RESEARCH Open Access
O e use o unde use? Use o heal hca e
se ices among i egula mig an s in a
no h-eas e n Spanish egion
Luis And és Gimeno-Feliu
1,2,3,4*
, Ma a Pas o -Sanz
3,5
, Bea iz Poblado -Plou
1,3
, Amaia Calde ón-La añaga
1,3,6
,
Espe anza Díaz
7,8
and Alexand a P ados-To es
1,3
Abs ac
Backg ound: The e is li le e i ied in o ma ion on global heal hca e u iliza ion by i egula mig an s.
Unde s anding how immig an s use heal hca e se ices based on hei needs is c ucial o es ablish e ec i e heal h
policy. We compa ed heal hca e u iliza ion be ween i egula mig an s, documen ed mig an s, and Spanish
na ionals in a Spanish au onomous communi y.
Me hods: This e ospec i e, obse a ional s udy included he o al adul popula ion o A agon, Spain: 930,131
Spanish na ionals; 123,432 documen ed mig an s; and 17,152 i egula mig an s. Heal hca e u iliza ion da a we e
compa ed be ween i egula mig an s, documen ed mig an s and Spanish na ionals o he yea 2011. Mul i a iable
s anda d o ze o-in la ed nega i e binomial eg ession models we e gene a ed, adjus ing o age, sex, leng h o
s ay, and mo bidi y bu den.
Resul s: The a e age annual use o heal hca e se ices was lowe o i egula mig an s han o documen ed
mig an s and Spanish na ionals a all le els o ca e analyzed: p ima y ca e (0.5 s 4 s 6.7 isi s); specialized ca e (0.2
s 1.8 s 2.9 isi s); planned hospi al admissions (0.3 s 2 s 4.23 pe 100 indi iduals), unplanned hospi al admissions
(0.5 s 3.5 s 5.2 pe 100 indi iduals), and eme gency oom isi s (0.4 s 2.8 s 2.8 pe 10 indi iduals). The
a e age annual p esc ip ion d ug expendi u e was also lowe o i egula mig an s (€9) han o documen ed
mig an s (€77) and Spanish na ionals (€367). These di e ences we e only pa ially a enua ed a e adjus ing o age,
sex, and mo bidi y bu den.
Conclusions: Unde condi ions o equal access, heal hca e u iliza ion is much lowe among i egula mig an s han
Spanish na ionals (and lowe han ha o documen ed mig an s), ega dless o coun y o o igin o leng h o s ay in
Spain.
Keywo ds: Emig a ion and immig a ion, Undocumen ed mig an , Heal hca e dispa i ies, Heal h se ices accessibili y,
Hospi aliza ion, P ima y heal hca e, Eme gency se ice, D ug u iliza ion, Heal hca e use, Spain
© The Au ho (s). 2021 Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e
app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i
changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle's C ea i e Commons
licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle's C ea i e Commons
licence and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain
pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he
da a made a ailable in his a icle, unless o he wise s a ed in a c edi line o he da a.
* Co espondence: lugi [email protected]
1
EpiCh on Resea ch G oup on Ch onic Diseases, A agón Heal h Sciences
Ins i u e (IACS), IIS A agón, Miguel Se e Uni e si y Hospi al, Za agoza, Spain
2
A agón Heal hca e Se ice, San Pablo Heal h Cen e, Za agoza, Spain
Full lis o au ho in o ma ion is a ailable a he end o he a icle
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41
h ps://doi.o g/10.1186/s12939-020-01373-3
Backg ound
Acco ding o he 2018 Uni ed Na ions Mig a ion Repo ,
he numbe o mig an s wo ldwide eached 244 million
in 2015 and is expec ed o inc ease u he [1]. A small
bu no able po ion o he gene al mig an popula ion
consis s o mig an s wi hou legal au ho iza ion o eside
in hei hos coun y. The e m i egula mig an e e s
o “a pe son who, owing o unau ho ized en y, b each
o a condi ion o en y, o he expi y o his o he isa,
lacks legal s a us in a ansi o hos coun y”[2].I in-
cludes pe sons who (a) lack he necessa y documen a-
ion o legally en e a coun y bu do so clandes inely;
(b) en e o s ay in a coun y using audulen documen a-
ion; o (c) a e en e ing a coun y wi h alid legal docu-
men a ion, s ay beyond he pe iod au ho ized o o he wise
iola e he e ms o en y and emain wi hou au ho iza ion
[2]. This e m is conside ed a synonym o “undocumen ed
mig an ”, which e e s o he indi idual’s adminis a i e
si ua ion and is one o he mos widely used and accep ed
e ms [2–4]. The e m encompasses isa “o e s aye s”,
hose who ha e los esiden s a us, ejec ed asylum seeke s,
and indi iduals who ha e en e ed a coun y illegally [3–5].
In 2008 an es ima ed 1.9–3.8 million i egula mig an s
we e li ing in he Eu opean Union [4].
