Implementing a Basic Package of Health Services in post-conflict Liberia: Perceptions of key stakeholders

Social Science & Medicine 78 (2013) 42e49

Contents lists available at SciVerse ScienceDirect

Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed

Implementing a Basic Package of Health Services in post-conflict Liberia: Perceptions of key stakeholders Dörte Petit a, Egbert Sondorp a, Susannah Mayhew a, Maria Roura b, Bayard Roberts a, * a b

Faculty of Public Health and Policy, London School of Hygiene and Tropical Medicine, 15-17 Tavistock Place, London WC1H 9SH, United Kingdom CRESIB e Barcelona Centre for International Health Research, Spain

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 1 December 2012

Recovery of the health sector in post-conflict countries is increasingly initiated through a Basic Package of Health Services (BPHS) approach. The country government and partners, including international donors, typically contract international and local NGOs to deliver the BPHS. Evidence from routine data suggests that a BPHS approach results in rapid increases in service coverage, coordination, equity, and efficiency. However, studies also show progress may then slow down, the cause of which is not immediately obvious from routine data. Qualitative research can provide insight into possible barriers in the implementation process, particularly the role of health workers delivering the BPHS services. The aim of this study was to explore perceptions of health service providers and policy makers on the implementation of the BPHS in post-conflict Liberia, using SRH services as a tracer and Lipsky’s work on “street-level bureaucrats” as a theoretical framework. In JulyeOctober 2010, 63 interviews were conducted with midwives, officers-in-charge, and supervisors in two counties of Liberia, and with policy makers in Monrovia. The findings suggest health workers had a limited understanding of the BPHS and associated it with low salaries, difficult working conditions, and limited support from policy makers. Health workers responded by sub-optimal delivery of certain services (such as facility-based deliveries), parallel private services, and leaving their posts. These responses risk distorting and undermining the BPHS implementation. There were also clear differences in the perspectives of health workers and policy makers on the BPHS implementation. The findings suggest the need for greater dialogue between policy makers and health workers to improve understanding of the BPHS and recognition of the working conditions in order to help achieve the potential benefits of the BPHS in Liberia. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Reproductive health Post-conflict Liberia Policy Implementation

Introduction The provision of health services is extremely challenging in countries emerging from prolonged and wide spread civil conflict. Providing health services in post-conflict countries is commonly impeded by damaged health infrastructure and limited government stewardship, domestic financial resources, and health workforce. This has often resulted in fragmented, uncoordinated and inefficient health service delivery by a range of stakeholders with limited national coverage (Kruk, Freedman, Anglin, & Waldman, 2009; Waters, Garrett, & Burnham, 2007). To address these challenges, the concept of a Basic Package of Health Services (BPHS) approach has become an important feature of post-conflict health policy making (Ameli & Newbrander, 2008). This has been seen in post-conflict settings such as Afghanistan, Liberia,

* Corresponding author. Tel.: þ44 (0)20 7927 2050; fax: þ44 (0)20 7927 2358. E-mail address: [email protected] (B. Roberts). 0277-9536/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.socscimed.2012.11.026

South Sudan, Somalia, the Democratic Republic of Congo, and Cambodia. In these countries, the BPHS consists of a limited list of cost-effective priority health services delivered at primary and secondary health care levels addressing the country’s major health problems, but typically including key services for communicable disease control, immunisation, sexual and reproductive health (SRH) including maternal health, and newborn and child health. The BPHS is financed from a combination of domestic revenues and aid from international donor agencies, with non-governmental organisations (NGOs) commonly contracted to deliver the agreed package of services based upon a competitive bidding process, with stewardship and monitoring provided by the government and international partners. Potential advantages of the national roll out of a BPHS in a post-conflict setting may be rapid increases in health care coverage, coordination, equity, and efficiency; and standardisation of services, facilities, staffing, drugs and equipment (Ameli & Newbrander, 2008; Loevinsohn & Sayed, 2008; WHO, 2008). The monitoring and evaluation of the implementation of BPHS in post-conflict settings relies predominantly on quantifiable

