Articles Frequency of health-care service use and severity of illness in undocumented migrants in Catalonia, Spain: a population- based, cross-sectional study Albert Dalmau-Bueno, Anna García-Altés, Emili Vela, Montse Clèries, Carles Vicenç Pérez, Josep Maria Argimon Summary Background In Spain, legislation was passed in 2012 excluding undocumented migrants from the public health-care Lancet Planet Health 2021; system. Catalonia was one of the Spanish regions that did not implement this legislation, and continued to guarantee 5: e286–96 access to health care to the whole population. We aimed to analyse health-care use and health status among Agència de Qualitat i Avaluació Sanitàries de Catalunya, undocumented migrants in Catalonia, and compare health-care use and health status with legal residents classified Barcelona, Spain according to their socioeconomic position (SEP). (A Dalmau-Bueno MSc, A García-Altés PhD); CIBER de Methods We did a population-based, cross-sectional study, with administrative individual data. The study included the Epidemiología y Salud Pública, Barcelona, Spain resident population in Catalonia, Spain, in 2017, aged younger than 65 years and with a maximum annual income (A García-Altés); Institut of less than €18 000 per year, and classified into three socioeconomic (SEP) groups—low SEP, very low SEP, and d’Investigació Biomèdica, undocumented migrants. Indicators regarding health-care service use (primary care, emergency care, mental health Barcelona, Spain care, acute care), drug prescriptions, and selected chronic and infectious diseases were analysed. (A García-Altés); Servei Català de la Salut, Barcelona, Spain (E Vela PSy, M Clèries MPH, Findings Between Jan 1 and Dec 31, 2017, 4 071 988 residents of Catalonia were included in this study; C V Pérez EDSEC); Institut Català undocumented migrants represented 2·8% (n=113 450) of this population. Of all undocumented migrants, de la Salut, Barcelona, Spain 25 942 (61·0%) female participants aged 15–64 years and 19 819 (46·0%) male participants aged 15–64 years (J M Argimon PhD) attended primary health-care centres: these rates were lower than in individuals with a very low SEP (84·8% in Correspondence to: Dr Anna García-Altés, Agència de female participants and 72·1% in male participants). Hospital admission rates among male participants aged Qualitat i Avaluació Sanitàries de 15–64 years in the very low SEP group were more than three times as high as in undocumented migrants Catalunya, Generalitat de (111·6 vs 35·7). The highest tuberculosis rate was found in undocumented male migrants (incidence rate 4·35 Catalunya, 08005 Barcelona, [95% CI 3·55–5·16]). Spain agarciaaltes@gencat.cat Interpretation Undocumented migrants made less use of health-care services than those in the low and very low SEP groups, but for some infectious diseases, incidence was higher in undocumented migrants. These results constitute an additional argument to support the maintenance of universal health coverage for all citizens. Funding None. Copyright © 2021 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license. Introduction control of its external borders and of migration flows. Throughout history, Europe has been a crossroads of Some countries have taken political action to limit the global migration systems. During the past 10 years, entry of undocumented migrants. In Spain, in 2012, Europe has received increasing numbers of refugees legislation to exclude undocumented migrants from the (from 200 000 in 2008 to 600 000 in 2019), reaching a peak public health-care system was approved.5 Catalonia in 2015, a time that has become known as the refugee and was one of the Spanish regions that did not implement migrant crisis.1 In 2019, countries in the European this legislation, and immediately announced that all Economic Area resettled 30 264 refugees through national people living in Catalonia would have access to and EU programmes.2 In addition to those refugees who health-care services, including undocumented migrants.6 obtain legal status in a new country, many undocumented In July, 2018, the Spanish legislation was repealed.7 migrants come to Europe in an irregular or clandestine However, the maintenance of universal coverage in manner. Once there, they seek to legalise their situation Catalonia over the intervening period obliged the Catalan and earn a living. Since 2015, Spain has greatly increased Government to register undocumented migrants and to its migrant intake. In 2019, Spain received 664 557 regular identify them as a special interest group. migrants,3 and 32 513 undocumented migrants.4 Previous studies done in Europe, using surveys or In parallel, the salience of migration as a political and administrative data from small regions, found a lower social concern has intensified in Europe. The European use of primary care and less contact with general Council has implemented measures to improve the practitioners by undocumented migrants than regular www.thelancet.com/planetary-health Vol 5 May 2021 e286 Articles Research in context Evidence before this study detailing socioeconomic characteristics, use of health-care The only quantitative (survey-based) study found in the services, and disease diagnosis, collected from administrative literature reported no differences in the use of health-care databases. We found that undocumented migrants made less services between legal and illegal migrants when ill; two relevant use of health-care services than those in the low and very low systematic reviews showed that undocumented migrants in socioeconomic position groups. Also, the tuberculosis rate is Europe faced particular problems when using health-care higher in undocumented migrants. services, especially regarding maternal and child health care. In Implications of all the available evidence Spain, in 2012, legislation was passed that excluded The results of this study challenge the claims that undocumented migrants from the public health-care system. undocumented migrants make higher than average use of Catalonia did not implement this legislation, and continued to health-care services, and thus support the maintenance of guarantee access to health care to the whole population. universal health coverage. Added value of this study The individual data of 4 071 988 residents of Catalonia (113 450 undocumented migrants) were included in this study, migrants or the native population.8–10 Maternal health-care comparison between the different population groups. The services were also underused by or inadequate for following SEP groups were defined for this study: (1) low undocumented migrants.11,12 SEP, which includes individuals receiving unemployment One of the arguments used by far-right political parties allowance, individuals receiving unemployment benefit, to oppose immigration is that foreign-born people use and individuals in employment earning less than €18 000 health-care services more intensively than the native per year; (2) very low SEP, which includes individuals population; at the other end of the political spectrum, receiving some form of universal benefit, individuals left-oriented parties have argued that migrants face who no longer receive unemployment allowance, and barriers restricting their access to these services.13 individuals who no longer receive unemployment benefit We aimed to analyse health-care use and