Multiple Indicator Cluster Survey Data Research Articles

Gender and child feeding practices in Belize- a trend analysis

The study aimed to determine the trend and gender differences in child feeding practices in Belize based on consecutive multiple indicator cluster surveys (MICS) using a retrospective quantitative study. The secondary analysis of the last consecutive MICS data from 2006, 2011 and 2015-16 was done using SPSS software version 23. Association between gender and feeding practices was determined by bivariate analysis using chi-square test and p value < 0.05 was considered statistically significant. There was increased gender differentials in optimal feeding practices in children over the study period. The inequalities between male and female increased over the study period for most of the feeding practices except for exclusive breastfeeding where the gender inequality diminished. The exclusive breastfeeding rate among female children was higher than among male children in the three surveys but significant only in 2006 and 2011 surveys, p<0.05. More female children were timely initiated on complementary feeding compared to male children in the three surveys and was found to be significant in 2006 and 2015 surveys, p<0.05. Minimal meal frequency and minimal dietary diversity was higher among female than among male but there was no significant difference(p>0.05). The study shows evidence of gender differential in feeding practices. There is need to improve gender sensitization in terms of feeding practices within the households; develop and implement culturally appropriate gender sensitive guidance and approaches on optimum infant and young child feeding practices in the country.

Are children with disabilities more likely to be malnourished than children without disabilities? Evidence from the Multiple Indicator Cluster Surveys in 30 countries

IntroductionA key Sustainable Development Goal target is to eliminate all forms of malnutrition. Existing evidence suggests children with disabilities are at greater risks of malnutrition, exclusion from nutrition programmes and...

Empowering Women through Education: A Study on Reducing Violence against Women in Lahore

This study using MICS data with 2411 observations explores factors influencing violence against women (VAW) such as women's education (WEDU), safety perceptions (WS), polygamous marriage (PM), and age (WA) to inform targeted interventions. Utilizing data from 2,411 observations in the Multiple Indicator Cluster Survey (MICS), a logistic regression model is employed to assess the effects of these variables on VAW. The key variables examined include women education (WEDU), women safety (WS), polygamous marriage (PM), and women's age (WA). The analysis reveals that women’s education (WEDU) and women safety (WS) significantly reduce the likelihood of experiencing violence, with coefficients of 0.516 and 0.441, respectively, both statistically significant at the 1% level. Polygamous marriage (PM) is associated with an increased risk of violence (coef. 0.939, p < 0.05). Women’s age (WA) has a negative effect on VAW, although it is marginally significant (coef. -0.114, p = 0.051). This research provides a nuanced understanding of the factors influencing violence against women in Lahore, highlighting the importance of education and safety measures in mitigating violence. The cross-sectional nature of the MICS data limits the ability to infer causality. Future research should employ longitudinal data to better capture the dynamics over time. The findings suggest that enhancing women’s education and safety measures can significantly reduce violence against women. Policymakers should prioritize these areas to create safer and more empowering environments for women.

Regional Disparities in Safe and Clean Environments in Uzbekistan: Analysis of 2021–2022 Uzbekistan Multiple Indicator Cluster Survey Data

Background: Water, sanitation, and hygiene (WASH) services and clean fuels are representative factors of a clean and safe residential environment. Uzbekistan faces environmental issues and ranks low among countries on the Environmental Performance Index. This study aimed to identify regional disparities and wealth inequalities in WASH services and clean fuels in Uzbekistan. Methods: We employed raw data from the 2021–2022 Uzbekistan Multiple Indicator Cluster Survey (MICS) and the WASH and clean fuels coverage were analyzed. For each of the variables with the lowest coverage within WASH services and clean fuels variables, we evaluated the disparities between urban and rural areas and calculated the concentration index (CI). Results: Among WASH services and clean fuels, basic handwashing and clean fuels for space heating demonstrated the lowest coverage. In most regions, urban areas had higher coverage of basic handwashing and clean fuels for space heating compared to rural areas. The CI of clean fuels for space heating was 0.2141 or higher in five areas. The CI was notably high in areas with low coverage of WASH services and clean fuel for space heating. Conclusions: Basic handwashing and clean fuels for space heating showed varied regional coverage patterns and wealth-related inequalities. The results of this study can provide evidence for policy formulation, particularly in addressing disparities.