In ecen yea s, se e al coun ies ha e es ic ed public
heal hca e access o his popula ion, a guing ha mi-
g an s mig a e o hos coun ies o a ail o ea men o
p e-exis ing medical condi ions [6–8]. Cuad a e al. classi-
ied EU coun ies in o h ee main g oups acco ding o he
a ailabili y o heal hca e se ices o mig an s [9](summa-
ized in Fig. 1). Immig an s in Spain accoun o 12.2% o
he popula ion (12.7% in A agon), and mig a e o Spain
p ima ily o economic easons [10]. The Spanish na ional
heal h sys em p o ides uni e sal co e age and is almos
ully unded by axes. Ca e p o ision is ee o cha ge a
he poin o deli e y, esul ing in a p ac ically ee sys em.
P ima y ca e cen es se e as ga ekeepe s and a e dis ib-
u ed o ensu e app op ia e geog aphical co e age. While
he Spanish public heal h sys em is decen alized and
managed by he egional go e nmen o each au onomous
communi y, he legisla ion go e ning access o he na-
ional heal h sys em applies o he en i e s a e [11]. Be-
ween 2000 and 2012, immig an s we e gua an eed legal
access o he same heal hca e se ices as Spanish na-
ionals, ega dless o legal s a us. Du ing ha pe iod, Spain
was one o he mos p og essi e coun ies in he wo ld in
e ms o heal hca e access o i egula mig an s [12]. In
2012, in esponse o he pe sis en e ec s o he 2008 eco-
nomic c isis, a cen al go e nmen dec ee wi hd ew his
igh , in alida ing he heal h ca ds o i egula mig an s
[13–15]. The go e nmen a gued, wi hou o e ing e i-
dence, ha his adical change in heal hca e access would
help limi alleged “heal h ou ism”and imp o e he sus-
ainabili y o Spain’s na ional heal h sys em. This measu e
was accompanied by a a o o he cu backs a ec ing
heal hca e and social se ices. Acco ding o epo s by he
Spanish go e nmen , his policy a ec ed 870,000 i egula
mig an s [13]. Since hen, he e has been con inued poli -
ical deba e o e he es o a ion o uni e sal heal hca e
co e age.
As ecommended by he WHO and o he Human
Righ s o ganiza ions, heal hca e p o ision is he minimum
Aus ia
Bulga ia
Czech Republic
Finland
I eland
La ia
Luxembou g
Mal a
Romania
Sweden
Belgium
Cyp us
Denma k
Es onia
Ge many
G eece
Hunga y
Li huania
Poland
Slo ak Republic
Slo enia
UK
F ance
I aly
Ne he lands
Po ugal
Spain
Less han minimum
igh s o access
Minimum
igh s o access
Mo e han minimum
igh s o access
Adap ed om Cuad a CB e al: Ri
g
h o access o heal h ca e o undocumen ed mi
g
an s in EU: a compa a i e s udy o na ional policies. Eu J Public Heal h 2012,22:267-271.
Fig. 1 Na ional policy ega ding igh s o access o heal h ca e o i egula mig an s in he EU
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 2 o 11
condi ion o gua an ee equi y in heal h [16]. In 2018, he
socialis go e nmen o Ped o Sánchez passed Royal De-
c ee Law 7/2018 on uni e sal access o he na ional heal h
sys em. While he esul ing changes imp o ed heal hca e
access, hey did no es o e uni e sal heal h co e age
ac oss Spain. Royal Dec ee Law 7/2018 is cu en ly being
e ised by he Spanish pa liamen , and has become a i s -
o de poli ical issue, gi en he widesp ead end in Eu ope
and beyond o wi hholding uni e sal heal h co e age om
i egula mig an s [7,17,18].
Published da a on he use o heal hca e se ices by i -
egula mig an s is sca ce [5,17]. A ecen sys ema ic e-
iew by Win e s e al. highligh ed he pauci y o such
s udies [8]. Mo eo e , he indings o he ew s udies
conduc ed o da e a e limi ed by small sample popula-
ions, a lack o adjus men o con ounding ac o s, and
an absence o con ol g oups, as emphasized in a scop-
ing e iew by Woodwa d e al. [19].
By acing i egula mig an s whose Spanish public
heal h ca ds we e in alida ed in 2012, i is possible o
examine heal hca e use du ing he p e ious yea , when
he e was no dis inc ion be ween i egula mig an s,
documen ed mig an s, and Spanish na ionals in e ms o
public heal hca e access. Ou s udy analyses he eal use
o he Spanish public heal hca e sys em by a la ge coho
o mig an s ( he la ges published o da e), including un-
documen ed mig an s, du ing a pe iod (2011) when mi-
g an s had unlimi ed heal h sys em access ega dless o
legal s a us. Gi en ha ew na ional heal hca e sys ems
o e uni e sal co e age o i egula mig an s, his ana-
lysis p o ides an impo an window in o heal hca e
usage in his speci ic con ex . The indings may he e o e
be o in e es o policymake s and esea che s seeking o
imp o e heal hca e sys ems and ensu e heal h equi y o
he mos disad an aged popula ions.