health status and do not qualify for unemployment allowance (benefits among undocumented migrants in Catalonia and to and allowances definitions have been previously make a comparison with the legally resident population published16); and (3) undocumented migrants, identified classified by socioeconomic position (SEP). as a special interest group. The need for written informed consent is waived for this study type. Methods Study design and participants Procedures We did a population-based, cross-sectional study, with A database including the population resident in Catalonia admin­istrative individual data in Catalonia, Spain. Health in 2017, up to the age of 64 years was created, containing care in Catalonia is organised as a National Health System, information on their use of health-care services, drug funded through taxation.14 All residents (7 348 275 as prescriptions, and selected chronic and infectious of 2017) are granted universal public health-care coverage diseases. SEP was defined according to the information by law. Public health-care spending represents 5·4% of obtained from the levels of pharmacy copayment and Catalan’s gross domestic product.15 The use of publicly undocumented migrants after their recognition as a funded health-care services is free; the sole exception is special interest group. Children were assigned the SEP of drug prescriptions, which are based on a copayment their parents or guardian. The following sources of system calculated according to the individual’s income (or, information were used: (1) The Catalan Government’s if appropriate, according to the social security benefits central registry of insured individuals, as of Jan 1, 2017, received). Each resident is assigned a unique personal for demographic information, SEP, and undocumented health-care identification number, which can be used to migrant status; (2) the databases for primary care trace their use of health-care services and prescribed (Conjunt Minim Basic de Dades [CMBD]-primary care), drugs, and link administrative databases. acute hospitals (CMBD-hospital), emergencies (CMBD- Because only 4·3% of undocumented migrants were emergencies), mental health outpatient centres (CMBD- aged 65 years or older, this population group was excluded ambulatory mental health), and psychiatric hospital­isation from the study. Individuals with an annual income (CMBD-psychiatric hospitalisation; 2017); and (3) the of €18 000 or more were also excluded, as the practice of pharmacy billing database (2017). contracting voluntary private insurance in higher income As these databases are used to monitor health-care groups entails a lower use of public health-care services quality and also for reimbursement purposes, data quality among these sectors, and would thus have introduced a is checked both at the provider level and systematically by bias in the study that might have compromised the the Catalan Health Department. e287 www.thelancet.com/planetary-health Vol 5 May 2021 Articles Outcomes Very low SEP Low SEP Undocumented migrants p value* We assessed the percentage of people who attended (n=269 915) (n=3 688 623) (n=113 450) primary care services and mean number of visits to Sex ·· ·· ·· <0·01 primary care services; the percentage of people who Female 143 345 (53·1%) 1 884 374 (51·1%) 56 154 (49·5%) ·· attended emergency services (acute hospital, primary Male 126 570 (46·9%) 1 804 249 (48·9%) 57 296 (50·5%) ·· care emergency centres, or emergency medical services) Age, years ·· ·· ·· <0·01 and the mean number of visits to emergency services; <15 59 524 (22·1%) 689 614 (18·7%) 29 100 (25·7%) ·· the percentage of people who attended mental health 15–64 210 391 (77·9%) 2 999 009 (81·3%) 84 350 (74·4%) ·· centres and the mean number of visits to mental health Region ·· ·· ·· <0·01 centres; the hospital admission rate; child­birth hospital Spain (Europe) 191 614 (71·0%) 2 743 251 (74·4%) NA ·· admission rate, surgical hospital admission rate, and psy­chiatric hospital admission rate (excluding children Latin America and 10 719 (4·0%) 232 829 (6·3%) 38 734 (34·1%) ·· Caribbean <5 years) per 1000 inhabitants; the percentage of Europe and central 9742 (3·6%) 209 366 (5·7%) 19 761 (17·4%) ·· individuals prescribed at least one drug, mean number Asia of prescriptions, percentage of individuals prescribed Middle East and 39 657 (14·7%) 162 665 (4·4%) 9015 (7·9%) ·· psychoactive drugs (anxiolytics, anti­depres­sants, or anti­ north Africa psychotics, excluding children <5 years); the percen­ Sub-Saharan Africa 358 (0·1%) 74 273 (2·0%) 23 309 (20·5%) ·· tage of attendant individuals diagnosed with diabetes, South Asia 5987 (2·2%) 61 432 (1·7%) 9632 (8·5%) ·· chronic obstructive pulmonary disease (COPD), asthma, East Asia and 1783 (0·7%) 54 607 (1·5%) 4802 (4·2%) ·· cirrhosis, or neoplasms (appendix pp 1–3 for International Pacific Classification of Diseases, ninth revision, clinical North America 40 (<0·1%) 2130 (0·1%) 354 (0·3%) ·· modification disease codes); and the percentage of Unknown 10 015 (3·7%) 148 070 (4·0%) 7843 (6·9%) ·· attendant individuals diagnosed with HIV, tuberculosis, Data are n (%), unless specified. SEP=socioeconomic position. NA=not applicable. *χ² test. or hepatitis C virus (HCV; see appendix pp 1–3 for disease Table 1: Population characteristics by SEP in Catalonia, 2017 codes). Statistical analysis less than €18 000 per year were included in this study. See Online for appendix We did a descriptive analysis of sex, age group, and Undocumented migrants represented 2·8% (n=113 450) nationality stratified by SEP, and the results are presented of the whole study population, those with very low with absolute and relative frequencies. Next, the number SEP 6·6% (n=269 915), and those with low SEP 90·6% of individuals and use of public health-care services in (n=3 688 623). Among undocumented migrants, each SEP group were calculated (percentages, means, or 57 296 (50·5%) were male and 84 350 (74·3%) were aged rates). Incidence rates were calculated taking into account 15–64 years; the most common origin was Latin American the exposure time (person-years) in each SEP, by means or Caribbean (34·1%), followed by sub-Saharan of the information on registrations and cancellations in African (20·5%), European or central Asian (17·4%), the central registry of insured individuals, to make the south Asian (8·5%), and Middle Eastern or north African SEP groups comparable (appendix p 4). 