The present investigation was a cross-sectional observational study in which a secondary analysis of the National Socioeconomic Characterization Survey (CASEN version 2017) was performed. The CASEN survey is regularly applied by the Ministry of Social Development to Chilean households and their residents; aiming to know their socioeconomic situation, multidimensional poverty and income distribution. As well as update the evidence of priority groups and detect their particular needs. This voluntary survey follows a structured interview answered by an adult who provides data of the other household members. This survey was designed with a probabilistic, stratified and multistage sampling; that is representative at each national, regional (16 regions), and urban/rural level, but excluding geographic areas with difficult access. The total sample was comprised by 70.947 households with 216.439 residents, which represented 16.843.471 Chilean population and 777.407 international migrants. The data base of CASEN survey has public access (27). This study is part of the Fondecyt Regular project 1201461 approved by the Ethics Committee of the Faculty of Medicine of The Universidad del Desarrollo and Ethics Committee of the Servicio de Salud Metropolitano Sur-Oriente. The study complied with ethical guidelines and regulations according to the principles of the Declaration of Helsinki.
Health status was examined using the framework of the small module on health from the European Statistics of Income and Living Condition (EU-SILC) as a reference. The instrument contains 3 different variables with its corresponding concepts (28). These concepts were used to create new variables from the questions available in the CASEN survey.
Negative Self-perceived health (NSPH)
new variable was created based on the question “from 1 to 7 how would you rate your current health status”. According to previous literature the seven-grade scale could be interpreted as 1 very poor health to 7 excellent health that cannot be improved (29). Like previous studies (30), the variable was dichotomized as positive health for scores ranging 4-7 and negative health for scores ranging 1-3. This study focused on negative self-perceived health as indicator in order to maintain consistency with the other negative health indicators included in the analysis.
Chronic morbidity (CM)
based on question “have you been receiving medical treatment for the past 12 months?”. Dichotomized as yes or no according to the presence of hypertension/dental Emergency, diabetes, depression, acute myocardial infarction, cataracts, chronic obstructive pulmonary disease, leukemia, bronchial asthma, cancer (gastric, cervical uterine, breast, testicular, prostate, colorectal), preventive cholecystectomy, chronic kidney failure, ischemic brain accident, bipolar disorder, lupus or other chronic condition.
Although the EU-SILC framework does not separate disability from the activity limitation variable. The CASEN survey includes a question focused on disability (31), from whom the new variable was created “¿Do you have any of the following permanent conditions? Dichotomized as yes or no according to the presence of one or more physical/speaking/psychiatric/mental/hearing/visual conditions.
Activity limitations (AL)
The variable was created using all types of daily living activities limitations asked by CASEN. “How much difficulty do you have for...”. This question was restricted to population over 6 years. Dichotomized as yes or no according to the presence of mild, moderate, severe, or extreme difficulty for one or more activities (eating, showering, displacing, bathroom use, lie down or get out of bed/ get dressed).
Social determinants of health
age as continuous variable and categorical (<6 years, 6-14, 15-64 and >64 years). Sex (male, female). Ethnicity for those belonging or being descendant of minority groups in Chile (yes, no), marital status (single, married/cohabitant, separated/divorced/annulled, widow), area (urban, rural).
educational level according to the highest level achieved or current level of the household informant (categorized as university, technical, high school, primary, kinder, special education, none). Household income categorized in five quintiles of equal size sorted in ascending order according to the autonomous per capita household income (I, II, III, IV, V). Occupation defined by the occupational activity of the household informant. The variable was created from questions related to current job/occasional job/ work license/search for a job/ attending to educational center (categorized as unemployed, does not study, study, employed, and study and work).
Access to health care
The variable affiliation to the health care system was used as a proxy of access and created from the question “Which health insurance system do you use?”. Further categorized as none, public health system affiliation, private health system affiliation, other.
the variable social support was created from available questions that were mainly related to instrumental social support network; dichotomized yes or no according to the presence of one or more supportive behaviors from someone at home and outside. Social capital variable was created from a question of belonging and participation in diverse organizations or organized groups over the last 12 months. Dichotomized as yes or no according to the participation in one or more of these groups.
Migratory related factors
Country of origin was created as a categorical variable based on the question “When you were born, ¿what country did your mother live in?”. The categories were selected according to the intraregional pattern reported in migratory statistics (4) (Venezuela, Perú, Haiti, Colombia, Bolivia, Argentina, Ecuador, other countries in South America and other). Time of residence was created based on the year period in which the migrant arrived and categorized (2015 or later, 2010-2014, 2005-2009, 2000-2004, 1999 or before).
Health status outcomes were analyzed descriptively for international migrants and Chilean born population. The crude and stratified prevalence by demographic, socioeconomic, access to health care and migratory related factors were presented as proportion. The Pearson’s chi-square test was used to test independence between migration and health status outcomes. Multivariate logistic regression was used to estimate the probability (odds ratio, OR) of reporting these health outcomes and adjusted by each set of SDH in international migrants and local population, separately. The association between migratory related factors and health outcomes was explored with multivariate logistic regression adjusted by sex and age. Then, the healthy migrant effect was examined using multivariate logistic regression sequentially adjusted for SDH, where NSPH, CM, DIS and AL were dependent variables and migrant status was the independent variable (reference Chilean born). In order to estimate the crude and adjusted probability of presenting these health outcomes if being international migrant. The Hosmer-Lemeshow goodness of fit test was used as post-estimation after logistic regression. Data analysis were performed with STATA 14 software (Stata Corp) and weighted according to the survey’s sampling design. Significance was set at 0.05 with 95% confidence interval (95% CI).