To see, hear, and live: 25 years of the vitamin A programme in Nepal

Abstract Nepal has a rich history of vitamin A research and a national, biannual preschool vitamin A supplementation (VAS) programme that has sustained high coverage for 25 years despite many challenges, including conflict. Key elements of programme success have included (a) evidence of a 26–30% reduction in child mortality from two, in‐country randomized trials; (b) strong political and donor support; (c) positioning local female community health volunteers as key operatives; (d) nationwide community mobilization and demand creation for the programme; and (e) gradual expansion of the programme over a period of several years, conducting and integrating delivery research, and monitoring to allow new approaches to be tested and adapted to available resources. The VAS network has served as a platform for delivering other services, including anthelmintic treatment and screening for acute malnutrition. We estimate that VAS has saved over 45,000 young lives over the past 15 years of attained national coverage. Consumption of vitamin A‐ and carotenoid‐rich foods by children and women nationally remains low, indicating that supplementation is still needed. Current challenges and opportunities to improving vitamin A status include lower VAS coverage among younger children (infants 6–11 months of age), finding ways to increase availability and access to dietary vitamin A sources, and ensuring local programme investments given the recent decentralization of the government.

and diarrhoea; anaemia; and hearing loss from severe ear infection (Ross, 1996;Schmitz et al., 2012;Sommer & West, 1996). Underlying these "vitamin A deficiency disorders" are numerous pathways and clinical observations that reveal roles for vitamin A in maintaining epithelial, immune, and haematopoietic function and provide a clear biological basis for both these consequences and their prevention (Palmer, Darnton-Hill, & West, 2017). In the early 1970s, national vitamin A programmes were launched in India and Bangladesh to prevent xerophthalmia and consequent blindness that involved the delivery of high-dose vitamin A capsules (200,000 international units [IUs] and half-dose for infants) twice a year to preschool-aged children (World Health Organization, 1976). Subsequently, the role of periodic vitamin A supplementation in preventing preschool child mortality became evident in the Aceh study, a community randomized controlled trial in 450 villages in northern Sumatra, Indonesia, from 1982Indonesia, from -1984 observed a 34% reduction in all-cause mortality attributable to the vitamin A programme (Sommer et al., 1986).
In the latter half of the 1980s, Nepal became the setting for replicative vitamin A supplementation trials, prompted by reports of high child mortality rates and endemic xeropthalmia in two national surveys in 1981 (Brilliant et al., 1985;Upadhyay, Gurung, Pillai, & Nepal, 1985), which suggested that vitamin A deficiency was a serious public health problem in the country. As the country hosted the 13th meeting of the International Vitamin A Consultative Group in November 1989, two community trials were launched to test whether periodic high-dose vitamin A supplementation could reduce child mortality in Nepal. In the Terai district of Sarlahi, under the auspices of Nepal Netra Jyoti Sangh (National Society for the Prevention of Blindness), a double-masked, placebo-controlled trial was carried out through the Johns Hopkins Nepal Nutrition Intervention Project, Sarlahi, enrolling, dosing, and following 28,630 children 6-72 months of age. Published in 1991, the trial revealed a 30% reduction in mortality among children who received the vitamin A versus placebo supplement (relative risk (RR) 0.70, 95% confidence interval [CI] [0.56, 0.88]; West et al., 1991). Shortly after, a second trial in the Far Western Hill district of Jumla reported a similar 26% reduction in mortality among children living in randomized programme subdistricts who received a large dose of vitamin A versus children in control subdistricts (RR 0.74, 95% CI [0.55, 0.99]; Daulaire et al., 1992). Additional trials of varied design were carried out in India, Sudan, and Ghana.
When combined in a meta-analysis, the findings from a total of eight large trials carried out in Southern Asia and sub-Saharan Africa indicated that vitamin A provision, via supplements or food fortification, could reduce preschool child mortality by, on average, 23% (Beaton et al., 1993). the programme has continued to function, be evaluated, and has been found to save numerous lives of children over the intervening years (Thapa, Choe, & Retherford, 2005). This manuscript, developed as part of a series of papers reflecting on 25 years of Nepal's progress in nutrition, has two main objectives.
The first is to describe the critical role that Nepal has played globally in scientific discovery related to vitamin A and how the country overcame clear challenges to create a programme that would be internationally lauded for its sustained high coverage over time. The second is to address the contemporary opportunities and challenges facing the country as it weighs future options to combat vitamin A deficiency. research, over the history of the programme, including FCHVs. Literature searches were conducted using PubMed and Google Scholar with the keywords "vitamin A" and "Nepal" to identify additional research studies and grey literature describing the programme. We also reviewed the findings of 28 minisurvey reports published from 1993 to 2010 that provided detailed analyses describing the challenges and successes encountered as the programme expanded to cover the entire country (NTAG, 1999).