Me hods
This c oss-sec ional popula ion-based e ospec i e s udy
analyzed clinical and adminis a i e da a om indi iduals
assigned o all public p ima y ca e (PC) cen es in A agon,
Spain, in 2011. These da a we e o iginally collec ed as pa
o he EpiCh on Coho , which ga he s demog aphic, clin-
ical, and pha maceu ical da a om elec onic heal h e-
co ds (EHRs) and he heal h insu ance da abase o
almos all inhabi an s o A agon, using a unique anon-
ymized pe sonal iden i ica ion code [20,21]. The A agon
heal h se ice is pa o he Spanish na ional heal h sys-
em, which p o ides ee ca e and medical es ing, and is
unded by axes. Ca e p o ision is ee o cha ge a he
poin o deli e y, esul ing in a p ac ically ee sys em [22].
PC cen es se e as ga ekeepe s and gua an ee app op ia e
geog aphical co e age. Seconda y ca e is p o ided h ough
ambula o y specialized ca e, hospi als, and eme gency
ooms. Pha maceu icals p esc ibed o hose unde 65
equi e a co-paymen o 40% o he e ail p ice (o less in
he case o ch onic medica ion); medica ion is o he wise
ee o cha ge a he poin o deli e y.
Fo each pa ien aged 18 and olde , demog aphic a i-
ables including age, sex, coun y o bi h, and leng h o
esidence in A agon we e ex ac ed om he heal h
insu ance da abase o he yea 2011. Immig an s we e
de ined as any o eign-bo n pe son, ega dless o na ion-
ali y o du a ion o esidence in Spain (2). The s udy
popula ion was ca ego ized as Spanish na ionals, docu-
men ed mig an s, o i egula mig an s, de ined as any
indi idual whose heal h ca d was in alida ed as a esul
o Royal Dec ee-Law 16/2012 [13]. The mig an
popula ion was classi ied acco ding o hei a ea o
o igin (A ica, Asia, Eas e n Eu ope, La in Ame ica,
and Wes e n Eu ope & No h Ame ica).
Diagnos ic da a we e ex ac ed om EHRs and om he
Hospi al Minimum Basic Da ase (Spanish ac onym,
CMBD). In he o me , diseases a e egis e ed acco ding
o he In e na ional Classi ica ion o P ima y Ca e, Ve -
sion 1 (ICPC-1). The la e ga he s he diagnoses o pa-
ien s discha ged om all public and p i a e hospi als,
coded using he Clinical Modi ica ion-Nin h Re ision o
he In e na ional Classi ica ion o Diseases (ICD-9-CM).
The Adjus ed Clinical G oup (ACG) Sys em® was used o
g oup all ICPC and ICD diagnos ic codes based on du -
a ion, se e i y, diagnos ic ce ain y, ae iology, and special-
ized ca e in ol emen . A unique ACG ca ego y was
assigned o each indi idual based on age, sex, and all diag-
noses egis e ed du ing he s udy pe iod. Indi iduals
wi hin a gi en ACG show simila pa e ns o mo bidi y
and esou ce u iliza ion o e a gi en yea . ACGs wi h
simila expec ed use o esou ces we e agg ega ed in o
one o he six so-called esou ce u iliza ion bands (RUB 0
= non use s; RUB 1 = heal hy use s; RUB 2 = low mo bid-
i y; RUB 3 = mode a e mo bidi y; RUB 4 = high mo bidi y;
and RUB 5 = e y high mo bidi y). Each indi idual was
hus addi ionally assigned a RUB ca ego y.
PC use was de ined as he numbe o isi s o he PC
doc o and nu se, including on-demand, scheduled,
eme gency, and home isi s. Specialized ca e u iliza ion
was measu ed as he o al numbe o isi s o any spe-
cialis . Hospi al ca e included planned and unplanned
admissions, and he o al numbe o hospi al days. The
use o eme gency oom se ices was measu ed as he
o al numbe o isi s and p io i y isi s. P io i y isi s
we e iden i ied based on he iage le el es ablished by
he heal h se ice o A agon; ou o he i e ca ego ies
lis ed, le els 1–3 a e assigned o p io i y isi s. P esc ip-
ion d ug use was measu ed as he o al annual expend-
i u e using ecommended e ail d ug p ices [23].
The s udy was app o ed by he Clinical Resea ch E h-
ics Commi ee o A agon (CEICA).
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 3 o 11
S a is ical analysis
The use o each le el o ca e and p esc ip ion d ug ex-
pendi u e we e de e mined acco ding o mig an s a us.
Gi en he o e -dispe sion in he dis ibu ion o he ou -
come a iables, nega i e binomial eg ession models we e
applied o de e mine he associa ion be ween he la e
and indi iduals' a ea o o igin h ough incidence a e a ios
(IRR) and 95% con idence in e als. E alua ion o he
o e -dispe sion pa ame e alpha allowed e i ica ion o he
adequacy o he nega i e binomial eg ession models wi h
espec o Poisson coun models. Because he obse ed
ou come da a o en con ained a highe ela i e equency
o ze os han is consis en wi h nega i e binomial model
speci ica ions, ze o-in la ed models we e used. The Vuong
closeness es was used o assess he app op ia eness o
he ze o-in la ed models. When he Vuong es was no
signi ican , and p oduced la ge nega i e alues, s anda d
nega i e binomial models we e employed. When non-
conca e egions epea edly appea ed, use o a s epping al-
go i hm as an al e na i e o he s anda d maximum likeli-
hood algo i hm was pe mi ed. No mal Poisson models
and ze o-in la ed Poisson models we e also pe o med (see
Supplemen a y Table 2) as a sensi i i y analysis. Mo eo e ,
we epea ed ou analysis o pha macy cos s using an o -
dina y leas squa es analysis (i.e. linea eg ession), he e-
sul s o which a e shown in Supplemen a y Table 4.