Individual- and community-level factors associated with the presence of adequate iodized salt in households in Bangladesh: a multilevel modelling approach.

The aim of this study is to estimate the factors at both the individual and community levels related to the adequacy of iodized salt in households in Bangladesh. For this study we utilized the 2019 Multiple Indicator Cluster Survey data. A total of 61242 households were chosen as samples from Bangladesh. In our study of socio-economic disparities, we applied a concentration indexing method. To identify the factors associated with the adequacy of iodine in salt at both the individual and community levels, we employed multilevel logistic regression. Aside from the multilevel regression used in the study, we also applied spatial analysis. The results indicated that the prevalence of iodine adequacy in household salt was found to be 57.8% (95% confidence interval 57.4 to 58.2). Rural areas have a higher concentration of iodine than urban areas. According to the multilevel model, younger women (adjusted odds ratio [aOR] 0.70), Muslim women (aOR 0.89), illiterate women (aOR 0.80) and those from poor households (aOR 0.33) were found to be less likely to consume iodine in concentrated salt compared with their counterparts. Disabled women and those with low media exposure have a lower likelihood of iodine adequacy in salt compared to their reference group. Furthermore, households in urban areas exhibited higher odds of having iodine adequacy in salt compared with households in rural areas. Barisal, Chattogram, Dhaka, Khulna, Mymensingh, Rajshahi and Rangpur Divisions have lower iodine adequacy in salt compared with Sylhet Division. The findings reveal that religion, physical disability and exposure to media exert an equal influence on the presence of iodized salt intake. Moreover, women's age, wealth status, education level and the educational background of the household head positively contribute to the adequacy of iodine in household salt. In light of these results, policymakers are advised to prioritize efforts aimed at enhancing iodine concentration, with a particular focus on mass media advertising, especially in rural areas (excluding Sylhet Division).

Height-for-age and weight-for-age growth charts for Pakistani infants under six months: derived from a novel case selection method using multiple indicator cluster survey data

BackgroundIn the past two decades, there has been a growing recognition of the need to establish indigenous standards or reference growth charts, particularly following the WHO multicenter growth study in 2006. The availability of accurate and reliable growth charts is crucial for monitoring child health. The choice of an appropriate model for constructing growth charts depends on various data characteristics, including the distribution’s tails and peak. While Pakistan has reported some reference growth charts, there is a notable absence of indigenous charts for children under two years of age, especially for infants aged 0–6 months who are exclusively breastfed. Additionally, acquiring data poses a significant challenge, particularly for low-income countries, as it demands substantial resources such as finances, time, and expertise. The Multiple Indicator Cluster Survey (MICS) constitutes a large-scale national survey conducted periodically in low-income countries under the auspices of UNICEF. In this study, we propose methods for generating selection variables utilizing the “Novel Case Selection Method,“ as previously published. Further our approach enables to select and fit appropriate model to the MICS data, selected, and to develop the standard growth charts.MethodsOut of the 11,478 children under 6 months of age included in MICS-6 (Pakistan), 3,655 children (1,831 males and 1,824 females) met the specified criteria and were selected using the “Novel Case Selection Method”. The sample was distributed across provinces as follows: 841 (23.0%) from KPK, 1,464 (40.1%) from Punjab, 819 (22.4%) from Sindh, and 531 (14.5%) from Balochistan. This sample encompassed both rural (76.4%) and urban (23.6%) populations. Following data cleaning and outlier removal, a total of 3,540 records for weight (1,768 males and 1,772 females) and 3,515 records for height (1,759 males and 1,756 females) were ultimately available for the development of standard charts. The Bayesian Information Criterion (BIC) was employed to determine the optimal degrees of freedom for L, M, and S using RefCurv_0.4.2. Three families within the gamlss class—namely, Box Cox Cole and Green (BCCG), Box Cox T (BCT), and Box Cox Power Exponential (BCPE)—were applied, each with three smoothing techniques: penalized splines (ps), cubic splines (cs), and polynomial splines (poly). The best-fitted model was selected from these nine combinations based on the Akaike Information Criteria.ResultsThe Novel Case Selection Method yielded 3655 cases as per criteria. After cleaning the data, this method lead to selection of 3540 children for “weight for age” (W/A) and 3515 children for “height for age” (H/A). The “BCPE” family and “ps” as smoothing method proved to be best on AIC for all four curves, i.e. the W/A male, W/A female, H/A male, and H/A female. The optimum selected degrees of freedom for the curve “W/A”, for both genders were (M = 1, L = 0, S = 0). The optimum degrees of freedom for H/A male were again (M = 1, L = 0, S = 0), but for females the selected degrees of freedom were (M = 1, L = 1, S = 1). The indigenous fitted standard curves for Pakistan were on lower trajectory in comparison to WHO standards.ConclusionThis study uses the Novel Case Selection Method with introduced algorithms to construct tailored growth charts for lower and middle-income countries. Leveraging extensive MICS data, the methodology ensures representative national samples. The resulting charts hold practical value and await validation from established data sources, offering valuable tools for policy makers and clinicians in diverse global contexts.

