, there was also disagreement on whether the new revision of antenatal care guidelines had been finalised and approved and whether they were in line with the WHO PCPNC guidelines:
'The process is sufficiently participatory.' (MD, government level, Burkina Faso)
'Heads of districts are involved although one may say that they are not adequately involved. It is for the validation of the document but not for the development.'(MD, district level, Burkina Faso)
'The revision was not participatory, that I have to tell you upfront...so basically those materials for the updated ANC need to be looked at to see if it is in line with this [WHO PCPNC].'(Program officer, international NGO, central level, Tanzania)
The WHO is a central partner for the development of guidelines in all three countries. No key informants believed that there were any major differences in content between WHO recommendations and national protocols. The recently updated Tanzanian Essential Newborn Care guidelines (not included in the document review) had been entirely based on WHO PCPNC. In Ghana, key informants stressed that the WHO guidelines constitute one but not the only source used as a basis for national protocols. In terms of the adaptation of the WHO generic guidelines, one key informant from Burkina Faso expressed:
'We are obliged to adapt to our context but not too far from what is advocated because we need quality care.'(MD, government level, Burkina Faso)
Health workers' access to guidelines
In Burkina Faso, distribution of guidelines was perceived as a problem and the general view among key informants was that health workers' access to guidelines is limited:
'There is an effort made to have partners reproduce documents... But, often, they are kept stored in storehouses.' (MD, international NGO, central level, Burkina Faso)
'Between planning and having means, there is a gap. There is a big gap. We plan, develop [guidelines], and when it comes to implementation, we are not able to sufficiently mobilize resources.' (MD, government level, Burkina Faso)
In Tanzania, several versions of maternal health guidelines circulate. There are separate guidelines for antenatal care, postnatal care, PMTCT (Prevention of Mother to Child Transmission of HIV) and family planning. Mostly, health workers' access to materials depends on if they have recently attended a training course and on what subject:
'They [health workers] use everything what they get! Because even the job aide it is not everywhere. So those who have been trained in LSS [Life Saving Skills Curriculum in Emergency Obstetric Care], those [course materials] are their job aides.' (Program officer, international NGO, central level, Tanzania)
In Ghana, health workers' access to guidelines was not brought up during the interviews.
One reason related to poor access to guidelines in Burkina Faso was identified as staff mobility, both at district and at primary care level. Health workers consider guidelines as personal possessions and there is no inventory of what guidelines should be in place in each health facility:
'People transferred move with the documents instead of making copies and leaving the original document. There is no service transfer with an inventory set up.' (MD, government level, Burkina Faso)
As an effort to solve this problem in Burkina Faso, 7,000 copies of Emergency Obstetric and Newborn Care guidelines were to be printed in 2010 in a pocket size format. The same idea was also brought forward by a Tanzanian NGO. Another strategy perceived as a solution to the limited availability of guidelines was the use of wall posters:
'We were asked to make posters to be posted in different rooms... It helps the worker to take the right decision only after a glance.' (MD, government level, Burkina Faso)
Health workers' use of guidelines
In all three countries, key informants expressed doubts as to whether guidelines are used and followed by health workers in practice. This was also confirmed by the few interviews with service providers:
'...my fear is, it [the guideline] may not be used at all.' (Reproductive health officer, international NGO, central level, Ghana)
'As for the conducting of the delivery and then after that the postnatal care, that one I know it as a midwife, but for the actual laid down guideline, I don't know.' (Health worker, district level health facility, Ghana)
In Burkina Faso, reasons identified for poor guideline adherence stem from health workers' initial vocational training as well as from their attitudes towards continuing education.
Having left pre-service training, new guidelines are not endorsed unless you attend a new course or orientation:
'But one must admit that the guidelines are not applied and this is due to the training of health workers themselves who come out of training schools with lacks.' (MD, district level, Burkina Faso)
'The issue of health workers requires a perpetual search of new knowledge or an updating. If at the end of the basic training at school, we do not open any document and we only expect workshops, there will be a problem.' (MD, government level, Burkina Faso)
The challenge of guideline implementation and the often disappointing outcomes of training courses were illustrated by a comment from a Tanzanian NGO worker:
'... there were people trained by the ministry before we moved there [district hospital] and yet when we came there, nobody had followed them [the guidelines], they had gone back to what they were doing before, they were not practising what they had been taught.' (Program manager, international NGO, central level, Tanzania)
In both Ghana and Tanzania, key informants expressed the lack of user-friendliness of guidelines as a probable reason for low adherence. The lack of flow charts, algorithms, and clear steps for how to manage different conditions is especially a problem with the training packages commonly used as guidelines by health workers in Tanzania:
'But unfortunately they [training curriculums] have not been followed up by flow charts, these are not there, so these need to be extracted and developed. For example, if you manage eclampsia, what are the steps?' (Program officer, international NGO, central level, Tanzania)
'Maybe the guidelines were not easy to follow.' (Reproductive health officer, international NGO, central level, Ghana)
In Burkina Faso, another reason for the perceived limited use of guidelines was negative beliefs about using guidelines during patient consultations:
'... We have a problem on how to have permanently documents laid on consultation table. First of all, people think that it devaluates the fact of consulting with a paper before the patient while one must ensure what they are going to do is right.' (MD, government level, Burkina Faso)