How to estimate small area health related characteristics of populations?
Health related characteristics of a population in a society are significant to health promotion programs and to the provision of better health services. The efforts of feasible health planning generally target local areas such as local health regions or small area health units, and the population specific health program planning often requires precise estimates of health behaviours of the population at a fine spatial scales e.g., statistical local areas (SLAs) in Australia or counties in US or Wards in UK. But health related data are not always available at such scales due the policy and privacy constraints.
Policy makers sometimes rely on national level or state level datasets to understand the health needs of their communities. However, the basic problem with these surveys is that they are not designed for efficient estimation at small areas. Any conventional estimators from these data mainly have the two major limitations: (i) prevalence estimates can only be computed for a subset of all areas which contain respondents to the survey; and (ii) for those small sampled areas the achieved sample size will usually be very small and the estimator will thus have low precision, and ultimately statistically unreliable.
The lack of a national dataset having detailed characteristics of individuals at fine spatial scales negatively impacts the ability of local and national agencies to manage serious health issues in the community and associated risks factors. For examples, youth and adults’ smoking behaviours, characteristics of overweight and obesity, risk factors and distribution of cardiovascular disease, etc at small areas are not readily available, but crucially important for policy making or evaluation purposes. In this study, an appraisal of the methodologies for the estimation of health related characteristics of populations at small area levels is provided.
The overall methodologies for SAE are depicted in Figure 1. Traditionally there are two types of SAE – direct and indirect model-based estimations. The direct methods include three estimators: H-T estimator, GREG estimator and modified direct estimator. Indirect methodologies of SAE are divided into two approaches – statistical and geographic. The statistical modelling is based on statistical models (i.e. implicit and explicit models) and techniques; while, the geographic approaches use microsimulation modelling.
Most of the developed nations are utilising SAE as an essential means to support the knowledgeable and effective decision making and policy analysis for various issues at local levels. There are three distinct sets of modelling methodologies utilised by practitioners: (i) indirect standardisation and individual level modelling; (ii) multilevel statistical modelling; and (iii) microsimulation modelling technology (MMT). A highlight of these three sets of methodologies is depicted in Table 1.
The findings suggest that although each of these three modelling approaches has its own strengths in relation to generating small area health related characteristics estimation, the microsimulation modelling technology has more robustness over other methods in the sense that further aggregation or disaggregation is possible on the basis of the choice of spatial scales or domains. As well, it is possible to use the microdata file from MMT for further analysis and updating, and to assess small area effects of policy changes. MMT approaches also allow what if scenarios in terms of policy changes. The other approaches do not have such utilities. Finally, from the overall assessment, it is apparent that MMT is comparatively a precise way to estimate small area health related characteristics and evaluate policy changes. Our future research will employ this new approach to produce the estimates of small area health related characteristics – in particular, the estimates of smoking behaviours of adults and/or estimates of the prevalence of obesity of adults at small area levels in Australia.
Discipline of Mathematics and Statistics, School of Computing and Mathematics,
Charles Sturt University, Wagga Wagga, NSW 2678 Australia
Estimating small area health-related characteristics of populations: a methodological review.
Geospat Health. 2017 May 8
|Prevalence and prevention of unwanted online sexual… Screens, smartphones, and devices are integral to the lives of young people and provide opportunities for positive social connection, learning, and entertainment. However, existing alongside the many rewards of technology…|
|AI Can’t Save Us From This Pandemic - But It Can… Much has been said about artificial intelligence (AI) and its vast potential to revolutionize life as we know it. But if we were to pinpoint a specific field that has…|
|California has changed social norms, dramatically… The rapid rise in the dependency on cigarette smoking in the United States occurred in the first half of the 20th century fueled by large mass media marketing campaigns and…|
|Prevalence of cardiovascular disease in Bangladesh:… Cardiovascular diseases (CVD) are among the leading causes of death globally and most deaths (80%) occur in low- and middle-income countries like Bangladesh. A significant increase in the prevalence (the…|
|Bell ringers: Concussion problems in youth football players Concussion has become an important topic in sports news and research. While researchers are studying the mental, physical, and emotional effects of concussion within professional, college, and high school athletes;…|
|How to build great AI Even though artificial intelligence (AI) is in full bloom and more and more businesses are developing their own solutions, there are several things around it that aren’t quite clear. One…|