All models we e s a i ied by sex and a ea o o igin and
adjus ed o age, mo bidi y bu den, and leng h o s ay.
Leng h o s ay in Spain was ca ego ized as < 5 o ≥5 yea s
as p e iously desc ibed [24]. When analyzing subg oups
based on sex and a ea o o igin, he small sample size o
some subg oups p e en ed con e gence o he models.
These esul s a e he e o e no p esen ed (i.e. pha macy use
among Asian immig an s). All s a is ical analyses we e pe -
o med using STATA / IC 12.
Resul s
Da a o 1,070,715 indi iduals we e analyzed: 930,131
Spanish na ionals; 123,432 documen ed mig an s; and
17,152 i egula mig an s. Table 1p esen s he main
demog aphic cha ac e is ics, mo bidi y bu den, and
heal hca e u iliza ion da a o Spanish na ionals and im-
mig an s in Spain, wi hou adjus ing o any a iable.
The dis ibu ion o he mig an popula ion acco ding o
a ea o o igin is shown in Supplemen a y Table 1.The
mean age o i egula and documen ed mig an s was simi-
la , and in bo h cases was lowe han ha o Spanish na-
ionals. The pe cen age o women was lowe among
mig an s han Spanish na ionals, and was lowes among
i egula mig an s. Use o he public heal h sys em by i -
egula mig an s was lowe han ha o documen ed mi-
g an s and Spanish na ionals, in e ms o bo h he numbe
o isi s and hospi aliza ions pe yea : PC, 0.5 s 4.0 s 6.7;
ou pa ien hospi al isi s, 0.2 s 1.8 s 2.9; planned
hospi al admissions (pe 100 indi iduals), 0.3 s 2 s 4.2;
unplanned hospi al admissions (pe 100 indi iduals), 0.5
s 3.5 s 5.2; eme gency oom isi s (pe 10 indi iduals),
0.4 s 2.8 s 2.8. Pha macy expendi u e was also lowe
among i egula mig an s han bo h documen ed mig an s
and Spanish na ionals: €8.7 s €77.4 s €366.5. Leng h o
s ay (< 5 yea s s ≥5 yea s) did no signi ican ly in luence
he use o heal hca e se ices by i egula mig an s.
Figu e 2p esen s heal h sys em u iliza ion da a ad-
jus ed o age and s a i ied by sex, aking Spanish na-
ionals as he e e ence popula ion. The same da a,
adjus ed o age and disease bu den, a e also p esen ed.
The age-adjus ed models e ealed lowe use o heal h-
ca e se ices o i egula immig an s a all le els o ca e
(IRR be ween 0.1–0.2 o bo h men and women). A e
adjus ing o disease bu den, hese di e ences dec eased,
bu emained signi ican , wi h IRR alues be ween 0.15–
0.3.
Figu e 3shows heal h sys em u iliza ion adjus ed o
age and disease bu den and s a i ied by a ea o o igin
and sex. In all cases, he use o heal hca e se ices by i -
egula mig an s was lowe han ha o Spanish na-
ionals. Among i egula mig an s, he use o heal hca e
se ices was highes among hose om A ica, ollowed
by Eas e n Eu ope and Wes e n Eu ope & No h Ame -
ica. Heal hca e use was lowes o hose om Asia.
Supplemen a y Table 2 depic s he use o heal hca e se -
ices by immig an s acco ding o legal s a us (IRR), as de-
e mined using no mal Poisson, ze o-in la ed Poisson, and
s anda d o ze o-in la ed nega i e binomial eg ession
models. The esul s ob ained we e e y simila ac oss
models. Supplemen a y Table 3 depic s he use o heal h-
ca e se ices by immig an s acco ding o legal s a us, ad-
jus ed o sex, age, mo bidi y bu den, and addi ionally by
a ea o o igin (IRR). The esul s ob ained we e e y simila
o hose shown in Fig. 1. The analysis o pha macy cos s
was epea ed using an o dina y leas squa es analysis (i.e.
linea eg ession) (Supplemen a y Table 4), which e ealed
indings compa able o hose ob ained using he s anda d
o ze o-in la ed nega i e binomial eg ession.
