Challenges in Implementing the Integrated Community-Based Outpatient Therapeutic Program for Severely Malnourished Children in Rural Southern Ethiopia

Currently, treatment of uncomplicated severe acute malnutrition is managed in the integrated Community based Outpatient Therapeutic Program (C-OTP) using ready-to-use therapeutic foods (RUTFs). The aim of this study was to determine challenges in implementing the critical steps in C-OTP and caregivers’ perceptions of service provision in southern Ethiopia. A total of 1048 caregivers of children admitted to the OTP and 175 Health Extension Workers (HEWs) from 94 selected health posts were included in the study. Program admission, follow-up and exit information was collected from caregivers during home visits. HEWs were interviewed at their respective health posts. Only 46.6% (481/1032) were given the recommended amount of RUTF and 19.3% (196/1015) were given antibiotics on admission. During C-OTP participation 34.9% (316/905) had uninterrupted provision of the recommended amount of RUTF. Of the children who left the program, 220/554 (39.7%) exited the program in line with the national recommendation. Caregivers (42.9% (394/918) and HEWs (37.1%, 62/167) perceive that RUTFs were being sold as a commodity. Inadequate provision and unintended usage of RUTFs, lack of antibiotics and inappropriate exit from the program were major constraints. For successful saving of lives, adequate resources must be allocated, and providers must be trained regularly, and supervised properly.


Introduction
Globally, 19 million children under five are suffering from Severe Acute Malnutrition (SAM) [1]. Three out of four children diagnosed with SAM are uncomplicated, and the majority have a good appetite for ready-to-use therapeutic foods (RUTF) and do not have co-morbidities [2][3][4]. The World Health Organization (WHO) recommends the Community-based Outpatient Therapeutic Program (C-OTP) as a standard treatment protocol for the management of uncomplicated SAM at the community level [4,5]. Based on achievements from small-scale programs, C-OTP has been scaled up and integrated into the lowest level government health systems in several developing countries [6][7][8][9] where children are diagnosed, classified and treated for SAM by Community Health Workers (CHWs). CHWs have limited training and have many other responsibilities; thus, one would expect SAM management is likely to differ from small-scale programs which are implemented by non-governmental organizations [3,5].

Study Design and Setting
This cohort study was part of a larger research project evaluating effectiveness of community-based management of an acute malnutrition program in southern Ethiopia. Children admitted to C-OTPs were identified through a weekly visit to all the health posts in the selected districts. Caregivers of children aged from 6 to 59 months, from whom the research team collected anthropometric data within seven days of admission, and their caregivers were eligible to participate in the study. Four home visits were carried out; the first visit was carried out within one week of admission, and follow-up visits took place at 4, 8 and 14 weeks after admission. Enrolment and data collection are carried out from July to December 2011.
Mostly rural, the residents of the four districts we included in this study engage in crop-livestock mixed farming both for household consumption and marketing. However, recurrent drought and subsequent crop failure in the area has resulted in cyclical nutritional emergencies that have resulted in the districts being recipients of food aid for decades [19,20]. Since 2004, C-OTP has been scaled up and integrated into the government health system in Ethiopia. Guidelines for providing services at community level by Health Extension Workers (HEWs) has been in place since 2007. The HEWs were trained on SAM management and simplified guidelines in local language were prepared which is a key element in the utilization of RUTF innovation [6,21]. HEWs are women CHWs, who have been trained for 10 months through the national Health Extension Program, and who provide basic curative and preventive health services to rural communities for which they are given a salary [3].

Participants and Data Collection
The study participants were caregivers of children admitted to C-OTP and HEWs who provide the service at the 94 health posts. A total of 1048 children admitted to C-OTP and their caregivers were included in this study. Of these, 57 caregivers and their children were lost to follow-up and 18 children died before discharge from the C-OTP ( Figure 1). Caregivers were interviewed using a structured questionnaire to collect information on socio-demographic and household characteristics and perceptions of the service provided by the HEWs. All 175 HEWs working in the selected health posts were included in the study. HEWs were interviewed using a structured interview questionnaire regarding their experience of SAM management. Data collectors were 23 women nurses who were not involved in C-OTP implementation and who were not employed by any government health system. They were trained in anthropometric measurement techniques and went through standardization sessions to ensure the accuracy and precision of the information gathered by the measurements according to established guidelines [3]. Responses from HEWs were triangulated with those of the caregivers to add to the validity of our results.