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49

indicators, such as the number of services provided in each of the facilities (MOHSW, 2009; Peters et al., 2007; WHO, 2008). While this information is crucial, it provides only one perspective on the implementation of the BPHS. Initially good results may be expected in response to the influx of substantial resources, but continued progress may be hampered by wider system issues or other barriers in implementation, as observed with the BPHS in Afghanistan (Palmer, Strong, Wali, & Sondorp, 2006; Strong, Wali, & Sondorp, 2005). Qualitative research can help explore these barriers and give insight on the perspectives of different actors involved in the translation of a policy into practice, particularly the front-line health workers implementing the policy through the delivery of services (Buse, 2007; Buse, Mays, & Walt, 2005; Pope & Mays, 1995; Pulzl & Treib, 2007). In this study we explore the implementation of the BPHS using a case study of Liberia, following a “bottomeup” approach to explore how “front-line” staff influence policy implementation (Hill & Hupe, 2002; Matland, 1995; Pulzl & Treib, 2007; Sabatier, 1986). This is in contrast to “topedown” models which view implementation as a more rational linear process from policy formulation to execution and may fail to adequately recognise the role of front-line staff and limited resources and policy understanding. More specifically, we use the work of Michael Lipsky as a theoretical framework for our study. Lipsky focuses on the ways in which individuals (street-level bureaucrats) delivering government policies adapt policies according to their working conditions. He notes how workers may not fully share the objectives and preferences of the policy makers and may not necessarily work towards the same goal (Lipsky, 2010). Health service providers therefore play a crucial role in the shaping of policies during their implementation as they exercise discretion in terms of types, quantity, and quality of services provided to beneficiaries. Lipsky’s approach was selected as it focuses upon workers delivering government policies in sub-optimal working conditions which accorded closely with the observed situation in Liberia. Lipsky’s focus on the role of individuals (rather than other bottomeup theorists such as Hjern and Porter who focused on networks (Hjern & Porter, 1981)) also reflected the isolated working conditions experienced by health workers in Liberia. We were also interested in how the perception of the policy implementation by the street-level bureaucrats compared with policy makers in Liberia and so included policy makers from the Liberian Ministry of Health and Social Welfare (MOHSW), international donors, NGOs and the UN in the study. The aim of our study was therefore to explore perceptions of health service providers and policy makers on the implementation of the BPHS in post-conflict Liberia. We use SRH services (including maternal health) as a tracer for the implementation process as they represent a major part of the services included in the BPHS and the promotion of SRH appears particularly high on the agenda of the MOHSW, reflecting the high maternal mortality and reproductive health needs in Liberia (MOHSW, 2010d,e). Liberia Liberia experienced civil war between 1989 and 2003, leading to the deaths of an estimated 270,000 people and the displacement of 500,000 people out of a population of around 3.7 million. Following a comprehensive peace agreement in 2003 and a transition period, in 2005 a civilian government was democratically elected, providing an opportunity for recovery and development. However, many challenges persist such as wide spread poverty, lack of basic infrastructure, weak government institutions, low literacy, and high unemployment (IMF, 2008). Maternal mortality in Liberia increased from an already high 578 maternal deaths per 100,000 deliveries in 1999 to 990 per 100,000 deliveries in 2008. The vast majority of

43

births take place without skilled staff. Approximately half of all women in Liberia have their first child before the age of 20, with early childbearing increasing between 2000 and 2007. The fertility rate remains high (estimated 5.2e6.2), and contraceptive use is low despite high demand for contraceptives (MOHSW, 2008b; Msuya & Sondorp, 2005; WHO, 2010). High levels of risky sexual behaviour are also reported among adolescents and young people (Kennedy et al., 2011). During the conflict most government owned health facilities had stopped functioning unless they received international assistance through humanitarian NGOs or faith-based organizations (MOHSW, 2007b). In the period following the peace agreement of 2003 these same organisations managed approximately 80% of all the health facilities in Liberia (MOHSW, 2007b) and managed to expand health service delivery, in particular to areas that had been most affected by the war and were home to many refugees and internally displaced people. However, due to the limited government leadership and absence of a uniform policy in this period, these expanded services were fragmented and skewed in coverage. The majority of health expenditure in this period came from donor sources (47%) and out of pocket spending (35%) (GoL, 2009). A major constraint to health service provision was a lack of qualified staff, with many health care workers, particularly nurse-aids, holding sub-standard qualifications while also having to take up senior positions in health facilities (Msuya & Sondorp, 2005). Most facilities also required substantial improvement to infrastructure and drug supply chains (MOHSW, 2007a). After the democratic elections of a new President and parliament, a new leadership was appointed at the MOHSW in 2006. Following an inclusive consultation process involving existing health cadres from throughout Liberia and the international partners including donors, UN agencies and NGOs, the MOHSW presented a new health policy and plan (MOHSW, 2007a). The policy and plan represented a shift from emergency humanitarian relief to development and from vertical programmes to an integrated health system based on four pillars: the BPHS, human resources, support systems, and infrastructure; with health service delivery through the BPHS considered the corner stone of the National Health Policy (MOHSW, 2007a, 2007b, 2008a, 2008c; Pavignani, 2009). The BPHS is financed from a combination of domestic revenues and aid from major bilateral and multilateral donor agencies such as USAID and the World Bank. The BPHS covers health services, health facility staffing and training, costing and planning, and drug supplies (with the drug supply system in the process of being centralised but stocks still being provided by international agencies at the time of the study). Staffing patterns and salary scales were set for the various types of health facilities. The core services reflect those found in other BPHS of SRH, newborn and child health, immunisation services, nutrition, and communicable disease control. The SRH services included in the Liberia BPHS are summarised in Box 1. Liberia’s new national health policy comes with a strong notion of decentralisation and sets the vision that it will be the health authorities at the next administrative level, the County Health Teams (CHTs), who will be responsible for the implementation of the BPHS in their counties. However, to bring in sufficient capacity, NGOs are contracted to assist the CHTs to deliver the BPHS while also strengthening the capacity of the CHTs. The NGOs therefore play a major role in the counties to ensure rehabilitation of health facilities, provision of drugs and other supplies, supervision and staff training. Health staff are employed by the government, but the NGOs directly pay the staff their salaries. To assess County and NGO performance, a monitoring and evaluation system has been set up to be able to measure progress on a range of indicators, such as number of women attending ANC

44

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49

Box 1. List of SRH services included in the Liberian BPHS. 1. Maternal and newborn health:  antenatal care  labour and delivery services (including increasing facility-based deliveries)  emergency obstetric care  postpartum care  newborn care  family planning. 2. Reproductive and adolescent health:  family planning  sexually transmitted infections. 3. Communicable diseases:  Sexually transmitted infections  HIV/AIDS. Note: Sexual and gender-based violence is addressed under “mental health” and “emergency medicine”.