5-year age- (7·9%; table 1). adjusted Poisson models were used to model the data, For primary care, of all undocumented migrants, 76·1% with individuals (or number of visits) as the dependent of female and male participants younger than 15 years variable, SEP as the exposure variable, and person-years as attended primary health-care centres and these rates were observation time. The average predicted incidence rate for lower than in individuals with low and very low SEP. The each SEP through marginal effect for each indicator with incidence rate for attendance at primary care centres was its respective 95% CI was calculated and results presented 92·25% (95% CI 90·48–94·03) of undocumented female by 100 person-years. Finally, incidence rate ratio, with low migrants younger than 15 years and 91·78% SEP as the reference group, was included with its (90·05–93·50) of undocumented male migrants younger respective 95% CI and p value. The results were stratified than 15 years (table 2). The corres­ponding figures for the by two age groups (<15 years and 15–64 years) and sex, and low SEP group were 88·42% (95% CI 88·09–88·74) for are presented for the different SEP groups (very low SEP, female participants younger than 15 years and 88·29% low SEP, and undocumented migrants). All the analyses (87·98–88·61) for male participants younger than 15 years were done with Stata, version 14.2, statistical software. (table 2). In the 15–64 years age group, the age-adjusted incidence rate of undocumented migrants who attended Role of the funding source primary care centres was 82·59% (95% CI 81·58–83·60) There was no funding source for this study. for female participants and 64·25% (63·35–65·15) for male partici­pants (table 3). The corresponding figures Results were lower in the low SEP group for female participants Between Jan 1 and Dec 31, 2017, 4 071 988 residents of (77·96%) and similar for male participants (66·89%) and Catalonia aged younger than 65 years and with an income higher in the very low SEP group (female www.thelancet.com/planetary-health Vol 5 May 2021 e288 Articles participants 84·82%, male participants 72·18%; table 3). participants and 28·54% for male participants, in The mean number of age-adjusted visits also differed whom female participants made a mean of 0·75 visits between socioeconomic groups, with undocumented and male participants made a mean of 0·60 visits migrants presenting the lowest values: a mean of (table 3). The very low SEP group presented the highest 6·35 visits for female participants younger than 15 years incidence rate of emergency service use, between 6 to and 6·41 in male participants younger than 15 years, 11 points higher than the undocumented migrants in 4·61 for female participants aged 15–64 years, and 3·37 for both age groups and for both sexes. In all SEP groups, male participants aged 15–64 years (tables 2, 3). more male participants younger than 15 years attended For emergency care, undocumented migrants made emergency services than female participants younger less use of emergency services than the very low SEP than 15 years. This relation was reversed in individuals group. Incidence rate was 35·59% of female partici­ aged 15–64 years, in which female participants attended pants younger than 15 years and 38·71% of male more than male participants. participants younger than 15 years; this population Regarding mental health care, the highest incidence group made the lowest mean number of visits (0·71 in rate of those younger than 15 years who attended mental female participants and 0·81 in male participants; health centres were found in the very low SEP group table 2). Among undocumented migrants aged (male participants 8·43%, female participants 4·35%). 15–64 years, these figures were 34·94% for female The values were lower among undocu­ mented male Female participants younger than 15 years Male participants younger than 15 years Number IR (95% CI) IRR (95% CI) p value Number IR (95% CI) IRR (95% CI) p value Individuals attended in primary care Low SEP 283 360 (84·8%) 88·42% (88·09–88·74) 1 (ref) ·· 301 096 (84·7%) 88·29% (87·98–88·61) 1 (ref) ·· Very low SEP 25 353 (87·7%) 90·67% (89·56–91·79) 1·03 (1·01–1·04) <0·001 26 814 (87·6%) 90·58% (89·50–91·67) 1·03 (1·01–1·04) <0·001 Undocumented 10 357 (76·1%) 92·25% (90·48–94·03) 1·04 (1·02–1·06) <0·001 10 838 (76·1%) 91·78% (90·05–93·50) 1·04 (1·02–1·06) <0·001 migrants Mean primary care visits Low SEP 2 187 475 (6·6%) 6·77% (6·77–6·78) 1 (ref) ·· 2 376 006 (6·7%) 6·91% (6·90–6·92) 1 (ref) ·· Very low SEP 200 337 (6·9%) 7·52% (7·48–7·55) 1·11 (1·10–1·11) <0·001 214 348 (7·0%) 7·63% (7·60–7·67) 1·10 (1·10–1·11) <0·001 Undocumented 71 273 (5·2%) 6·35% (6·30–6·39) 0·94 (0·93–0·94) <0·001 76 147 (5·3%) 6·41% (6·36–6·45) 0·93 (0·92–0·93) <0·001 migrants Individuals attended at emergency services Low SEP 119 239 (35·7%) 36·99% (36·78–37·20) 1 (ref) ·· 138 664 (39·0%) 40·47% (40·26–40·68) 1 (ref) ·· Very low SEP 11 225 (38·8%) 41·58% (40·81–42·35) 1·12 (1·10–1·15) <0·001 12 895 (42·1%) 44·99% (44·21–45·77) 1·11 (1·09–1·13) <0·001 Undocumented 3999 (29·4%) 35·59 (34·49–36·70) 0·96 (0·93–0·99) 0·02 4587 (32·2%) 38·71% (37·59–39·83) 0·96 (0·93–0·98) 0·003 migrants Mean emergency services visits Low SEP 247 353 (0·7%) 0·76% (0·76–0·77) 1 (ref) ·· 295 546 (0·8%) 0·86% (0·86–0·86) 1 (ref) ·· Very low SEP 24 872 (0·9%) 0·95% (0·93–0·96) 1·24 (1·22–1·25) <0·001 29 129 (0·9%) 1·05% (1·04–1·06) 1·22 (1·21–1·24) <0·001 Undocumented 7995 (0·6%) 0·71% (0·70–0·73) 0·93 (0·91–0·95) <0·001 9649 (0·7%) 0·81% (0·80–0·83) 0·94 (0·93–0·96) <0·001 migrants Individuals attended at mental health centres Low SEP 10 304 (3·1%) 3·30% (3·24–3·37) 1 (ref) ·· 20 705 (5·8%) 6·20% (6·12–6·29) 1 (ref) ·· Very low SEP 1341 (4·6%) 4·35% (4·12–4·58) 1·32 (1·24–1·39) <0·001 2765 (9·0%) 8·43% (8·12–8·75) 1·36 (1·31–1·41) <0·001 Undocumented 323 (2·4%) 2·91% (2·59–3·23) 0·88 (0·79–0·98) 0·02 538 (3·8%) 4·64% (4·25–5·03) 0·75 (0·69–0·81) <0·001 migrants Mean number of mental health centre visits Low SEP 62 952 (0·2%) 0·20% (0·20–0·20) 1 (ref) ·· 134 892 (0·4%) 0·40% (0·40–0·41) 1 (ref) ·· Very low SEP 8331 (0·3%) 0·27% (0·26–0·28) 1·34 (1·31–1·37) <0·001 17 901 (0·6%) 0·54% (0·54–0·55) 1·35 (1·32–1·37) <0·001 Undocumented 1902 (0·1%) 0·17% (0·16–0·18) 0·85 (0·81–0·89) <0·001 3173 (0·2%) 0·27% (0·26–0·28) 0·68 (0·65–0·70) <0·001 migrants Hospital admission rate, per 1000 inhabitants Low SEP 16 276 (48·7%) 50·10% (49·33–50·87) 1 (ref) ·· 23 906 (67·2%) 69·25% (68·37–70·13) 1 (ref) ·· Very low SEP 1501 (51·9%) 58·83% (55·85–61·81) 1·17 (1·11–1·24) <0·001 1 968 (64·3%) 72·10% (68·91–75·29) 1·04 (0·99–1·09) 0·09 Undocumented 601 (44·2%) 53·50% (49·22–57·78) 1·07 (0·98–1·16) 0·11 833 (58·5%) 69·87% (65·13–74·62) 1·01 (0·94–1·08) 0·80 migrants (Table 2 continues on next page) e289 www.thelancet.com/planetary-health Vol 5 May 2021 Articles Female participants younger than 15 years Male participants younger than 15 years Number IR (95% CI) IRR (95% CI) p value Number IR (95% CI) IRR (95% CI) p value (Continued from previous page) Surgical hospital admission rate, per 1000 inhabitants Low SEP 7112 (21·3%) 22·07% (21·55–22·58) 1 (ref) ·· 12 571 (35·4%) 36·68% (36·04–37·32) 1 (ref) ·· Very low SEP 736 (25·5%) 27·00% (25·05–28·95) 1·22 (1·13–1·32) <0·001 1103 (36·0%) 38·14% (35·89–40·40) 1·04 (0·98–1·11) 0·21 Undocumented 305 (22·4%) 27·12% (24·07–30·16) 1·23 (1·10–1·38) <0·001 458 (32·1%) 38·61% (35·08–42·15) 1·05 (0·96–1·16) 0·28 migrants Psychiatric hospital admission rate, per 1000 inhabitants Low SEP 348 (1·0%) 1·14% (1·02–1·26) 1 (ref) ·· 214 (0·6%) 0·65% (0·57–0·74) 1 (ref) ·· Very low SEP 38 (1·3%) 1·27% (0·87–1·67) 1·11 (0·80–1·56) 0·53 41 (1·3%) 1·28% (0·89–1·68) 1·97 (1·41–2·75) <0·001 Undocumented 12 (0·9%) 1·11% (0·48–1·74) 0·97 (0·55–1·73) 0·92 9 (0·6%) 0·79% (0·27–1·31) 1·21 (0·62–2·37) 0·57 migrants Mean number of prescriptions Low SEP 1 075 413 (3·2%) 3·34% (3·33–3·34) 1 (ref) ·· 1 318 797 (3·7%) 3·85% (3·84–3·85) 1 (ref) ·· Very low SEP 147 478 (5·1%) 5·46% (5·43–5·48) 1·64 (1·63–1·65) <0·001 170 859 (5·6%) 5·98% (5·95–6·01) 1·55 (1·55–1·56) <0·001 Undocumented 32 664 (2·4%) 2·91% (2·87–2·94) 0·87 (0·86–0·88) <0·001 38 305 (2·7%) 3·23% (3·20–3·26) 0·84 (0·83–0·85) <0·001 migrants Individuals prescribed at least one drug Low SEP 205 731 (61·6%) 63·98% (63·70–64·25) 1 (ref) ·· 