Discussion
Summa y o he indings
Global use o heal hca e se ices
The use o heal hca e se ices by i egula mig an s was
e y low, ega dless o he le el o ca e and a e adjus -
ing o age and disease bu den (IRR be ween 0.15–0.3,
a e no maliza ion o he co esponding alues o
Spanish na ionals). The use o heal hca e se ices among
i egula mig an s was also much lowe han ha ob-
se ed o documen ed mig an s. These esul s a e con-
sis en wi h he indings o a sys ema ic e iew by Win e s
e al. [8]. Two ques ionnai e-based s udies conduc ed in
Po ugal by Dias e al. epo ed indings simila o ou s oo,
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 4 o 11
Table 1 Demog aphics, mo bidi y bu den, and heal hca e u iliza ion o Spanish na ionals and immig an s in Spain
Spanish
na ionals
Fo eign-bo n (all) A ica Asia Eas e n Eu ope La in Ame ica Wes e n Eu ope & No h
Ame ica
DM IM DM IM DM IM DM IM DM IM DM IM
Demog aphic in o ma ion
N930,131 123,432 17,152 29,643 2563 4621 327 39,678 7429 39,829 4643 9661 2190
Women, % 51.42 48.55 33.76 31.96 18.77 43.24 37.61 51.23 32.12 59.09 45.53 47.49 31.37
Age
Mean (SD) 51.91 (0.020) 38.18 (0.034) 38.06 (0.083) 37.39 (0.063) 37.85 (0.204) 38.49 (0.184) 39.71 (0.664) 36.71 (0.053) 37.23 (0.118) 38.51 (0.063) 38.89 (0.170) 45. 12 (0.147) 39.16 (0.248)
18–44 yea s, % 39.71 74.11 76.10 78.27 80.53 72.80 69.11 78.79 78.58 71.39 72.73 53.97 70.68
45–64 yea s, % 32.14 22.94 22.11 19.48 17.17 23.31 26.91 20.26 20.62 25.14 24.6 35.26 26.94
65+ yea s, % 28.16 2.95 1.79 2.25 2.3 3.90 3.98 0.95 0.79 3.47 2.67 10.78 2.37
Leng h o s ay in Spain
≥5 yea s, %
─68.61 63.23 62.77 73.55 80.65 85.32 53.65 52.4 75.73 78.35 57.55 52.51
Mo bidi y bu den
Heal hy use s/non-use s, % 21.15 36.02 75.29 36.64 73.16 47.72 79.51 41.15 79.26 29.92 68.49 32.61 78.13
Low/mode a e mo bidi y, % 72.03 60.54 24.31 59.65 26.3 49.49 20.18 56.01 20.35 66.39 31.12 63.02 21.55
High/ e y high mo bidi y, % 6.81 3.44 0.4 3.71 0.55 2.79 0.31 2.84 0.39 3.68 0.39 4.38 0.32
Use o P ima y Ca e
No isi s o doc o , % 22.83 34.19 86.5 33.35 79.17 46.09 90.83 38.32 87.66 29.18 85.89 34.71 91.78
No isi s o nu se, % 48.55 71.42 95.56 70.8 93.64 80.09 97.86 74.69 95.76 69.3 95.48 64.47 96.94
Mean (SD) # o isi s o
doc o , no mal ca e
6.7 (8.4) 4.0 (5.6) 0.5 (2.0) 4.1 (5.6) 0.8 (2.2) 2.7 (4.6) 0.2 (0.8) 3.3 (5.1) 0.5 (1.9) 4.5 (5.8) 0.6 (2.2) 4.5 (6.6) 0.3 (1.8)
Mean (SD) # o isi s o
doc o , u gen ca e
0.4 (1.0) 0.4 (1.1) 0.001 (0.1) 0.6 (1.3) 0.004 (0.1) 0.2 (0.9) 0 (0) 0.4 (1.0) 0.0003 (0.2) 0.3 (0.9) 0.0006 (0.04) 0.3 (1.0) 0.002 (0.1)
Mean (SD) # o isi s o nu se 3.7 (7.7) 1.0 (3.5) 0.1 (0.7) 1.1 (3.5) 0.1 (0.8) 0.8 (3.4) 0.6 (5.1) 0.8 (3.3) 0.1 (0.8) 1.1 (3.3) 0.1 (0.6) 1.8 (4.8) 0.1 (0.7)
Use o Specialized Ca e
No isi s, % 46.86 61.15 93.48 66.52 92.51 69.88 97.25 65.37 94.18 52.33 91.19 59.57 96.48
Mean (SD) # o isi s 2.9 (4.9) 1.8 (3.7) 0.2 (1.3) 1.5 (3.6) 0.3 (1.6) 1.4 (3.5) 0.1 (0.7) 1.4 (3.3) 0.2 (1.2) 2.2 (4.0) 0.3 (1.5) 2.1 (4.3) 0.1 (0.9)
Use o Hospi al Ca e
Mean (SD) # o planned
admissions/100 ind.
4.2 (26.0) 2.0 (17.3) 0.3 (5.4) 1.5 (14.0) 0.3 (6.8) 1.4 (13.2) 0 (0) 1.8% (18.3) 0.2 (4.6) 2.6 (18.6) 0.4 (6.7) 2.8 (18.6) 0.1 (3.0)
Mean (SD) # o unplanned
admissions/100 ind.