Correlations among the maternal healthcare services in Bangladesh: An application of joint modelling technique

BackgroundCaesarean section (C-section) in Bangladesh have received great attention as the number has been amplified during the last two decades. The question arises whether this rise has a correlation with other maternal healthcare services and/or has been influenced by their predictors. ObjectiveThe main objectives of this study are to assess correlations among the maternal healthcare indicators—antenatal care use, childbirth in private facilities, and childbirth through C-section—and identify their associated predictors in Bangladesh through the development of an appropriate cluster-adjusted joint model that accounts for inter-correlation among the indicators in the same cluster. DesignThe 2019 Bangladesh Multiple Indicator Cluster Survey data have been utilized in this study. Separate generalized linear mixed models developed for the three outcome variables are combined into a joint model by letting cluster-specific random effects be in association. FindingsThe joint model shows that the number of antenatal cares is fairly positively correlated with delivery in private facilities and C-section, while the latter two are strongly positively correlated. Household socio-economic condition, women and their partners' education, women's exposure to mass media, place of residence, religion, and regional settings have significant influence on the joint likelihood of receiving antenatal care, choosing a private health facility for birth, and opting for C-section birth. Key conclusions and implicationsThe rising rate of C-section delivery over time is alarming for Bangladesh to achieve the World Health Organization target of 10–15 %. The joint model reveals that the rising rate of C-sections may be correlated with the choice of a private health facility as the delivery place. The study findings also suggest that maternal childbirth care is private-dominant and predominantly utilized by urban women with better education and higher socio-economic status. The policy should focus on strengthening the public health sector while also keeping importance in increasing coverage of maternal care services among the less well-off.

Sustained and Universal Fertility Recuperation in Kazakhstan

The fertility rates of Kazakhstan have reversed to levels not seen for several decades. The striking fertility increase poses questions regarding the extent to which this new development is shared across socio-demographic groups and the nature of fertility recuperation. The current study employs UNICEF Multiple Indicator Cluster Survey data and event-history modelling to analyse parity progressions to one, two, three, and four children. The results suggest a sustained fertility increase that is not merely associated with the recuperation of delayed first births, but a genuine increase across all birth orders. This pattern is evident for both main ethnicities in Kazakhstan and across educational groups. The gradual increase of higher-order births, especially among ethnic Kazakhs, indicates a reversed fertility transition and also that the previous fertility decline in the 1990s was not part of a general transition towards below-replacement fertility but rather a reflection of economic crisis after the collapse of the Soviet Union.