223 338 (62·8%) 65·28% (65·01–65·55) 1 (ref) ·· Very low SEP 20 432 (70·7%) 74·14% (73·12–75·16) 1·16 (1·14–1·18) <0·001 21 887 (71·5%) 75·05% (74·05–76·04) 1·15 (1·13–1·17) <0·001 Undocumented 6923 (50·9%) 61·59% (60·14–63·04) 0·96 (0·94–0·99) 0·002 7445 (52·2%) 62·87% (61·44–64·30) 0·96 (0·94–0·99) 0·001 migrants Individuals prescribed psychoactive drugs Low SEP 13 497 (4·0%) 4·17% (4·10–4·24) 1 (ref) ·· 15 972 (4·5%) 4·66% (4·58–4·73) 1 (ref) ·· Very low SEP 1542 (5·3%) 5·74% (5·46–6·03) 1·38 (1·31–1·45) <0·001 1985 (6·5%) 6·90% (6·59–7·20) 1·48 (1·41–1·55) <0·001 Undocumented 469 (3·5%) 4·17% (3·79–4·54) 1·00 (0·91–1·09) 0·96 552 (3·9%) 4·65% (4·26–5·04) 1·00 (0·92–1·09) 0·98 migrants Data are n (%) unless specified. IR=incidence rate. IRR=incidence rate ratio. SEP=socioeconomic position. Table 2: Use of health-care services in those younger than 15 years by sex and SEP in Catalonia, 2017 migrants younger than 15 years (4·64%) and undocu­ lower than those in the very low SEP group (41·29%; mented female migrants (2·91%; table 2). Among table 3). individuals aged 15–64 years, the differences between the Regarding sex differences, hospital admission incidence lowest SEP group and undocu­ mented migrants were rates per 1000 inhabitants among those younger than greater—eg, 9·19% of female participants in the very low 15 years were higher in male participants (69·87% in SEP group attended mental health centres compared undocumented migrants and 72·10% in the very low with 1·45% of female participants in the undocumented SEP group) than in female participants (53·50% in migrant population (table 3). Similarly, the mean number undocumented migrants and 58·83% in the very low SEP of visits was highest in the very low SEP group, in both group; table 2). Female participants aged 15–64 years had age groups and for both sexes. In all SEP groups, more the highest hospital admission incidence rates across all male participants younger than 15 years attended mental groups. Among those younger than 15 years, no differ­ health centres and made more visits than female ences in surgical hospital admission incidence rates participants younger than 15 years. The relationship was between very low SEP and the undocumented migrant reversed in the attended adult population, in which group were observed, but the rate was higher in male female participants attended more often than male participants than in female participants. By contrast, participants. This pattern was observed in all SEP groups. among individuals aged 15–64 years, the very low SEP For acute care, hospital admission rates among male group had higher rates (58·73% in female participants participants with a very low SEP were more than three and 51·99% in male participants) than undocumented times those of undocumented migrants and male migrants (female participants 39·33%, male partici­ participants with a low SEP. Among female participants, pants 36·64%), who had the lowest rates. hospital admission rates were 69% higher in the very low Individuals with a very low SEP had the highest SEP group (136·52%) than in the undocumented migrant psychiatric hospital admission incidence rates per group (80·98%). Childbirth hospital admission incidence 1000 inhabitants (10·01% [95% CI 9·39–10·63]), while rates among undocumented migrants (34·78%) were rates were similar in the low SEP (4·62% [4·49–4·74]) and www.thelancet.com/planetary-health Vol 5 May 2021 e290 Articles undocumented migrants groups (5·63% [4·34–6·92]; groups, and the lowest proportion of prescriptions was table 3). In the case of male participants younger than found in undocumented migrants (table 3). 15 years, the psychiatric hospital admission incidence rate Individuals younger than 15 years in the very low SEP ranged between 0·65% (95% CI 0·57–0·74) in those with group had the highest percentage of prescriptions of a low SEP and 1·28% (0·89–1·68) in those with a very low psycho­active drugs, with higher values in male partici­ SEP (table 2). Among individuals aged 15–64 years, pants (6·5%) than in female participants (5·3%). In the incidence rates of psychiatric hospitalisation were higher other groups, the rates ranged from 3·5% in female in male participants than in female participants across all participants who were undocumented migrants to 4·5% SEP groups. in male participants with a low SEP. Among individuals Approximately 72% of individuals younger than aged 15–64 years, psychoactive drug prescription was 15 years in the very low SEP group had been prescribed higher in female participants than in male participants at least one drug, compared with 61% in the low SEP across all SEP groups. The number of individuals group and 51% in the undocumented migrant group; the prescribed psychoactive drugs in the very low SEP group values for male participants and female participants were (female participants 37·4%, male partici­ pants 25·7%) similar (table 2). Among individuals aged 15–64 years, was between three or four times higher than in undocu­ the proportion of drug prescriptions was higher in mented migrants (female participants 8·5%, male female participants than in male participants for all SEP participants 4·6%). Female participants aged 15–64 years Male participants aged 15–64 years Number IR (95% CI) IRR (95% CI) p value Number IR (95% CI) IRR (95% CI) p value Individuals attended in primary care Low SEP 1 188 697 (76·7%) 77·96% (77·82–78·10) 1 (ref) ·· 940 639 (64·9%) 66·89% (66·75–67·02) 1 (ref) ·· Very low SEP 97 245 (85·0%) 84·82% (84·29–85·36) 1·09 (1·08–1·10) <0·001 69 138 (72·1%) 72·18% (71·64–72·73) 1·08 (1·07–1·09) <0·001 Undocumented migrants 25 942 (61·0%) 82·59% (81·58–83·60) 1·06 (1·05–1·07) <0·001 19 819 (46·0%) 64·25% (63·35–65·15) 0·96 (0·95–0·97) <0·001 Mean primary care visits Low SEP 8 271 791 (5·3%) 5·44% (5·43–5·44) 1 (ref) ·· 5 654 627 (3·9%) 4·13% (4·13–4·14) 1 (ref) ·· Very low SEP 896 395 (7·8%) 7·49% (7·47–7·50) 1·38 (1·37–1·38) <0·001 528 230 (5·5%) 5·31% (5·30–5·33) 1·29 (1·28–1·29) <0·001 Undocumented migrants 134 258 (3·2%) 4·61% (4·59–4·64) 0·85 (0·84–0·85) <0·001 87 691 (2·0%) 3·37% (3·35–3·40) 0·82 (0·81–0·82) <0·001 Individuals attended at emergency services Low SEP 494 016 (31·9%) 32·46% (32·37–32·55) 1 (ref) ·· 408 375 (28·2%) 28·69% (28·60–28·78) 1 (ref) ·· Very low SEP 46 932 (41·0%) 43·16% (42·77–43·56) 1·33 (1·32–1·34) <0·001 34 613 (36·1%) 37·31% (36·92–37·71) 1·30 (1·29–1·31) <0·001 Undocumented migrants 11 783 (27·7%) 34·94% (34·31–35·57) 1·08 (1·06–1·10) <0·001 9517 (22·1%) 28·54% (27·96–29·11) 0·99 (0·97–1·02) 0·61 Mean emergency services visits Low SEP 1 047 855 (0·7%) 0·69% (0·69–0·69) 1 (ref) ·· 805 092 (0·6%) 0·57% (0·57–0·57) 1 (ref) ·· Very low SEP 131 912 (1·2%) 1·24% (1·23–1·24) 1·79 (1·78–1·80) <0·001 92 336 (0·96%) 0·99% (0·98–0·99) 1·74 (1·73–1·75) <0·001 Undocumented migrants 25 902 (0·6%) 0·75% (0·74–0·76) 1·09 (1·08–1·10) <0·001 19 393 (0·5%) 0·60% (0·59–0·61) 1·06 (1·04–1·07) <0·001 Individuals attended at mental health centres Low SEP 57 207 (3·7%) 3·83% (3·80–3·86) 1 (ref) ·· 44 079 (3·0%) 3·17% (3·14–3·20) 1 (ref) ·· Very low SEP 11 223 (9·8%) 9·19% (9·02–9·36) 2·40 (2·35–2·45) <0·001 8321 (8·7%) 8·30% (8·12–8·48) 2·62 (2·56–2·68) <0·001 Undocumented migrants 402 (0·9%) 1·45% (1·31–1·59) 0·38 (0·34–0·42) <0·001 417 (0·9%) 1·41% (1·27–1·54) 0·44 (0·40–0·49) <0·001 Mean mental health centre visits Low SEP 366 298 (0·2%) 0·24% (0·24–0·24) 1 (ref) ·· 335 892 (0·2%) 0·24% (0·23–0·24) 1 (ref) ·· Very low SEP 87 695 (0·8%) 0·71% (0·71–0·72) 2·95 (2·92–2·97) <0·001 76 742 (0·8%) 0·76% (0·75–0·76) 3·22 (3·19–3·24) <0·001 Undocumented migrants 2013 (0·1%) 0·07% (0·07–0·07) 0·29 (0·28–0·31) <0·001 2754 (0·1%) 0·09% (0·09–0·09) 0·38 (0·37–0·40) <0·001 Hospital admission rate, per 1000 inhabitants Low SEP 137 376 (88·6%) 88·86% (88·39–89·34) 1 (ref) ·· 98 847 (68·2%) 74·95% (74·47–75·42) 1 (ref) ·· Very low SEP 15 622 (136·5%) 134·09% 1·51 (1·48–1·53) <0·001 10 709 (111·6%) 104·19% 1·39 (1·36–1·42) <0·001 (131·98–136·21) (102·20–106·18) Undocumented migrants 3445 (81·0%) 109·66% 1·23 (1·19–1·28) <0·001 1536 (35·7%) 74·09% 0·99 (0·94–1·04) 0·66 (105·98–113·33) (70·37–77·82) Childbirth hospital admission rate, per 1000 inhabitants aged 15–49 years Low SEP 31 331 (28·2%) 27·62% (27·31–27·93) 1 (ref) ·· ·· ·· ·· ·· Very low SEP 2392 (34·9%) 41·29% (39·63–42·95) 1·49 (1·43–1·56) <0·001 ·· ·· ·· ·· Undocumented migrants 924 (41·9%) 34·78% (32·53–37·02) 1·26 (1·18–1·34) <0·001 ·· ·· ·· ·· (Table 3 continues on next page) e291 www.thelancet.com/planetary-health Vol 5 May 2021 Articles Female participants aged 15–64 years Male participants aged 15–64 years Number IR (95% CI) IRR (95% CI) p value Number IR (95% CI) IRR (95% CI) p value (Continued from previous page) Surgical hospital admission rate, per 1000 inhabitants Low SEP 70 460 (45·5%) 46·26% (45·92–46·61) 1 (ref) ·· 60 208 (41·6%) 45·35% (44·98–45·72) 1 (ref) ·· Very low SEP 7623 (66·6%) 58·73% (57·41–60·06) 1·27 (1·24–1·30) <0·001 5316 (55·4%) 51·99% (50·58–53·40) 1·15 (1·11–1·18) <0·001 Undocumented migrants 979 (23·0%) 39·33% (36·86–41·80) 0·85 (0·80–0·91) <0·001 796 (18·5%) 36·64% (34·08–39·20) 0·81 (0·75–0·87) <0·001 Potentially avoidable hospital admission rate, per 1000 inhabitants Low SEP 4469 (2·9%) 3·01% (2·92–3·09) 1 (ref) ·· 5702 (3·9%) 4·62% (4·49–4·74) 1 (ref) ·· Very low SEP 1069 (9·3%) 8·31% (7·81–8·81) 2·77 (2·59–2·96) <0·001 1031 (10·7%) 10·01% (9·39–10·63) 2·17 (2·03–2·32) <0·001 Undocumented migrants 98 (2·3%) 4·22% (3·38–5·06) 1·40 (1·15–1·72) <0·001 74 (1·7%) 5·63% (4·34–6·92) 1·22 (0·97–1·54) 0·09 Psychiatric hospital admission rate, per 1000 inhabitants Low SEP 5345 (3·5%) 3·48% (3·38–3·57) 1 (ref) ·· 7 308 (5·0%) 4·94% (4·83–5·06) 1 (ref) ·· Very low SEP 1718 (15·0%) 14·47% (13·77–15·16) 4·16 (3·94–4·39) <0·001 2194 (22·9%) 23·03% (22·05–24·02) 4·66 (4·44–4·89) <0·001 Undocumented migrants 93 (2·2%) 3·06% (2·44–3·69) 0·88 (0·72–1·08) <0·001 188 (4·4%) 5·35% (4·58–6·12) 1·08 (0·94–1·25) 0·29 Mean number of prescriptions Low SEP 17 805 981 (11·5%) 11·95% (11·94–11·95) 1 (ref) ·· 13 989 629 (9·7%) 10·80% (10·79–10·80) 1 (ref) ·· Very low SEP 3 265 954 (28·5%) 24·97% (24·95–25·00) 2·09 (2·09–2·09) <0·001 1 990 103 (20·7%) 18·95% (18·92–18·98) 1·75 (1·75–1·76) <0·001 Undocumented migrants 176 748 (4·2%) 7·68% (7·65–7·72) 0·64 (0·64–0·65) <0·001 99 657 (2·3%) 5·67% (5·64–5·71) 0·53 (0·52–0·53) <0·001 Individuals prescribed at least one drug Low SEP 1 076 223 (69·4%) 70·63% (70·49–70·76) 1 (ref) ·· 801 898 (55·4%) 57·30% (57·17–57·43) 1 (ref) ·· Very low SEP 94 035 (82·2%) 81·25% (80·73–81·78) 1·15 (1·14–1·16) <0·001 64 655 (67·4%) 66·17% (65·66–66·68) 1·15 (1·15–1·16) <0·001 Undocumented migrants 21 519 (50·6%) 69·63% (68·69–70·56) 0·99 (0·97–1·00) <0·001 15 297 (35·5%) 52·17% (51·34–53·00) 0·91 (0·90–0·93) <0·001 Individuals prescribed psychoactive drugs Low SEP 325 812 (21·0%) 21·21% (21·13–21·28) 1 (ref) ·· 179 671 (12·4%) 12·81% (12·75–12·87) 1 (ref) ·· Very low SEP 427 66 (37·4%) 32·77% (32·46–33·09) 1·55 (1·53–1·56) <0·001 24 667 (25·7%) 23·64% (23·35–23·94) 1·85 (1·82–1·87) <0·001 Undocumented migrants 3605 (8·5%) 14·26% (13·79–14·73) 0·67 (0·65–0·69) <0·001 1957 (4·6%) 7·67% (7·33–8·01) 0·60 (0·57–0·63) <0·001 Data are n (%) unless specified. IR=incidence rate. IRR=incidence rate ratio. SEP=socioeconomic position. Table 3: Use of health-care services in individuals aged 15–64 years by sex and SEP in Catalonia, 2017 For chronic diseases, the same SEP pattern was For all infectious diseases, male participants had higher observed in COPD, cirrhosis, and neoplasms: the highest diagnosis rates than female participants (table 4). diagnosis incidence rate was observed in the very low SEP group, followed by the low SEP group, and then by Discussion the undocumented migrants (table 4). The highest The results of this study showed that undocumented absolute differences between the very low SEP and migrants made less use of health-care services than undocumented migrants were observed in asthma for individuals in the low and very low SEP groups. The female participants (very low SEP 9·28%, undocumented diagnosis rate of chronic diseases was also lower among migrants 2·71%). In terms of relative differences, COPD undocumented migrants than in their peers with a low or diagnosis incidence rate was 76% lower in female very low SEP. By contrast, the diagnosis rate of HIV and participants who were undocumented migrants (0·55 per HCV in undocumented migrants was higher than in the 100 person-years, incidence rate ratio 0·24 [95% CI low SEP group but lower than in the very low SEP group, 0·20–0·29]) than in female participants with a low SEP and tuberculosis rates in male undocumented migrants (2·29 per 100 person-years). COPD and cirrhosis were were the highest compared with the other groups. more prevalent in male participants than in female Sex and age differences were also observed in most of participants, but asthma, diabetes, and cancer were more the indicators and for all SEP groups. These results are prevalent in female participants. similar to those observed in previous work by our group For infectious diseases, the very low SEP group had the for the general population, in different age and SEP highest diagnosis rate of HIV and HCV, followed by groups.16 It was found that although undocumented undocumented migrants and the low SEP group. In migrants made less use of health-care services than legal tuberculosis, male undocumented migrants presented the residents, their patterns of use of health-care services highest diagnosis incidence rate (4·35 per 100 person-years), with regard to sex and age were similar. For example, followed by the very low SEP group (1·90 per 100 person- male participants younger than 15 years attended years), and the low SEP group (0·77 per 100 person-years). emergency care more frequently than female participants; www.thelancet.com/planetary-health Vol 5 May 2021 e292 Articles Female participants aged 15–64 years Male participants aged 15–64 years Number IR (95% CI) IRR (95% CI) p value Number IR (95% CI) IRR (95% CI) p value Diabetes Low SEP 54 698 (3·5%) 3·76% (3·73–3·79) 1 (ref) ·· 81 125 (5·6%) 6·51% (6·47–6·56) 1 (ref) ·· Very low SEP 9015 (7·9%) 6·63% (6·50–6·77) 1·76 (1·72–1·80) <0·001 8720 (9·1%) 8·08% (7·91–8·25) 1·24 (1·21–1·27) <0·001 Undocumented migrants 1043 (2·5%) 5·49% (5·16–5·83) 1·46 (1·37–1·55) <0·001 798 (1·9%) 6·22% (5·79–6·66) 0·96 (0·89–1·02) 0·20 Chronic obstructive pulmonary disease Low SEP 33 765 (2·2%) 2·29% (2·27–2·32) 1 (ref) ·· 45 605 (3·2%) 3·66% (3·63–3·70) 1 (ref) ·· Very low SEP 5702 (5·0%) 4·27% (4·16–4·38) 1·86 (1·81–1·91) <0·001 6291 (6·6%) 6·02% (5·87–6·17) 1·64 (1·60–1·69) <0·001 Undocumented migrants 118 (0·3%) 0·55% (0·45–0·65) 0·24 (0·20–0·29) <0·001 192 (0·5%) 1·35% (1·16–1·54) 