Can ITN distribution policies increase children’s ITN use? A DHS analysis

Long-lasting insecticidal nets (LLINs) are among the most effective tools for preventing malaria. As a result, LLINs have become the primary malaria prevention strategy recommended and championed by the World Health Organization (WHO) global malaria control programme [1]. Early mosquito net distribution efforts focused on pregnant women and children, because of their particular vulnerability to severe malaria morbidity and mortality. More recent strategies have focused on “universal coverage”, with the intention of providing ITNs for all people in areas with a high malaria burden. The WHO recommends household ownership of nets equal to one net for every 2 household members [2]. The Roll Back Malaria (RBM) strategic plan’s 2015 target stated that at least 80% of all members of populations at risk for malaria should be sleeping under an ITN on any given night [3, 4].

Since the mid-2000s, mass campaigns of free insecticide-treated nets (ITNs) have been the most common distribution strategy [5‐7], accounting for 89% of the nets distributed in Africa [8]. The provision of free nets has dramatically increased ownership across Africa in the last decade.

Recognizing that campaigns are not enough to ensure high ITN coverage over time, new continuous distribution strategies have been developed to improve and maintain coverage in concert with campaigns. These strategies often focus on providing ITNs to the most vulnerable groups: pregnant women, infants and children. One such strategy, now recommended by the WHO, is the routine facility-based distribution of ITNs via both antenatal care (ANC) and childhood vaccination (EPI) programmes [9]. These recommendations specifically state that an ITN should be distributed to every woman at her first antenatal care visit in conjunction with other health interventions to assess and support the needs of a pregnant woman. By comparison, the recommendation for ITN distribution through EPI states that every child attending vaccination services should receive an ITN, but it does not specify at which vaccination visit within the first year of life the distribution should take place, leaving it to countries to make a decision for themselves. In most countries, there are at least five vaccination visits between birth and 1 year of age which could be used as a platform for ITN distribution.

Despite these recommendations, many countries have yet to implement routine facility-based ITN distribution, or have only implemented distribution via ANC. Distributing ITNs via ANC has been a significant topic of research, which finds that ANC distribution, in combination with campaigns, increases ITN ownership [10‐12]. ANC has been the focus of routine facility-based distribution with the assumption that newborn babies will share a net with their mother, especially while they are breast feeding. ITNs distributed via ANC should, therefore, cover both the pregnant mother and the newborn child.

Some models developed to assess the benefit of different ITN distribution do not account for the increasing household size that results from a new birth [11]. While ITNs distributed via ANC may successfully replace campaign nets that have worn out between campaigns, new-born babies will increase the size of a household, thus increasing the total number of ITNs needed to cover all the household members. These growing households might benefit from additional ITN distribution through EPI. Research has yet to look at the benefits of ITN distribution through ANC and EPI together, as compared to ANC alone, a distribution strategy which could address some of these issues.

Analyses clearly show that ITN use is a function of ITN ownership: the more ITNs in a household, the more people sleep under an ITN [13]. There are also clear patterns by age across countries, with the highest use being by new-borns and pregnant women, and the lowest use by young adults 11–19 years of age [14]. The decline in net use with age begins in infancy. This analysis focuses on children under five, as one of the groups targeted by both ANC and EPI-based ITN distribution.

This research looks at the net use by children under-five comparing countries with no routine distribution through ANC or EPI, to countries with only ANC distribution, and countries with both ANC and EPI based distribution of ITNs.

Across all countries, the odds of ITN use generally decreased with age up to five (Table 1). A pooled analysis of all countries showed that children zero and 1-year-old had roughly the same level of ITN use, which decreased for 2, 3, and 4-year-olds respectively (Table 1). Gender in under-five year olds did not have any effect on ITN use, with the possible exception of Mali, Guinea and Rwanda. In Mali female children were less likely than their male counterparts to use ITNs: OR = 0.84 (0.74–0.95), while in Guinea and Rwanda female children were slightly more likely to use ITNS: OR = 1.24 (1.07–1.43) and OR = 1.11 (1.01–1.23) respectively (Table 1). However, recognising that this analysis involves multiple comparisons, there is an expectation that at least one will produce a significant difference by chance alone. A heterogeneity Chi squared for gender and ITN use across all countries was χ² = 13.62, p = 0.96, supporting that conclusion and suggesting that there is no evidence of a significant difference between countries.

Net use across countries decreases with age in children under 5 years (Table 1) [15‐17]. The decline in net use early in life is a concern for malaria prevention programmes, as these children are more vulnerable to severe malaria than their adult counterparts. The results from this analysis suggest that countries implementing both ANC and EPI based ITN distribution have higher household ITN ownership, higher household universal access, higher ITN use in all children under-five, and higher ITN use in children under-five in houses with at least one ITN, on average, compared with countries with fewer routine facility-based distribution channels. These data suggest that the combined impact of these two facility-based distribution strategies together far surpasses the benefit of ANC-based distribution alone.