| METHODS
We generated estimates of (a) the number of lives saved of children aged 6-59 months from 2002-2017 due to the preventive vitamin A supplementation programme from 2002, the year that the programme achieved national coverage, to 2017; and (b) mortality rates among children aged 6-59 months had the vitamin A supplementation programme not been put into place ( Figure S1). To generate these estimates, we relied on modelled mortality data over time from the United Nations Interagency Group for Mortality Estimation (2018) estimates of vitamin A supplementation coverage from Demographic and Health Surveys and national micronutrient surveys and an approximate all-cause mortality efficacy estimate of 30% on the basis of the findings from the efficacy trials in Nepal (Daulaire et al., 1992;West et al., 1991). Specific assumptions are presented in the supporting information.
The effectiveness of vitamin A supplementation in preventing long-term hearing loss associated with episodes of purulent ear discharge during the preschool years used estimates taken from a 16year follow-up study of children who participated in the original vitamin A trial (West et al., 1991) conducted by Schmitz et al. (2012), which found a 42% reduction in children with prospectively documented and validated (Katz et al., 1998) ear discharge in their preschool years in Nepal, with other assumptions provided in the supporting information.
Looking ahead, we estimated the potential survival benefit of introducing a neonatal vitamin A prophylaxis programme in Nepal, assuming an average all-cause mortality reduction of 13% for South Asian settings, as derived from a recent individual participant analysis of trials (Neonatal Vitamin A Supplementation Evidence Group, 2018) and adapted with assumptions about vitamin A supplementation coverage and UNICEF estimates of population size (supporting information).

| RESULTS AND DISCUSSION
In this section, we describe the origins of the National Vitamin A Program in Nepal, its evolution over time (Table 1), and the elements that were critical to its sustained success. We then describe the present-  Trial in Sarlahi (1989-1990 shows that biannual VAS reduces mortality among children aged 6-59 months by 30% 1992 Trial in Jumla (1989) shows that biannual VAS reduces mortality among children aged 6-59 months by 26% Eighth National Plan (1992-1997) includes VAS as a child mortality prevention strategy and 10-year National Programme of Action sets a target of achieving virtual elimination of vitamin A deficiency 1993 Implementation of the Nepal vitamin A programme begins in eight districts with financial and technical assistance from USAID and UNICEF, with a plan to scale up in phases. mountainous Jumla and the population-dense, low-lying plains (Terai) of Sarlahi and informed the development of implementation guidelines and a strategy.

| Development of the Guidelines for Implementation of the National Vitamin A Deficiency Control Program
First produced by the Nutrition Section in the Child Health Division of the MOH in 1992 and subsequently revised in 1996, the Guidelines for Implementation of the National Vitamin A Deficiency Control Program in Nepal (MOH Nepal, 1996) outlined the objectives, protocols for implementation, a monitoring and evaluation plan, and the roles that different organizations were assigned in combatting deficiency.
The guidelines identified 32 priority districts, largely located in the western region of the country and the Terai, as being at particularly high risk of xerophthalmia and to be prioritized during the roll-out of the programme.
The strategy encompassed multiple facets of a national strategy to address vitamin A deficiency (MOH Nepal, 1996) including

Preventive vitamin A supplementation twice per year, once in
March/April (Baisakh) just prior to the time when measles and xeropthalmia were known to peak; and again in October/November (Kartik) in the periharvest season, prior to a minor seasonal peak in xerophthalmia (Sinha & Bang, 1973). A dosage of 100,000 IU was recommended for children 6-11 months of age and 200,000 IU for children 12-59 months of age.
2. Nutrition education and efforts to increase the production and consumption of vitamin A-rich foods.
3. Training of health care workers to treat cases of xeropthalmia, measles, severe malnutrition, and prolonged diarrhoea with vitamin A and supporting efforts to ensure adequate supply of vitamin A capsules at health posts and clinics.

| Providing technical guidance as the programme scaled up
Acknowledging the potential difficulty in bringing such a programme to scale, the decision was made to gradually expand the programme by two to eight districts at 6-month intervals over 9 years. This

| The important role of FCHVs
On the basis of the experiences of the Jumla trial, which had delivered vitamin A supplements to children using a cadre of trained local community workers (both male and female) that had previously been involved in case management of acute respiratory infections (Pandey, Daulaire, Starbuck, Houston, & McPherson, 1991) (Devkota, 2005). Although the conflict disrupted logistics, demand from communities was so strong that a "day of tranquility" was negotiated every 6 months to facilitate the national vitamin A campaign (Houston, Mathema, Adikari, & Pokharel, 2006). Knowing the popularity of the programme in communities, Maoist leaders even attended the vitamin A training in many remote areas.

