A carregar...
Projeto de investigação
Sem título
Financiador
Autores
Publicações
Picturing inequities for health impact assessment : linked electronic records, mortality and regional disparities in Portugal
Publication . Nicolau, Leonor Bacelar; Rodrigues, Teresa; Fernandes, Elisabete; Lobo, Mariana F.; Nisa, Cláudia Fernandes; Azzone, Vanessa; Pinto, Armando Teixeira; Pereira, Altamiro Costa; Normand, Sharon-Lise Teresa; Miguel, José Pereira
Health impact assessment (HIA) focuses on minimizing inequities when studying the effects of a policy on the population’s health. Nevertheless, it is seldom simultaneously quantified, multivariate, and visually graphically comprehensible for non-statisticians. This paper aims to address that gap, assessing a policy promoting the quality of Electronic Health Records, linking hospital and primary health care data (Blood Pressure, Cholesterol, Triglycerides, Waist Circumference, Body Mass Index) to mortality outcomes and regional inequities. Acute Myocardial Infarction patients admitted in the hospital are then followed regularly in Portuguese NHS Primary Care. Regional disparities regarding recorded information are observed and different association patterns with mortality identified, ranked, and visualized through adjusted ORs for sex, age, and indicators of severity of hospital admission, complemented with multivariate correspondence analysis. A pathway to handling equity within quantitative HIA shows that complexity in data and methods may generate simplicity and clarity through visual graphical aids. Tackling Big Data with Data Science in HIA may even be at the center of future health reforms, assessing impacts of health promotion and chronic disease policies.
Avaliação de impacte na saúde : quantificar e modelizar para melhor
Publication . Nicolau, Leonor Bacelar; Miguel, José M. Pereira, 1947-; Saporta, Gilbert
Context Health Impact Assessment (HIA) is a combination of procedures, methods and tools by which a policy may be judged as to its potential effects and its distribution on a population’s health, with thus a particularly strong concern regarding equity issues. It is at its core a decision-making support tool, havin as main output recommendations to help decision-makers assure that the policies they implement minimize negative effects on health and equity and maximize positive ones. Throughout the various HIA steps, it is still generally and very often a qualitative approach, very seldom using multivariate statistical methodologies. Main Aim To show the usefulness of applying quantified multivariate statistical methodologies to enrich HIA practice, while making the decision-making process easier, in terms of issuing understandable outputs even for non-statisticians, but still in a deeply evidence based context. Exploring and Profiling through the State of the Art HIA is born out of two distinct areas of knowledge: environmental assessment and public health. Two questions arise: Are consensus guidelines possible? Are quantitative methods applied throughout both areas? Profiling the 45 HIA guides studied by Herbert (2012) with hierarchical cluster analysis shows a consensus universal guide would be plausible and 9 candidates are pinpointed to potentially serve as its foundation. The titles of 247 HIA Journal Papers listed in the Health Impact Assessment Section of the HIA Gateway Bibliography associated to Public Health England from 2012 to 2014, and additionally the available keywords for 170 of these papers, were studied with multivariate text analysis, regarding profiles of year, being published in a more environmental or public health journal and using quantitative methods. It is shown that results based on titles and on keywords are coherent, that 2014 tends to be less related with environmental journals, that HIA in environmental fields is more quantitative (especially in what concerns related risk assessments) and in public health is more qualitative and mainly related to social health determinants and equity concerns. It is thus pertinent and an added value to current knowledge to encourage the use of quantified, multivariate statistical methods (such as cluster analysis or multivariate text analysis) when assessing health impacts of public policies and related equity issues with HIA. Contributing with Hierarchical Cluster Analysis to the Health Impact Assessment Screening Step The screening step of HIA helps us here to choose the area and focus of our research. Hierarchical Cluster Analysis is used as a basic methodology to contribute for improving the HIA screening step, regarding two situations. A total of 76 Health Reform Policies in Portugal was rated by a panel of 7 international renowned public health experts from Nova University in 2011 on a 10-point scale (1-Very low to 10-Very high) regarding Potential Impact, Ease of implementation and Implementation costs. Cluster analysis allowed us to make scenario analysis regarding all possible cluster priorities and to pinpoint a group of policies with high potential impact and low implementation costs, focusing policies linked to management issues setting a legal background and operational benchmarks improving the link between hospital and primary care services. A total of 30 National Health Programs from the National Health Plan 2004-2010 were described by datasets of variables in 7 domains. Each domain is ranked according to a previously determined set of criteria for being a potential priority for HIA. Hierarchical cluster analyses results and complementary multivariate methodologies enabled us then to rank each cluster in each partition, within each domain and then taking into account all 7 domains. The policies related to the Cerebro-cardiovascular Program are thus chosen as first priority to apply HIA to, according to various criteria of regional span, concerns of health gains, health systems needs and effectiveness gains, being high-priority for National Health Plans, among others. These analyses lead us to the choice of the empirical setup used throughout the study of equity below, namely regarding cardiovascular disease and policies concerning hospital-primary care links. Picturing Inequities for Health Impact Assessment A conceptual quantitative path of analysis to study the association between a public policy and health impacts while taking equity into account is proposed, showing how one may put it into practice. Data includes 3.776 adults admitted in hospitals for acute myocardial infarction during the second semester of 2012, followed regularly in primary care during 2013, always within the National Health Service in Portugal. The policy whose impact is be assessed in a HIA context is the registry of information regarding indicators Blood Pressure, Cholesterol, Triglycerides, Waist Circumference and Body Mass Index. The health impact of interest is the mortality status at the end of the period of study (December 31 2013). The equity aspect under analysis concerns regional differences in Portugal, since not assuring an even application of the policy throughout all the 5 health regions of residence (Norte, Centro, LVT, Alentejo and Algarve) may increase regional health inequities. Crude and adjusted OR (for sex, age, sex*age, 4 severity hospital admission indicators) are calculated with Logistic Regressions, complemented by exploratory multivariate data analysis such as Principal Components and Multiple Correspondence Analysis. Firstly, differences are found among regions regarding the registration of health indicators under study. This assures the relevancy of studying the association between the mortality outcome and the registration of health indicators stratified by region. Distinct mortality-registration associations are then found among regions in their association patterns regarding mortality and the registration of the health indicators under study. This hints that the registration of health indicators linking information from Hospital and Primary healthcare, throughout the different regions, may contribute for better health outcomes. The pursuit of delivering research results in a very visually comprehensible approach, given the elected multivariate methodologies used, intends to facilitate decision makers’ understanding of outcomes reached. Discussion, Conclusions and Future Perspectives Applying data mining and data science methodologies means that even if multivariate statistical methods used are complex, graphical outputs may simplify the understanding of results by decision makers. The future of healthcare reforms shifts the center of evaluation of health systems from providers to people’s individual needs and preferences, by reducing health inequities in access and health outcomes, adapting health systems to new health technologies, using big data linking information from providers to social and economic health determinants. New or innovative statistical and assessment methodologies are needed to set this transformation into motion. Quantified multivariate HIA thus represents a valuable tool to assure health impacts of public policies are indeed measured taking into consideration health determinants and equity and bringing citizens to the center of the decision-making process.
Unidades organizacionais
Descrição
Palavras-chave
Contribuidores
Financiadores
Entidade financiadora
Fundação para a Ciência e a Tecnologia
Programa de financiamento
3599-PPCDT
Número da atribuição
HMSP-ICT/0013/2011