Beyond the project: Building a strategic theory of change to address dementia care, treatment and support gaps across seven middle-income countries.

Evidence from middle-income countries indicates high and increasing prevalence of dementia and need for services. However, there has been little investment in care, treatment or support for people living with dementia and their carers. The Strengthening Responses to Dementia in Developing Countries (STRiDE) project aims to build both research capacity and evidence on dementia care and services in Brazil, Indonesia, India, Jamaica, Kenya, Mexico and South Africa. This article presents the Theory of Change (ToC) approach we used to co-design our research project and to develop a strategic direction for dementia care, treatment and support, with stakeholders. ToC makes explicit the process underlying how a programme will achieve its impact. We developed ToCs in each country and across the STRiDE project with researchers, practitioners, people living with dementia, carers and policymakers at different levels of government. This involved (1) an initial ToC workshop with all project partners (43 participants); (2) ToC workshops in each STRiDE country (22-49 participants in each); (3) comparison between country-specific and overall project ToCs; (4) review of ToCs in light of WHO dementia guidelines and action plan and (5) a final review. Our experiences suggest ToC is an effective way to generate a shared vision for dementia care, treatment and support among diverse stakeholders. However, the project contribution should be clearly delineated and use additional strategies to ensure appropriate participation from people living with dementia and their carers in the ToC process.

Principales sesgos en la investigación epidemiológica en salud ocupacional / Main biases in epidemiological research in occupational health

El objetivo de este documento es mostrar una síntesis de los principales sesgos que pueden aparecer en el estudio de la salud del trabajo, sus orígenes y consecuencias. Para ello, se realizó una revisión documental de libros y artículos publicados en los últimos 10 años en inglés y español. Se dejó fuera aquellos documentos que no estuvieran publicados en revistas con comité editorial y en caso de los libros, que no tuvieran registro ISBN. Se incluyeron 7 libros y 14 artículos. Un sesgo es un error sistemático, que sucede en el proceso de investigación y que pueden darse durante la planeación, recolección y procesamiento de datos. Se revisarán dos tipos principales de sesgos: 1) Selección y 2) Información y/o medición, en cada caso se describe la definición, potencial impacto en los resultados y se plantean una serie de ejemplos que ayudan a su compresión. Uno de los efectos principales de la presencia de sesgos en investigación es que ocasiona un error de clasificación en los participantes, que puede ser diferencial o no, esto impacta los resultados al sobre o subestimarlos, en otras palabras, se encuentran efectos mayores a los reales o peor, no los encontramos cuando si existen. Para ilustrar el impacto de los sesgos, se revisan distintos escenarios que ejemplifican los casos de sobreestimación y subestimación que originan resultados poco confiables(AU)

This document aims to show a synthesis of the main biases that can be present in occupational health studies, their origins and their consequences. For this purpose, a documentary review was conducted of books and articles published in the last 10 years in English and Spanish. Excluding those documents that were not published in journals with an editorial board and, in the case of books, that did not have an ISBN registry. A total of 7 books and 14 articles were included. A bias is a systematic error that occurs in the research process and can occur during the data planning, collection, and processing. Two main types of biases will be reviewed: 1) Selection and 2) Information and/or measurement biases, in each case their definition and, their possible impact on the results are described, a series of examples that may help to understand them are provided. One of the main effects of the presence of biases in research is that they cause an error in the classification of participants, which can be differential or not, and this has an impact on the results by overestimating or underestimating them, i.e., finding stronger effects than the real ones, or worse, not finding them when they exist. To illustrate the impact of biases, different scenarios were reviewed that provide examples of cases of overestimation and underestimation that lead to unreliable results(AU)