0·37 (0·32–0·42) <0·001 Asthma Low SEP 109 760 (7·1%) 7·24% (7·19–7·28) 1 (ref) ·· 86 600 (6·0%) 6·15% (6·11–6·19) 1 (ref) ·· Very low SEP 10 271 (9·0%) 9·28% (9·10–9·46) 1·28 (1·26–1·31) <0·001 5706 (5·9%) 6·65% (6·47–6·82) 1·08 (1·05–1·11) <0·001 Undocumented migrants 882 (2·1%) 2·71% (2·53–2·89) 0·37 (0·35–0·40) <0·001 636 (1·5%) 1·73% (1·59–1·86) 0·28 (0·26–0·30) <0·001 Cirrhosis Low SEP 4287 (0·3%) 0·29% (0·28–0·30) 1 (ref) ·· 8640 (0·6%) 0·68% (0·67–0·70) 1 (ref) ·· Very low SEP 881 (0·8%) 0·63% (0·59–0·67) 2·15 (2·00–2·32) <0·001 1559 (1·6%) 1·36% (1·29–1·43) 1·99 (1·89–2·10) <0·001 Undocumented migrants 49 (0·1%) 0·26% (0·19–0·33) 0·89 (0·67–1·18) 0·41 76 (0·2%) 0·58% (0·45–0·71) 0·85 (0·67–1·06) 0·15 Neoplasms Low SEP 67 616 (4·4%) 4·52% (4·49–4·56) 1 (ref) ·· 42 565 (2·9%) 3·39% (3·36–3·43) 1 (ref) ·· Very low SEP 6857 (6·0%) 5·02% (4·90–5·14) 1·11 (1·08–1·14) <0·001 3478 (3·6%) 3·41% (3·29–3·52) 1·00 (0·97–1·04) 0·81 Undocumented migrants 486 (1·1%) 2·31% (2·10–2·52) 0·51 (0·47–0·56) <0·001 244 (0·6%) 1·60% (1·40–1·80) 0·47 (0·42–0·53) <0·001 HIV Low SEP 3730 (0·2%) 0·23% (0·22–0·24) 1 (ref) ·· 11 590 (0·8%) 0·79% (0·78–0·81) 1 (ref) ·· Very low SEP 1394 (1·2%) 1·00% (0·95–1·06) 4·34 (4·08–4·62) <0·001 2718 (2·8%) 2·31% (2·22–2·40) 2·91 (2·78–3·03) <0·001 Undocumented migrants 155 (0·4%) 0·58% (0·49–0·68) 2·53 (2·15–2·97) <0·001 509 (1·2%) 1·97% (1·80–2·15) 2·48 (2·27–2·71) <0·001 Tuberculosis Low SEP 759 (0·5%) 0·48% (0·45–0·52) 1 (ref) ·· 1096 (0·8%) 0·77% (0·72–0·82) 1 (ref) ·· Very low SEP 121 (1·1%) 1·02% (0·84–1·20) 2·11 (1·74–2·56) <0·001 198 (2·1%) 1·90% (1·63–2·17) 2·47 (2·12–2·88) <0·001 Undocumented migrants 34 (0·8%) 1·09% (0·72–1·46) 2·25 (1·60–3·18) <0·001 119 (2·8%) 4·35% (3·55–5·16) 5·66 (4·67–6·86) <0·001 Hepatitis C virus Low SEP 5419 (3·5%) 3·50% (3·40–3·59) 1 (ref) ·· 10 293 (7·1%) 7·43% (7·28–7·57) 1 (ref) ·· Very low SEP 1527 (13·3%) 10·67% (10·13– 3·05 (2·88–3·23) <0·001 3359 (35·0%) 26·55% (25·64–27·46) 3·57 (3·44–3·72) <0·001 11·21) Undocumented migrants 112 (2·6%) 5·01% (4·08–5·94) 1·43 (1·19–1·73) 0·001 405 (9·4%) 22·35% (20·15–24·55) 3·01 (2·72–3·32) <0·001 Data are n (%) unless specified. IR=incidence rate. IRR=incidence rate ratio. SEP=socioeconomic position. Table 4: People attended with chronic and infectious diseases aged 15–64 years by sex and SEP in Catalonia, 2017 female participants aged 15–64 years attended primary services by undocumented migrants matches their needs. care and emergency care more frequently than male In this study, we restricted our analysis to the population participants; and male participants younger than 15 years with earnings less than €18 000 per year. We did so because attended mental health services more frequently than in Catalonia, 25% of the population has additional private female participants and had a higher prescription of insurance, a proportion that rises to 47% in the high- psychoactive drugs.16 income bracket.18 People with higher incomes tend to use Therefore, the results of this study challenge the claims private services that offer greater comfort and faster access often made in Spain by some political parties that to care for minor conditions; for serious diseases and undocumented migrants make greater use of health-care interventions, all SEP groups access the public health-care services than the native population.17 In general, as the system. Thus, because our aim was to compare undocu­ results show, the use of undocumented migrants is lower mented migrants with the native population according to than that of the native population. This finding has SEP group, we decided to exclude middle and upper two implications: undocumented migrants are accessing socioeconomic groups because their lower use of public the health-care system, but, their lower usage might health-care services would probably skew the results. We indicate some barriers to access. Related to this point, we should also emphasise that individuals are assigned to a have no way of knowing whether the use of health-care particular pharmacy copayment group on the basis of e293 www.thelancet.com/planetary-health Vol 5 May 2021 Articles their income tax declarations, a procedure that is not housing situation.24 Regarding the undocumented guaranteed to provide an accurate assessment of their migrant population, the only quantitative (survey-based) income. Regarding undocumented migrants, because of study found in the literature reported no differences in their particular characteristics, we decided to treat them as the use of health-care services between legal and illegal a unique group and compare them with legal residents of migrants when ill, but a significantly lower use of health diverse SEPs. Moreover, as they do not have legal services was associated with lower levels of education.27 employment, they cannot be classified in an earnings Two relevant systematic reviews in the international group. literature showed that undocumented migrants in The study has important strengths. Firstly, it analysed Europe faced particular problems when using health- the health and health-care use of undocumented migrants, care services, especially with regard to maternal and whose data are exhaustively recorded in administrative child health care.28,29 databases, as is the case for all Catalan citizens. The Health status, disease spectra, and use of prescribed Catalan Health Department does not routinely cross-check medication have been reported to vary in different groups registration requests with other Catalan Governmental of migrants and according to the different phases of institutions, nor with the Spanish Government, as migration.30,31 Migration is an independent determinant undocumented migrants are considered full citizens with of health, but it also interacts with other socioeconomic a right to health. We do not know the exact percentage of factors.30 Migrants are usually reported to be healthier those who are registered because there were no alternative than their peers, a phenomenon known as the healthy data sources available with which to validate our figures, migrant effect.32 Other factors that interact are the but we believe it to be high based on the fact that the precarious living and working conditions, legal barriers, objective of the Catalan Health Department is to maintain cultural and language barriers, and even severe forms of universal coverage. Registration as a special interest group exclusion, fear, and stigma.33 Earlier studies assessed the is very easy: it can be achieved by approaching the health- use of primary health services by different groups of care system or by means of third-sector mediators and migrants,34,35 including children.36–38 Similar studies of citizen services officers. undocumented migrants are scarce, as their legal status Second, this paper analysed individual information from makes it impossible to identify them in administrative more than 4 million people in Catalonia, with no additional databases, or even to quantify them.39 cost and without jeopardising data anonymisation, and Regarding chronic diseases, the statistics for COPD thus shows the value of reusing administrative data for and asthma suggest that these conditions are likely to be research purposes, something that European authorities underdiagnosed among undocumented migrant adults recommend. Especially relevant is the fact that individual