Assessment of linkage between Productive Safety Net Program and Health Services in Somali Region, Ethiopia: Lessons, Challenges and Missed Opportunities

Background: Ethiopia’s Productive Safety Net Programme (PSNP) which has been implemented since 2005 is a large-scale, social protection intervention aimed at improving food security. The fourth phase of the PSNP included a system of integrated health and nutrition service delivery for its categories of beneciaries especially the creation of a temporary direct support(TDS) category for clients that are pregnant and lactating women (PLW) or caretakers of malnourished children, who are exempted from public work but expected to comply with co-responsibilities which counts towards their public works requirement aimed at improving utilization of health services. Methods: The study was a cross-sectional descriptive survey and used qualitative methods, in-depth interviews and focus group discussions (FGDs), conducted in two woredas( districts) (Gursum and Kebribayah) in Farfan zone of Somali region. The study population were key individuals involved in the linkages of PSNP with health services and the beneciaries. The study assessed the implementation of the linkage between PSNP4 and health services Results: The study observed that the stakeholders have adequate knowledge and understood their roles in the linkages between PSNP and the health services, in addition the beneciaries also are aware of their rights and the process for exemption from public work. However, the major issues identied included poor coordination among the implementing actors, poor knowledge , monitoring and compliance with the co-responsibilities by the stakeholders and the beneciaries. Conclusion: Considering the size of the program in the region which is targeted to the poor, the Productive Safety net program being the main tool to help forward Ethiopia’s Social Protection Policy and Strategy, has the potential to improve access and utilization of health and nutrition services if more efforts are put to strengthen integration and linkages with the health sector and monitoring of compliance of co-responsibilities by the beneciaries of the program.


Introduction
Ethiopia's Productive Safety Net Programme (PSNP) which has been implemented since 2005 is a large-scale, social protection intervention aimed at improving food security. It involves a mix of public works, employment and unconditional cash and food transfers for chronically food-insecure households identi ed through a mix of geographic and community-based information. The main objective of the PSNP is to increase livelihoods and resilience to shocks and to improve food security and nutrition for vulnerable rural households.
The fourth phase of the PSNP 1 was launched in 2015 and included a system of integrated health and nutrition service delivery and the creation of a Temporary Direct Support(TDS) category for clients that are pregnant and lactating women (PLW) or caretakers of malnourished children, who are public works clients as part of the several innovations introduced on PSNP4 to enhance programme outcomes, speci cally in terms of child nutrition.
The pregnant and lactating women (up to one year after birth) and primary caregivers of malnourished children designated as temporary direct support (TDS) clients are to comply with co-responsibilities which counts towards their public works requirement aimed at improving the health and nutrition of themselves and their children. The co-responsibilities have been introduced for TDS and PDS clients, particularly to strengthen linkages between PSNP with health care services.
These co responsibilities include attendance of four ante-natal consultations, one post-partum health facility visit, uptake of routine immunization, attendance at growth monitoring and promotion, behavioural change communication sessions, monthly check up of the malnourished child at the closest health facility for appropriate treatment.
The effective linkage of PSNP with health services which is aimed at improving maternal and newborn and child health and nutrition outcomes is very relevant and essential in Somali region of Ethiopia which has shown less improvement in key maternal and child health and nutrition outcomes compared to the national average. The 2019 Mini DHS reports that the region has the lowest percentage of fully vaccinated children, with only18.2% having received all basic vaccinations, and 48.8% had received no vaccinations at all compared to national average of 43.1% and 19.2% respectively while the proportion of pregnant women who received at least four antenatal care services(ANC4) and delivered by skilled birth attendant is 11.1% and 26% compared to national average of 43% and 49.8% respectively and post-natal care rate (10.3 per cent). Malnutrition continues to be a serious issue in the region with wasting rate of 22.7 percent compared to 9.9% national average and contribute to approximately 25% of the total number of children with wasting in the country. Stunting prevalence in children under 5 years of age stands at 30.5 per cent and 40 per cent of children are exclusively breastfed during the rst six months of life, and