5.2 (28.4) 3.5 (21.1) 0.5 (8.1) 3.8 (21.8) 0.6 (7.9) 3.6 (21.5) 0.6 (7.8) 3.1 (19.8) 0.5 (8.4) 3.5 (21.1) 0.5 (7.5) 3.6 (24.1) 0.4 (8.8)
Mean (SD) hospi al s ay, days 6.7 (8.4) 4.7 (9.7) 6.0 (1.2) 4.4 (6.4) 8.5 (17.9) 5.0 (6.3) 4 (2.8) 4.6 (14.4) 4.6 (6.5) 4.5 (6.4) 6.5 (15.7) 6.1 (10.1) 6.7 (5.5)
Use o Eme gency Ca e
No isi s, % 81.83 81.95 96.93 83.06 96.06 84.48 96.64 82.8 96.92 79.2 96.64 85.16 98.63
Mean (SD) # o isi s/10 ind. 2.8 (8.0) 2.8 (7.8) 0.4 (3.2) 2.8 (8.2) 0.5 (3.3) 2.4 (7.1) 0.4 (2.2) 2.7 (7.5) 0.4 (3.4) 3.3 (8.0) 0.5 (3.2) 2.2 (6.9) 0.2 (2.5)
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 5 o 11

Table 1 Demog aphics, mo bidi y bu den, and heal hca e u iliza ion o Spanish na ionals and immig an s in Spain (Con inued)
Spanish
na ionals
Fo eign-bo n (all) A ica Asia Eas e n Eu ope La in Ame ica Wes e n Eu ope & No h
Ame ica
DM IM DM IM DM IM DM IM DM IM DM IM
High p io i y isi s, % 51.2 39.6 44.5 36.7 46.6 37.2 27.3 37.5 41.8 42.2 50.3 46.0 33.3
Pha macy expendi u e
Ze o expendi u e, % 24.0 39.4 89.6 36.0 83.8 50.6 93.6 44.8 90.6 35.3 88.8 38.6 94.4
Mean (SD) expendi u e, €366.5 (763.4) 77.4 (302.4) 8.7 (90.1) 70.4 (314.3) 10.4 (70.0) 64.3 (248.2) 2.7 (16.0) 53.9 (225.8) 9.1 (115.2) 82.9 (279.9) 9.7 (73.8) 179.3 (543.0) 4.6 (34.9)
Abb e ia ions: DM documen ed mig an , ind indi iduals, SD s anda d de ia ion, IM i egula mig an
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 6 o 11
Fig. 3 Use o heal hca e se ices by immig an s acco ding a ea o o igin, legal s a us, and sex (exp essed as incidence a e a ios, IRR): esul s o
s anda d o ze o-in la ed nega i e binomial eg ession models. Abb e ia ions: DM, documen ed mig an ; IM, i egula mig an . Models adjus ed
o age and mo bidi y bu den. Da a a e no malized o co esponding alues o Spanish na ionals. In hose cases whe e he Vuong es was
s a is ically non-signi ican , s anda d nega i e binomial models we e used
Fig. 2 Use o heal hca e se ices by immig an s acco ding o legal s a us, and sex (exp essed as incidence a e a ios, IRR): esul s o s anda d o
ze o-in la ed nega i e binomial eg ession models. Abb e ia ions: DM, documen ed mig an ; IM, i egula mig an . Da a a e no malized o
co esponding alues o Spanish na ionals. In hose cases whe e he Vuong es was s a is ically non-signi ican , s anda d nega i e binomial
models we e used
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 7 o 11
al hough he au ho s did no obse e as g ea a di e ence
be ween i egula and documen ed mig an s [25,26].
Howe e , ou esul s di e om hose o To es-Can e o
e al., who in e iewed 380 Ecuado ians in Mad id and ob-
se ed no di e ence in heal hca e u iliza ion be ween i -
egula and documen ed mig an s [27]. I is possible ha
in ha s udy, limi ing he s udy popula ion o a e y spe-
ci ic g oup ha was in e iewed on he s ee in a single
dis ic in Mad id may ha e led o some deg ee o selec-
ion o memo y bias.
P ima y ca e
PC u iliza ion, a e adjus ing o disease bu den, was
much lowe among i egula mig an s han Spanish na-
ionals (72% and 69% ewe isi s o men and women,
espec i ely). Recen s udies pe o med in he
Ne he lands also epo ed educed PC u iliza ion among
i egula mig an s, al hough he di e ences wi h espec
o documen ed mig an s (37% [28] and 36% [29] ewe
isi s pe yea ) we e no as g ea as hose obse ed in
ou popula ion. I should be no ed ha , despi e hese
ma ked di e ences, PC ( oge he wi h eme gency ca e)
was he le el o ca e o which he smalles p opo ional
di e ences we e obse ed be ween i egula mig an s
and Spanish na ionals. This sugges s g ea e equi y o
PC u iliza ion in Spain, which has al eady been
highligh ed in o he publica ions [30].
Hospi aliza ions, hospi al isi s, and eme gency ca e
The numbe s o o al and unplanned hospi aliza ions
and hospi al ou pa ien isi s we e much lowe among
i egula mig an s han ei he documen ed mig an s o
Spanish na ionals. Al hough i egula mig an s made
ewe eme gency oom isi s han ei he documen ed
mig an s o Spanish na ionals, he di e ences obse ed
o his le el o ca e, as o PC, we e ela i ely small.
