Last update: Nov 2009
Country-level coverage estimates are derived from a variety of sources, which can be grouped into four broad categories.
Tally sheets are used during campaign-style events – which account for approximately three quarters of delivery efforts – to capture the total number of capsules delivered to children. Tallies are summed up and presented as a proportion of the targeted population, most often derived from census projections. Data quality may be affected by inaccuracies in numerators (e.g., children outside of the targeted age range may receive capsules) and denominators (e.g., outdated census data).
Health information systems in most affected countries capture the total number of capsules delivered through routine contacts. These data are generally rolled up to the central level in conjunction with information on vaccines or essential drugs and reported as a proportion of the targeted population. Data quality varies widely by country, and these systems may be inadequate where capsules are delivered through campaigns.
Nationally representative household surveys, such as the UNICEF-supported Multiple Indicator Cluster Surveys or USAID-supported Demographic and Health Surveys (DHS), include questions designed to capture the proportion of children receiving supplements within six months prior to data collection. The reliability of survey-derived estimates may be affected by maternal recall, as timing between dosing and data collection often exceeds several months and mothers may not have been present at dosing or timing of the survey (i.e. often before or during a vitamin A distribution round). Timing of major household surveys becomes important because coverage is reported for each of two semesters in one year (January–June and July–December) while in many countries, major efforts to distribute prophylactic vitamin A supplements in a given semester are confined to a few days or at best, a few weeks only (e.g. a Child Health Day may occur from 10–15 March for semester 1 and 10–15 September for semester 2). As large household surveys cannot logistically be timed to accommodate such events, coverage estimates from this source are often not accurate for any given event or semester. For example a DHS, in Uganda that was enumerated between 28 September 2000 and 3 March 2001 was not consistent with any one semester and therefore may not best represent coverage of vitamin A in that country.
Rapid coverage assessments – generally adhering to Expanded Programme on Immunization (EPI) methodology – are fielded immediately following capsule distribution to specifically capture coverage in a limited number of countries. Rapid coverage assessments reduce the potential for recall problems, as they follow within two months of an event; however, mothers may not have been present at dosing, and there are also problems with sampling methods, which render most data from such sources to be non-representative of the population they are intended to measure.
Information from these sources is reported to UNICEF headquarters via Nutrition Section Reporting Forms, WHO/UNICEF EPI Joint Reporting Forms or survey reports for quality review, entry into the vitamin A monitoring database and publication in The State of the World's Children.
Selection of coverage for vitamin A programme database
The following series of rules was used during the annual review of coverage data received from country offices in order to select coverage figures for entry into the vitamin A programme database and publication in The State of the World's Children.
Only nationally representative coverage data are considered. Coverage figures are therefore excluded for country programmes that are subnational in scope (i.e. geographically targeted). For example, coverage data have been excluded for China, where targeted supplementation efforts regularly reach upwards of 90 per cent coverage in areas of the country with an elevated prevalence of vitamin A deficiency.
Coverage for programmes targeting children other than those aged 6–59 months are reported as targeted (i.e., the proportion of the targeted age range receiving supplements) and marked with a footnote in The State of the World's Children (e.g. Viet Nam supplements children aged 6–36 months in some provinces and 6–59 months in other provinces).
Existence of multiple data sources and reporting pathways may yield more than one coverage estimate for the same distribution. In this case, coverage data points are accepted in the following order: 1) data from tallies and/or health information systems reported directly to UNICEF; 2) data from tallies and/or health information systems reported through the WHO/UNICEF Joint Reporting Form; 3) data from standardized monitoring surveys fielded within six months of the distribution (considered least reliable due to concerns about maternal recall); and 4) data from rapid coverage surveys fielded immediately following a distribution. Where WHO/UNICEF Joint Reporting Form data were more complete than those reported directly to UNICEF, the former were given priority.
Coverage as determined by nationally representative household surveys is only accepted where survey data collection was carried out within six months of a vitamin A distribution.
Where coverage data from both a distinct event and routine delivery were available for the same country and timing of the routine distribution could not be determined relative to distinct events, data from routine delivery were excluded. For example, Malawi achieved 57 per cent coverage through Child Health Day activities in January 2004, to be reported in The State of the World's Children. Data from the country's health information system suggest that 13 per cent of targeted children were reached through routine health system contacts over the course of the year; however, it is not possible to determine the timing of delivery and to know therefore whether this coverage represents a second dose.
In calculating coverage based on tallied data, the denominator of targeted children is compared with official United Nations estimates for quality assurance. Where denominators reported by the country office were inconsistent with United Nations figures: If the country office denominator was =90 per cent of the UN estimate, the country office denominator was used; if the country office denominator was <90 per cent of the UN estimate, the UN denominator was used. Numerators and denominators from the same data source were used to avoid introducing additional error.
Coverage data points exceeding 100 per cent were capped at 100 per cent.
Data must be available from at least 80 per cent of districts: Starting in 2007, each country was asked to indicate the lowest administrative level for which coverage data were available (e.g. district, subdistrict, health post). To be included in UNICEF databases, the national coverage figure required final coverage reports from at least 80 per cent of all districts/health posts in a country. Data points not meeting this criterion are not included in the UNICEF database.
Derivation of global and regional coverage estimates
Global vitamin A supplementation coverage is calculated as the average of country-level coverage, weighted by the country's under-five population. Only countries with at least one accepted data point in the given year are considered. Regional coverage is similarly calculated; however, to ensure that estimates fairly represent the situation in a given region, coverage estimates must be available from countries comprising at least 50 per cent of the region's under-five population in order to be published. For example, an estimate of 2008 regional coverage for the Middle East and North Africa is not reported, as only Djibouti, Egypt and Sudan had acceptable data points in that year, and together, the populations of these three countries accounted for less than 50 per cent of the population of the entire region.
Coverage data would ideally be available from all priority countries such that global coverage estimates could be representative of the full population of at-risk children worldwide. However, almost one third of the 106 priority countries were excluded in 2008 due to non-reporting or data quality concerns. As the countdown advances towards international development goals, further efforts will be needed to ensure accurate and complete reporting on coverage with this critical child survival intervention.
Estimating coverage with two annual doses
Although international recommendations call for vitamin A supplementation every four to six months, current monitoring efforts are unable to capture the proportion of children covered who are receiving both annual doses of vitamin A. Approximations of two-dose coverage presented here assume that in countries providing more than one round of supplementation, the same group of children – those with poor access to health services – is probably missed by both distributions. The proportion of children reached by one campaign but not by the other would be minimal. Therefore, the lower of two coverage data points for a given year is assumed to be roughly equivalent to the proportion of children receiving two doses of vitamin A. UNICEF and its partners are working to develop more refined methods for measuring the proportion of children fully protected in the near future.