Page 3/11
Study population and setting: The study was conducted in 2 woredas( districts) (Gursum and Kebribayah) in Farfan zone of the region being the zone with highest number of PSNP bene ciaries in the region.
Kebribeyah woreda has estimated total population of 130,763 , 5 health centers and 33 health posts while Gursum woreda has estimated total population of 39,375, 3 health centers and 16 health posts. The total number of PSNP 4 bene ciaries in the region is 1,673,009 ; Kebribayah woreda has the largest caseload of bene ciaries of 84,977 while Gursum woreda in the same zone has one of the lowest caseloads of 18,245 bene ciaries.
The study population were key individuals involved in the linkages of PSNP with health services at the regional, woreda and kebele levels and the bene ciaries identi ed based on the World Bank's framework on accountability. 14 Study Design: The study was a cross-sectional descriptive survey and used qualitative methods Sampling Technique: Multistage sampling technique was used. In the rst stage, two woredas and regional level bureaus were selected purposely, Kebribeyah woreda which has the highest caseload of bene ciaries in the region and Gursum woreda which has one of the lowest caseloads in the same zone( Farfan) with Kebribeyah woreda.
In the second stage, three kebeles(sub-districts) from each of the two purposely selected woredas (districts) were selected randomly from among the total list of PSNP kebeles in each of the two selected woredas.
The participants interviewed were purposefully identi ed and recruited based on their positions and roles in the implementation of PSNP and linkages with health services in the study sites.
Study participants were the agriculture and health sectors focal persons at the regional, woreda and kebele levels: Sectoral focal points for health and agriculture at the Regional Level Data was collected through in-depth interviews and focus group discussions (FGDs). The in-depth interviews aimed at assessing the linkage between PSNP and health services.
Interview guide with semi-structured interview questionnaire was used to obtain information on a range of issues about PSNP implementation and linkages with health services from the participants.
The study participants during the Key informant interviews were asked about their knowledge and roles in the transition of eligible households to TDS clients from public works, the co-responsibilities of the bene ciaries and barriers to effective linkages. This was done at the region, two woredas and six kebeles.
The FGDs was conducted in the six selected kebeles among the TDS clients and assessed their knowledge about their rights, selection process and co-responsibilities and their participation as required.
The study was conducted in between March and April 2021 and The FGDs and KII were audio-taped and notes were also taken with prior consent from the participants.
Data Analysis: The FGD and KII data collected from different categories of respondents and responses to the same questions were triangulated and transcribed verbatim to produce transcripts of narrative text for thematic analysis. The data were coded according to the types of themes and issues and thematic analysis was used which comprised a mix of inductive and deductive coding. Lactating women with a child less than one year old: attendance at one post-partum health facility visit; follow recommended immunization schedules for infants and attendance at growth monitoring and promotion/behavioural change communication sessions.
Primary caregiver of a malnourished child under ve years old during treatment are attend the clinic monthly to complete the treatment (e.g. community management of acute malnutrition or targeted supplementary feeding) as advised.

Results
The results are organized in two major themes: Process involved in the linkages between PSNP and health services for the temporary direct support (TDS) clients and barriers to effective linkages. Section 1: Stakeholders knowledge about the linkages between PSNP and health services -this focused on the knowledge and roles of the stakeholders at the regional, woreda and kebele levels about the process for transition of eligible households to temporary direct support (TDS) clients from public works and the co-responsibilities of the bene ciaries.

Knowledge and role in transition of eligible households to temporary direct support (TDS) clients
The Key informants were asked about their knowledge and roles in the process for transition of eligible households to temporary direct support (TDS) clients. All the key informants interviewed were familiar with the recommended process and their roles in the transferring of pregnant and lactating women and households with children with malnutrition from the public work to temporary direct support clients which exempt them from the public work. However, there was variation in their knowledge about the exact duration the TDS clients are to be exempted from public work as detailed in the project implementation manual (PIM).

Regional perspective :
Even though actual implementation of linkage is done at the kebele level, however the regional and woreda team are involved in the process and expected to provide needed monitoring and support to ensure this is effectively done.
'PSNP is a multi-sectoral project in its nature and has relations with health, and when a mother is around three months pregnant, she contacts the DA who completes a form and that allows the pregnant mother to get a leave from the public work until her return when her child gets 10months old. This linkage exists at regional, woreda and Kebele level.' Generally, the role of health extension workers (HEWs) is to identify pregnant women, lactation women and households of children with malnutrition who are engaged in public work and provide letter upon con rmation in the health facilities to the Development Agents so they can be exempted from public works. The Development agents who keep the names of all PSNP bene ciaries are to effect their transfer to temporary direct support groups and continue to provide them with the support throughout the duration until they return back to public work. However, the DAs and HEWs interviewed provided different expected duration of exemption from public works. ' When pregnant women tell us that they are pregnant or we notice it ourselves, we asked them to bring con rmation letter from the health facility and based on that we exempt them from public work.' (Development Agent Key Informant 2) 'As a Health extension worker, my role is to register the lactating, pregnant mothers and malnutrition children and send them to the DA to transfer them until 6 months after they deliver, and the child is cured from malnutrition.' ( Health Extension Worker Key Informant 1) 'As HEW my responsibility is that when mothers realize they are pregnant they come to me ,I give them a con rmation letter which will be used to exempt her from working during her pregnancy and six months after delivery.'