The use of mass media by mothers and its association with their children’s early development: comparison between urban and rural areas

BackgroundMass media can play critical roles in influencing parents’ attitudes and practice toward the healthy upbringing of children. ObjectiveThis study examined the association between the use of five types of mass media among mothers living in rural and urban areas and the early childhood development (ECD) of their children.MethodsWe analysed nationally representative and internationally standardized Multiple Indicator Cluster Survey data collected in 2013 and 2019 in Bangladesh. The ECD was calculated using four domains of development: physical health, literacy-numeracy, learning and social-emotional. Mothers’ use of newspapers/magazines, radio, television, internet and mobile phones was the study factor. We used Poisson regression with robust variance. The dataset included 27,091 children aged three or four years.ResultsAlmost 21% of the children were living in urban and 78% in rural areas. Mothers/caretakers of 30% of the children used none, 39% used one, 25% used two, and approximately 6% used three or more of the five types of media. Mobile phones and television were the dominant types of media, both in terms of the number of users and the frequency of use. Overall, 68.87% of the children were on track in terms of their ECD and 31.13% were not. A significantly larger proportion of urban children (74.23%) than rural children (67.47%) were on track in their ECD. The prevalence of children being on track of ECD increases by 4% (aPR 1.04; 95%CI: 1.01–1.06) for each additional media use among women who lived in urban areas and increases by 7% if women live in rural areas. In terms of the individual formats of media, using newspapers, television and internet was found to be significantly associated with the children in rural areas being on track in terms of their ECD. In the urban sample, only radio use was found to be significant.ConclusionsTargeted and well-designed child development campaigns that are delivered through popular media types are likely to help mothers to take better care of their children.

DETERMINANTS OF CONTRACEPTIVES USE AMONG MARRIED WOMEN IN THE AGE GROUP 15-49

This article presents an analytical reading of the survey results with multiple indicators on clusters 2019: the case of Algeria.The use of contraceptives among married women is closely associated with a range of factors, whether demographic, social or economic, that affect its usage decision. Therefore, this study endeavours to identify the determinants of contraceptives use among married women in the age group (15-49) based on Multiple Indicator Cluster Survey data 2019, the case of Algeria in order to reveal the magnitude of the impact of socio-demographic and economic determinants on the use of contraceptives as well as to explain that based on the particularity of society. Furthermore, the descriptive analytical method was adopted as well as the technique of analysing ready-made statistical data. The results demonstrated the existence of a different effect between demographic, social, economic determinants and the use of contraceptives among married women aged (15-49 years) in Algeria.

Novel Methodology for the Selection of Cases from MICS data for Growth Charts Development Based on WHO Criterion

Background: The growth charts are very important for monitoring a child growth and for estimation of stunting and wasting in a country. This can be done efficiently if the references and standards for growth monitoring are correct. Aim: To devise the methodology of selection of cases from multiple indicator cluster survey (MICS) data available worldwide under UNICEF for most of the developing and non-developed countries. Methods: This methodology was devised during 1st January 30th June 2022 in College of Statistical Sciences, University of the Punjab, Lahore. The three rounds of MICS, i.e. the 4th, 5th and 6th, were evaluated for their eligibility to develop a standard growth chart for children under two year’s age. The criteria needed to be met were, singleton, full term birth up to 4thorder, exclusive breastfeeding for 4-6 months, continuing breastfeeding till one year age. Belonging to affluent socioeconomic status, not having any severe illness and born to non-smoking mothers. The MICS data is available in the format of SPSS data files. These files have been explored for the variables answering these questions directly or indirectly. Results: The MICS rounds 4 and 5 were discarded after having a detailed view, as they were not fulfilling the WHO criteria due to missing information for several important variables. The round 6 was found to have direct information on many of the variables, i.e., singleton, full-term, birth order, family size, mother smoking status while some other variables i.e. exclusive breast feeding, severe illness, and not belonging to deprived class were objectively created using some other variables existing in MICS data Conclusion: This methodology for selection of cases, can provide us children with height, weight, age and gender of children as per WHO criteria required to meet for development of the standard growth chart for children under two years age. Keywords: WHO criteria, Pediatric growth standards, MICS

Factors affecting contraceptive use among currently married women in Iraq in 2018.