Outcome Measures and Statistical Analysis
The outcome measures in this study were both program-and caregivers-related. The program-related outcomes include the critical steps in SAM management as outlined in the national and WHO guidelines: (1) application of diagnostic criteria for identifying and classifying SAM; (2) provision of recommended amount of RUTF and antibiotics; (3) follow-up during children's stay in the program; (4) length of C-OTP stay; and (5) application of C-OTP exit procedures. The SAM diagnostic elements were: measuring MUAC; assessing the presence of edema; co-morbidities and carrying out appetite test for RUTF. The former two measurements are to identify SAM while the latter two are for classifying SAM as being either complicated or not [3,22]. Children who were assessed for all the four diagnostic elements were considered as being diagnosed appropriately. RUTFs provision was measured against the recommended amount on admission, and at the 4th and 8th week follow-ups [23]. Key elements of follow-up during the C-OTP stay were measured by performing a check-up of nutritional status (measuring weight and assessing edema status) at the 4th week of admission and length of admission to the program in weeks. The length of C-OTP admission was considered appropriate if it was within the national recommended length of 8 weeks. Follow-up was considered to be appropriate when nutritional status was assessed and the stay in the C-OTP was ď8 weeks. Unlike the above program-related outcomes, the application of C-OTP exit procedures were measured only for children who had SAM on admission and an anthropometric assessment was done within 7 days of discharge (n = 554). Exit from C-OTP was considered appropriate when discharged children achieved the program recovery criteria (a gain of 15% of admission weight for children without edema on admission and resolution of edema for those with edema on admission) while those who failed to achieve the recovery criteria were transferred to inpatient care.
Caregiver-related outcomes were: (1) perception of RUTF usage; (2) experience of attending scheduled health post visits and feeding RUTF to their children; and (3) perception of the service provided by the HEWs. These outcomes were assessed using responses recorded on a four-scale rating (never, few, often and always). The responses were later dichotomized in to "never/few" and "often/always" during analysis. Data were entered into Epi-Info (Version 6.0) [24], cleaned and exported to SPSS for Windows (Version 20.0), [25] for analysis. Percentage was used for categorical variables and differences in proportion were compared using Chi squared test. Mean ± standard deviation (SD) and median (25th; 75th percentile) were used for continuous variables during analysis and for presentation of the findings.

Ethical Considerations
The Ethical Review Board at Addis Continental Institute of Public Health, Addis Ababa, and the regional ethical review board in Uppsala, Sweden approved the research protocol. A letter of support was obtained from the Regional Health Bureau of South Ethiopia and district health bureaus of the four districts. The purpose and procedures of data collection, confidentiality and voluntary Data were entered into Epi-Info (Version 6.0) [24], cleaned and exported to SPSS for Windows (Version 20.0), [25] for analysis. Percentage was used for categorical variables and differences in proportion were compared using Chi squared test. Mean˘standard deviation (SD) and median (25th; 75th percentile) were used for continuous variables during analysis and for presentation of the findings.

Ethical Considerations
The Ethical Review Board at Addis Continental Institute of Public Health, Addis Ababa, and the regional ethical review board in Uppsala, Sweden approved the research protocol. A letter of support was obtained from the Regional Health Bureau of South Ethiopia and district health bureaus of the four districts. The purpose and procedures of data collection, confidentiality and voluntary participation was explained to caregivers of children admitted to OTP and verbal consent was obtained from all caregivers as most of the caregivers of admitted children were not able to read and write. All interviews and anthropometric measurements were conducted in privacy during home visits. In order to protect the study participants' right to confidentiality, the exact location of the study setting is concealed.