and skilled birth attendance rates. In addition, an accreditation system was established, with all health facilities annually assessed and scored on indicators relating to: human resources and facility management, drugs and supplies (including equipment), laboratory and diagnostic services, health services, operational space and utilities, infection prevention and environmental sanitation, medical records, confidentiality, and referrals (Cleveland et al., 2011). The roll out of the BPHS started at the end of 2009. Both the accreditation reports and the programme service data show good improvements in service delivery and availability during the first years of implementation (MOHSW, 2010a, 2011). However, to detect potential barriers to continued progress in implementation, complementary qualitative data is needed. Such research can provide new insight on underlying processes, barriers or facilitators to SRH service provision in the Liberian BPHS (as has been shown with the BPHS implementation in Afghanistan (Strong et al., 2005)). Since the implementation of the BPHS is dependent on compliance by front-line health workers, understanding their perceptions of the BPHS and motivations for supporting or stalling its delivery is essential to improving the delivery of services and the implementation of the BPHS more broadly. Moreover, a review of current assessment methods and service delivery in Liberia indicated that the outcomes of the accreditation report may provide a limited perspective of the BPHS implementation (Kruk et al., 2010). Concern about the limited information on the BPHS implementation was also expressed in preliminary discussions with policy makers prior to the main research.

each visited clinic, the officer in charge and the midwife in each visited health centre, and the officer in charge, medical officer and one midwife at the visited hospitals. However, this was not possible for all facilities, as often there was no certified midwife in place, or the midwife was at the same time the officer in charge. 23 interviews were conducted with front-line health workers at the three health facility levels of clinics, health centres, and hospitals. They consisted of 8 interviews with officers-in-charge and 15 with certified midwives. The second level involved 16 interviews with health supervisors at the county level: 10 from County Health Teams and 6 from NGOs (Table 1). The interviews with health workers and their county-level supervisors were conducted between July and August 2010 in two counties of Liberia (Grand Cape Mount and Nimba), with two counties selected to capture some potential differences at the county level. Purposive sampling was applied for the selection of the counties to ensure that the BPHS had been implemented in them for at least one year, that they had at least one health centre to allow for assessment of all levels of the health system (i.e. along with health clinics and hospitals), and that they were accessible during the rainy season. Both Grand Cape Mount and Nimba counties are rural, with agriculture the dominant means of livelihood. Grand Cape Mount, in comparison to Nimba County, is closer to Monrovia and smaller in size and population (129,005 compared to 468,088 in 2008). Results in the accreditation report were better for Nimba (84% overall score) than for Grand Cape Mount county (73% overall score) (Government of Liberia, 2008). It was not feasible to visit all the county clinics and so approximately one out of four clinics was visited. In order to guarantee an even distribution across the county, at least one clinic in each county district was selected. Clinics with varying degrees of local accessibility were also chosen to get a balanced picture. The selection included both clinics supported by the MOHSW/county health team only and those supported by NGOs. This resulted in data collection conducted in July and August 2010 in 19 health facilities: 2 hospitals, 6 health centres and 11 clinics (see Table 1). In Nimba County, a large number of clinics were not accessible due to bad road conditions, particularly during the rainy season. Also, many clinics were either temporarily closed or run by a student, due to absence of staff. The third level of interviews consisted of 24 interviews with policy-makers based in Monrovia between September and October 2010 (8 government officials, 3 donor officials, 7 international NGOs, 3 local NGOs and 3 UN agencies). Policy-makers were

Table 1 Information on research locations. Location

Clinics visited Health centres Total hospitals Respondents (total clinics) visited (total visited (total interviewed health centres) hospitals)

Grand Cape Mount

7 (27)

2 (2)

1 (1)

Nimba

4 (37)

4 (4)

1 (1)

Methods This study used semi-structured interviews with three levels of respondents at the policy implementation process (health workers, supervisors, and policy makers). Health staff at facility level are considered the front-line actors in the implementation of the BPHS. Therefore they were the main focus for data collection. In addition, supervisors from the county health team as well as the implementing partner NGOs were included to expand understanding of the implementation process at the county level. At the health facilities, staff involved in management and provision of SRH services were included, focussing particularly on certified midwives as they are considered the main providers of SRH services (MOHSW, 2010b). The intention was to interview the midwife at

Monrovia (policy makers)

4 officers in charge 8 certified midwives 3 county health team members 2 NGO supervisors 4 officers in charge 7 certified midwives 4 county health team members 4 NGO supervisors 8 MOHSW staff 3 UN staff 7 INGO staff 3 local NGOs staff 3 donors staff