in primary care settings. The low rates observed might be SEP could be obtained from administrative registries.19,20 due to the lower frequency of their visits to primary care Government policies related to health care, but primarily services, or to lower disease severity.40 Access barriers due those related to migration, have a profound influence on to sociocultural norms or fear of deportation might also migrant health, and restrictive entry and integration play a role.41,42 policies have been linked to poor migrant health outcomes In the case of HCV and HIV, the results showed slightly in high-income countries.21 Between 2012 and 2018, the higher rates in undocumented migrants than in the low Catalan Government continued to allow undocumented SEP group, and similar rates to those in the very low SEP migrants the same access to health care as the native group. In 2014 in Europe, around 40% of people diagnosed population. There were three reasons for this decision: to with HIV were migrants. Sub-Saharan African and Latin avoid damaging migrants’ health status; to avoid American or Caribbean migrants are the most at-risk increasing their susceptibility as a group at risk of social migrant populations for HIV,43 and these two groups exclusion; and to prevent public health issues caused by accounted for most of the undocumented migrant the spread of some infectious diseases. Some preliminary population in our study. In addition, HCV diagnosis rate evaluations of the Spanish legislation have shown is higher in low-income and middle-income countries substantial reductions in specialist visits and hospital than in Europe, and studies have shown rates ranging admissions among undocumented migrants,22 but also between 9% in migrants from sub-Saharan Africa and increased mortality in this group.23 However, these studies 15% in migrants from Eastern Europe.44 have been unable to accurately identify the undocumented Our results showed a much higher tuberculosis rate migrant population. among undocumented migrants than in all the other Studies done in Spain of the regular migrant population SEP groups. Approximately three out of ten cases of have not identified any clear patterns in health-care use, tuberculosis in the EU and European Economic Area are as the differences recorded depend mainly on the reported among individuals born outside of Europe, and migrants’ country of origin.24–26 Although migrants make Spain, France, Germany, and the UK account for less use of specialised care and more use of emergency 74·9% of cases of foreign origin.45 The country of origin care,25 these differences often disappear when adjusting of migrants also influences tuberculosis rates, due to the for other confounders, such as employment status or high prevalence (>250 per 100 000 population in some www.thelancet.com/planetary-health Vol 5 May 2021 e294 Articles countries), which might explain the higher diagnosis rate 8 Dias S, Gama A, Cortes M, de Sousa B. Healthcare-seeking patterns observed in our study, at least in men. among immigrants in Portugal. Health Soc Care Community 2011; 19: 514–21. For WHO, universal health coverage is, by definition, a 9 Brindicci G, Trillo G, Santoro CR, Volpe A, Monno L, Angarano G. practical expression of the concern for health equity and Access to health services for undocumented immigrants in Apulia. the right to health, and is an essential component of its J Immigr Minor Health 2015; 17: 618–23. 10 Goossens MCM, Depoorter AM. Contacts between general new Sustainable Development Goals.46,47 To quote the practitioners and migrants without a residence permit and the use declaration on universal health coverage by economists of “urgent” medical care. Scand J Public Health 2011; 39: 649–55. from 44 countries published in The Lancet in 2015, “the 11 Castañeda H. Paternity for sale: anxieties over “demographic theft” and undocumented migrant reproduction in Germany. intrinsic value of improved health…points to maintaining Med Anthropol Q 2008; 22: 340–59. and expanding commitment to health through investing 12 Fedeli U, Alba N, Lisiero M, Zambon F, Avossa F, Spolaore P. in pro-poor pathways to universal health coverage”.48 Obstetric hospitalizations among Italian women, regular and Health care is not universal if undocumented migrants irregular immigrants in North-Eastern Italy. Acta Obstet Gynecol Scand 2010; 89: 1432–37. are excluded.49 13 Natter K, Czaika M, de Haas H. Political party ideology and Cohort studies taking into account country of origin, immigration policy reform: an empirical enquiry. Pol Res Exchange time spent in the host country, individual characteristics, 2020; 2: 1735255. 14 Bernal-Delgado E, García-Armesto S, Oliva J, et al. Spain: Health and SEP group would be valuable to analyse the relative system review. Health Syst Transit 2018; 20: 1–179. effect of such variables on the use of health-care services 15 CatSalut. Despesa sanitària a Catalunya. Despesa sanitària pública and disease diagnosis over time. More sound evidence is sobre el PIB. Barcelona, 2019. https://catsalut.gencat.cat/ca/coneix- catsalut/informacio-economica/despesa-sanitaria-catalunya needed to disprove populist arguments.50 (accessed July 26, 2019). Contributors 16 García-Altés A, Ruiz-Muñoz D, Colls C, Mias M, Martín Bassols N. All authors had full access to all the data in the study and had final Socioeconomic inequalities in health and the use of healthcare responsibility for the decision to submit for publication. AD-B did all the services in Catalonia: analysis of the individual data of 7.5 million data management, statistical analysis, and table preparation. AG-A did residents. J Epidemiol Community Health 2018; 72: 871–79. the final write-up of the Article and coordinated the project. EV and MC 17 Bruquetas-Callejo M, Perna R. Migration and healthcare reforms in helped in the planning of the project and supervised the statistical Spain: symbolic politics, converging outputs, oppositions from the analysis and the manuscript preparation. CVP did the data extraction field. South Eur Soc Polit 2020; 25: 75–98. and data curation. JMA had the idea for the article and led the project. 18 Barcelona: Ministry of Health. Government of Catalonia. Health AD-B, EV, and CVP have verified the underlying data. No medical writer Survey of Catalonia Health status, health-related behaviours and use of health services in Catalonia. ESCA 2018 main results. Executive or editor has been involved in the creation of the manuscript. summary. 2019. http://salutweb.gencat.cat/web/.content/_ Declaration of interests departament/estadistiques-sanitaries/enquestes/Enquesta-de-salut- We declare no competing interests. de-Catalunya/Resultats-de-lenquesta-de-salut-de-Catalunya/ documents/2018-resultats/summary-esca-2018.pdf (accessed Data sharing July 26, 2019). Requests for data should be sent to the corresponding author. 19 Jutte DP, Roos LL, Brownell MD. Administrative record linkage as a References tool for public health research. Annu Rev Public Health 2011; 1 Asylum statistics. Eurostat Statistics Explained. Luxembourg: 32: 91–108. Eurostat. European Commission. https://ec.europa.eu/eurostat/ 20 EUR-Lex. Access to European Union law. Directive 2013/37/EU of statistics-explained/index.php/Asylum_statistics (accessed the European Parliament and of the Council of 26 June 2013 Dec 14, 2020). amending Directive 2003/98/EC on the re-use of public sector 2 OM’s Regional Office for the EEA. the European Union and NATO. information Text with EEA relevance. 