( Health Extension Worker Key Informant 2)
'When mothers claim to be pregnant it's our task as HEWs to con rm their pregnancy and then write con rmation report to the kebele DA for the project to transfer the status of the pregnant women to TDS group and exempt them from work until a year after delivery.' ( Health Extension Worker Key Informant 3) 1.2 Knowledge about the co-responsibilities of temporary direct support (TDS) clients/ The Key informants were aware the Temporary direct support (TDS) clients are supposed to participate in some activities referred to as co-responsibilities in replacement of the public work. However, they did not appear to have more precise knowledge of co-responsibilities about clients' speci c obligations other than the general advice of coming to the health facilities, but the schedules and number of visits expected not well articulated The development agents and health extension workers are expected to orientate the TDS bene ciaries about their co-responsibilities as detailed in the PIM and follow them up to monitor compliance.
However , after linking them up with the development agents who registers them, they do not follow strictly to monitor the co-responsibilities but manage them like other patients who are to come to the health facilities routinely.
Kebele perspective "Once we transfer them and exempt them from public work, they are to be seen and monitor by the health workers in the clinics to ensure they go to the clinic regularly to receive services." 'There is no special clinic for them, we are busy, so we attend to them like other patients when they come and sometimes, they don't come to the clinic again.
We don't have a special register for them.' The knowledge of the pregnant women, lactating mothers and mothers of children with malnutrition seen during the various focus group discussions( FGDs) show that they are aware of the process for identi cation of households eligible to be excluded from the public work and transferred to the TDS categories and still receive their monthly bene ts both money or food or other materials.
"When a woman gets pregnant, she comes to the hospital for the nurse to con rm her pregnancy and give her a letter to the DA and then she will not work again until baby reaches two years and she will still continue to collect the money even though not working.' ( FGD participant I) "When we become pregnant, we become free from the public work until we give birth and that child become one year, after that we return back to the public work.' (FGD participant 2) "When the women is lactating and breastfeeding her children she is given a free service and not expected to work in the public work, also she gets the payment and other supportive and need materials, she has been getting even though she is not working in the public work." (FGD participant 3) "When a malnourished child is seeing in the clinic, the mother will be given a letter so the family will be free from the public work until the child recovers so they can have time to treat their child, and they will get the payment without working.' (FGD participant 4)

Knowledge of the temporary direct support(TDS) clients about their Co-responsibilities
Most of participants in the focus group discussions were not aware of the speci c obligation and required number of visits to the health facility for their co-responsibilities in lieu of the public work other that the general advice on health seeking behaviour and clinic attendance for relevant services. 'We are expected to participate in different community awareness related activities and attend immunizations before and after delivery for our children and Most of the barriers/challenges to the linkages were expressed by the woreda key informants and this ranges from poor coordination and communication between the key actors( HEWs and DAs), poor knowledge due to lack of training on the project especially clear guideline to clarify the expected role of each of the actors especially in reference to the monitoring of the bene ciaries. Other barriers mentioned included lack of dedicated budget for the health sector either to follow up and monitor the compliance, no clear guideline on reporting template or mechanism for the co-responsibilities and limited access to health services for TDS bene ciaries who live far from health centers or in kebele with no health facilities. Some also mentioned that the project is seen and being implemented more as belonging to the agricultural sector and not as a key element of local development planning as designed.
Regional perspective: ' There is poor awareness and capacity regarding linkage between the two sectors. There is lack of clear and simpli ed linkage guides for DA, HEWs and woreda steering committee on implementation of the linkages.' ( Regional key Informant 1) "The DAs and HEWs are expected to monitor the TDS clients to ensure they ful l the obligation even tough without any sanction, but the record is kept at the kebele level, we don't include this in our report at the regional level but we have noticed that there is poor documentation system at woreda and kebele level about the TDS record of compliance.' ( Regional Key Informant 1) 'Except participating the meeting there is no intersectoral collaboration between PSNP and regional/ woreda health o ces, there is no regular weekly, monthly, or quarterly coordination platform to discuss PSNP performance at regionally.'