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49

purposively selected based upon their knowledge of the BPHS in Liberia, including in the two selected counties. The final number of interviews at the policy making level was determined using a saturation approach. The results of interviews with health workers and supervisors were analysed preliminarily to help refine the interview guide for policy makers. All interviews were conducted by DP, with a Liberian researcher present to translate if necessary. All respondents gave informed consent and all the interviews were anonymous and confidential. Ethical approval was provided by the Liberian MOHSW and the London School of Hygiene and Tropical Medicine. In-depth analysis was conducted using a thematic approach to group data into key themes emerging from the data and looking at possible relations between the themes. A sub-set of interviews at each level (health workers, supervisors, policy makers) was initially analysed to develop the coding framework. Based on this framework, all interviews were then analysed. The analysis was guided by Lipsky’s theoretical approach (Lipsky, 2010). The analysis followed an iterative process with the coding scheme and grouping of data as the analysis progressed. The main themes (and sub-themes) were: firstly, an understanding of the BPHS policy (knowledge of the BPHS, service quality and uptake, BPHS expansion); and secondly, the working conditions (salaries, motivation, staff turnover, sense of isolation, supervision, training). Cross-cutting themes included individual SRH services and similarities and differences between the perception of health workers and policy makers. The analysis did not show any meaningful differences between the counties and so the findings are not separated out by county. Results A number of themes were highlighted as influencing the work of health workers and the way in which they implemented the BPHS. These related to the broad issues of understanding of the BPHS and the working conditions for health workers. Understanding of the BPHS policy There appeared to be limited understanding of the BPHS and what it was trying to achieve among the health staff at the facilities, which appeared to stem from limited information provided to them by policy makers about the BPHS. “When we came, people used to be coming from the ministry to tell us about the basic health package, but they can’t really go into detail about it.” [Nimba Certified Midwife] While the health workers associated the BPHS with positive aspects such as increased training and improvements in buildings and supplies, they commonly associated it with a reduction in salaries. “First of all, the BPHS from the introduction, it creates a little problem for us health workers. The NGOs e IRC, MSF e were here paying the health workers with high salaries; but during the introduction of the basic health package, the salaries were paid by the NGOs to health workers according to the policy of the basic package, it was cut down. So, that’s it, reduction in the salaries and reduction in the manpower.” [Nimba Officer-in-Charge] The limited understanding of what the BPHS is trying to achieve combined with changes perceived as detrimental to their working conditions suggested that the changes introduced by the BPHS are less likely to be accepted and fully implemented.

45

The responses of health workers contrasted to the supervisors who appeared to better understand the concept of the BPHS and were more ready to accept the changes resulting from the implementation of the BPHS. “We are ok with it [BPHS] because now we know what we expect of our facilities, and we know what the ministry expects of us in the line of basic package.” [Cape Mount Supervisor] The limited knowledge of the purpose and role of the BPHS among health care providers was acknowledged by some policy makers who also recognised that a full understanding of the policy was necessary. “You make a good policy [the BPHS], but it is with the ministry of health. But the midwife doesn’t know. Who is going to implement it? There is a barrier between the policy maker and the implementer.” [NGO Policy Maker] “I think they should know the BPHS, because it is a national government policy, they are the implementers, they are expected to articulate and provide to the people who they serve. But we need to have a better language and to be more consistent in disseminating that language.” [Government Policy Maker] However, not all policy makers were aware of this lack of understanding and thought that the BPHS made sure that staff at all levels understood the standard setting through the BPHS. Policy makers also recognised how the BPHS implementation was associated with salary reductions. “But then it [BPHS] came with a salary cut. So it is not about the services. It is about cutting their salaries. So people don’t understand what are the services in the BPHS. If you send a nurse out, and all she knows is that the BPHS is about cutting salaries, then she will wonder what she is doing out there.” [NGO Policy Maker] There was also a divergence in perception between health care workers and policy makers of key elements of the BPHS implementation such as the quality of SRH services and service uptake. Most health workers and the county supervisors were satisfied with the quality of services and felt standards were improving as a result of increased training and improved drug supply. “The BPHS helps us provide good quality of services because we are trained.” [Certified Midwife] In contrast, policy makers felt training and supervision was ineffective. They also felt that health workers had a poor comprehension of the standards of care expected in the BPHS and that it was mainly the poor service quality, lack of skills, and negative attitudes of health care workers that kept people from taking up services, particularly delivery services at health facilities. Indeed, while the BPHS seeks to increase the number of facility-based deliveries (traditional midwives who had been trained during the conflict are no longer considered part of the health system and discouraged from conducting deliveries due to their poor delivery outcomes), policy makers felt the poor standards of care among health care works were deterring women from delivering at the health facilities. “I am totally understanding that this lady doesn’t want to deliver at the facility, because the midwife is rude, she doesn’t have light, she is not paid for this extra time and in the night there is no ambulance to pick her up. She is alone. So, I understand that this woman decides to deliver at home with her neighbour, who is also a trained traditional midwife, she will take care of the kids, at home, be comfortable.” [Government Policy Maker]