Many studies have found ANC to be a useful and/or cost effective ITN delivery strategy [10, 18‐26]. A few studies define ANC and EPI based distribution as the same thing, referring to them collectively as “facility-based distribution”, without differentiating between them [11, 27]. This is the first study looking at the separate and cumulative effects of ANC and EPI-based routine distribution compared to the singular effect of just one. One important function of ANC-distributed ITNs is “catch-up and keep-up”: the idea that between campaigns, ANC nets maintain ITN coverage as older nets fall out of use [19, 28]. This may not take into account the fact that in households with pregnancies and births, the household size is increasing, so a “keep-up” strategy might replace used nets, but may not be sufficient to provide additional ITNs to cover the growing household size. Countries providing ITNs via both ANC and EPI are providing additional nets for the increased household size resulting from births as well as ensuring there are enough nets for household members in-between campaigns. This may be why there is such a difference seen between the average country with both policies, compared to the average country with only ANC-based distribution.

Most efforts to increase household ITN ownership have focused on mass-distribution campaigns, while routine facility-based distribution has been seen as an appropriate way to “keep-up” [19]. Mass campaigns are an essential tool for increasing ownership of ITN for vector control, but these findings show that routine facility-based distribution can also be used to increase household ITN ownership. Not only does household ownership, and household universal access, improve with ITN distribution through both ANC and EPI (Table 2), but there is also an increase in ITN use by under-fives (Fig. 1). On a population level, the proportion of total children sleeping under ITNs is higher with the addition of an ITN distribution policy through EPI. More interestingly, even in households with only one net, ITN use among under-five children is higher in countries with than without both ANC and EPI distribution (Fig. 1). This may be a result of health facility training and behaviour change messaging that comes with a policy to distribute ITNs via EPI. Even if ITN availability is not consistent through these channels [8], health workers gain training and implement messaging about net use in infancy, which may be enhanced in countries with EPI-based distribution.

The benefits of ANC-based distribution seem to decrease as children age, but the added benefit from EPI-based distribution increases as children age (Table 3). This may be the result of ANC-distributed ITNs wearing out by the time children reach older ages. If ITNs are expected to last for 3 years, on average [29, 30], a child of 3 or 4 years of age is more likely to have a usable ITN if it was given to them within their first year of life, than if it was given to their mother at some time before they were born.

This analysis is not without limitations. Understanding the consistency and extent of routine facility-based distribution programmes through ANC and EPI is challenging. There are limited data available on national programme implementation, and there may be inconsistencies in reporting to WHO [8]. In countries with reported facility-based distribution, research suggests that ITNs may not be available for the majority of women and children attending these services [8]. While pooled analyses and summary findings have been presented, there is diversity in national ITN coverage and use, within each distribution policy category, especially between countries with no routine facility-based distribution. Countries that have implemented policies for ITN distribution through ANC and/or EPI may differ from countries without these policies.

For countries with only an ANC distribution policy, Nigeria serves as a significant outlier. Net ownership in Nigeria is very similar to that in other countries in this category, but net use in children under-five in Nigeria is appreciably lower. For the pooled estimates, presented in Fig. 3 and Table 3, the low net use in the “Only ANC distribution” group is partially the result of the low ITN use in Nigeria. There were no countries with EPI based distribution, but not ANC based distribution, making it impossible to compare the two programmes individually.

This analysis considered country level patterns in net use based on country self-reporting of policies to the WHO. In order to understand the added benefit of ITNs distributed through ANC vs EPI on an individual or household level, it would be useful to examine ITN ownership within countries, and compare households and families that did and did not receive ITNs through these channels. Unfortunately, the DHS does not currently collect information on ITN distribution via ANC and EPI, for analysis, making that type of analysis impossible.

The logistic analyses controlled for household size and maternal education between countries, but data on broader programmatic or implementation strengths were not available as potential confounders within the DHS dataset. More consistent and robust data on these channels are needed to understand fully the impact they have on ITN ownership and use.

Many countries are still lacking any policy or active distribution of LLINs via ANC and EPI programmes despite WHO recommendations for routine distribution through these channels [8]. These findings suggest that there may be a significant benefit to ITN distribution through both channels, beyond the benefit from ANC distribution alone. Both these distribution channels can be implemented across countries to improve ITN ownership and use.

As supplements to mass-distribution campaigns, ITN distribution through ANC and EPI, together, can increase net ownership, universal access, and net use in children under-five. These routine distribution programmes, when implemented together greatly improve net ownership and use, and provide nets to vulnerable children who may not otherwise be covered. A second facility-based distributed ITN, via EPI, beyond the one given at ANC, has the potential to increase the total ITNs in a household, increasing the number and proportion of homes with universal access, and improving ITN use in children under five. Beyond “keeping-up” ITN coverage, the combination of these services can improve coverage, which is an important tool for the control and elimination of malaria.