Hakunawadi Alexander Pswarayi, Edward J M Joy, Dawd Gashu, Fanny Sandalinas, Adamu Belay, R Murray Lark
{"title":"Analysis of data from a national micronutrient survey with a linear mixed model: estimates, predictions and lessons for future surveys.","authors":"Hakunawadi Alexander Pswarayi, Edward J M Joy, Dawd Gashu, Fanny Sandalinas, Adamu Belay, R Murray Lark","doi":"10.1177/22799036241274962","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Because micronutrient deficiencies affect public health, countries monitor population status by national-scale, multi-stage, micronutrient surveys (MNS). In design-based surveys, inclusion probabilities are specified for sample units and the corresponding sample weights allow design-unbiased estimates to be made of population parameters. Corrections may be possible on departures from the design; an alternative is to use linear mixed models (LMM), with an estimated covariance structure reflecting the sampling design, to obtain model-based estimates.</p><p><strong>Design: </strong>The Ethiopia National Micronutrient Survey (2016) specified inclusion probabilities at enumeration area (EA) and household (HH) levels, and sample weights are provided. However, the design was not followed as it would have resulted in insufficient sampling from women of reproductive age.</p><p><strong>Results: </strong>Having found no evidence that sample weights were informative for target serum micronutrient concentrations (Zn), we estimated LMM parameters, with Regions as fixed effects, and the variation of individuals nested within households, households within EA, and EA within regions, random effects. We obtained LMM standard errors, Best Linear Unbiased Estimates (BLUEs) of regional means, and empirical Best Linear Unbiased Predictions for sampled/unsampled EA and HH. The probability that each true regional mean exceeded the sufficiency threshold <math> <mrow><mrow><mo>(</mo> <mrow><mn>65</mn> <mi>μ</mi> <mi>g</mi> <mspace></mspace> <msup><mrow><mi>dL</mi></mrow> <mrow><mo>-</mo> <mn>1</mn></mrow> </msup> </mrow> <mo>)</mo></mrow> </mrow> </math> was evaluated. The variances of BLUEs of regional means, under alternative sampling designs, were bootstrapped from LMM variance components.</p><p><strong>Conclusions: </strong>We demonstrate use of LMM to obtain model-unbiased estimates and predictions when surveys deviate from the original design; and the use of LMM variance components to evaluate alternative designs for further sampling, or for sampling comparable populations.</p>","PeriodicalId":45958,"journal":{"name":"Journal of Public Health Research","volume":null,"pages":null},"PeriodicalIF":1.6000,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11526165/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Public Health Research","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1177/22799036241274962","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/10/1 0:00:00","PubModel":"eCollection","JCR":"Q3","JCRName":"PUBLIC, ENVIRONMENTAL & OCCUPATIONAL HEALTH","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Because micronutrient deficiencies affect public health, countries monitor population status by national-scale, multi-stage, micronutrient surveys (MNS). In design-based surveys, inclusion probabilities are specified for sample units and the corresponding sample weights allow design-unbiased estimates to be made of population parameters. Corrections may be possible on departures from the design; an alternative is to use linear mixed models (LMM), with an estimated covariance structure reflecting the sampling design, to obtain model-based estimates.
Design: The Ethiopia National Micronutrient Survey (2016) specified inclusion probabilities at enumeration area (EA) and household (HH) levels, and sample weights are provided. However, the design was not followed as it would have resulted in insufficient sampling from women of reproductive age.
Results: Having found no evidence that sample weights were informative for target serum micronutrient concentrations (Zn), we estimated LMM parameters, with Regions as fixed effects, and the variation of individuals nested within households, households within EA, and EA within regions, random effects. We obtained LMM standard errors, Best Linear Unbiased Estimates (BLUEs) of regional means, and empirical Best Linear Unbiased Predictions for sampled/unsampled EA and HH. The probability that each true regional mean exceeded the sufficiency threshold was evaluated. The variances of BLUEs of regional means, under alternative sampling designs, were bootstrapped from LMM variance components.
Conclusions: We demonstrate use of LMM to obtain model-unbiased estimates and predictions when surveys deviate from the original design; and the use of LMM variance components to evaluate alternative designs for further sampling, or for sampling comparable populations.
期刊介绍:
The Journal of Public Health Research (JPHR) is an online Open Access, peer-reviewed journal in the field of public health science. The aim of the journal is to stimulate debate and dissemination of knowledge in the public health field in order to improve efficacy, effectiveness and efficiency of public health interventions to improve health outcomes of populations. This aim can only be achieved by adopting a global and multidisciplinary approach. The Journal of Public Health Research publishes contributions from both the “traditional'' disciplines of public health, including hygiene, epidemiology, health education, environmental health, occupational health, health policy, hospital management, health economics, law and ethics as well as from the area of new health care fields including social science, communication science, eHealth and mHealth philosophy, health technology assessment, genetics research implications, population-mental health, gender and disparity issues, global and migration-related themes. In support of this approach, JPHR strongly encourages the use of real multidisciplinary approaches and analyses in the manuscripts submitted to the journal. In addition to Original research, Systematic Review, Meta-analysis, Meta-synthesis and Perspectives and Debate articles, JPHR publishes newsworthy Brief Reports, Letters and Study Protocols related to public health and public health management activities.