46

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49

In contrast, health workers attributed this low delivery rate in health facilities to clinic opening times, the travelling times involved for pregnant women, and particularly to trained traditional midwives discouraging women from delivering at health facilities because traditional midwives are not included in the BPHS and so are not paid to refer women to health facilities. “Delivery is a problem for us. The trained traditional midwives are all over in the community. So most often the people are used to them. So we really have to look at how to encourage delivery at the facility.” [Cape Mount Clinic Officer-in-Charge] Another important discrepancy between the policy making and implementation level was the perception regarding the need to expand the list of services, with the MOHSW planning to expand its range of BPHS services after the first five years of implementing to include reproductive cancer and treatment of obstetric fistula. Staff at facility level felt that a sufficient range of SRH services was provided and that the focus should be on increasing uptake of existing services rather than expanding them. “Even the ones they told us to provide, for the people to accept it, it was not easy. We are still trying to convince the people so that they can accept it.” [Cape Mount Clinic Officer-in-Charge] Supervisors at the county level held similar opinions. “They have been increasing the number of services. It is the utilization of the services. That’s where the problem is, getting the people to really use the services, the services that are offered.” [Cape Mount Supervisor] While almost all policy respondents felt that certain existing SRH services remained limited in their quality and uptake (particularly family planning, adolescent SRH, SGBV management and emergency obstetric care), opinions varied as to whether the list of services should also be expanded due to concerns of limited staff skills and numbers. One NGO policy respondent felt that now “there are higher expectations” from community members and a donor policy maker noted that “if you refer to the needs of the population, they should expand it pretty fast... I think they can still expand on it”, while at the same time acknowledging that “certainly, they don’t have the full package down at this point.” A donor respondent was more sceptical. “You could add more services, but if you don’t have the workforce to address it, it will be chaos. There will be a demand and then in the end it will have less impact.” [Donor Policy Maker] Working conditions Key issues related to the theme of working conditions regarded salaries and lack of recognition of the work and conditions of health workers by policy makers, high workload and a sense of isolation. The reduced salaries for health workers associated with the BPHS increased discouragement among health workers, with three health workers simply stating that there was nothing to motivate them. This feeling was even stronger in facilities managed by the government, where salaries were often paid with great delay, compared to those run by NGOs. Some staff perceived the low incentive in the BPHS as a lack of recognition by the ministry of health and not feeling valued by the government. This was also linked to the perceived deterioration in working conditions compared to the humanitarian relief phase when salaries were higher and more staff were working at the facilities. “I was motivated at the time that MSF, the NGOs here at that time they were paying more. But unfortunately for us MSF left. IRC came,

so they cut down everything. They said they going by government policy [BPHS].” [Nimba Certified Midwife] Low health worker motivation and high staff turnover was a concern among supervisors, and the supervisors advocated for better recognition of certified midwives who were the key providers of SRH services but received lower salaries than nurses and doctors. Several supervisors from both county health teams and NGOs mentioned the difficulties of keeping staff at the facilities, and the constant need to encourage them to stay. The staff turnover also meant that attempts to improve capacity and standards were affected: “Change is a gradual process so in order to reach the standard for BPHS is gradual so we are on a day to day basis trying to teach, to supervise, to coach our people to standard but it becomes difficult because staff are always withdrawing. So after you have gone like maybe 50e60% they will stop and you have to move to new person.” [Nimba Supervisor] Among the policy makers there was generally agreement that the health worker salaries were too low and staff turnover was too high, especially in rural areas. “You send a certified midwife to Bambala, she has no additional incentive, no hardship payment, why should she stay?” [NGO Policy Maker] Respondents from NGOs and donors also highlighted the need for increased recognition of health workers, particularly midwives, by the contracting NGOs and the MOHSW, commenting that MOHSW staff from Monrovia paid insufficient visits to rural facilities. It was noted that midwives also worked in private settings on the side to earn more money. “If you are not satisfied, you do not put an effort, you do not provide good quality services. We can see this in the staff that works in two jobs. A government, B private. The same person can be so different! In the private, they greet the patient and all those things.” [Government Policy Maker] Health workers also felt that the workload, particularly of certified midwives, was too heavy because it included a wide range of SRH services and because they could be called upon at night despite the official working hours of clinics from 8 am to 4 pm. “One person will work the whole night...like the other night I did two deliveries here one night I came to work 8 o’clock in the morning stay there until the next morning then I went home I still went back to work and stay there until in the evening. One person cannot work like that.” [Nimba Clinic Certified Midwife] However, government respondents were less sympathetic to the needs of health workers. Four respondents from the government stated that it was part of a midwife’s job to provide emergency services as required, even at night and during weekends (for no extra payment). “The health workers understand, this is what it entails to be a health worker, you have to work after hours.” [Government Policy Maker] Another important working condition issue raised by health workers was a sense of isolation, feeling “left alone in the bush”, with limited support and opportunities to consult with colleagues. For this reason, the role of supervisors within the BPHS was generally perceived positively by the health care workers who felt that the supervisors provided guidance and supported them by reducing the sense of isolation. However, this support role of the

