Routine vitamin A supplementation and other high impact interventions in Sierra Leone

Abstract In 2017, transition to routine vitamin A supplementation (VAS) commenced as an integrated reproductive and child health service including vaccinations, Albendazole for deworming, complementary feeding demonstrations, ‘quality’ family planning counselling and provision of modern contraceptives. After 10 months, a lot quality assurance sampling survey evaluated coverage of these interventions. Each of three districts was divided into five supervision areas (lots), and 19 villages were randomly selected in each lot proportional to population size. Households were randomly selected, and a questionnaire was administered to a caregiver of a child 6–11, 12–23 and 24–59 months in each village. Overall, caregivers of 855 children were interviewed, and 19 questionnaires were completed for each age group (6–11, 12–23 and 24–59 months) in each of the five lots in each district. All lots in one district passed the threshold of 80% for VAS and 75% coverage for Albendazole, and two lots failed for either VAS/Albendazole in the other two districts. Overall, weighted VAS coverage for children 6–59 months was 86.9%, and weighted Albendazole coverage for children 12–59 months was 80.9%. Most caregivers (77.2%) knew that complementary feeding should be introduced at 6 months, 44.9% were providing three or more (of six) food groups, 84.9% were aware of family planning and 37.5% were using a modern contraceptive. Integration of reproductive and child health services appears to be a suitable platform for routine VAS and Albendazole whilst improving complementary feeding practices and access to family planning.

access to and utilization of reproductive and child services (Sierra Leone Food assistance fact sheet 2017).
From 2006 to 2017, Sierra Leone held mass campaigns during biannual maternal and child health weeks to achieve universal coverage of high-impact and cost-effective maternal and child survival interventions. These interventions include vitamin A supplementation (VAS), Albendazole for deworming and vaccines such as polio and measles. The mass campaigns achieved effective, equitable coverage of VAS, Albendazole and vaccinations by district, age group, mothers' religion and occupation .
Mass campaigns are expensive, donor driven and highly reliant upon funding for mass vaccination campaigns. They require extensive macro planning at the national and international level for timely ordering, shipment and distribution of commodities and micro planning at all levels: national, district and at every peripheral health unit (PHU). Campaigns do not encourage health-seeking behaviours within vulnerable communities or build the capacity/resilience of routine services to respond to that demand.
Children may also receive services later than indicated on their health cards, because the campaigns only take place every 6 months. For example, an infant who is 5 months of age during a campaign would receive their first VAS at 11 months rather than at 6 months when it is recommended. After 6 months of age, breastmilk needs to be complemented by other foods to meet the child's nutritional needs for rapid growth. Appropriate complementary feeding provides key nutrients (e.g., micronutrients, essential fatty acids, protein and energy). Inadequate complementary feeding can restrict growth and jeopardize child survival and development (©United Nations Children's Fund, 2011).
In 2010, the primary donor funding VAS, the Canadian International Development Agency, requested Helen Keller International (HKI) develop a platform that could achieve effective VAS coverage at 6 months of age in preparation for transitioning from mass to routine VAS (rVAS), as it was anticipated that funding for polio campaigns would decline as polio-elimination goals were met in the region. In 2011-2012, a 6-month contact point (6MCP) was piloted that included rVAS, infant and young child nutrition (IYCN) counselling with mother's participation in the preparation of complementary food made from a preroasted blend of locally available ingredients and confidential family planning (FP) counselling and provision of long-term hormonal implants (Jadelle), which are the most commonly used long acting reversible contraceptives methods as few health workers have been trained in intrauterine contraceptive devices (IUCDs).
An evaluation of the pilot found that at 6-7 months of age, 75% of infants had received rVAS, 96% had been fully vaccinated, 64% of mothers had participated in the preparation of complementary food and 75% had been counselled on FP with 45% .
During the Ebola emergency, health facilities offering the integrated package of services at the 6MCP were able to retain significantly higher health-seeking attendance than those that had not integrated and were able to offer rVAS and vaccination when a mother attended for participatory IYCN and quality confidential FP services .
By mid-2017, the 6MCP had been integrated into 340 (of 1,280) (PHUs nationwide and national overall rVAS coverage for children 6 to 59 months had increased from under 10% to approximately 35% (HMIS reports 2018). In the catchment communities reached by the 6MCP, complementary feeding practices for children 6-23 months of age had improved significantly compared with the previous national surveys: minimal meal frequency (58%), minimal dietary diversity (49%) and consumption of vitamin A-rich foods (17%). In addition, 97% of mothers were aware of FP and 53% were taking modern contraceptives (Koroma et al., 2019). In early 2018, the full transition to rVAS and Albendazole commenced in three districts (Bo, Kenema and Koinadugu), and mass VAS and Albendazole during mother and child health weeks ceased. Services were delivered at the facility level, with outreach sessions organized by PHU staff for remote communities in their catchment area. The district health management teams (DHMTs) and HKI worked with all PHUs to reconfirm the remote communities in their catchment area and developed a schedule of monthly outreach child services for immunization, VAS and deworming with Albendazole. Monthly in-charges meeting held by the DHMTs were regularly attended by the HKI-focal person to report supervisory findings and reinforce best practices. These meetings were held every 2 months in Koinadugu due to the longer distances involved and time required. After 10 months, a lot quality assurance sampling survey (LQAS) was conducted to evaluate the coverage of these interventions and is presented in this manuscript.

