To examine geographical variations, trends and projections in the prevalence of childhood anemia at national and subpopulation levels.

Repeated cross-sectional Demographic and Health Survey (DHS) conducted during 2000–2018.

Fifty-three low- and middle-income countries (LMIC)

Totally, 776 689 children aged 6–59 months.

During the latest DHS rounds between 2005 and 2018, the prevalence of child anemia was > 20 % in fifty-two out of fifty-three countries and ranged from 15·9 % in Armenia in 2016 to 87·8 % in Burkina Faso in 2010. Out of thirty-six countries with at least two surveys during 2000–2018, the prevalence of child anemia decreased in twenty-two countries, highest in Zimbabwe (–4·2 %) and increased in fourteen countries, highest in Burundi (5·0 %). Based on the trend, eleven and twenty-two out of thirty-six countries are projected to experience, respectively, moderate and severe public health problem according to the WHO criteria (moderate problem: 20–39·9 % and severe problem: ≥ 40 %) due to child anemia in 2030, with the highest prevalence in Liberia (87·5 %, 95 % credible interval 52·0–98·8 %). The prevalence of child anemia varied across the mother’s education and age, child sex, wealth quintiles, and place of residence, with the highest rate of child anemia among the poorest, rural and low-educated mothers. These scenarios are projected to continue. The probability of reducing child anemia at < 0·5 % by 2030 is 0 % for all study countries.

The prevalence of child anemia varied between and within countries. None of the thirty-six LMIC is likely to eradicate child anemia by 2030.

We assessed the role of home visits by Shasthya Shebika (SS) – female volunteer community health workers (CHWs) – in improving the distribution of micronutrient powder (MNP), and explored the independent effects of caregiver–provider interaction on coverage variables.

We used data from three cross-sectional surveys undertaken at baseline (n 1927), midline (n 1924) and endline (n 1540) as part of an evaluation of a home fortification programme. We defined an exposure group as one that had at least one SS visit to the caregiver’s household in the 12 months preceding the survey considering three outcome variables – message (ever heard), contact (ever used) and effective coverage (regular used) of MNP. We performed multiple logistic regressions to explore the determinants of coverage, employed an ‘interaction term’ and calculated an odds ratio (OR) to assess the modifying effect of SS’s home visits on coverage.

Sixty-eight sub-districts from ten districts of Bangladesh.

Children aged 6–59 months and their caregivers.

A home visit from an SS positively impacts message coverage at both midline (ratio of OR 1·70; 95 % CI 1·25, 2·32; P < 0·01) and endline (ratio of OR 3·58; 95 % CI 2·22, 5·78; P < 0·001), and contact coverage both at midline (ratio of OR 1·48; 95 % CI 1·06, 2·07; P = 0·021) and endline (ratio of OR 1·74; 95 % CI 1·23, 2·47; P = 0·002). There was no significant effect of a SS’s home visit on effective coverage.

The households visited by BRAC’s volunteer CHWs have better message and contact coverage among the children aged 6–59 months.

BRAC, an international development organization based in Bangladesh, engages community health workers called Shasthya Shebikas (SS) to implement home fortification of foods with micronutrient powders (MNP). We identified factors associated with home visits by SS, at different levels of the BRAC programme-delivery hierarchy, to implement home-fortification interventions.

We conducted a cross-sectional survey, semi-structured interviews, and collected programme-related data from sub-districts included in the caregiver survey of BRAC’s home-fortification programme and performed multilevel logistic regression modelling to investigate factors associated with home visits by SS.

Sixty-eight sub-districts in Bangladesh.

Caregivers of children aged 6–59 months (n 1408) and BRAC’s SS (n 201).

Households with older children (0·55; 0·42, 0·72; P < 0·001) and located >300 m from the SS’s house (0·67; 0·50, 0·89; P = 0·006) were less likely to have been visited by the SS, whereas those with caregivers who had ≥5 years of schooling (1·53; 1·10, 2·12; P = 0·011) were more likely to have been visited by the SS (adjusted OR; 95 % CI). Households in the catchment area of older SS aged >50 years (0·44; 0·21, 0·90; P = 0·025) were less likely to have been visited by the SS, whereas those with SS who received incentives of >800 BDT (3·00; 1·58, 5·58; P = 0·001) were more likely to have been visited by the SS (adjusted OR; 95 % CI).

The number of home visits is a function of the characteristics of SS, factors that characterize the households they serve and characteristics of their organizational context, particularly to implement home fortification of foods with MNP.