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49

supervisors was not raised by most policy makers, and they instead questioned the improvements in quality resulting from the supervision. “People in the facility do not know what they need in terms of supervision. There is no correlation between supervision and quality of services.” [Donor Policy Maker] More encouragingly, responses from health staff at facilities about SRH training were positive in terms of building SRH knowledge, skills and also interaction with other health workers. “They help a lot. It is the workshop that helps me to go forward. When you go to workshop it gives you some experience than sitting here using one man’s idea. It motivates me to do my work.” [Cape Mount Clinic Officer-in-Charge] However, the perception of SRH training was more mixed among the policy makers. While it was seen as necessary, most policy makers criticised the large quantity of training which interrupted the provision of services, but also the limited effectiveness of the training. Discussion The BPHS approach is increasingly used in post-conflict countries. The strong emphasis on SRH in the basic packages is a very positive development in improving SRH in post-conflict settings and results from elsewhere suggest the ability of the BPHS to make notable improvements in access to SRH services and improved SRH outcomes (Hansen et al., 2008; Kim et al., 2012). However, there remains limited evidence on the implementation of these basic packages beyond the use of routine monitoring data. This study helps to address this evidence gap by exploring perceptions of the implementation of SRH services in the BPHS in Liberia through interviews with stakeholders at different levels of the implementation process. It also contributes evidence on the needs of health workers in post-conflict countries (Kruk et al., 2009; Leather et al., 2006; Smith & Kolehmainen-Aitken, 2006; WHO, 2005). The study followed a bottomeup perspective, informed by Lipsky’s street-level bureaucrats theory, to help understand policy implementation e most particularly recognising health workers as active participants in shaping how a policy is implemented. Our findings illustrate and confirm the usefulness of Lipsky’s streetlevel bureaucrats theory for understanding how front-line health workers shape the implementation of policy. The findings suggest that health workers respond to a perceived lack of recognition and support for working in difficult conditions by sub-optimal delivery of certain services (particularly facility-based deliveries), delivery of parallel private services, and leaving their posts. These responses distort and undermine the implementation of the BPHS. As noted by Lipsky, health workers have little incentive to fully implement the policies intended by policy makers when they are working in non-conducive conditions and so will change the content of policies in order to control their stress and the complexity of their every day work so as to minimise their own discomfort (Lipsky, 2010). The interview findings suggest that health workers might be willing to provide better care if they felt more motivated and recognized. In addition, the street-level bureaucrats in our study do not necessarily lose anything if they are not satisfying their clients, since those clients have little genuine choice about where to obtain health services. Providers therefore have even less incentive to perform. While it may not be surprising that front-line providers do not fully implement the new policy and feel de-motivated by low salaries, this study identified important issues regarding the

47

implementation process that had not been previously raised, and which also did not appear to be fully recognised by policy makers. The study findings highlighted important differences in the perception between policy makers and health workers. While health workers attributed low uptake of SRH services to low demand arising from socio-cultural factors and the role of trained traditional midwives, policy makers attributed it largely to a supply-side problem because of poor quality of services and poor attitudes towards patients e particularly to women delivering at health facilities. The limited common understanding of quality of care is also an important barrier to the implementation of services. While the criticisms of poor staff attitude and public desire for improved service quality in Liberia have been reported in other studies of Liberia (Kruk, Rockers, Varpilah, & Macauley, 2011a,b), there was a lack of recognition among a number of policy makers of the social and cultural barriers that exist in Liberia regarding the use of health services (Lori & Boyle, 2011), and also limited engagement in understanding why health workers may have a poor attitudes. This limited engagement by policy makers with the perspectives of health workers is also reflected by the fact that policy makers focussed more on the need to expand the list of BPHS services, and rather less on improving the working conditions and constraints to this expansion. A larger list of services in the BPHS could pose a further burden on staff and potential decreases in quality, particularly given the training requirements to provide services for reproductive cancer and fistula to required standards. It may also exacerbate the risk of high staff turnover as reported by health workers. Indeed the limited rise in the number of certified midwives from 297 in 2006 to 412 in 2010 (despite substantially scaled up recruitment and training and a target of 725 for 2011) could potentially be attributable to midwife dissatisfaction and turnover (MOHSW, 2006, 2010c). The influence of low skills and poor attitude of health care workers in dissuading women from service uptake is widely reported in other settings and evidence has shown how improvements in salaries to health workers has helped to improve the uptake of facility-based deliveries (De Brouwere, Richard, & Witter, 2010; Hadi, Rahman, Khuram, Ahmed, & Alam, 2007; Kandeh et al., 1997). Improvements in health care staff salaries, training, professional support and recognition may help to increase health worker commitment to the BPHS and so help improve the quality and uptake of facility-based delivery services. This clearly requires complex and costly political and technical responses (Kruk et al., 2010). But the observed differences in perceptions between health workers and policy makers show the importance of the relationship between central, regional and local actors and the need to have a more common understanding, particularly on the challenges faced by health workers in implementing the BPHS. By putting health workers in the centre of health policy analysis, this study confirms their critical position as they interpret and deliver policies to beneficiaries according to the constraints facing them, as theorised by Lipsky (2010). A further theme influencing the implementation of the BPHS relates to limited understanding of the BPHS itself. In addition to their working routines, values and interests, health workers tend to shape the policy and its implementation in response to their understanding of it. Therefore, the successful implementation of a policy requires clear understanding of the policy by the streetlevel bureaucrats who are involved in its implementation in order to ensure sufficient information about the entire program (Gilson & Erasmus, 2008; Kamuzora & Gilson, 2007; Lipsky, 2010; Spratt, 2009; Walker & Gilson, 2004). In the case of Liberia, health workers were unclear about the BPHS and associated it with low salaries, staffing levels and increased workloads. The limited understanding of the concept of the BPHS combined with changes