| METHODS
A LQAS survey was conducted to ascertain whether performance thresholds of 80% for VAS and 75% for Albendazole had been reached in each lot and to estimate coverage for VAS, Albendazole, complementary feeding and FP practices. The availability of essential

Key messages
• The lot quality assurance sampling survey found the integration of reproductive and child health services at a routine six monthly contact points to be a suitable platform for the transition of mass vitamin A supplementation into routine services.
• Overall, by caregivers' recall, vitamin A supplementation and Albendazole coverage was greater than 80% and 75%, respectively, 77% of caregivers knew that complementary feeding should be introduced at 6 months, 45% recalled having provided three or more (of six) food groups over the last 24 h, 85% were aware of family planning and 38% were using a modern contraceptive. commodities (vitamin A, Albendazole, male condoms, oral contraceptive pills, Depo-Provera and hormonal implants), and the training status of health workers and community health workers (CHWs) was also evaluated. The LQAS survey was conducted from September 27 to October 1, 2018, in Bo and Kenema districts, and from October 8 to 12 in Koinadugu district.
A total of 285 surveys were completed in each district. Nineteen surveys were completed for each age group (6-11, 12-23 and 24-59 months) in each of the five lots in each district. The population size of each lot and district is shown in Figure 1.

| Questionnaire development
Households were randomly selected, and questionnaires were pretested to determine suitability, sequencing of questions and amount of needed time per interview. Both English and Krio versions of the surveys were then programmed into Open Data Kit, uploaded to a web-based platform, Ona.io., then downloaded onto Samsung tablets. Three slightly different questionnaires were developed for mothers/caregivers of infants aged 6-11 months, mothers/caregivers of children aged 12-23 months and mothers/caregivers of children aged 24-59 months as summarized in Table 1 and available in full in Data S1.

| Training
In October 2018, 38 enumerators (undergraduate/postgraduate students) with local language skills and experience in mobile data collection were trained on the LQAS survey, digital data collection and interviewing. One training was held in Bo Town for enumerators who would be administering the survey in Bo and Kenema districts. A second training was held in Kabala, for enumerators who would be administering the survey in Koinadugu district. A posttest was administered, and the 30 enumerators with the best performance on the posttest were retained to conduct the survey (10 per district). Trainings consisted of 1 day in the classroom, 1 day of field practice and 1 day for debriefing. Each pair of interviewers received a tablet to enter data and hard copies of the questionnaires.
F I G U R E 1 Map of districts, chiefdoms and lots Upon arrival at each village, the enumerators introduced themselves to the local head man and explained the purpose of their visit, asking for one 'helper' to map the village and estimate the number of houses.
After the map of each village was drawn, the village was divided into four sectors of similar population size using landmarks (e.g., roads, rivers, schools, mosques and churches). One sector was randomly selected by ballot, and each house was numbered with chalk. If that selected sector had less than 20 households, enumerators randomly selected one household using a random number table and used it as their starting point.

| Identifying respondents
At the selected household, the enumerators introduced themselves, explained the purpose of their visit and requested consent to proceed. If no eligible children resided in the household, the next household on the left was selected. If a household had more than one child in an age group, the names of the children were written down on small pieces of paper, one was selected at random, and the caregiver of that child was interviewed. If children of different age groups resided in the household, one survey was administered for each age group.

| Survey administration
Children's ages were verified with health cards whenever possible.
When health cards were unavailable, children's ages were estimated using a calendar of events. Vitamin A capsules (red and blue) and Albendazole tablets were shown by enumerators to assist caregivers with recall The child health card was also requested and inspected to assess whether VAS or Albendazole had been recorded by the health worker. Interviewers recorded data directly into the tablets using the Ona application.