Results
Most of the caregivers (88.6%, 903/1019) of children admitted to OTP were biological mothers and on average, were 30.6 years old. The majority of the children were below 24 months of age (62.3%) and the female-to-male ratio was 1.3. Among children included in the study, 7.3% (77/1048) had edema, 71.5% (749/1048) were severely wasted and the remaining 21.2% (222/1048) were not severely wasted. HEWs were on average 24.7 years old and married (69.9%). The majority, (73.4%) were permanent residents of their work area and almost half of the HEWs (51.8%) lived within <30 min walking distance from the health post where they worked.  Figure 2). The HEWs also reported lack of RUTF as the biggest problem (32.9%, 55/167) they face in the OTP and 62.9% (100/159) reported that they never had antibiotics available at their health posts for children admitted to OTP ( Table 1). The proportion of children who achieved program recovery criteria was significantly higher for those children who received antibiotics compared to those treated with RUTFs alone (47.4%, 46/97 vs. 52.6, 51/97; p = 0.002). The adjusted hazard ratio to achieve program recovery criteria of children who took antibiotics was 1.60 (1.60, 95%-CI: 1.13, 1.16).

Length of Admission, Follow-Up of Nutritional Status and Exit from C-OTP
The majority of the children (71.4%, 690/966) had a length of admission to C-OTP within the national limit of maximum stay in C-OTP which is ď8 weeks, while 44.0% (425/966) had a length of admission within the Sphere standard limit of maximum stay which is 6 weeks. Almost all children (98.1%, 934/952) had their weight measured and 83.6% (789/942) had their edema status checked during follow-up visits. However, only 33.2% of the caregivers and their children had a home visit either by an HEW or community volunteer. A total of 554 children who had SAM on admission and for whom anthropometric measurements were done within one week of discharge from the program (Figure 1) were included for analysis of appropriateness of exit from C-OTP. From the 554 children who exited the program, 216/554 (39.0%) achieved program recovery criteria (gain of 15% of admission weight for children without edema and resolution of edema for those with edema on admission) and of the 338 who did not achieve the recovery criteria only 4 were transferred to inpatient care. Thus, only 39.8 (220/553) left the program as per the exit options recommended by the national guideline (discharge if program recovery criteria achieved and referral to inpatient care if not).

Perceptions of RUTF Usage and OTP Service Provision
Most of the caregivers (98.1%, 941/959) and 78.2% (129/165) of the HEWs mentioned that getting the malnourished children eat RUTF was not a problem. However, both the caregivers (42.9%, 394/918) and HEWs (37.1%, 62/167) reported that RUTFs were sold as a commodity and 26.4% (245/941) of the caregivers reported that RUTFs were available in shops in their areas. Moreover, 31.6% (300/949) of caregivers mentioned that avoiding sharing RUTF with other children in their household was a problem, even though they were recommended by the HEWs to feed it to only the children admitted to the program ( Table 2).

Discussion
In the studied C-OTPs both service provision at the health posts and caregivers' management at their homes were constrained. Only 1 in 4 children admitted to the program had adequate and uninterrupted RUTF provision and only 1 in 35 children who failed to achieve program recovery criteria at exit from the program were transferred to inpatient care. Only 1 in 3 had a home visit either by an HEW or community volunteer. However, caregivers perceived that HEWs adequately listened, understood their situation and had sufficient time with them during the scheduled health post visits. Moreover, RUTFs were perceived to be sold and shared to meet other purposes than the program's intention.