Level of contraceptive use is one of the primary proximate determinants of fertility level. Among Arab countries, Iraq experiences high fertility rates, especially among adolescents. According to the 2018 Iraq Multiple Indicator Cluster Survey (IMICS 2018), 52.8% of currently married Iraqi women used a contraceptive method (36.1% used modern methods, and 16.7% used traditional methods). A multiple multinomial logistic regression model was adopted to investigate the demographic, socioeconomic, fertility, and cultural determinants of contraceptive use among currently married, nonpregnant Iraqi women based on the 2018 Iraq Multiple Indicator Cluster Survey data. The main findings indicated that women using modern contraceptive methods had more children, did not experience the death of a child, and had no desire for more children. Moreover, they were married before age 21, had at least an upper secondary education, had access to mass media, used the internet, belonged to a low or middle wealth category, and lived in urban Kurdistan. Kurdish women prefer to use traditional methods. The results suggest that family planning programmes should target women who live in southern/central areas of Iraq, as they were the least likely to use contraceptives, and motivate Kurdish women to use modern contraceptives rather than traditional contraceptives.

Trends of HIV/AIDS knowledge and attitudes among Nigerian women between 2007 and 2017 using Multiple Indicator Cluster Survey data

BackgroundGlobally, Nigeria ranks third among the countries with the highest number of People Living with HIV (PLHIV). Given that HIV/AIDS knowledge is a key factor that determines the risk of transmission and certain attitudes towards PLHIV, there is a need to understand the trend of HIV knowledge within the population for the purpose of assessing the progress and outcome of HIV prevention strategies. The aim of the study was to understand the trends of HIV/AIDS knowledge and attitude towards PLHIV between 2007 to 2017 among Nigerian women, and to investigate change in the factors associated with HIV/AIDS knowledge and attitude towards PLHIV over years.MethodsData were derived from three Nigerian Multiple Indicator Cluster Surveys (2007, 2011 and 2016–2017) among women aged 15–49 years old from each geo-political zone (South South, South East, South West, North East, North West, North Central) in Nigeria. Participants who did not answer questions related to HIV/AIDS knowledge and attitude were excluded from the study. The final sample sizes were 17,733 for 2007, 26,532 for 2011 and 23,530 for 2017. In descriptive statistics, frequencies represented the study sample, while percentages represented weighted estimates for the population parameters. Rao-Scott chi-square test for complex survey design studies was used to assess bivariable associations. Factors associated with outcome variables were examined using the survey-weighted multivariable logistic regression models for the complex survey design while controlling for potential confounding variables.ResultsThere was a relatively high level of HIV/AIDS knowledge level in 2007 and 2016–2017 surveys (64.6 and 64.1%, respectively), however a decrease in HIV/AIDS knowledge trend was observed in 2011 (45.6%). The positive attitude towards PLHIV progressively increased across the years (from 40.5 to 47.0% to 53.5%). Multivariable analysis revealed that women who had a higher educational level, higher wealth index, and lived in urban areas had higher odds for HIV/AIDS knowledge and positive attitude towards PLHIV across the years. In addition, the Northern zones had predominantly higher knowledge and attitude levels.ConclusionsOur study found increasing tendency for high HIV/AIDS knowledge and positive attitude towards PLHIV over the years. Women’s age, wealth index, education level and residence were consistently associated with knowledge and attitude over the years. There is a need for more pragmatic HIV/AIDS-related knowledge action plan to target to cover all age groups, all geo-political zones while paying close attention to the rural areas and the less educated women. In addition, more replicative studies of HIV/AIDS knowledge and attitude trends is crucial in monitoring of the progress of HIV interventions in the country in the coming years.

Topological Data Analysis and UNICEF Multiple Indicator Cluster Surveys

Multiple Indicator Cluster Surveys (MICS), supported by UNICEF, are one of the most important global household survey programs that provide data on health and education of women and children. We analyze the Serbia 2014–2015 MICS dataset using topological data analysis which treats the data cloud as a topological space and extracts information about its intrinsic geometric properties. In particular, our analysis uses the Mapper algorithm, a dimension-reduction and clustering method which produces a graph from the data cloud. The resulting Mapper graph provides insight into various relationships between household wealth—as expressed by the wealth index, an important indicator extracted from the MICS data—and other parameters such as urban/rural setting, ownership of items, and prioritization of possessions. Among other uses, these findings can serve to inform policy by providing a hierarchy of essential amenities. They can also potentially be used to refine the wealth index or deepen our understanding of what it captures.