48

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49

perceived as detrimental to their working conditions suggest that the changes introduced by the BPHS may not be fully accepted or fully implemented. Greater communication and consultation between central and local actors on the policy and perceived constraints can be important drivers in increasing health care staff motivation and strengthen the services they provide (Walker & Gilson, 2004). This would include ensuring better understanding by front-line providers of the policy objectives and the rationale behind policy but also policy makers taking into account the challenges felt by health workers (Bennett & Miller Franco, 1999; Hornby & Sidney, 1988; Walker & Gilson, 2004). Study limitations The research at health facilities was conducted in two counties and so the findings cannot be generalised across Liberia. The study excluded hard-to-reach counties and hard-to-reach facilities within Grand Cape Mount and Nimba for logistical reasons. As a result, implementation issues may vary between and also within in the selected counties (although the results of the BPHS accreditation monitoring activities do not show significant differences between the facilities included in our study and the others in the two study counties (MOHSW, 2010a, 2011)). In addition, certain relevant key informants who were knowledgeable about the development and implementation of the BPHS in Liberia were not interviewed as they were no longer in the country and could not be contacted. Conclusions Confirming the usefulness of Lipsky’s street-level bureaucrat theory, our qualitative findings have helped to provide a better understanding of the influence of health care workers implementing the BPHS policy in Liberia. The findings highlight potential mechanisms by which policy implementation may be distorted due to an important disconnect between health care workers and policy makers and limited understanding of the policy. The findings suggest the need for greater dialogue between policy makers and health care workers to improve understanding by health workers of the BPHS and recognition by policy makers of the working conditions of health workers in order to help further achieve the potential benefits of the BPHS in Liberia. Acknowledgements This study was financially and logistically supported by the World Health Organization in Geneva and Liberia, and financially supported by the Reproductive Health Access, Information and Services in Emergencies Initiative (RAISE). References Ameli, O., & Newbrander, W. (2008). Contracting for health services: effects of utilization and quality on the costs of the Basic Package of Health Services in Afghanistan. Bulletin of the World Health Organization, 86(12), 920e928. Bennett, S., & Miller Franco, L. (1999). Health worker motivation and health sector reform: a conceptual framework. In Partnership for health reform projects. (Ed.), Major applied research: Technical paper. Bethesda: United States Agency for International Development. Buse, K. (2007). How can the analysis of power and process in policy-making improve health outcomes? In Overseas Development Institute. (Ed.), Briefing paper London: Overseas Development Institute. Buse, K., Mays, N., & Walt, G. (2005). Making health policy. Berkshire: Open University Press. Cleveland, E. C., Dahn, B. T., Lincoln, T. M., Safer, M., Podesta, M., & Bradley, E. (2011). Introducing health facility accreditation in Liberia. Global Public Health, 6(3), 271e282. De Brouwere, V., Richard, F., & Witter, S. (2010). Access to maternal and perinatal health services: lessons from successful and less successful examples of