| Government and donor commitment and overcoming bottlenecks
In the early days of the programme, a consortium of bilateral and mul- 3.5 | The role of monitoring and learning in the success of the programme • Collection of survey data within 1 month after each round of distribution to minimize recall bias and quickly turnaround data to improve the next round of supplementation.
• Collection of process information from FCHVs, which served to identify logistical and other challenges and to provide a feedback loop enabling their voices to be heard.
• Collection of data on the reasons why children were not supplemented and analysis of the factors associated with being missed.
• Collection of data on night blindness.
In the early days of the supplementation programme, these surveys provided rapid data that could be used to assess programme performance. As the programme scaled up, more information was sought from the minisurveys to assess the effectiveness of multiple aspects of community mobilization including use of radio and TV, loud speakers, and leaflets in terms of the proportion of people at the community level that reported hearing about the campaign from each of these different sources.
Over time, the country also showed its willingness to pilot new approaches and adjust policies to improve coverage and protect children from vitamin A deficiency. When Demographic and Health Survey findings showed that infants aged 6-11 months tended to have much lower coverage than older children (MOH, 2002;MoHP, 2007;Thapa, 2010), the decision was made to provide FCHVs with instructions to supplement infants with 100,000 IU as soon as they turned 6 months of age rather than wait for the next round of the 6-monthly campaigns. If children were missed at 6 months of age, they were dosed at 9 months by FCHVs during the routine measles vaccination.
This policy was possible because the FCHVs were based in the communities themselves. In the three areas where this approach was piloted, coverage among children aged 6-8 months increased from 56% in 2012 to 70% in 2013 (Public Health and Infectious Disease Research Center, 2014).
Early programme reviews also raised the concern that children missed by the programme tended to come from poorer households and emphasized the need to improve the equity of the programme (Fiedler, 2000). Such findings were of particular concern given epidemiologic evidence that xeropthalmia tends to cluster in poorer households .

Multivariable analysis of 2006
Demographic and Health Survey data found that children from the poorest quintile were 35% less likely to receive a vitamin A supplement in the previous 6 months compared with those from the richest quintile after adjusting for factors such as education, urban/rural residence, gender, and child age (Thapa, 2010). Although a similar multivariable analysis has not been conducted over the past 9 years, unadjusted analyses of the Demographic and Health Survey from 2006 suggest similar coverage for the wealthiest and poorest quintiles has been maintained over time ( Table 2).
The urban programme began much later in Dharan in 1995. The community awareness observed in rural areas was much more difficult to achieve in urban areas because there were no FCHVs at the start of the urban programme. Instead the vitamin A programme relied on volunteers recruited for the National Polio Programme, who had not received the same training as the FCHV's. Additional challenges faced by the programme in urban areas included (a) perceptions by urban residents that they did not need to receive vitamin A due to their better diet and (b) less intensive promotion campaigns. Despite these challenges, the rural-urban coverage gap has also narrowed over time (Table 2).
3.6 | What has the impact of the programme been?
A number of prior efforts have quantified the effectiveness of the programme in terms of the lives saved and the cost effectiveness of the programme. An analysis of data from the 2001 Demographic and Health Survey estimated that the mortality reduction among children aged 6-59 months was 53% in programmes, a number that surprisingly exceeded the~30% efficacy observed in trials (Thapa et al., 2005). The author of the study attributed such findings to the inclusion of other inventions and services provided by FCHVs alongside the vitamin A supplements. A separate analysis conducted using data from the start-up of the programme in Nepal found that the programme cost ranged from $327 to $397 per death averted under different coverage assumptions, placing it among the most cost-effective health interventions (Fiedler, 2000). An even lower estimate of $11 per death averted was calculated on the basis of experiences in the Jumla trial (Daulaire et al., 1992).
As illustrated in Figure 1, assuming that the underlying risk envi- Additional benefits of the programme may also be present in terms of preventing disabilities such as blindness and hearing loss. A recent finding on the basis of the follow-up of children who participated in the vitamin A supplementation trial in Sarlahi as preschoolers was a 42% reduction in the risk of hearing loss attributable to otitis media in adolescence and young adults (Schmitz et al., 2012). Applying these findings to modern day population estimates suggests that each year the vitamin A supplementation programme may be saving 4,593 cases of hearing loss from middle ear infections (supporting information).