Socio-economic Correlates of Marital Status and Marriage Timing Among Adult Men in Nigeria

There is a gradual tendency towards prolonged bachelorhood among men in Nigeria. Studies have linked this to socio-economic factors, but this evidence is sparsely explored in the context of Nigeria. Hence, this study fills the knowledge gap. The 2016/17 Nigeria Multiple Indicator Cluster Survey data of 7803 adult men (aged 18–34 years) was analysed by using descriptive and fitting binary logitic regression and Cox regression models. Results show that slightly more than one-third of adult men in Nigeria (35%) had a marriage history and their median age at first marriage was about 24 years. Educated men (than the uneducated) and those in middle wealth group (than the poor men) were less likely to have ever been married and to delay marriage, respectively. Wealthy men were more likely to delay marriage. Employed men were more likely to have a marriage history and to delay marriage.

Child Development in Low- and Middle-Income Countries.

The United Nations (UN) created the Multiple Indicator Cluster Surveys (MICS) to monitor progress toward achieving goals of the World Declaration on the Survival, Protection, and Development of Children and its plan of action. The MICS is nationally representative and internationally comparable. In this study, we use MICS data from 51 low- and middle-income countries on 159 959 children between 36 and 59 months of age. To index national development, we used the 2013 UN Human Development Index (HDI), which provides data on country-level life expectancy, education, and income. To index child development, we used the Early Childhood Development Index (ECDI), which assesses literacy and numeracy, socioemotional development, physical health, and approaches to learning. Children's literacy and numeracy, socioemotional development, and approaches to learning all increase linearly as national development on the HDI (especially education) increases. Overall, the HDI revealed a positive association (r = 0.40) with the ECDI: the HDI explained 16% of variance in children's ECDI scores and was the most influential predictor of ECDI scores examined. HDI-ECDI relations are robust, even when we control for multiple demographic aspects of children (age, sex), mothers (age, education), and households (size variables) as covariates. No family demographic variable was a stronger predictor of child development than national development. To promote child development, low- and middle-income countries need to develop and implement policies that ensure national health and wealth and, particularly, the educational achievements of children's caregivers. These findings are faithful to the World Summit for Children and inform the UN Sustainable Development Goals, which drive the international development agenda through 2030.

Socio-economic inequalities in children's nutritional status in Democratic Republic of the Congo in 2017-2018: an analysis of data from a nationally representative survey.

The Democratic Republic of the Congo (DRC) has one of the highest levels of child undernutrition globally; however, little information exists on the underlying socio-economic inequalities resulting in undernutrition. This study aims to examine the differences in the nutritional statuses of children across different wealth quintiles and explores the association between malnutrition in children and related factors. We utilised the 2018 Multiple Indicator Cluster Survey data. We estimated the prevalence of malnutrition across all twenty-six provinces. The study used the WHO 2006 child growth standards to measure stunting, underweight and wasting. We employed a mixed-effect linear model to analyse the association between nutritional status and healthcare accessibility, domestic sanitation, and socio-demographic factors. Twenty-six provinces in the DRC. 21 477 children under 5 years of age and 21 828 women of childbearing age in the DRC. The national prevalence of underweight, stunting and wasting was found to be 23·33 %, 42·05 % and 5·66 %, respectively. Household wealth and mother's education level were significantly positively associated with the nutritional statuses of children. Among households in the lowest wealth quintile, residence in urban areas was a protective factor against undernutrition. The findings of this study indicate considerable socio-economic inequalities in the nutritional statuses of children under 5 years of age in the DRC, highlighting the need for nutrition promotion as part of maternal and child healthcare. Interventions and policies should include improving nutrition education for less-educated mothers, in particular, in the central provinces of the DRC.