Discussion
The transition of the eligible pregnant and lactating women and households with children with malnutrition from public work to temporary direct support category and implementation of the co-responsibilities( soft conditions) are key components of the linkage between PSNP and health services.
The study observed that the stakeholders have adequate knowledge and understand their roles in the linkages between PSNP and the health services through the temporary direct support clients, in addition the bene ciaries also are aware of their rights and the process for transitioning .However the major issues are the poor coordination at the kebele level among the implementing actors, poor knowledge, monitoring and compliance with the co-responsibilities by the stakeholders and the TDSs clients which is the crux of project so as to improve demand and utilization of services for better maternal newborn children health and nutrition outcomes.
The study shows that the selection of the bene ciaries was adequately done and no problem or complaint by the bene ciaries this is unlike in a study 16 on safety net program in some countries in Latin America and the Caribbean, which reviewed linkage between conditional cash transfer program and health services and reported poor identi cation of bene ciaries as one of the major problem in the project.
Our study found weak coordination among the key actors at all level in terms if planning, joint monitoring, report sharing and participation in coordination meetings. This seems to be a major problem in many interventions that requires multisectoral collaboration, the nding in our study is similar to ndings from other studies in interventions that involved multisectoral collaboration that also observed poor coordination among the implementers especially at the community level even though they reported better coordination at the central level. A study 17 in Malawi that looked at integration of intervention to improve nutrition outcomes found strong multidisciplinary interaction exists at central levels but not at the community level, this is similar to study 16 on improved linkages between conditional cash transfers and reproductive health programs in Latin America and the Caribbean that reported limited integration at the subdistrict level among the various actors and suggested increase supervision by Regional team needed to ensure effective linkages. A land scape analysis 18 done in 19 countries that assessed the extent of agricultural investments contribution to nutrition outcome noted that strong collaboration observed at the higher policy level ,national coordination and district coordination levels does not extend to the community level in most of the countries studied and provided practical guidance on how to initiate and manage multi-sectoral approaches and improve coordination and collaboration across a range of stakeholders.
The focus of the PSNP in linkage can only be achieved if the co-responsibilities are adhered to which is aimed at ensuring pregnant and lactating women and children with malnutrition have access to regular health services to improve maternal newborn child health and nutrition and reduce defaulter among children seeking nutrition and immunization services especially in a region like Somali with poor health and nutrition indices. However, our study found out that knowledge of the stakeholder and TDS bene ciaries about the co-responsibilities( soft conditions) is poor. In addition, there is no established follow up or monitoring, or reporting mechanism to monitor compliance. Some of the factors mentioned responsible for this included work overload by HEWs, limited or no budget/operational/logistic support for the HEWs for follow up and monitoring.
As found in our study, these respondents both the health workers, the development agents and the TDS clients did not appear to have more precise knowledge of co-responsibilities in terms of the speci cs apart from the general advice for health seeking behaviour, this is similar to nding from earlier studies 15,19 in Ethiopia which reported imperfect understanding among woreda and kebele staffs and bene ciaries did not have adequate knowledge on co-responsibilities especially on the speci c obligations of the bene ciaries. A similar review study 16 on conditional cash transfer and linkage with reproductive health reported that in many of the countries evaluated ,a large proportion of program bene ciaries were not aware of, or did not know about, all or even some of the conditions for to be ful lled in receiving payments. However, an intervention study 20 in Ethiopia where social workers were engaged to follow up and monitor compliance to co-responsibilities of bene ciaries, found that the knowledge and compliance to co-responsibilities was higher among those in intervention area compared to the control group with no active follow up or monitoring of compliance.
Our study found out that there is no or limited monitoring of compliance to co-responsibilities by the TDS clients with associated risk of not being able to achieve the objective of exemption of the TDS bene ciaries from public work and help improve maternal newborn and child health and nutrition outcomes. A study 21 on conditional cash transfer shows that only if the conditionality is monitored and compliance enforced will it have an effect and found that similar to our study in some conditional cash transfer programs, compliance were not necessarily being monitored. The study noted that in conditional cash transfer programs where there was effective monitoring, compliance was found to be extremely high by up to 94 percent of households and reported that in cases without monitoring the evidence of the impact of the program was mixed. A study by Morris et al (2004) 22 found null effects in the Honduran conditional cash transfer program where conditionality was not enforced and noted that unless adequate measures are put in place to monitor compliance the program bene ciaries will not comply with program conditions and suggested joint review adherence to the co-responsibilities by all stakeholders to be part of the activities in the project cycle.