| Changes in vitamin A status over time
The 1998 National Micronutrient Status Survey suggested that nearly 17% of children aged 6-59 months had subclinical vitamin A deficiency, as measured using serum retinol ( (<3% had a modified relative dose response assay > 0.060) observed among women in the 2016 survey are difficult to explain given that this group did not receive supplementation. Evidence from other settings also suggests that serum retinol rises in response to supplementation only transiently (within 2 months following receipt) but does reflect underlying dietary consumption of vitamin A-rich foods (Palmer, West, Dalmiya, & Schultink, 2012). As such, the findings of the most recent survey are somewhat paradoxical, given the evidence outlined below that vitamin A-rich foods are not consumed frequently in Nepal. Additional research may be needed to contextualize these findings in relation to the dietary intake of vitamin A in Nepal.

| Progress made on fortification and dietary diversification
Many country-level and global plans to combat vitamin A deficiency describe three complementary strategies: vitamin A supplementation, fortification, and dietary diversification. From its inception, the National Vitamin A Deficiency Control Program of Nepal outlined two of these three strategies: vitamin A supplementation and the need for behaviour change to "increase dietary vitamin A intake of the target group through nutrition education, increased home production, consumption and preservation of vitamin A rich foods, proper breastfeeding and child feeding practices" (MOH Nepal, 1996).

| Additional scientific discoveries related to vitamin A made in Nepal
Research in Nepal has contributed to other scientific contributions on vitamin A, including findings that maternal supplementation with vitamin A in pregnancy was associated with lower maternal mortality in Nepal and that maternal night blindness in pregnancy increases the risk of mortality and morbidity and is treatable with vitamin A (Christian, West, Khatry, Katz, LeClerq, Pradhan, & Shrestha, 1998a;Christian, West, Khatry, Katz, Shrestha, Pradhan, et al., 1998b;Katz et al., 1995). These findings have drawn attention to the need to address maternal as well as child vitamin A deficiency as a public health problem. • Rapidly growing mobile phone access and opportunities for new channels of behaviour change communication.
• Expanding roads and market access by many but not all.
• Demographic shifts including a rise in single female-headed households and men sending remittances from a distance.
• Growing availability and consumption of highly processed foods.
• Newly available fortification technologies including biofortified foods and rice fortification.
Reflecting on potential policies to pursue going forward, it is worth noting that even in high-income countries, multiple strategies have been needed to alleviate micronutrient deficiencies. As described above, the MSNP to guide action from 2018-2022 contains elements related to each of the three pillars of vitamin A deficiency control (supplementation, fortification, and dietary diversification). In this section, we examine the current status and future opportunities related to these deficiency control strategies, and implications given current realities, with a view towards the next 25 years.

| Dietary diversification
The MSNP-II sets concrete targets for the increased production of vitamin

| Supplementation programmes
The national vitamin A supplementation programme has maintained impressively high coverage over time. This coverage has been sustained in recent years despite less investment in training of FCHVs, promotional activities, minisurveys, or other aspects of the programme. As a result of decentralization, the programme faces both new opportunities and challenges. Given the community-based delivery platform for vitamin A supplementation, implementation of the programme may not be adversely affected by decentralization, provided supplements are delivered to the field on time. One challenge in the short term is that the responsibility for supply chain management, including the procurement of vitamin A supplements, has been transferred from the central government to the province level. There are likely to be capacity challenges at the provincial level, at least in the short term, to procuring vitamin A supplements. It may make more sense to readjust responsibilities so that the provinces procure vitamin A capsules from the central government or UNICEF using government funds until they are able to independently manage this process.

| CONCLUSIONS
The National Vitamin A Program of Nepal is viewed by many in Nepal as the most successful public health intervention, likely contributing substantially to the acceleration of child mortality reductions and the achievement of the mortality MDG. What facilitated the success of the National Vitamin A Program? Clearly there were multiple elements that contributed, but in the end, it boils down to the motivation, trust, and drive of people at all levels of the country, from families and communities to the very top, to achieve the result of getting a capsule in the mouths of every eligible child twice a year. Certainly, it would not have been possible without government commitment and leadership as well as the external backstopping and technical support provided by the international community and donors over two decades. Also key was the fact that programmers were able to create a demand for the capsule that compelled those responsible for managing the vitamin A capsule supplies to ensure that sufficient capsules were delivered to all distribution sites on time. Integral to creating this demand was the placement of FCHVs at the heart of the programme and the successful advocacy and community mobilization that turned vitamin A supplementation days into a platform for the delivery of a package of health and nutrition interventions. Going forward, as the country expands fortification efforts and dietary diversification, it will be important to ensure that sufficient dietary intake of vitamin A exists prior to any scaling down of the programme and that careful monitoring is done during such a transition to ensure that rises in the signs and symptoms of vitamin A deficiency (which may be a sentinel indicator of increased mortality) do not occur.

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.