Regional variations of child development index in Bangladesh

Early development is a vital phase in childhood life. The study aimed to identify factors that were associated with the early development of 36–59 months children in Bangladesh. The findings of this study will formulate the design of appropriate policy and programmed responses.Utilizing Multiple Indicator Cluster Survey data, influencing components of child development status were evaluated for both rural and urban areas of Bangladesh. A total of 23,099 children under the age of five were included in this analysis. Chi-square analysis was conducted to assess the association between outcome variables and selected covariates. At the same time, this study uses two separate multivariate binary logistics regression models (respectively for urban areas and rural areas) to determine the risk factors that are primarily related to child development.Our research estimates that more than 70 percent of children develop early throughout the country. The multivariate analysis on the determinants of child development index among children aged between 36 and 59 months old regarding residence discovered a significant impact on child age and sex, maternal education, child education, wealth status, reading children's books. The adjusted odds of child nutrition status, playthings, and maternal functional difficulties have had a major impact on early child development in rural Bangladesh.Based on the findings, educational status, nutritional status, wealth-status, and some determinants of children care the most noteworthy findings in this study. Hence, policymakers should emphasize on such factors for improving children's development in residence.

Using DHS and MICS data to complement or replace NGO baseline health data: an exploratory study

Background: Non-government organizations (NGOs) spend substantial time and resources collecting baseline data in order to plan and implement health interventions with marginalized populations. Typically interviews with households, often mothers, take over an hour, placing a burden on the respondents. Meanwhile, estimates of numerous health and social indicators in many countries already exist in publicly available datasets, such as the Demographic and Health Surveys (DHS) and the Multiple Indicator Cluster Surveys (MICS), and it is worth considering whether these could serve as estimates of baseline conditions. The objective of this study was to compare indicator estimates from non-governmental organizations (NGO) health projects’ baseline reports with estimates calculated using the Demographic and Health Surveys (DHS) or the Multiple Indicator Cluster Surveys (MICS), matching for location, year, and season of data collection. Methods: We extracted estimates of 129 indicators from 46 NGO baseline reports, 25 DHS datasets and three MICS datasets, generating 1,996 pairs of matched DHS/MICS and NGO indicators. We subtracted NGO from DHS/MICS estimates to yield difference and absolute difference, exploring differences by indicator. We partitioned variance of the differences by geographical level, year, and season using ANOVA. Results: Differences between NGO and DHS/MICS estimates were large for many indicators but 33% fell within 5% of one another. Differences were smaller for indicators with prevalence <15% or >85%. Difference between estimates increased with increasing year and geographical level differences. However, <1% of the variance of the differences was explained by year, geographical level, and season. Conclusions: There are situations where publicly available data could complement NGO baseline survey data, most importantly when the NGO has tolerance for estimates of low or unknown accuracy.

Using DHS and MICS data to complement or replace NGO baseline health data: an exploratory study.

Background: Non-government organizations (NGOs) spend substantial time and resources collecting baseline data in order to plan and implement health interventions with marginalized populations. Typically interviews with households, often mothers, take over an hour, placing a burden on the respondents. Meanwhile, estimates of numerous health and social indicators in many countries already exist in publicly available datasets, such as the Demographic and Health Surveys (DHS) and the Multiple Indicator Cluster Surveys (MICS), and it is worth considering whether these could serve as estimates of baseline conditions. The objective of this study was to compare indicator estimates from non-governmental organizations (NGO) health projects' baseline reports with estimates calculated using the Demographic and Health Surveys (DHS) or the Multiple Indicator Cluster Surveys (MICS), matching for location, year, and season of data collection. Methods: We extracted estimates of 129 indicators from 46 NGO baseline reports, 25 DHS datasets and three MICS datasets, generating 1,996 pairs of matched DHS/MICS and NGO indicators. We subtracted NGO from DHS/MICS estimates to yield difference and absolute difference, exploring differences by indicator. We partitioned variance of the differences by geographical level, year, and season using ANOVA. Results: Differences between NGO and DHS/MICS estimates were large for many indicators but 33% fell within 5% of one another. Differences were smaller for indicators with prevalence <15% or >85%. Difference between estimates increased with increasing year and geographical level differences. However, <1% of the variance of the differences was explained by year, geographical level, and season. Conclusions: There are situations where publicly available data could complement NGO baseline survey data, most importantly when the NGO has tolerance for estimates of low or unknown accuracy.