Service Provision-Related Constraints
The major constraints in service provision were the inability to provide the recommended amount of RUTF and the inappropriate exit of admitted children from C-OTPs. Less than half of the children (46.6%) on admission and only 49.9% during follow-up received the recommended amount of RUTFs. Health posts should get RUTFs from their respective district health offices monthly [26], however, during data collection we observed a severe shortage of transportation means to get RUTFs to the health posts although the district health office had it in stock. Both caregivers of admitted children and HEWs mentioned that the supply of medications and other program supplies to this area is inadequate and interrupted. The same was reported about the program in Bangladesh, where the supplies are provided through the health system rather than NGOs [27]. Moreover, reviews of various research findings from several community-based health services showed that shortage of resources was the main reason for the failure of the CHWs to fulfil their work responsibilities [28,29].
The second challenge in the implementation of a scaled-up and integrated C-OTP is the exit of severely malnourished children from C-OTP who have neither achieved program recovery criteria nor have been transferred to inpatient care. Failure to achieve program recovery criteria can result from inadequate consumption of RUTF by the admitted children because of inadequate provision both at the health post and at home. Moreover, studies from the same area reported that HEWs were seen discharging children on the basis of the maximum stay (8 weeks) in the C-OTPs [13]. Previous studies also showed that discharge from C-OTP is perceived as temporary by caregivers and they expect their children to be back in C-OTP in 2-3 weeks [13]. Limited availability of inpatient care and a weak referral network between health posts and higher health institutions are barriers to the transfer of children who have failed to achieve the recommended program criteria upon their exit from the program [12].
Thirdly, only 19.3% of admitted children were given antibiotics and the majority (62.9%) of the HEWs reported that they never had antibiotics in their OTPs. The WHO as well as the Ethiopian national guideline recommends the use of broad spectrum antibiotics for all children admitted to C-OTP with a diagnosis of uncomplicated SAM [3,5]. In Ethiopia, HEWs are allowed to provide antibiotics to children admitted to the program [3,26]. However, the C-OTPs we studied suffered a severe shortage of antibiotics. Previously, researchers were skeptical about the use of antibiotics to manage of children with uncomplicated SAM and highlighted the risk of antibiotic resistance that can result from such blanket use [30,31]. In contrast, a recent study by Trahan et al. that had large a sample size showed significant improvement in the children's recovery and a reduction in the mortality rate with the use of antibiotic [32]. The 2013 updated WHO guideline for the management of SAM recommends the use of antibiotics for the management of SAM in C-OTPs [4]. In this study, the proportion of children achieving the program recovery criteria was significantly higher for those who received antibiotics than those who were treated with RUTF alone. Amoxicillin is the most commonly used antibiotic in the management of uncomplicated SAM in Ethiopia [3].

Constraints Related to Home-Based Management
RUTFs that were given for consumption by children admitted to C-OTPs were perceived to be used as a commodity and RUTFs were available in shops in the study area. Avoiding the sharing of RUTF with other children in the household was difficult. Our study was conducted among a population that has been facing household food insecurity and extreme poverty. In such a context available resources, including RUTF and other nutritional and food interventions, are used by the community for collective preservation of their precarious livelihood [13,33,34]. Moreover, the very characteristics of RUTFs that makes it suitable for the program also make it easily bartered for a reasonable amount of money that enables extremely poor households to obtain cash for purchasing basic household commodities and cheap food that can be shared by all family members [13]. Thus, unintended use of RUTF has the potential to undermine the effectiveness of a C-OTP which is already constrained by inadequate and/or interrupted RUTF provision. The SAM management protocol in Ethiopia recommends home visits by HEWs and/or community volunteers to support caregivers and address conditions that delay/hinder the recovery of SAM children [3,5]. In our study, almost two-third of caregivers never had any such home visits during their children's stay in C-OTP. The home visit element was found to be the weakest activity of the C-OTP service [12].
A strength of this study is that data on the elements of service delivery at admission, during stay and exit from C-OTP were collected in a timely manner (i.e., within the first week of admission and while the children were still admitted). The study was completed by an independent team that was not involved in any aspect of the program or any other health and nutrition interventions in the study area. A limitation of this study is that diagnostic criteria were used but were not observed for correctness of techniques used for assessment thus, the proportion of children with proper diagnosis might not reflect appropriateness of the diagnostic criteria assessed.

Conclusions
The implementation of an integrated C-OTP was constrained, both from the program-and home-based management sides. Inadequate provision of RUTF by the service providers and unintended usage of RUTFs by caregivers of admitted children have the potential to delay and/or hinder the nutritional recovery of SAM children and thus lead to chronic nutritional deprivation of the children. Efficient provision and proper use of RUTFs is critical for reducing childhood severe acute malnutrition. Thus, availability of RUTFs should be ensured through application of appropriate supply chain and monitoring policy. In addition, nutrition interventions that address household food insecurity has the potential to reducing sharing and selling of RUTFs, thus improve effectiveness of SAM management at integrated C-OTPs. This might need a policy platform that brings together different stakeholders such as local government, non-government organization and international organizations. The other challenges in integrated C-OTP implementation were lack of antibiotics, lack of transfer of children who fail to achieve program recovery criteria to inpatient care, and inadequate home visits by the HEWS were major challenges. To ensure the success of C-OTPs in saving lives, adequate antibiotics need be made available, providers must be trained regularly, and regular supportive supervision must be in place.