One po en ial explana ion o his minimal di e ence is
ha eme gency ca e is a ailable 24 h a day, 7 days a
week, and he e o e i egula mig an s wi h job insecu -
i y can mo e easily a ail o his ype o ca e.
Pha macy expendi u e
Analysis o pha macy expendi u e e ealed he g ea es
di e ences among g oups o all ou comes analyzed. A
he ime a which his s udy was conduc ed, pa ien s
wi h a p esc ip ion om he Spanish na ional heal h se -
ice paid 40% o he cos o acu e medica ion and 10% o
he cos o ch onic medica ion, wi h a maximum limi o
€2.64 pe package. Medica ion was ee o inpa ien s and
“exemp ”g oups (i.e. e i ees and hose who ha e disabil-
i ies o ha e had occupa ional acciden s) [22]. The exis -
ence o his co-paymen may ha e dissuaded i egula
mig an s om e illing hei p esc ip ions.
In e p e a ion o he indings
Se e al possible scena ios could help explain ou ind-
ings. On he one hand, al hough he heal hca e access
o i egula mig an s was equal o ha o Spanish na-
ionals and documen ed mig an s, heal hca e
u iliza ion by i egula mig an s may be lowe due o
ba ie s o accessibili y ela ed o hei wo k ci cum-
s ances. Fo ins ance, hose wo king in he unde -
g ound economy a e less likely o be able o a ail o
wo k lea e [14,31]. A quali a i e s udy o i egula
La in Ame ican women wo king in Spain ound ha
he main de e minan o hei lowe heal hca e
u iliza ion was p eca ious wo k and exploi a ion by
employe s [32]. In a s udy o i egula mig an
women in he Ne he lands, 40% epo ed ba ie s o
heal hca e access ha included ea , shame, a lack o
in o ma ion, and inancial di icul ies [33].
Ano he po en ial ba ie o heal hca e access is he
ea ha he use ’s da a will be ans e ed o he au ho -
i ies, po en ially esul ing in depo a ion om he coun-
y. Al hough such da a sha ing was p ohibi ed in Spain
a he ime hese da a we e ga he ed, he ac ha his
has occu ed in o he coun ies may ha e in luenced i -
egula immig an s' way o hinking [8,14,31,34]. Ye
ano he po en ial ba ie is di icul y communica ing due
o a lack o knowledge o he local language [31,33,35],
al hough his e ec would be expec ed o wane wi h in-
c eased leng h o s ay in he hos coun y, and his was
no obse ed in ou s udy popula ion. Fu he mo e, in
such case, a g ea e use o heal hca e se ices would be
expec ed in he La in Ame ican popula ion (who mos ly
speak Spanish), and his was no bo ne ou by ou ind-
ings (Fig. 2).
One o he impo an ac o is he eme gence in a i-
ous coun ies o “an i-immig a ion”policies, blaming im-
mig an s, and especially i egula mig an s, o
nume ous social and economic p oblems [31,36]. These
policies can ein o ce eelings o ejec ion pe cei ed by
i egula mig an s, making hem mo e hesi an o ully
exe cise hei igh s, ei he ou o ea o pe cei ed “un-
wo hiness”[14,31,36]. La chanche e al. a ibu e his
si ua ion o he exis ence o “in angible obs acles”, such
as s igma iza ion, s uc u al iolence, and ea [14]. Fi-
nally, “an i-immig a ion” he o ic can in luence heal h
se ice pe sonnel, gi ing ise o disc imina o y and p e-
judiced a i udes ha cause i egula mig an s o s op
a ailing o hese se ices [14,36].
In any case, he a ailable e idence indica es ha he
lowe use o heal hca e se ices among i egula mi-
g an s is mo e closely linked o social ac o s ela ed o
hei “illegal”s a us (e.g. a lack o wo ke ’s igh s, ea o
depo a ion, economic di icul ies, e c.) han o simply
“being an immig an ”, which is likely o ha e a lesse im-
pac [1,8,14,31].
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 8 o 11
Fo bo h i egula and documen ed mig an s, heal hca e
u iliza ion was highes among hose om A ica and low-
es among hose om Asia. The majo i y o Asians in ou
s udy popula ion we e om China, whe e he e is a s ong
cul u e o he use o adi ional Chinese medicine [37].
Fu he mo e, in Spain many immig an s o Asian o igin
wo k in small businesses wi h e y long hou s ha may
limi hei access o heal hca e se ices. The high ela i e
use o heal hca e se ices among immig an s o A ican
o igin is in line wi h he indings o a small Du ch s udy o
emale i egula mig an s [33], and hose o wo s udies
ha did no di e en ia e mig an s acco ding o legal s a-
us, one conduc ed in Spain by ou g oup using adminis-
a i e da a [35] and ano he Po uguese ques ionnai e-
based s udy [26]. We ha e no clea explana ion o his
pa icula inding. A s udy conduc ed in Po ugal ound
ha A ican mig an s exp essed g ea e sa is ac ion wi h
he Po uguese heal hca e sys em han hose om Eas e n
Eu ope [25]. This di e ence wa an s u he speci ic
esea ch.