| Supportive supervision
Supportive supervision was conducted during the first 2 days of the 5-day survey. The training teams, which comprised representatives from the Directorate of Food and Nutrition, MoHS and HKI, supervised the enumerators in their use of the survey protocol, village identification, mapping, segmentation, random sampling and interviewing technique. The Ona account administrator crosschecked the number of interviews performed, the GPS location and the time taken to conduct interviews. Some data fields were mandatory and programmed to alert for errors.

| Health workers questionnaire
Staff from at least five PHUs were interviewed from each lot to evaluate the level of training, stock out of commodities and number of trained CHWs who were actively working for the PHU.

| Debriefing review meetings with the enumerators, HKI and the DHMTs
A day after the survey work, a 1-day debrief meeting was held with the interviewers and members of the DHMTs, the Directorate of Food and Nutrition and HKI to report on the results and discuss challenges met during the transition from mass to rVAS and during the survey.  (Table 3). There were no significant differences in VAS, Albendazole or Pentavalent 3 coverage by the child's sex, mothers' religion, educational status or occupation.
Caregivers of children age 6-11 months were significantly more likely to report that their child received VAS from a PHU compared with caregivers of children age 12-23 and 24-59 months (p < 0.001 and 0.0001 respectively). Caregivers of children 12-23 months and 24-59 months were more likely to report their child received VAS from outreach services (Figure 2). Most caregivers had heard about VAS from health workers/CHWs (40.3%) or from a health talk at the PHU (30.3%) (Figure 3).   Table 4). These were also the most commonly used commodities. In general, Bo district performed better than Kenema district and both did better than Koinadugu district with higher coverages Each PHU has a micro plan to schedule their outreach services and defaulter tracing in their remote communities. The outreaches are frequently scheduled for Fridays and Saturdays when mothers and children are more accessible due to school closures on these days in Muslim communities. Outreach services might be more effective if they were scheduled to coincide with local market days when many mothers who are trading and will be in attendance with their infants.

| Health worker interviews to ascertain training and stock-out status
A higher proportion of coverage was achieved by outreach in older age groups as the older age groups do not attend PHUs as often, having completed their immunization schedules.
Emphasis and funding for outreach services for communities far from the PHU and for DHMTs to provide supportive supervision needs to be maintained if effective coverage is to be sustained. In addition, more funding is required to improve the ratio of PHUs per population and per km 2 .
The DHMTs in Kenema have recommended an additional 10 buildings built/allocated by remote communities of sufficient size and characteristics but are unable to proceed due to lack of funding for staff salaries.
Monthly in-charges meeting held by the DHMTs and regularly attended by the HKI-focal person were identified as an opportunity to reinforce best practices on the 6MCP and help improve data collection through the health management information system (HMIS). The It has previously been demonstrated that the major barriers to diversity of complementary feeding are availability, affordability and accessibility even when caregivers are aware of the advice given by health workers (Turay H et al., 2013). In the post-Ebola context, 50% of households in Sierra Leone was found to be food insecure. Access to modern contraception is a MoHS priority and its uptake is usually associated with female educational status and further contributes to female empowerment (Sierra Leone Demographic Health Survey, 2015). Teenage pregnancies are a major concern in Sierra Leone with 26% of girls giving birth before 18 years of age (Government of Sierra Leone, 2015). The rate of modern contraceptive use in this study (37.5%) was higher than the current national prevalence for all women of reproductive age (27%) and for married women (20%) and higher than the national target (30%) for 2020 (Sierra Leone Commitment Maker, 2019). This is reflected by their uptake and few health workers have been trained to counsel for and insert an IUCD. The next phase of scale up of the integrated package of reproductive and child health services will begin to train various cadres of health workers on the safe provision of IUCD as part of the training on FP.
This study is limited in that it did not include baseline or control from non-6MCP catchment communities. Both this study and the SLNNS, MoHS (2017), were conducted before the 'lean season' when access to stored crops from the previous harvest had been depleted. Comparison regarding dietary diversity will be limited due to different sample sizes, locations and age groups, which may have introduced bias.

| CONCLUSION
The integration of reproductive and child health services appears to be a suitable platform for transitioning from mass to rVAS and Albendazole whilst improving complementary feeding practices and access to FP despite the challenge of frequent commodity stock outs.
District-specific performance appears to be related to their cohort characteristics and the population density served by the PHUs.

ACKNOWLEDGMENTS
The

CONFLICTS OF INTEREST
The authors state that there is no competing interest and the contents are the responsibility of the authors and do not necessarily reflect the views of the Irish Aid, UNICEF, UNFPA or the Canadian Government.
The 6MlyCP is co-funded by Irish Aid and Global Affairs Canada. The