2013. https://eur-lex.europa. OM Resettlement Support in the EEA 2019. Brussels: International eu/legal-content/EN/TXT/?uri=CELEX%3A32013L0037 (accessed Organization for Migration (IOM), 2020. https://eea.iom.int/sites/ Oct 16, 2019). default/files/publication/document/IOM%20resettlement%20 21 Juárez SP, Honkaniemi H, Dunlavy AC, et al. Effects of non-health- support%20EEA%20-%202019.pdf (accessed Dec 14, 2020). targeted policies on migrant health: a systematic review and 3 Madrid: Instituto Nacional de Estadística. Flujo de inmigración meta-analysis. Lancet Glob Health 2019; 7: e420–35. procedente del extranjero por año, sexo y edad. 2019. https://www. 22 Jiménez Rubio D, Vall Castelló J. Effects on utilisation, health and ine.es/dyngs/INEbase/es/operacion.htm?c=Estadistica_C&cid= user satisfaction when access to health care is limited. Barcelona: 1254736177000&menu=ultiDatos&idp=1254735573002 (accessed Centre de Recerca en Economia i Salut (CRES). Universitat Dec 14, 2020; in Spanish). Pompeu Fabra (UPF), 2018. CRES-UPF Working Paper #201803- 4 Madrid: Gobierno de España. Ministerio del Interior. Immigración 105. https://www.upf.edu/documents/3223410/7582912/ irregular. Informe quincenal. Datos acumulados del 1 de enero a CRESWP201803105.pdf/fdd31a8c-903b-8527-9abf-f902d949d914 31 de diciembre 2019. 2019. http://www.interior.gob.es/ (accessed July 26, 2019). documents/10180/11261647/informe_quincenal_acumulado_01-01_ 23 Juanmarti A, López G, Vall J. The deadly effects of losing health al_31-12-2019.pdf/97f0020d-9230-48b0-83a6-07b2062b424f (accessed insurance. Eur Econ Rev 2021; 131: 103608. Dec 14, 2020; in Spanish). 24 Saurina C, Vall-Llosera L, Sáez M. Factors determining access to 5 Madrid: BOE Official State Gazette. Royal Decree 1192/2012, and use of primary health care services in the Girona Health Region of August 3rd, on the status of insured and beneficiary for the (Spain). Eur J Health Econ 2012; 13: 419–27. purposes of health care in Spain, with public funds, through 25 Llop-Gironés A, Vargas Lorenzo I, Garcia-Subirats I, Aller MB, The National Health System is regulated. 2012. https://www.boe.es/ Vázquez Navarrete ML. Acceso a los servicios de salud de la eli/es/rd/2012/08/03/1192/con (accessed Dec 14, 2020). población inmigrante en España. Rev Esp Salud Publica 2014; 6 Barcelona: DOGC Official Gazette of Catalonia. Law 9/2017, of 88: 715–34 (in Spanish). June 27th, on universal healthcare with public funds of the Catalan 26 Cainzos-Achirica M, Vela E, Cleries M, et al. Cardiovascular risk Health Service. 2017. https://portaljuridic.gencat.cat/eli/ factors and disease among non-European immigrants living in es-ct/l/2017/06/27/9 (accessed Dec 14, 2020). Catalonia. Heart 2019; 105: 1168–74. 7 Madrid: BOE Official State Gazette. Royal Decree 7/2018, of July 27th, 27 Torres-Cantero AM, Miguel AG, Gallardo C, Ippolito S. Health care on the universal access to the Spanish National Health System. 2018. provision for illegal migrants: may health policy make a difference? https://www.boe.es/eli/es/rdl/2018/07/27/7 (accessed Dec 14, 2020). Eur J Public Health 2007; 17: 483–85. e295 www.thelancet.com/planetary-health Vol 5 May 2021 Articles 28 Winters M, Rechel B, de Jong L, Pavlova M. A systematic review on 39 Jandl M. The estimation of illegal migration in Europe. Studi Emigr the use of healthcare services by undocumented migrants in 2004; XLI: 141–55. Europe. BMC Health Serv Res 2018; 18: 30. 40 Carballo M, Maclean EC, Gudumac I, Van Damme P. Hepatitis C 29 de Jong L, Pavlova M, Winters M, Rechel B. A systematic literature and migration: a public health challenge. J Fam Med 2016; 3: 1065. review on the use and outcomes of maternal and child healthcare 41 Mladovsky P, Ingleby D, McKee M, Rechel B. Good practices in services by undocumented migrants in Europe. Eur J Public Health migrant health: the European experience. Clin Med 2012; 2017; 27: 990–97. 12: 248–52. 30 Rechel B, Mladovsky P, Ingleby D, Mackenbach JP, McKee M. 42 Willen SS. Migration, “illegality,” and health: mapping embodied Migration and health in an increasingly diverse Europe. Lancet vulnerability and debating health-related deservingness. Soc Sci Med 2013; 381: 1235–45. 2012; 74: 805–11. 31 Balcazar AJ, Grineski SE, Collins TW. The Hispanic health paradox 43 Bednarek M, Maciejewski J, Wozniak M, Kuca P, Zielinski J. across generations: the relationship of child generational status and Prevalence, severity and underdiagnosis of COPD in the primary citizenship with health outcomes. Public Health 2015; 129: 691–97. care setting. Thorax 2008; 63: 402–07. 32 Gimeno-Feliu LA, Calderón-Larrañaga A, Diaz E, Poblador-Plou B, 44 Benayoun S, Sondey J, Rojas-Castro D. Euro HIV EDAT Project Macipe-Costa R, Prados-Torres A. The healthy migrant effect in WP8. Access to HIV testing and linkage to care for migrant primary care. Gac Sanit 2015; 29: 15–20. populations in Europe. Results of the qualitative study. Pantin, 33 Lebano A, Hamed S, Bradby H, et al. Migrants’ and refugees’ health France: AIDES, 2017. status and healthcare in Europe: a scoping literature review. 45 European Centre for Disease Prevention and Control/WHO BMC Public Health 2020; 20: 1039. Regional Office for Europe. Tuberculosis surveillance and 34 Uiters E, Devillé W, Foets M, Spreeuwenberg P, Groenewegen PP. monitoring in Europe 2016. Stockholm: ECDC, 2016. https://ecdc. Differences between immigrant and non-immigrant groups in the europa.eu/en/publications-data/tuberculosis-surveillance-and- use of primary medical care; a systematic review. monitoring-europe-2016 (accessed July 26, 2019). BMC Health Serv Res 2009; 9: 76. 46 Ooms G, Hammonds R. Anchoring universal health coverage in 35 Klein J, von dem Knesebeck O. Inequalities in health care the right to health: What difference would it make? Policy brief. utilization among migrants and non-migrants in Germany: Geneva: World Health Organization, 2015. https://www.who.int/ a systematic review. Int J Equity Health 2018; 17: 160. gender-equity-rights/knowledge/anchoring-uhc.pdf?ua=1 (accessed 36 Gimeno-Feliu LA, Armesto-Gómez J, Macipe-Costa R, July 26, 2019). Magallón-Botaya R. Comparative study of paediatric prescription 47 The Lancet Public Health. High ambitions for health equity in the drug utilization between the Spanish and immigrant population. Americas. Lancet Public Health 2019; 4: e538. BMC Health Serv Res 2009; 9: 225. 48 Summers LH. Economists’ declaration on universal health 37 Ballotari P, D’Angelo S, Bonvicini L, et al. Effects of immigrant coverage. Lancet 2015; 386: 2112–13. status on Emergency Room (ER) utilisation by children under age 49 Legido-Quigley H, Pocock N, Tan ST, et al. Healthcare is not one: a population-based study in the province of Reggio Emilia universal if undocumented migrants are excluded. BMJ 2019; (Italy). BMC Health Serv Res 2013; 13: 458. 366: l4160. 38 Fadnes LT, Møen KA, Diaz E. Primary healthcare usage and 50 Wickramage K, Annunziata G. Advancing health in migration morbidity among immigrant children compared with governance, and migration in health governance. Lancet 2018; non-immigrant children: a population-based study in Norway. 392: 2528–30. BMJ Open 2016; 6: e012101. www.thelancet.com/planetary-health Vol 5 May 2021 e296