The lack o in luence o leng h o s ay on heal hca e
u iliza ion among i egula mig an s is no able. A p e ious
s udy o his same coho , in which he e was no dis inc-
ion be ween documen ed and i egula mig an s, ound
ha heal hca e u iliza ion inc eased wi h leng h o s ay
[35]. In line wi h ou p esen indings, a Du ch s udy o 80
i egula mig an women (o mainly A ican and Eas e n
Eu ope o igin) ound ha leng h o s ay had no e ec on
he sel - a ed heal h s a us o i egula mig an s, al hough
he s udy in ques ion lacked s a is ical powe [34]. One
possible eason is ha documen ed mig an s ind i easie
o in eg a e and assimila e o he cul u al and socie al
no ms o hei hos coun y han i egula mig an s. This
ac o could con ibu e o he less medicalized heal h cul-
u e o i egula mig an s as compa ed wi h he na i e
popula ion o hei hos coun y [23,35]. Fu he mo e,
documen ed mig an s gene ally imp o e hei economic
si ua ion o e ime, allowing hem g ea e access o ma-
e ial goods such as medicines. This can lead o di e -
ences in expec a ions and p io i ies. I is also possible ha
documen ed mig an s, who ha e a legal igh o employ-
men , may ha e access o less p eca ious jobs wi h co e-
sponding wo ke ’s igh s, enabling hem o ake lea e
om wo k o medical consul a ions. Fu he mo e, as dis-
cussed abo e, i egula mig an s may be mo e ea ul o
being de ained o ejec ed by heal h o icials and less
awa e abou heal h ela ed “dese ingness”[14].
Analysis o sex- ela ed di e ences e ealed ha he
ela i e di e ence in heal hca e u iliza ion be ween mi-
g an s and Spanish na ionals ended o be smalle o
women han o men. This inding con adic s ha o a
Du ch s udy in which i was hypo hesized ha i egula
mig an women ha e poo e access o heal hca e se -
ices han hei male coun e pa s [33].
S eng hs
These da a, despi e hei e ospec i e na u e, a e pa -
icula ly impo an as hey cons i u e he la ges se ies
o heal hca e da a om undocumen ed mig an s pub-
lished o da e [8], in a e y speci ic con ex : a na ional
heal hca e sys em o e ing uni e sal co e age. The e-
sul s ob ained om his analysis can he e o e be o
in e es o policymake s and esea che s seeking o im-
p o e uni e sal heal hca e sys ems and ensu e heal h
equi y o disad an aged popula ions based on he bes
a ailable scien i ic e idence. These da a, ex ac ed om
adminis a i e da abases, we e p e iously collec ed as
pa o he EpiCh on coho [20]. Mo eo e , because
he Spanish public heal h sys em p o ided mig an s
wi h uni e sal co e age ega dless o hei legal s a us
du ing he pe iod o which hese da a co espond, he
isk o selec ion bias in ou da ase is limi ed. The da a-
se includes all indi iduals en i led o heal hca e, e-
ga dless o whe he hey had any con ac wi h he
heal h sys em ha yea . The use o EHR da a a oids
biases associa ed wi h in o ma ion p o ided di ec ly by
he pa ien .
Ano he key s eng h o ou s udy is he compa ison o
heal hca e u iliza ion be ween i egula and documen ed
mig an s wi h a con ol g oup (i.e. Spanish na ionals). The
use o adminis a i e da a allowed us o s udy he impac
o he leng h o s ay. Mo eo e , hanks o he la ge sample
size and di e si y o he mig an popula ion in A agon, i
was possible o u he classi y hem by a ea o o igin. By
measu ing indi idual-le el mo bidi y bu den using an
in e na ionally alida ed ool and da a om EHRs, we en-
su ed a b oad and eliable assessmen o he heal hca e
needs o he immig an popula ion [35]. Fu he mo e, by
simul aneously analysing he use o p ima y ca e, special-
ized ca e, hospi aliza ions, eme gency oom isi s, and
p esc ip ion d ug expendi u e, ou analysis allowed a glo-
bal assessmen o he Spanish heal h sys em. This global
app oach p o ides a comp ehensi e o e iew o he main
heal h esou ces in ol ed, allowing us o de e mine
whe he unde use o one esou ce can en ail o e use o
o he s [35,38].
Weaknesses
Se e al limi a ions o he p esen s udy should be no ed.
Ou analyses did no conside socio-economic a iables
such as income o educa ion le el, inclusion o which
could ha e helped iden i y some o he complex ac o s
ha condi ion he use o heal hca e se ices [39,40]. This
impo an pe sonal in o ma ion is no eco ded in Spanish
heal hca e da abases and could no be ob ained in any
o he way while p ese ing anonymi y. The quali y o
heal hca e da a ex ac ed om EHRs can also be called
in o ques ion. Howe e , he da abases om which ou
da a we e acqui ed ha e been p e iously alida ed o use
Gimeno-Feliu e al. In e na ional Jou nal o Equi y in Heal h (2021) 20:41 Page 9 o 11