One of the major barriers to monitoring of the compliance to co-responsibilities and poor intersectoral collaboration at the kebele level by the HEWs is the workload especially in health posts where mostly only one health extension worker is assigned as result of increase demand and responsibilities expected from HEWs( monitor co-responsibilities, provide services for the PSNP client,, follow up ) in addition to the routine activities in the clinics. A previous study 19 in Ethiopia also identi ed human resource shortage with too few staff with heavy workloads as factors that constrained the program implementation by reducing the ability of the Health extension workers to conduct home visits and monitor compliance to the co-responsibilities. This is similar to ndings in previous studies 23,24 which found that community health workers performance is affected by workload, number of tasks, the size of the catchment area, organization of tasks and proposed that community health workers have manageable workload in terms of a realistic number of tasks and clients, an organized manner of carrying out these tasks and a reasonable geographic distance to cover.
A study 17 in Malawi on the integration of community health workers from different sectors reported the integration was affected/limited by workload and suggested a need to have a liaison o cer who should track integration of the activities.
The use of social workers to track and monitor implementation of integration was found to be effective in as previous study 20 in Ethiopia which showed improved multisectoral collaboration among social workers and local development agents, health extension workers, which improved client ful lment of co-responsibilities in health compared to where social workers were not engaged. Access to health services by the temporary direct support bene ciaries especially those who live in kebeles/sub kebeles where there are no health facilities or those who live far from the nearest health facilities and requiring to travel from a long period of time was reported to be another barrier to compliance with coresponsibilities. This is similar to a study 17 on conditional cash transfer program which identi ed access to health services by bene ciaries as a major challenge in effective linkage between the program and health services and implemented various strategy including use of mobile team and use of NGOs to provide services to the population including home visits and follow up which improved access to services and ensure better results.
A study 20 in Ethiopia shows that temporary direct support bene ciaries who were aware and able to ful l their co-responsibilities said that they had been able to do so because their health posts were near their homes and provided good services.
Other concern raised by the health extension workers affecting their performance and for effective intersectoral collaboration and follow up, monitoring and reporting compliance to co-responsibilities was lack of funding allocated to the health sector for logistic or operational support including inadequate transportation unlike the agriculture sector. A previous study 19 in Ethiopia also identi ed inadequate transport support as being responsible for infrequent visits by the HEWS to conduct home visit of monitor compliance especially in remote communities. This is similar to other studies 25, 26 that review the practical implication of linkages between agriculture and health sector which reported that funding and budget controls as the core reason why there was little interaction between the two sectors. A similar study 27 in Ethiopia also identi ed lack of resources within individual sectors as barriers to effective coordination between health and agriculture and these studies recommended development of joint work plan with budget allocated for all activities to be implemented by both sectors for effective integration of intervention.
Another barrier to effective linkage and multisectoral collaboration reported in our study is the perception by the respondents in health sector about the PSNP being seen as a project that solely belong to the agriculture sector and that the success will be attributed to the sector because this is managed as vertical program and not included as part of the responsibilities of the health workers in the woreda plan. This is similar to nding in a study 26 that assessed the role of agriculture program in improving nutrition outcome which identi ed sector mandate and priorities as one of the barriers to integration because for the individual civil servant working within the con nes of a sector ministry or agency, personal incentives like career advancement tend to revolve around their contribution to that attainment of narrowly sector-speci c objectives.
Another study in 27 Ethiopia on linkages between health and agriculture sectors also identi ed presence of competing priorities within individual sectors as barriers to effective coordination between the two sectors and the studies recommended inclusion of the integrated activities in the workplan or deliverables of the staffs in the health and agriculture sectors.

Limitation of the study
The study couldn't report on the number of temporary direct clients who adhered to co-responsibility and effect on utilization of health services because of no available data at the time of the study. The ndings in our study is also subject to response biases since we rely on the information provided by the respondents.
We suggest further study to look at the effect of compliance with co-responsibilities by the bene ciaries on the utilization of services and health and nutrition outcomes we suggest a future research to look at the effect of the on utilization of services and also cost effective.

Conclusion
The Productive Safety net program which is the main tool to help forward Ethiopia's Social Protection Policy and Strategy, has the potential to improve maternal newborn and child health and nutrition outcomes if more efforts are put to strengthen integration and linkages with the health sector and monitoring of compliance of co-responsibilities by the temporary direct support bene ciaries of the program. Considering the size of the program in the region targeted to the poor, the bene ts of improving linking the bene ciaries clients with health services could be decisive for the government and all stakeholders in the region . The focus should be beyond coordination but deliberate pursue of collaboration among agriculture and health sectors especially at the kebele level which will involve sharing of resources and enhancing one another's capacity for mutual bene t and to achieve a common purpose in addition to exchanging information and altering/ synchronization of activities. The development and deployment of a management information system to provide timely data about temporary direct support bene ciaries and improved monitoring systems would help strengthen the linkages.