improving access to safe delivery and care for the newborn. Tropical Medicine and International Health, 15(8), 901e909. Gilson, L., & Erasmus, E. (2008). Tackling implementation gaps through health policy analysis. In Equinet. (Ed.), Policy series. Harare: Equinet. GoL. (2009). Liberia national health accounts 2007/2008. Government of Liberia and Health Systems 20/20 Project. Government of Liberia. (2008). 2008 National population and housing census: Preliminary results. Monrovia, Liberia: Government of the Republic Of Liberia. Hadi, A., Rahman, T., Khuram, D., Ahmed, J., & Alam, A. (2007). Raising institutional delivery in war-torn communities: experience of BRAC in Afghanistan. Asia Pacific Family Medicine, 6(1). Hansen, P. M., Peters, D. H., Niayesh, H., Singh, L. P., Dwivedi, V., & Burnham, G. (2008). Measuring and managing progress in the establishment of basic health services: the Afghanistan health sector balanced scorecard. International Journal of Health Planning and Management, 23(2), 107e117. Hill, M., & Hupe, P. (2002). Implementing public policy. London, Thousand Oaks, New Delhi: Sage Publications. Hjern, B., & Porter, D. O. (1981). Implementation structures. A new unit of administrative analysis. Organization Studies, 2, 211e227. Hornby, P., & Sidney, E. (1988). Motivation and health worker performance. In World Health Organization. (Ed.), WHO-EDUC 88.196. Geneva: World Health Organization. IMF. (2008). Liberia: Poverty reduction strategy paper. Poverty Reduction Strategy Papers. Washington DC: International Monetary Fund. Kamuzora, P., & Gilson, L. (2007). Factors influencing implementation of the Community Health Fund in Tanzania. Health Policy and Planning, 22, 95e102. Kandeh, H. B., Leigh, B., Kanu, M. S., Kuteh, M., Bangura, S., & Seisay, A. L. (1997). Community motivators promote use of emergency obstetric care services in rural Sierra Leone. The Freetown/Makeni PMM team. International Journal for Gynecology and Obstetrics, 59(Suppl. 2). Kennedy, S. B., Atwood, K. A., Harris, A. O., Taylor, C. H., Gobeh, M. E., Quaqua, M., et al. (2011). HIV/STD risk behaviors among in-school adolescents in postconflict Liberia. Journal of the Association of Nurses in AIDS Care, 23(4), 350e360. Kim, Y. M., Zainullah, P., Mungia, J., Tappis, H., Bartlett, L., & Zaka, N. (2012). Availability and quality of emergency obstetric and neonatal care services in Afghanistan. International Journal of Gynaecology and Obstetrics, 116(3), 192e196. Kruk, M. E., Freedman, L. P., Anglin, G. A., & Waldman, R. J. (2009). Rebuilding health systems to improve health and promote statebuilding in post-conflict countries: a theoretical framework and research agenda. Social Science & Medicine, 70(2010), 89e97. Kruk, M. E., Rockers, P. C., Varpilah, S. T., & Macauley, R. (2011a). Population preferences for health care in Liberia: insights for rebuilding a health system. Health Services Research, 46(6pt2), 2057e2078. Kruk, M. E., Rockers, P. C., Varpilah, S. T., & Macauley, R. (2011b). Which doctor?: determinants of utilization of formal and informal health care in postconflict Liberia. Medical Care, 49(6), 585e591. Kruk, M. E., Rockers, P. C., Williams, E. H., Varpilah, S. T., Macauley, R., Saydee, G., et al. (2010). Availability of essential health services in post-conflict Liberia. Bulletin of the World Health Organization, 88(7), 527e534. Leather, A., Ismail, E. A., Ali, R., Abdi, Y. A., Abby, M. H., Gulaid, S. A., et al. (2006). Working together to rebuild health care in post-conflict Somaliland. Lancet, 368(9541), 1119e1125. Lipsky, M. (2010). Street-level bureaucracy. Dilemmas of the individual in public services. London: Sage Publications. Loevinsohn, B., & Sayed, G. D. (2008). Lessons from the health sector in Afghanistan: how progress can be made in challenging circumstances. Journal of the American Medical Association, 300(6), 724e726. Lori, J. R., & Boyle, J. S. (2011). Cultural childbirth practices, beliefs, and traditions in postconflict Liberia. Health Care for Women International, 32(6), 454e473. Matland, R. E. (1995). Synthesizing the implementation literature: the ambiguityconflict model of policy implementation. Journal of Public Administration Research and Theory, 2, 145e174. MOHSW. (2006). Emergency human resources for health plan. Monrovia. MOHSW. (2007a). Liberia national health policy and plan 2007e2011. Monrovia. MOHSW. (2007b). Roadmap for the implementation of the basic package of health services. Monrovia. MOHSW. (2008a). Basic package of health and social welfare services. Monrovia. MOHSW. (2008b). Liberia demographic and health survey 2007. Monrovia: MOHSW/Liberia Institute of Statistics and Geo-Information Services/Macro International Inc. MOHSW. (2008c). National conference on the review of the first year of implementation of the national health plan. Monrovia. MOHSW. (2009). September 2009 BPHS accreditation: Final results report. Monrovia. MOHSW. (2010a). January 2010 BPHS accreditation final results report. Monrovia. MOHSW. (2010b). National family planning program strategy. Monrovia. MOHSW. (2010c). The national health census of health workers in Liberia. Monrovia. MOHSW. (2010d). National sexual and reproductive health policy. Monrovia. MOHSW. (2010e). Roadmap for accelerating the reduction of maternal and newborn morbidity and mortality in Liberia. Monrovia. MOHSW. (2011). January 2011 BPHS accreditation final results report. Monrovia. Msuya, C., & Sondorp, E. (2005). Interagency health evaluation Liberia. Interagency Standing Committee. Palmer, N., Strong, L., Wali, A., & Sondorp, E. (2006). Contracting out health services in fragile states. British Medical Journal, 332(7543), 718e721.

D. Petit et al. / Social Science & Medicine 78 (2013) 42e49 Pavignani, E. (2009). Liberia case study: aid effectiveness during transition from relief to development funding. In E. Pavignani, & S. Colombo (Eds.), Analysis of disrupted health sectors. Geneva: World Health Organization. Peters, D. H., Noor, A. A., Singh, L. P., Kakar, F. K., Hansen, P. M., & Burnham, G. (2007). A balanced scorecard for health services in Afghanistan. Bulletin of the World Health Organization, 85(2), 146e151. Pope, C., & Mays, N. (1995). Reaching the parts other methods cannot reach: an introduction to qualitative methods in health and health services research. British Medical Journal, 311(6996), 42e45. Pulzl, H., & Treib, O. (2007). Implementing public policy. In F. Fisher, G. J. Miller, & M. Sidney (Eds.), Handbook of public policy analysis. Theory, politics and practice. Boca Raton, FL: Taylor & Francis Group. Sabatier, P. A. (1986). Top-down and bottom-up approaches to implementation research: a critical analysis and suggested synthesis. Journal of Public Policy, 6, 21e48. Smith, J., & Kolehmainen-Aitken, R. (2006). Establishing human resource systems for health during postconflict reconstruction. Occasional Papers(3). Boston: Management Sciences for Health.

49

Spratt, K. (2009). Policy implementation barriers analysis: Conceptual framework and pilot test in three countries. Washington, DC: Future Group, Health Policy Initiative, Task Order I. Strong, L., Wali, A., & Sondorp, E. (2005). Health policy in Afghanistan: Two years of rapid change. London: LSHTM. Walker, L., & Gilson, L. (2004). We are bitter but we are satisfied. Nurses as street level bureaucrats in South Africa. Social Science & Medicine, 59, 1251e1261. Waters, H., Garrett, B., & Burnham, G. (2007). Rehabilitating health systems in postconflict situations. In World Institute for Development Economics Research. (Ed.), Research paper no. 2007/06. United Nations University. WHO. (2005). Guide to health workforce development in post-conflict environments. Geneva: World Health Organization. WHO. (2008). Essential health packages: What are they for? What do they change? Geneva: World Health Organization. WHO. (2010). Trends in maternal mortality 1990e2008. Geneva: WHO, UNICEF, UNFPA, World Bank.